From 65ccb7a1daf3cea9fc4c1289907b195c5555aba5 Mon Sep 17 00:00:00 2001 From: Cursor Agent Date: Sun, 20 Sep 2026 18:05:39 +0000 Subject: [PATCH] Fold hourly 1143 HIGH MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Catalog and skill fold for notes.md §121 / composition items 369-384 / findings batch #104. Primary lenses: open recreation is not a calibrated replica; semantic lint is a sensor not a proof; third-party benches stay theirs. uniqueness_gate now checks 0843+0915+jcr+0922+0940+0947+1049+1143. Does not bump 0.5.0. Does not reopen PR #23-#43. Co-authored-by: Basit Mustafa <24601@users.noreply.github.com> --- .agents/skills/augustus/SKILL.md | 18 +- .../references/agent-self-assessment.md | 4 + .../augustus/references/applied-mappings.md | 4 + .../references/composition-algebra.md | 78 ++ .agents/skills/augustus/references/faq.md | 22 + .../augustus/references/formal-methods.md | 4 + .../augustus/references/formal-semi-formal.md | 4 + .../augustus/references/judgment-class.md | 4 + .../skills/augustus/references/mappings.md | 4 + .../augustus/references/mental-models.md | 16 + .../augustus/references/methods-catalog.md | 4 + .../augustus/references/mixed-architecture.md | 4 + .../augustus/references/question-design.md | 4 + .../augustus/references/toolbox-mapping.md | 4 + .../skills/augustus/references/validation.md | 4 + .../augustus/scripts/evaluate_decisions.py | 37 + .../augustus/scripts/uniqueness_gate.py | 36 +- CHANGELOG.md | 28 + CONTRIBUTING.md | 4 +- README.md | 5 + docs/ecosystem.md | 4 + research/archive/findings.md | 38 + .../2026-09-20T17/novel_high_this_run.json | 1127 +++++++++++++++++ .../hourly/2026-09-20T17/run_digest.json | 13 + research/changelog-hourly.md | 12 + research/notes.md | 312 +++++ research/refresh-log.md | 13 + research/sources.json | 74 +- 28 files changed, 1872 insertions(+), 9 deletions(-) create mode 100644 research/archive/hourly/2026-09-20T17/novel_high_this_run.json create mode 100644 research/archive/hourly/2026-09-20T17/run_digest.json diff --git a/.agents/skills/augustus/SKILL.md b/.agents/skills/augustus/SKILL.md index 52869d55..3987ff67 100644 --- a/.agents/skills/augustus/SKILL.md +++ b/.agents/skills/augustus/SKILL.md @@ -54,7 +54,7 @@ classical method you already trust, substitute it, classify the win "paraphrase brittleness", "allowlist then judge", "TOCTOU-of-Noul", "Jev inside the database / sqlite-jev", "Jev picks bitrate / join order / the model", "wait for Archer", "lint the request / missing - other", "training confronts Choice other / none-of-the-above", "soft AGENTS.md rules vs the linter", "screenshot Choice / omni System One", "extractive quotes / pointer not generator", "compaction summarize vs pointer", "encoder vs Jev compaction backend", "shadow-mode compaction rollout", "CI flaky-vs-real merge gate", "fail-open VOI wake/resume", "claim vs session evidence", "S1 indexer escalate-S2", "Harbor on/off routing", "fail-open vs fail-closed wake vs CI gate", "encoder vs Jev computer-use backend", "hybrid local decide + remote fill", "DONE vs verified success", "stdout prune vs session compaction", "OpenCode jev-pruner vs Claude jev-pruner", "zen-chat vs jev-zen Noul", "hard envelope then Noul prune", "Cua-S1 vs TypeSafe Jev", "plan vs execute dry-run", "specialist computer-use vs general agent", "local drop-in vs stub scorer", "route vs memory", "when does it hold / extractable from state", "decision model vs constrained LLM", "dual-process S1/S2", "combinatorial grid vs extractive", "uncalibrated local likelihoods", "decision-native RAG", "classify-first / read selectively", "living applied-mappings atlas / class patterns", "silence as safer / draft-gate heartbeat", "robotics text-state vs pixels", "verbatim ledger vs summary", "judgment as language primitive", "Stagehand extract pick-and-copy", "harness observe-score-act vs demo loop", "public judgment wall / six parallel questions", "meaning-search without embeddings", "attention ≠ correctness", "skills→oxlint / AST prove ∩ remainder", "session-sticky first-prompt routing", "measured RAG rerank vs generative rerank", "capability kernel / secrets never in the agent", "Jev is SENSOR not policy", "type-safe ≠ correct", "typed control plane around DSPy", "native vs verbalized confidence", "engine owns truth / Jev owns judgment", "human-confirmed kill gate", "train specialist vs few-shot hosted", "decide→policy→LLM leftover", "Noul 0.5 cannot-tell never rounded", "calibration ≠ sortable / ORDER BY", "pairwise inversion / Score ordinality / two-decimal ties", "wire-compat GLiFormer /v1/systemone", "class-backend economics", "loopback gateway hosted + local", "do not distill Jev as teacher", "active-learning triage", "evidence-packet explorer", "meaning-grep AND/OR/NOT", "closed-vote-only / no planner LLM", "Jev vs PCD Harbor", "PCD O(1) ≠ Noul", "host-owned handlers × System One", "OMP/pi fail-open gate", "permission vs probability / operator owns thresholds", "judgment ≠ permission / Jev never grants access", "eval integrity / instrument not score", "constrained optimizer + S1 features / never sole hot-path gate", "privilege ≠ verdict / effect contracts not tokens", "attention filter / VOI for human review / never blocks / never green unless sure", "measurement owns endorsement / evidence-gated question packs", "Jev supplies evidence / code owns authority", "ranking ≠ calibration / never hard-threshold raw p as frequency", "hot-click CU / indexed element table", "Jev judges relevance / code decides structure", "local rules first then remainder / never auto-train on own hides", "combinators / System One as control plane", "receipts not leaderboard / type-safe ≠ correct jaggedness", "VOI over skill library / skillranker abstention", "OOD calibration / AUC ≠ ECE", "Jev vs thinking-budget small models", "turnstile / replayable evidence≠authority", "MLX one-pass schema→JSON / Apple Silicon replica economics", "memory leases ended by new evidence", "never confidently wrong / TLA+ compose / escalate instead of hard-gate", "no seal no advance / coverage ledger / mint ≠ product brain", "skill-broker sibling / judgment ≠ permission", "sureness bands / max_prob is generous", "JevBench / calibration not in Main Score", "CI typed gate before expensive review", "Codex MCP host adapter", "judgment as attention redirect / jev-preflight", "compress-before-first-send / dizk jev-lens", "tools≠use / SessionStart over hoping", "observational memory / pi-om keep-kind", "open-Jev class / openvons / JevPick", "physical-world System One / HA-Jev / not for locks", "judgment outside the store / jevql", "landed-script trust / headless≠auto-approve", "digital-design combinators / extended five", "VOI cache admission / same-intent skip LLM", "BM25 vs Jev skill routing Harbor harness", "zeroshot vs BERT / contamination DiD", "typed escalate continue abort baton / inverted loop", "worth-your-attention VOI / ThinkyMiner Winnow", "Jev WHETHER Python HOW LLM WHAT", "conflict vs ignorance / named Choice escape", "Playwright executes Jev chooses", "OpenJev /v1/decide not drop-in", "SemIf wire-compat runoff; SemIf rename densify / MLX backend / 5.21× systems≠semantic / Softmax ≠ Noul (`notes.md` §117)", "decision-as-memory flywheel", "record/replay CI / jevassert", "failure-finding arena / jevarena ≠ jev-arena", "BBQ not a bias cert", "decider≠executor", "sentence-as-rule lint / jevlint", "sentence-as-rule lint / jev-lint is jevlint rename", "VOI hunk prune", "whole-repo intent VERIFIED/VIOLATION/UNKNOWN", "GLiNER2 spec ≠ replica", "open replica substrates / grande / laya-jolt / JEV-CPU", "ONNX local-jev not equivalent", "persist constraints across compaction / pi-heed", "calibration+cost first-class gates", "Harbor-shaped Jev vs SGR LLM-as-judge / jev-judge-bench ≠ jevarena ≠ jevbench", "hand no-text steps / jev-use / Vercel drops confidence", "Pi System-One control plane / pi-jev-control", "never free-generates / jev-gpt tree of Choices", "OpenRouter recipe atlas / samples not benches", "personal history feed / jevfeed / no social graph", "competing NAR claims / dual-channel ECE / openJev-verdict ≠ OpenJev", "empty compaction-proxy skip / IPECTER", "throughput ≠ latency / like-for-like ECE", "1-token logprob endpoint ≠ Noul / coverage ≠ correctness", "open replica engine / jevinf", "unofficial Elixir SDK ≠ OTP peer", "jevex n=16 files-to-read VOI", "commit pre-review attention≠verdict / middle band", "Hermes plugin is Agnes not TypeSafe", "pi-jev-compact ≠ pi-jev-compaction", "empty Codex-proxy skip / IPECTER runway", "decision-native inbox / mailordinal", "unofficial jev-cli not ready / ≠ jevql", "laya-multilingual / English checkpoint confident-wrong OOD", "schema-scorer peaked ranking ≠ calibration", "HF 401 / GitHub 404 Hub-only", "productized System One HTTP / classifier.dev", "escalate-under-threshold / smart tier / multi-label ignores", "silent FALLBACK / granite 0.546 vs advertised 0.800", "vs_jev tracked JSON / read eval/README", "choxos/jev-reviewer ≠ egma-ai / systematic-review pointer", "two-pass Choice+Noul evidence extraction", "not-found is an answer", "human check as productized judgment", "githubnext/localjev ≠ kunchenguid/local-jev", "wire-compat ≠ logit-equiv / prompted JSON ≠ structured read", "self-reported probs / entropy confidence", "GitHub Next local /v1/systemone", "LM Studio runner gap / structured-read primitives", "NandhaKishorM/laya packaging ≠ Hub-only / Router script-before-p", "post-T ECE ≠ raw ECE / Banking77 token-budget", "0.85 still soft / not TypeSafe drop-in", "external census ≠ scored bake-off", "GLiNER2+routers class-boundary", "incomplete openjev census vs watch", "Harbor honesty watch / silent fallback", "JevBench v1.2 geometric mean / cal ON rank / weight sensitivity", "option-order 72→21 / instruction models in the class table", "self-host latency ×2 assumption / est. costs", "Laya absent is a gap not a named exclusion", "Qwen3.8 27B ≠ Archer", "hourly already-folded watch / apply-the-five / skip thin noise", "hard-gate Noul as PR/quality gate is soundness theater", "S1 never stalls waiting / S2 one-use advisory", "Local controller ≠ githubnext/localjev", "purple telemetry = consumed not arrived", "seed = geometry not async replay", "20% starting gate still soft", "no pixels to either provider", "OCR+AX observe-score-act / typesafe-computer-use", "never send screenshot to frontier for the decision", "overlapping CU options = false low confidence", "split kind/item/site", "155× one-screenshot ≠ Harbor taskset", "decision ≠ answer-reader capture", "ASR observe-score-act / jev-voice-browser", "partial-speech VOI / free-text waits", "spoken confirm ≠ hard auth", "numbered overlay without another model", "wrap-as-execution / AgentGhost ALLOW ASK DENY", "rules first then Jev remainder / ASK throws / fail-closed", "reddpy/AgentGhost ≠ jwen5419807/agentghost ≠ vventirozos", "JP genre atlas / studio_yebisu / stars ephemeral ≠ eval", "Jev Clearly Explained / akshay_pachaar / LLM hammer", "schema-safe ≠ correct / 200× 400× TypeSafe ceiling", "questions-as-code / shadow first / not a TypeSafe how-to", "proposition ≠ embedding / contrast-set", "boolean composition of soft Nouls / AND OR NOT", "uehaj/jev-semgrep ≠ semgrep.dev", "meaning-grep dedicated fold / not a gate", "decision-validated UI / Jev never authors text", "decision-as-assert / jevtest ambiguous band", "typed decisions drive UI / jev2ui", "hybrid S1 closed verb menu / anima3", "pointer-not-generator search / JevFind", "jev-frontier-bench ≠ frontier-100", "product bakeoff ≠ architecture duel / GLiClass", "four engines same questions / majority floor", "authorship named escape / not evidence", "ha-switchboard HA remains execution", "n8n classify/route/score / Low Confidence", "fast-jev-compaction-pi ≠ pi-jev-compact ≠ pi-jev-compaction", "jevloop full-distribution optimizer / no LLM in the loop", "laya-vision SmolVLM / score untrained", "Cerebellum-2B /v1/decide ≠ TypeSafe / wire-compat vs agent-routing", "laya-grounded not drop-in / Platt not temperature", "GestaltLabs/Jeff-1 ≠ logan-markewich/jeff / acc vs ECE n=9730", "stanley-code empty findings ≠ approval / human promote", "findme ≠ JevFind / NL memory beam-search FS", "jevsubrouter price workers not conversation / counts ≠ dollars", "feelings .feels() default 0.5 is Noul-0.5-never-rounded / ≠ hunch ≠ Probably", "apa-agent-harness ≠ AntonioCoppe/jev-harness / unpublished npm", "grok-bot-jev skill cannot force a bot that ignores it / A/B proxies not tokens", "Essentiel-Jev never authority / human every action", "enzo-mcp independently falsifiable claims / ≠ jev-sift", "pigeonhole OTHER skip / decision-as-filing", "jev-reliability Nothing about accuracy", "clduab11/jev-test ≠ realZachi/jevtest / Nothing runs yet", "jev-rag-benchmark Jev wins is not an assumption", "dairui1/jev-lab ≠ BrendanH18/jev-lab", "jevmail gmail.readonly / mailjay archive/trash", "ZHUBoer/ego-jev reserved __none__", "runWorkflow completed ≠ success", "jsort scores are relative", "Noul not Choice for scale", "groundedness-judge-bench native vs schema-guided", "implicit_true included in yes", "jev_playground 0 promotions", "routing-backtest 0.0447%", "yuyang2230/jev-agent-skill jev-1.13-free", "jev-techstack-classifier stack_config.json", "s1_ruby collapse late", "undecided? abstain", "2389-research/judgement license null", "confidence ≠ winner p", "typesafeai-sdk-community not a new species", "tpellet/hunch exit 3", "never-execute list", "jev-file-search scores not calibrated accuracy", "jev-linkmap Jev never sees S2 prose", "muhammedilyasy/jev-mail metadata only", "tidy none-of-folders stay", "tab-bouncer pinned/audio/current never closed", "lkclean Show fail-open", "jev-yt-time-saver Show anyway", "ORIGIN pause-if-no-Jev", "validResponse sums-to-1", "jev-crawlers risk bands never raw boolean", "jevbrain AUTO_ACT is not a Noul", "judgekit YAML classify/score/route/verify", "typed-judge-kit verdict-in-code", "alsoleg89/decide packing VOI", "0.8 ≠ 80% accuracy", "Jev-Calibration Platt ECE 0.117→0.052", "jev-calibration-arena never acts", "ctmx/openrouter-jev-mcp Decision-as-Plugin", "FrancoisChastel/jev-code ≠ npm jev-code", "claudecode-jev-marketplace fail-open not hot path", "pedroknigge/mcp_jev packs not ask_jev", "cyrusasco/typesafe-mcp noul deadband 0.35–0.65", "codaaiteam/jev-skill jevtypesafeai.com ≠ TypeSafe", "hermes-switchyard ≠ hermes-jev-router ≠ hermes-plugin-jev", "nanoprune 2.8MB ECE 2.58%", "smartdio/jev-browser-agent ≠ ZHUBoer/ego-jev", "Dakai/omp-jev-web DONE ≠ proof", "hari007sh/jev ≠ dannote/jev", "0thernet/system-one-skills deterministic verify", "typed-gate band [0.40,0.60] is refusal", "pi-jev-gate fail-closed; choice is the verdict", "Foq ~25ms/2.2GB local", "rev prefill-only + HF jev-0.5b", "robfrase/jev planning memo", "typesafe_agent_gates 27/27 / 31/31", "EpicEric/safe-sh static remainder", "pastepilot Confirm before act", "Jev-Reranker live Jev not yet measured", "sessionwise opt-in relevance", "jev-search pointer sieve", "400ms Salesforce WebMCP", "typesafe-scheduler-diagnostics advisory", "droidjev screenshot-free", "Tewoto1 jevcu planner still writes", "ha-conversation-jev Jev→Grok", "dsh-jev can only gate", "jev-classification-benchmark specified not run", "jev-luna-pagerduty p≥0.50", "meldltd/meldecision laya-go ONNX", "laya-doom never pixels", "logixism/laya-api empty README", "akpsahan/laya ≠ Archer", "choxos/jevchess engine owns truth", "jev-drive sim not AV", "story-arc Jev never authors", "jev-hs-assistant HS6", "golergka/jev-plays-starcraft-2 UI-verified ≠ API Victory", "awesome-jev-use-cases catalog", "Nibir1/typesafe-go ≠ official", "fingerprint after redact", "recall vs decide", "publish fingerprints+answers", "CI replay as Harbor cousin", "Cache hit ≠ correctness", "hyperspaceai/jevcache ≠ kushals256/jevcache", "human labels only", "score never auto-accepts", "production capture flywheel", "sutro-sh/jev-align ≠ caiovicentino/jev-align", "guidance ≠ hook", "catalysts ≠ summaries", "compile-time System One", "unofficial ≠ TypeSafe", "format_version modernbert-jev/1", "Argos1111/jev_local ≠ us/jev-local ≠ kunchenguid/local-jev", "LFM default ≠ ModernBERT backend", "Nemotron ≠ TypeSafe Jev", "not a calibrated replacement", "djev-dev complements djev-spark", "images as Choice options", "Laya essay numbers *theirs*", "Router/OOD confidence", "hosted bootstrap ≠ silent TypeSafe", "difficulty + policy thresholds + JSONL trace", "jev-codex-pilot model + reasoning depth", "keep/shadow/hybrid/reject", "quarry evidence projection", "Frank-ZY-Dou/awesome-jev robotics/3D/control", "one-dollar-tahoe TypeSafe Jev defense eval", "jevguard calibrator/cache/escape", "jev-ci-selector CI shadow mode", "llama-jev llama.cpp replica", "petercr/jev-orchestrator ≠ FleeexCorp/jev-orchestrator", "seb4ez/jevguard ≠ AseemPrasad/JevGuard ≠ pablozr/JevGuard", "webNeat/llama-jev ≠ WiktorB2004/llama-index-jev", "OpenCode jev-pruner context sieve", "observe→score-candidates→prune", "jev-zen / jev-1.13-free", "zen-chat ≠ Noul", "fail-open original", "keepScore >0.1 floor", "host port of tamaratran/jev-pruner", "indiejoseph/opencode-jev-pruner ≠ nrdz-labs/fast-jev-opencode", "jev-webagent-bench empty stub", "Kiln-AI/jev_jsonschema noul_threshold 0.5", "NSStudent/JevSwiftSDK unofficial", "GLiNER2 native Apple path", "unofficial Swift/Core ML GLiNER 2.5-small", "entity spans + confidence", "not Choice/Score/Noul", "not TypeSafe", "label descriptions as schema", "on-device ANE economics", "honesty locks", "shershah1024/gliner-native-runtime ≠ Fastino", "≠ gliner25-compaction ≠ gliner2-ultrafast ≠ Eran-BA/Jev_from_GLiNER2 ≠ NSStudent/JevSwiftSDK ≠ jevmlx", "default threshold 0.1 still soft", "soft Noul ≠ hard safety", "Decision Graph Protocol frame→assess→commit", "app retains permissions/effects", "Jev-first assessor-neutral", "guarded commit / receipt/next frame", "assessment batching", "hard-gating DGP as safety theater", "numerous-com/dgp ≠ TypeSafe official", "jegrep calibrated path+range Nouls", "no embeddings/index/daemon", "~$0.01–0.03 typical", "agent --json", "can1357/jegrep ≠ Bentlybro/jevgrep ≠ uehaj/jev-semgrep", "Archer-arch fidelity", "kev family OOD 0.76–0.77 vs Jev 0.86", "block-causal isolation", "pointer/readout CE-trained", "/v1/systemone drop-in", "replica honesty", "cost-sensitive decision theory × System One probabilities → control flow", "thresholds derived from costs not hard-coded", "YES / NO / UNSURE from cost_false_yes / cost_false_no / cost_human", "auto-batching same-object questions", "Kungie/gut ≠ tpellet/hunch ≠ carldaws/hunch", "judgment vs generation", "deterministic execution after probabilistic judgment", "exactly one app-owned callback", "explicit uncertain branch", "Illusion47586/judge ≠ lexingtonhibiki/judgekit ≠ Ascurse/typed-judge-kit", "variable-N option scoring as the trainable object", "dynamic candidate bags not fixed label sets", "zwliJay/jev-forge ≠ NanoJev", "open replica economics / latency vs closed Jev", "NAR local drop-in", "wfzyx/von late-catch HIGH", "competing NAR claims / replica honesty", "typed judgments vs chat judges on guardrailing", "ishaannk/llm-vs-jev cross-note only", "deeper integrity fold is rh-guard", "nothing wins outright", "can be argued out of guarding"", "Jev IS the if-statement", "judgments/probabilities drive branches", "text model only writes prose", "interpreter owns variables/loops/budgets/replay", "otherwise maybe / confidence gate", "chaos samples after the gate", "southpolesteve/probably ≠ carldaws/hunch ≠ feelings ≠ Kungie/gut ≠ Illusion47586/judge ≠ tidymodels/probably", "133★ / forks 10 live", "build calibrated classifiers from human feedback", "retrieve by relevance not resemblance", "one calibrated yes/no per memory in one request", "pointer mode 17/18 19/20 *theirs*", "embedding resemblance misses the allergy", "samdotmak/jev-recall ≠ jev-search ≠ jev-sift ≠ carryforward ≠ chopratejas/invalidate", "memory leases ended by new evidence", "six Nouls then fixed rules in code", "0 of 157 false invalidations", "questions/plans/directives are not evidence", "unsure → review queue", "host keeps the store", "name↔body / comment truth / test-claims", "mizchi/jev-lint is mizchi/jevlint rename", "no shipped rule has severity error", "~1 in 5 findings wrong *theirs*", "mizchi/jev-lint ≠ huntedman/JevLint ≠ MichitoSugawara/jev-lint", "JSON Schema → typed JSON via Jev", "noul_threshold 0.5 decoder not a proof", "IncompatibleSchemaError lists every bad property", "on-device Laya CoreML ANE", "~5 ms P50 short decisions", "189/189 FP16 checkpoint parity", "10× not achieved", "mizorewww/laya-coreml ≠ gliner-native-runtime ≠ jevmlx ≠ NandhaKishorM/laya", "softmax over allowed tokens ≠ Noul", "question-first cache", "Micha0827/snapjudge ≠ githubnext/localjev ≠ jevmlx ≠ cendress/SnapJudge", "Jev-first Pi agent loop", "slow-LLM fallback", "explicit action menu / CandidateSource unimplemented", "62 tests wiring not quality", "direwolfiy/JevPi ≠ standardagents/jevpilot ≠ pi-jev-control", "resume-screening bias audit methodology", "name×resume factorial independent Nouls", "callback determined by resume quality", "mean-probability name gaps operationally negligible", "natemoo-re/bias-bench ≠ BBQ", "Plan/PRD panel → code-owned pass|review|block", "cheerleading out of scope", "austindixson/planalyzer ≠ single-goodness Noul", "cost-aware multi-model routing/escalation", "decide vs do", "successful-task cost", "cannacre8ive/switchboard-ai ≠ ha-switchboard ≠ hermes-switchyard", "frozen-protocol zero-shot bench", "TypeSafe Jev vs PrismNLI vs Laya", "contamination caveat", "elcronos/jev-vs-open-decision-models ≠ JevBench ≠ DMB", "context-window admission control", "VOI gate which tokens are worth the expensive model", "fail polarity per lens", "on small inputs lenses lose money", "cvsgireesh/jevusher ≠ jev-sift ≠ winnow", "typed decision control plane", "receipt ≠ authorization", "historical-v0 zero retained cases", "MokiMeow/jev-fabric ≠ jev-forge ≠ dgp", "live 15-dim typed rubric re-score per pause", "scoring economics exemplar", "OpenJev/Codiv ≠ TypeSafe hosted", "jose-troche/live-rubric ~$0.000004 desc / ~$0.000006 README", "adversarial pre-registered Jev eval", "28 predictions before data", "123,805 requests", "confidence does not track ignorance", "polite injection 65% / crude 0%", "willkelly/jev-evaluation ≠ jevals ≠ jev-baselines-eval", "provider-neutral Elixir/BEAM Noul/Choice/Score SDK", "class infrastructure", "nshkrdotcom/system_one_sdk ≠ typesafe_sdk ≠ dannote/jev", "question-linting of Jev questions themselves", "nine jaggedness rules, no API key, no labelled data", "static lint ≠ measured separation", "yodablocks/jevq ≠ tenbin ≠ JevLint ≠ commitjev", "open-weights Laya as class exemplar (binding)", "Nx/Bumblebee runtime", "host chooses backend", "ChristianAlexander/laya_ex ≠ system_one_sdk ≠ dannote/jev ≠ NandhaKishorM/laya", "on-chain/edge Laya deploy", "parity_verified stays false", "model output never grants Tx", "humandebri/IC-Laya ≠ laya_ex", "auditable weekend replica", "Jev outputs never used for training", "soft human-vote distributions", "unpaired 0.577 vs 0.727", "agilabs-ai/jev48 ≠ JevBench ≠ Mapika/decider", "adversarial dual-judge / framing attack surface", "comparative framing is the usable judgment", "prior injection crowds out evidence", "copyleftdev/ember ≠ ember.js", "Laya specialist fine-tune pipeline", "training still GPU-pending", "PIXELZX0/XERON ≠ convaiinnovations/laya", "Hub Laya replica drop", "daliborsb/laya ≠ convaiinnovations/laya ≠ NandhaKishorM/laya", "System One student distillation corpus", "gold is programmatic", "teacher is closed-API clone", "do not distill Jev as teacher of record", "MagaBitmex/jev-4b-distill-data ≠ missing student checkpoint", "non-LLM VIN System One", "planning depth not chat", "lewislululu/jevon ≠ douglance/jevon", "source-bound evidence checks", "local quote mismatch needs no API", "exit 0 ≠ claim truth", "WaynezProg/jev-kit ≠ jonathanavis96/jev-kit (Airlock) ≠ jev-use ≠ jev-mcp", "independent System One evidence catalog", "scores not one leaderboard", "no external record currently reproduced", "TokenTrim no-Jev matched hybrid 62.4%", "reachjalil/system-one-bench ≠ mallahyari/system-one-benchmark", "21 tasks · 134 items · 208 questions", "scenes from public GitHub contracts, not production logs", "SivletLabs/jev-eval ≠ willkelly/jev-evaluation ≠ 4esv/jev-eval ≠ xxkuboxx/jev-eval ≠ onlyoneaman/jev-eval ≠ dayhaysoos/jevals", "option isolation (sibling-blind)", "permutation-equivariant", "Hub OWNER not published", "nafisazizir/hev ≠ jaredpalmer/kev", "frozen local LLM logits, no trained decision head", "residual-head 9,222-param decreased 73/96→67/96", "confidence = 1−normalized entropy, not P(correct)", "yuki-oshio/mini-jev ≠ r-ms/mini-jev", "Jev classifier as autoregressive next-token predictor", "ChatJev-style soundness theater", "erik-dunteman/ChatJev ≠ dannote/jev ≠ jev-gpt", "calibrated decision head × AlphaProof value head", "implementation-layer isomorphism, semantic difference", "timeout = censoring", "do not launder Noul as proof", "parallel rank-prediction vs serial selection", "independent questions can conflict", "zzzzzec/jevsort ≠ keltokhy/jsort", "curated open System One ecosystem catalog", "rupeshpoojary9/awesome-open-system-one ≠ AnotiaWang/awesome-jev", "arXiv paper radar with Jev relevance scoring", "ranking ≠ calibration / 0.5 still soft", "fail-open failed evals not marked seen", "train calibrated ~27M from scratch", "typed Q→prob dist / one forward pass / no LLM decode", "hyusi2003/MiniSystemOne ≠ Colvin0315/MiniSystemOne", "description-only stub / size 5", "ESCI hard probe fails four of six", "jev_bool ECE 0.242 inversion 0.255", "do not re-fold §60 six-gates as new", "jobbyjev one-request-per-company from batch-size result", "find/design/evaluate TypeSafe Jev decision loops", "karanb192/jev-architect ≠ samtay32/jev-system-architect", "Jairik/jev-distiller size 1", "distill-Jev UI stub / do not distill Jev as teacher of record", "post-launch scored use-case map / Jev self-scores then human curation", "licensedsaucer9-web/jev-opportunities", "Jev-inize a use case into classifier/router", "gavinHuang/jevinize → simple-jev not TypeSafe", "featherless-ai/simple-jev", "compare saved decisions / same label can still change the branch", "VihaanAgarwal/jev-diff ≠ Saik0s/diffusiongemma-jev-macos", "not tested with a live Jev API key", "constrained logprob + temp/Platt ≠ Noul", "OpenJevPro pastes openjev-sglang JevBench as own", "zhangcy122/OpenJevPro ≠ IamBusy/OpenJev ≠ ekzhang/openjev-sglang", "PolyForm Noncommercial", "SmolLM-135M / sub-70ms / 0 output tokens", "demo P(True) 0.5052 / Choice conf 0.2872 / Score conf 0.0055", "README claims MIT / GitHub license null / no LICENSE file", "patelvishwa112/jev-system-one-rlcd ≠ arnabgho/rlcd-lite ≠ blackwood-rlcd", "source-backed Awesome Jev radar / 306+ commit-pinned", "logicrw/awesome-jev-projects ≠ AnotiaWang/awesome-jev ≠ yibie/awesome-jev ≠ cobanov/awesome-jev ≠ rupeshpoojary9/awesome-open-system-one", "auto GitHub sync / Issue-only submissions", "hashed n-gram encoder / rival-aware attention", "olanotolu/jevbetter vs jevlike starter", "synthetic hard menus top-1 0.916 vs 0.873 / ECE 0.0182 vs 0.0367 / 40 vs 4608 menus/sec", "shuffled-context control 0.335", "Turn any open LLM into System-One Jev", "uspraveen/Jevify ≠ Mintzs/jevify ≠ gulagala001/jevify", "Jevify-any-LLM architecture probe", "description-only stub / size 0", "Train encoder-only calibrated decision models from a task sentence", "Exu is a toolkit, not a method", "strictly proper scoring rule", "Pre-alpha", "Ruivalim/exu-base", "scratch-trained calibrated decision model", "typed Q → probability dists", "Colvin0315/MiniSystemOne ≠ hyusi2003/MiniSystemOne", "no published weights download URL", "90.5 seconds / 29.2% pipeline evidence", "p_i/p_j independent of other candidates", "Recipe for calibrated decision models — small model out", "init → synth → train → eval → serve", "91.1 % / ECE 0.022 *theirs*", "Jev zero-shot 75.1", "scienthoon/luce", "Put Jev's three headline claims on trial", "0.5B local GPU", "46x speedup / accuracy identical", "ECE 0.624 sentiment catastrophe", "bigger model worse calibration", "RichardoMrMu/jev-mini ≠ yuki-oshio/mini-jev ≠ r-ms/mini-jev", "System-1 decision engine for local LLMs", "structured choices only", "JSON parse of generated text ≠ Noul", "TypefAI JEV / Journal Entry Voucher", "tapsin/jev-local ≠ us/jev-local ≠ Argos1111/jev_local", "Jev 1.13 reward-model eval across 8 benchmark tracks", "40,940 examples / 0 API errors", "RewardBench v1 92.58%", "Precise IF 50.63%", "goya4140/jev-reward-model-evaluation", "Scaffolding in progress", "Jev vs LLM support-ticket routing", "static + live decision bench", "TypeSafe's own published benchmark", "illustrative simulations, not live API calls", "JevBench v1 — smart/cheap/fast/reliable", "I/C/S/K 25% geometric mean", "classifier.dev fast tier 84.8 is Jev behind its own API", "do not re-fold §78 v1.2 board as new", "Laya (421M) 70.1 now on board", "Zero-shot/few-shot LLM routing", "hard budget filter before Jev", "Jev never asked to perform budget arithmetic", "Jev judges the next state, XState enforces transitions", "simulation uses synthetic keyword fixtures", "catalog gravity", "v-modal/awesome-jev-tools", "★339 live REST", "curation is not endorsement", "crawler-maintained directory", "Daily GitHub + npm sweep, human-merged", "RadRebelSam/awesome-jev ≠ AnotiaWang ≠ yibie ≠ cobanov ≠ logicrw ≠ v-modal", "HF peft SPLADE/BGE reranker", "rdxtremity/jev-reranking ≠ carlaiau/jev-reranking", "query-side encoders, not a Jev replica", "ONNX System One Qwen3.5-4B scorer", "source:pngwn/system-one-qwen3.5-4b-scorer", "CC-BY-NC-4.0", "temperature 1.75", "transformers.js AutoModel cannot load this graph", "Consistency benchmark Space", "This Space contains no benchmark result yet", "12-case plumbing fixture", "Benchmark-driven Jev router and judge", "cheap alone is not success", "Jev does not write, sum prices, or claim accuracy %", "Sol 94.2 / Luna 83.9 / Jev path 89.7", "19.2% Sol / 62.3% cost save / 4.5pp miss of 2pp non-inferiority", "p50 latency worse than Sol due to routing overhead", "erendikmenn/jev-llm-router-benchmark ≠ jev-rag-benchmark ≠ ryantsai/jev-llm-router", "Express + node:sqlite", "mock and Jev decision engines", "previous_ticket_count >= 3 is code", "MIN_CONFIDENCE 0.6 still soft", "substring false positives", "aesaganda/jev-ticket-router ≠ SarathChandraBellam/jev-vs-llm-ticket-router", "Universal Figure & Diagram Router", "confidence ≥ 0.85 hard-gate is theater", "generative AI banned from scientific plots", "six visual branches", "hoangngochuong24947-gif/jev-figure-router", "human-labeled (state, question, label)", "166,054 rows / 22 configs", "soft_label for human uncertainty", "Praveenrajus/jev-bench ≠ fstandhartinger/jevbench", "ternary bonsai System One GGUF", "openjev's mechanism, Bonsai's weights", "Hub does not ship weights", "100/100 easy T/F is not Harbor", "label_mass ≠ correctness", "stock llama.cpp Q2_0 silently gibberish", "NicolaiMTLassen/open-bonzi-jev ≠ NicolaiLassen", "transformers.js DeBERTa ONNX", "source:com-kotobalabs/open-jev-deberta-v3-large", "temperature 1.05", "AutoModel from_pretrained works", "onnx-community/open-jev-deberta-v3-large-ONNX ≠ system-one-qwen3.5-4b-scorer-ONNX", "107★ densify", "GH 151M vs README 149.6M", "PR #1 now closed unmerged", "do not re-fold §71 claim-audit as a beat", "typed decisions, RLCD, confidence-gated routing", "structured ≠ correct", "mock not live API", "26 tests", "wjdjdakf17/jev-study ≠ baekenough/jev-study", "bonzi-27b-v2 / ternary-8b / 27b-v1 GGUF family densify", "WANLI-256 74.6% / 65.2% / 71.1% *theirs*", "Bonsai 1 27B Q1_0 runs on stock llama.cpp", "ternary still needs PrismML fork", "hf:heman10x/openJev-verdict-2.0 twin tokenizer-only", "OpenJev Vision image classification + uncertainty", "CLEVR-4 held-out joint 0%", "hfdataset:IamBusy/OpenJev-Vision-Research-v0.1 12,832", "294,912 derived targets not independent samples", "Laya multilingual ONNX WebGPU typed-decisions port", "63/63 selected answers / 5.1e-4 CPU / 1.2e-2 WebGPU", "UpHash-Network/mini-jev is yuki-oshio transfer", "jev-injection-bench 11,900 labelled prompts", "Jev best ranking / Haiku better ECE 0.021 vs 0.058", "0.5–0.9 band is where Jev's numbers do not mean what they say", "Prompt wording moves panic 28%", "manojlds/jev-dspy-bench ≠ dspachos/jev-dspy ≠ jmanhype/jev-dspy-lab", "Jev agreement is similarity, never ground truth", "no aggregate quality grade or merge gate", "AbstentionBench-on-Jev rank 1 of 20 vs 2025 field", "question-asymmetry", "forward-looking 0.465 never extreme", "openkev calibration layer not a runtime", "ECE vs coverage independent", "select_threshold returns inf", "escalation catches uncertainty not ignorance", "misakaikato/openkev ≠ jaredpalmer/kev", "pdf-race Docling→Jev vs Gemini", "parser owns the wall clock", "12/12 tie is a tie", "titles selected not generated", "flopcheck 16 calibrated tweet judgments", "mechanical tells in code", "ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas", "Laya calibration lab Gradio MCP", "T never changes argmax", "confidence ≠ top-label p", "easy probe set refused", "40–48 rows too small to ship T", "Gemma-4 26B-A4B jevify classification+calibration", "LoRA adapter twin not independent eval", "Gemma-4 E4B jevify", "E4B LoRA stub card", "kushalpatil/jevify-gemma4 ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify", "GH kushalpatil07/jevify 404", "PAWS 0.580/ece 0.288 is the weak cell", "smaller E4B slightly better OOD ECE than 26B-A4B", "Hub jevify merged LoRA ships weights", "bonzi Bonsai-8B v1 GGUF densify", "Bonsai-1.7B v1", "Bonsai-4B v1", "WANLI-256 64.5% / 60.2% / 52.0% *theirs*", "rank #4 / #5 / #6 of 6", "JulesHuisman/jev-eval scaffolding / README SHA c356a584 (was empty e69de29b)", "JulesHuisman/jev-eval ≠ SivletLabs/jev-eval ≠ willkelly/jev-evaluation ≠ 4esv ≠ xxkuboxx ≠ onlyoneaman ≠ dayhaysoos/jevals", "7 bands 6/10 vs 40 bands 0/10", "source receipts + confidence slider re-policy without re-inference", "32/32 synthetic is smoke not production", "classify HF datasets across typed semantic dimensions", "roadus2 watch misspelling; lock roadius2/ultra_laya", "ultra_laya REVIEW defects", "default branch claude/laya-jev-review-gg5ppo", "XNLI EN 88.3% ECE 0.032 → RU 77.3% ECE 0.096", "Δ −11.0 pp [−14.2,−7.8]; ECE +0.063", "MASSIVE no detectable difference at n=600", "confidence is function of p_max (r=1.000)", "pointer-not-generator 400 human-authored responses", "proposed ≠ authorized", "FewRel 160: Jev 85.0% vs lexical 13.125%", "gated 100% (95/95) coverage 59.375%", "J++ composable semantic computation language", "judge-jev 0.5 still soft", "947 repos scored; A 273 / B 302 / C 372", "LLM rubric ≠ benches", "No benchmark winner is claimed", "phishing: naive 62.6% vs regex 91.8%; 5-atomic + LR 95.0% *theirs*", "AITuber tension ±15", "README npm global; repo is Rust", "git-confess code owns counting/blame/ratio", "httpx exhibit 11% (13/119) *theirs*", "90d trend +12.40% vs random +12.75% vs BH +41.71%", "5m win rate 25%", "Awesomejev 656 entries / 38,160 stars", "tracker likes 64 (+4) lastModified UNCHANGED", "Laya present; Blackwood ABSENT; Archer still promised_not_landed", "Blackwood tracker ABSENT; likes 2 gated manual", "r = c - p_a", "ECE 0.021; acc 0.807 vs warmup 0.746", "Independent primitive", "11.57s vs 54.10s · 4.67× · 120/128 *theirs*", "default path is pretrained Gemma probs not trained RLCD head", "GH Meanblock 404; lock leesk212/JEV-CPU", "softmax over letter slots ≠ Noul", "WANLI 0.741 vs openjev v2 0.77 *theirs*", "3-way NLI ≠ Noul", "priority 0.464 = majority floor", "banking77 contaminated", "raw margins not probabilities", "do not distill Jev as teacher of record (they distilled Haiku)", "“0.9 is not one number”", "ranking ≠ calibration", "banking77 0.8–0.9 stated 0.86 actual 0.73 over-confident *theirs*", "≠ Praveenrajus/jev-bench ≠ fstandhartinger/jevbench", "$0.0000153–$0.0000226 vs circulating $0.0004 (~20×)", "Score is 0..n-1 expectation not 0–1", "Noul has no confidence field", "TCP floor 198.8 ms", "type reliability is not a reason to choose Jev (json_schema 5/5)", "gateway tax not one number", "Function-only 5/8 vs hybrid 8/8", "4/8 without Jev", "8 designed cases not conversion lift", "200-row pilot Jev 86.5% 173/200 vs Gemini Flash-Lite 86.0% 172/200 vs Pro 87.0% 174/200 *theirs*", "not a ranking", "情緒測謊器", "8-example Jev vs GPT-5.6 Sol ~64× cost 5.4× latency *theirs*", "synthetic; no inference", "≠ JevBench v1.2 §78", "Judged 3317 / listed 2560", "Jev judges, code applies policy", "APA “microsecond policy / zero hallucination” overclaim", "Client-side quiz; pointer from held docs; scanned-PDF warn", "Jev judges / agent reasons / user decides", "selecting an option is not permission to implement", "pattern exact, judgement must clear floor", "no matching pattern → no model call", "not a correctness oracle", "Spec vs artifact remainder", "treating 0.85 as 85% / minProbability hard-gate as Harbor", "VERIFY acquires discriminating evidence, never same-pool confidence-only rescoring", "fast/full/max are ceilings not sizes", "Solar writes, Jev chooses NEXT ACTION", "do not reopen or amend PR #23 or #24 or #25 or #26 or #27", , "Calibration is not alpha", "NO CURRENT ALPHA CANDIDATE", "ΔR² approximately +0.00084", "Brier 0.2131387", "ECE 0.0421875", "Adding Jev probability to deterministic volatility improved Brier by only 1.4058e-05", "default 0.5 keeps zero non pinned", "keepResult median 0.14 to 0.17", "keepCall median 0.28 to 0.35", "usable range is about 0.10 to 0.25", "7.8% to 57.9%", "judges results it never sees", "task-finish eval not built yet", "$0.002 per compaction", "slavadubrov/sgr-judge-bench ≠ slavadubrov/jev-judge-bench", "Jev 108/120 $0.083 0.34 s", "Luna SGR 114/120", "paired Jev accuracy-difference intervals include zero", "not evidence of equivalence", "GLM SGR 26/120 93 format failures", "Terra-planned Jev hybrid 55/120", "rule-based by default, optionally Jev-backed", "empty README", "missing key cannot break the experience", "prefill plus exactly one decode", "softmax over A/B/C ≠ Noul", "BBQ 9,053/10,000 (90.53%)", "ECE 0.0890", "Mean confidence 0.9943", "overconfident", "score and noul not implemented", "DGUI 12 rows (was 6)", "INSTRUCT 119 rows likes 2", "encode the state once, decide everything in parallel", "0.740 accuracy against a 0.508 majority", "ECE 0.047", "fine-tune's advantage ends where its 384-token training data does", "jasonkneen/open-jev ≠ pngwn/open-jev", "same sha d41dc3cd", "Space does not call Jev", "recomputes routing from saved probabilities", "200-case Jev 97.0% / 100.0% / 95.0% / MAE 9.22", "synthetic repository benchmark", "Jev evaluations are advisory", "YehuiTang0316/jev-nlgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep", "default threshold 0.8 still soft", "40-line windows cannot prove whole function", "token-native sequential start/end Choice", "Gemini/Haiku stubs not configured yet", "handful of hand-written examples, not a benchmark", "Jev judged exactly what it was given", "laguagu/jev-skills ≠ laguagu/jev-evidence-lab ≠ Pleo2/awesome-jev-agent-skills", "contract_passed is not a claim of guaranteed factual truth", "Wilson lower bound 0.85 floor", "fixture mode no savings claim", "SemIf 2207★ (+21 vs §110 2186)", "jevlike 1043★ (+5 vs 1038)", "TypeAR 15★ (+1 vs 14)", "AnotiaWang 97★ (+1 vs 96)", "yibie/awesome-jev 506★ (+16 vs 490)", "Laya likes 822 (was 802)", "tracker likes 64 flat, lastModified UNCHANGED", "do not reopen or amend PR #23/#24/#25/#26/#27/#28", "Heman10x-NGU/Verdict-open-jev ≠ Heman10x-NGU/openJev-verdict-2.0", "TF-IDF + LogReg ECE 0.0207 vs Jev 0.1440", "Verdict-open-jev 48.07% vs Jev 90.80%", "abstention combined recall 10.00%", "p50 35.58 ms", "K=25 (maximum capacity) 72.00%", "0.85 coverage 84.60% selective risk 1.18%", "26.1× faster than standard Qwen JSON generation", "Jevify 90.0% / 167 ms CUDA graphs disabled", "Finding 1: Brier on stated confidence alone is a trap", "grpo_rlcr 0.78 / ECE 0.084", "reliability 0.007 but resolution 0.000", "27 900 schema-driven decisions", "13 600 / 13 600 questions", "candidate mass min 0.99999624", "22 configs · 166,054 rows · 4 calibration-gold", "sha a39eba3f", "Student B MAE 0.148 / Pearson 0.836 / 86.0%", "pngwn/open-jev-laya-bench README 404", "sha 9f69c742 likes 2", "HDFS 0.9933 (745/750) / retain 0.0084", "BGL ERROR/FATAL protection 1.0000", "2,479 / 2,500 HDFS uncertain", "cache hit 0.9648 (2412/2500)", "$0.153936 estimated", "E2 recomputes from saved probabilities", "Space sha eda59e0a", "MASSIVE English 0.783 / Khmer 0.033 / Hindi 0.133", "40–48 rows too small to ship T", "T never changes argmax", "siren2345/jev-single-decode-transformers ≠ siren2345/jev-single-decode", "Split Transformers experiment from llama.cpp runtime", "tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab", "Four experiments stress-testing TypeSafe's Jev: calibration, bundle bias, label bias, and ensembling", "second pass must be $0.00 from cache", "The pages never call Jev", "Gemma 4 31B 77.0% / Jev 1.13.0 61.4% / Laya 322M 0.0%", "restriction state 95.0% against 84.4%", "None of the systems are particularly good at knowing when to stop and ask", "They skip the question and call a tool directly", "100% schema pass", "six-field joint 48.8% vs 72.8%", "ywchiu/jev_benchmark ≠ Running-Dolphins/jev-bench ≠ Praveenrajus/jev-bench", "ACT / REVIEW / FALLBACK", "A provider failure, timeout, malformed output, or missing answer is **not** a policy outcome", "confidence is descriptive provider output, not a substitute for probability", "Quality denominators include only valid scored answers", "an exact halfway tie chooses the lower level", "aiwithenoch/Jev-Skill ≠ simplosophy/jev-skill ≠ laguagu/jev-skills", "The local path does not claim to turn a smaller checkpoint into Jev", "Low support becomes decision: \"review\"", "MIT-0 SPDX NOASSERTION", "current-llm", "结构兼容，不是 Jev 模型能力", "altryne/jevify ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify", "Find where Jev belongs. Design the questions. Measure the difference", "TypeAR-AI/TypeAR 301 → TypeLLM/TypeLLM", "TypeLLM/TypeLLM 16★", "SemIf 2241★ (+34 vs §111 2207)", "jevlike 1051★ (+8 vs 1043)", "AnotiaWang 98★ (+1 vs 97)", "yibie/awesome-jev 525★ (+19 vs 506)", "Laya likes 864 (was 822)", "tracker likes 67 (+3 vs 64)", "lastModified UNCHANGED `2026-09-20T04:29:16.000Z`", "do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#32", "hysteresis enter/exit / replay policy without inference", "calibration does not compose / hop-ECE permutation-invariant", "equal-width vs quantile ECE / ranking ≠ calibration", "Qwen2.5 ≠ Archer / Qwen 3.8 sparring ≠ Archer / Qwen/Qwen3.8-27B ≠ Archer", "Deferred Crispification / TCE / AMS", "g0runmezadam/what-is-jev IS tunahansahin897/what-is-jev", "pd.cut equal-width vs jeval quantile", "A hunch is a probability with a policy attached", "soundness theater / measurement theater / hourly 0843", , "Jev Capability Resolver / NiazMorshed2007/jcr", "one tool nested capability tree / returns context / does not execute", "skills vs capabilities / workflow+judgment vs operations", "format independent of Jev / proposed open standard", "JCR_BAND_RATIO 0.6 is application policy / soft scores ≠ hard gates", "routing ≠ permission / docs ≠ authority to run", "sol-vs-opus5-20 lookup+explain / n=1 / Not Harbor task-execution", "wall-time mixed / Sol slower with JCR in 19/20", "NiazMorshed2007/jcr ≠ skill-broker ≠ skillranker ≠ jev-sift ≠ jev-lens ≠ jevusher ≠ jev_select_capability", "do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34", "notes.md §116", "copy the SemIf/MLX installer?", "quote 5.21× as beating Jev?", "treat 0.845 as a TypeSafe replica?", "collapse SemIf into kw2828/zhihz/semif-rs/semif-serve", "softmax over options as a Noul", "llm prompt to jev primitives", "conversion assistant not equivalent behavior", "heuristic conversion ≠ calibrated Noul", "alexwestco/llm-to-jev ≠ altryne/jevify", "user-provided 0940 / notes.md §118", "judge ≠ actuator", "candidate_mass", "softmax over A–H ≠ Noul", "hourly 0947 / notes.md §119", "ggmlc GGUF is not llama.cpp", "serving substrate ≠ calibrated replica", "Qwen3.5-9B ≠ Archer", "planner writes JEV selects", "hourly 1049 / notes.md §120", or "cascade sign-flip / calibration theater": read `references/faq.md`, + other", "training confronts Choice other / none-of-the-above", "soft AGENTS.md rules vs the linter", "screenshot Choice / omni System One", "extractive quotes / pointer not generator", "compaction summarize vs pointer", "encoder vs Jev compaction backend", "shadow-mode compaction rollout", "CI flaky-vs-real merge gate", "fail-open VOI wake/resume", "claim vs session evidence", "S1 indexer escalate-S2", "Harbor on/off routing", "fail-open vs fail-closed wake vs CI gate", "encoder vs Jev computer-use backend", "hybrid local decide + remote fill", "DONE vs verified success", "stdout prune vs session compaction", "OpenCode jev-pruner vs Claude jev-pruner", "zen-chat vs jev-zen Noul", "hard envelope then Noul prune", "Cua-S1 vs TypeSafe Jev", "plan vs execute dry-run", "specialist computer-use vs general agent", "local drop-in vs stub scorer", "route vs memory", "when does it hold / extractable from state", "decision model vs constrained LLM", "dual-process S1/S2", "combinatorial grid vs extractive", "uncalibrated local likelihoods", "decision-native RAG", "classify-first / read selectively", "living applied-mappings atlas / class patterns", "silence as safer / draft-gate heartbeat", "robotics text-state vs pixels", "verbatim ledger vs summary", "judgment as language primitive", "Stagehand extract pick-and-copy", "harness observe-score-act vs demo loop", "public judgment wall / six parallel questions", "meaning-search without embeddings", "attention ≠ correctness", "skills→oxlint / AST prove ∩ remainder", "session-sticky first-prompt routing", "measured RAG rerank vs generative rerank", "capability kernel / secrets never in the agent", "Jev is SENSOR not policy", "type-safe ≠ correct", "typed control plane around DSPy", "native vs verbalized confidence", "engine owns truth / Jev owns judgment", "human-confirmed kill gate", "train specialist vs few-shot hosted", "decide→policy→LLM leftover", "Noul 0.5 cannot-tell never rounded", "calibration ≠ sortable / ORDER BY", "pairwise inversion / Score ordinality / two-decimal ties", "wire-compat GLiFormer /v1/systemone", "class-backend economics", "loopback gateway hosted + local", "do not distill Jev as teacher", "active-learning triage", "evidence-packet explorer", "meaning-grep AND/OR/NOT", "closed-vote-only / no planner LLM", "Jev vs PCD Harbor", "PCD O(1) ≠ Noul", "host-owned handlers × System One", "OMP/pi fail-open gate", "permission vs probability / operator owns thresholds", "judgment ≠ permission / Jev never grants access", "eval integrity / instrument not score", "constrained optimizer + S1 features / never sole hot-path gate", "privilege ≠ verdict / effect contracts not tokens", "attention filter / VOI for human review / never blocks / never green unless sure", "measurement owns endorsement / evidence-gated question packs", "Jev supplies evidence / code owns authority", "ranking ≠ calibration / never hard-threshold raw p as frequency", "hot-click CU / indexed element table", "Jev judges relevance / code decides structure", "local rules first then remainder / never auto-train on own hides", "combinators / System One as control plane", "receipts not leaderboard / type-safe ≠ correct jaggedness", "VOI over skill library / skillranker abstention", "OOD calibration / AUC ≠ ECE", "Jev vs thinking-budget small models", "turnstile / replayable evidence≠authority", "MLX one-pass schema→JSON / Apple Silicon replica economics", "memory leases ended by new evidence", "never confidently wrong / TLA+ compose / escalate instead of hard-gate", "no seal no advance / coverage ledger / mint ≠ product brain", "skill-broker sibling / judgment ≠ permission", "sureness bands / max_prob is generous", "JevBench / calibration not in Main Score", "CI typed gate before expensive review", "Codex MCP host adapter", "judgment as attention redirect / jev-preflight", "compress-before-first-send / dizk jev-lens", "tools≠use / SessionStart over hoping", "observational memory / pi-om keep-kind", "open-Jev class / openvons / JevPick", "physical-world System One / HA-Jev / not for locks", "judgment outside the store / jevql", "landed-script trust / headless≠auto-approve", "digital-design combinators / extended five", "VOI cache admission / same-intent skip LLM", "BM25 vs Jev skill routing Harbor harness", "zeroshot vs BERT / contamination DiD", "typed escalate continue abort baton / inverted loop", "worth-your-attention VOI / ThinkyMiner Winnow", "Jev WHETHER Python HOW LLM WHAT", "conflict vs ignorance / named Choice escape", "Playwright executes Jev chooses", "OpenJev /v1/decide not drop-in", "SemIf wire-compat runoff; SemIf rename densify / MLX backend / 5.21× systems≠semantic / Softmax ≠ Noul (`notes.md` §117)", "decision-as-memory flywheel", "record/replay CI / jevassert", "failure-finding arena / jevarena ≠ jev-arena", "BBQ not a bias cert", "decider≠executor", "sentence-as-rule lint / jevlint", "sentence-as-rule lint / jev-lint is jevlint rename", "VOI hunk prune", "whole-repo intent VERIFIED/VIOLATION/UNKNOWN", "GLiNER2 spec ≠ replica", "open replica substrates / grande / laya-jolt / JEV-CPU", "ONNX local-jev not equivalent", "persist constraints across compaction / pi-heed", "calibration+cost first-class gates", "Harbor-shaped Jev vs SGR LLM-as-judge / jev-judge-bench ≠ jevarena ≠ jevbench", "hand no-text steps / jev-use / Vercel drops confidence", "Pi System-One control plane / pi-jev-control", "never free-generates / jev-gpt tree of Choices", "OpenRouter recipe atlas / samples not benches", "personal history feed / jevfeed / no social graph", "competing NAR claims / dual-channel ECE / openJev-verdict ≠ OpenJev", "empty compaction-proxy skip / IPECTER", "throughput ≠ latency / like-for-like ECE", "1-token logprob endpoint ≠ Noul / coverage ≠ correctness", "open replica engine / jevinf", "unofficial Elixir SDK ≠ OTP peer", "jevex n=16 files-to-read VOI", "commit pre-review attention≠verdict / middle band", "Hermes plugin is Agnes not TypeSafe", "pi-jev-compact ≠ pi-jev-compaction", "empty Codex-proxy skip / IPECTER runway", "decision-native inbox / mailordinal", "unofficial jev-cli not ready / ≠ jevql", "laya-multilingual / English checkpoint confident-wrong OOD", "schema-scorer peaked ranking ≠ calibration", "HF 401 / GitHub 404 Hub-only", "productized System One HTTP / classifier.dev", "escalate-under-threshold / smart tier / multi-label ignores", "silent FALLBACK / granite 0.546 vs advertised 0.800", "vs_jev tracked JSON / read eval/README", "choxos/jev-reviewer ≠ egma-ai / systematic-review pointer", "two-pass Choice+Noul evidence extraction", "not-found is an answer", "human check as productized judgment", "githubnext/localjev ≠ kunchenguid/local-jev", "wire-compat ≠ logit-equiv / prompted JSON ≠ structured read", "self-reported probs / entropy confidence", "GitHub Next local /v1/systemone", "LM Studio runner gap / structured-read primitives", "NandhaKishorM/laya packaging ≠ Hub-only / Router script-before-p", "post-T ECE ≠ raw ECE / Banking77 token-budget", "0.85 still soft / not TypeSafe drop-in", "external census ≠ scored bake-off", "GLiNER2+routers class-boundary", "incomplete openjev census vs watch", "Harbor honesty watch / silent fallback", "JevBench v1.2 geometric mean / cal ON rank / weight sensitivity", "option-order 72→21 / instruction models in the class table", "self-host latency ×2 assumption / est. costs", "Laya absent is a gap not a named exclusion", "Qwen3.8 27B ≠ Archer", "hourly already-folded watch / apply-the-five / skip thin noise", "hard-gate Noul as PR/quality gate is soundness theater", "S1 never stalls waiting / S2 one-use advisory", "Local controller ≠ githubnext/localjev", "purple telemetry = consumed not arrived", "seed = geometry not async replay", "20% starting gate still soft", "no pixels to either provider", "OCR+AX observe-score-act / typesafe-computer-use", "never send screenshot to frontier for the decision", "overlapping CU options = false low confidence", "split kind/item/site", "155× one-screenshot ≠ Harbor taskset", "decision ≠ answer-reader capture", "ASR observe-score-act / jev-voice-browser", "partial-speech VOI / free-text waits", "spoken confirm ≠ hard auth", "numbered overlay without another model", "wrap-as-execution / AgentGhost ALLOW ASK DENY", "rules first then Jev remainder / ASK throws / fail-closed", "reddpy/AgentGhost ≠ jwen5419807/agentghost ≠ vventirozos", "JP genre atlas / studio_yebisu / stars ephemeral ≠ eval", "Jev Clearly Explained / akshay_pachaar / LLM hammer", "schema-safe ≠ correct / 200× 400× TypeSafe ceiling", "questions-as-code / shadow first / not a TypeSafe how-to", "proposition ≠ embedding / contrast-set", "boolean composition of soft Nouls / AND OR NOT", "uehaj/jev-semgrep ≠ semgrep.dev", "meaning-grep dedicated fold / not a gate", "decision-validated UI / Jev never authors text", "decision-as-assert / jevtest ambiguous band", "typed decisions drive UI / jev2ui", "hybrid S1 closed verb menu / anima3", "pointer-not-generator search / JevFind", "jev-frontier-bench ≠ frontier-100", "product bakeoff ≠ architecture duel / GLiClass", "four engines same questions / majority floor", "authorship named escape / not evidence", "ha-switchboard HA remains execution", "n8n classify/route/score / Low Confidence", "fast-jev-compaction-pi ≠ pi-jev-compact ≠ pi-jev-compaction", "jevloop full-distribution optimizer / no LLM in the loop", "laya-vision SmolVLM / score untrained", "Cerebellum-2B /v1/decide ≠ TypeSafe / wire-compat vs agent-routing", "laya-grounded not drop-in / Platt not temperature", "GestaltLabs/Jeff-1 ≠ logan-markewich/jeff / acc vs ECE n=9730", "stanley-code empty findings ≠ approval / human promote", "findme ≠ JevFind / NL memory beam-search FS", "jevsubrouter price workers not conversation / counts ≠ dollars", "feelings .feels() default 0.5 is Noul-0.5-never-rounded / ≠ hunch ≠ Probably", "apa-agent-harness ≠ AntonioCoppe/jev-harness / unpublished npm", "grok-bot-jev skill cannot force a bot that ignores it / A/B proxies not tokens", "Essentiel-Jev never authority / human every action", "enzo-mcp independently falsifiable claims / ≠ jev-sift", "pigeonhole OTHER skip / decision-as-filing", "jev-reliability Nothing about accuracy", "clduab11/jev-test ≠ realZachi/jevtest / Nothing runs yet", "jev-rag-benchmark Jev wins is not an assumption", "dairui1/jev-lab ≠ BrendanH18/jev-lab", "jevmail gmail.readonly / mailjay archive/trash", "ZHUBoer/ego-jev reserved __none__", "runWorkflow completed ≠ success", "jsort scores are relative", "Noul not Choice for scale", "groundedness-judge-bench native vs schema-guided", "implicit_true included in yes", "jev_playground 0 promotions", "routing-backtest 0.0447%", "yuyang2230/jev-agent-skill jev-1.13-free", "jev-techstack-classifier stack_config.json", "s1_ruby collapse late", "undecided? abstain", "2389-research/judgement license null", "confidence ≠ winner p", "typesafeai-sdk-community not a new species", "tpellet/hunch exit 3", "never-execute list", "jev-file-search scores not calibrated accuracy", "jev-linkmap Jev never sees S2 prose", "muhammedilyasy/jev-mail metadata only", "tidy none-of-folders stay", "tab-bouncer pinned/audio/current never closed", "lkclean Show fail-open", "jev-yt-time-saver Show anyway", "ORIGIN pause-if-no-Jev", "validResponse sums-to-1", "jev-crawlers risk bands never raw boolean", "jevbrain AUTO_ACT is not a Noul", "judgekit YAML classify/score/route/verify", "typed-judge-kit verdict-in-code", "alsoleg89/decide packing VOI", "0.8 ≠ 80% accuracy", "Jev-Calibration Platt ECE 0.117→0.052", "jev-calibration-arena never acts", "ctmx/openrouter-jev-mcp Decision-as-Plugin", "FrancoisChastel/jev-code ≠ npm jev-code", "claudecode-jev-marketplace fail-open not hot path", "pedroknigge/mcp_jev packs not ask_jev", "cyrusasco/typesafe-mcp noul deadband 0.35–0.65", "codaaiteam/jev-skill jevtypesafeai.com ≠ TypeSafe", "hermes-switchyard ≠ hermes-jev-router ≠ hermes-plugin-jev", "nanoprune 2.8MB ECE 2.58%", "smartdio/jev-browser-agent ≠ ZHUBoer/ego-jev", "Dakai/omp-jev-web DONE ≠ proof", "hari007sh/jev ≠ dannote/jev", "0thernet/system-one-skills deterministic verify", "typed-gate band [0.40,0.60] is refusal", "pi-jev-gate fail-closed; choice is the verdict", "Foq ~25ms/2.2GB local", "rev prefill-only + HF jev-0.5b", "robfrase/jev planning memo", "typesafe_agent_gates 27/27 / 31/31", "EpicEric/safe-sh static remainder", "pastepilot Confirm before act", "Jev-Reranker live Jev not yet measured", "sessionwise opt-in relevance", "jev-search pointer sieve", "400ms Salesforce WebMCP", "typesafe-scheduler-diagnostics advisory", "droidjev screenshot-free", "Tewoto1 jevcu planner still writes", "ha-conversation-jev Jev→Grok", "dsh-jev can only gate", "jev-classification-benchmark specified not run", "jev-luna-pagerduty p≥0.50", "meldltd/meldecision laya-go ONNX", "laya-doom never pixels", "logixism/laya-api empty README", "akpsahan/laya ≠ Archer", "choxos/jevchess engine owns truth", "jev-drive sim not AV", "story-arc Jev never authors", "jev-hs-assistant HS6", "golergka/jev-plays-starcraft-2 UI-verified ≠ API Victory", "awesome-jev-use-cases catalog", "Nibir1/typesafe-go ≠ official", "fingerprint after redact", "recall vs decide", "publish fingerprints+answers", "CI replay as Harbor cousin", "Cache hit ≠ correctness", "hyperspaceai/jevcache ≠ kushals256/jevcache", "human labels only", "score never auto-accepts", "production capture flywheel", "sutro-sh/jev-align ≠ caiovicentino/jev-align", "guidance ≠ hook", "catalysts ≠ summaries", "compile-time System One", "unofficial ≠ TypeSafe", "format_version modernbert-jev/1", "Argos1111/jev_local ≠ us/jev-local ≠ kunchenguid/local-jev", "LFM default ≠ ModernBERT backend", "Nemotron ≠ TypeSafe Jev", "not a calibrated replacement", "djev-dev complements djev-spark", "images as Choice options", "Laya essay numbers *theirs*", "Router/OOD confidence", "hosted bootstrap ≠ silent TypeSafe", "difficulty + policy thresholds + JSONL trace", "jev-codex-pilot model + reasoning depth", "keep/shadow/hybrid/reject", "quarry evidence projection", "Frank-ZY-Dou/awesome-jev robotics/3D/control", "one-dollar-tahoe TypeSafe Jev defense eval", "jevguard calibrator/cache/escape", "jev-ci-selector CI shadow mode", "llama-jev llama.cpp replica", "petercr/jev-orchestrator ≠ FleeexCorp/jev-orchestrator", "seb4ez/jevguard ≠ AseemPrasad/JevGuard ≠ pablozr/JevGuard", "webNeat/llama-jev ≠ WiktorB2004/llama-index-jev", "OpenCode jev-pruner context sieve", "observe→score-candidates→prune", "jev-zen / jev-1.13-free", "zen-chat ≠ Noul", "fail-open original", "keepScore >0.1 floor", "host port of tamaratran/jev-pruner", "indiejoseph/opencode-jev-pruner ≠ nrdz-labs/fast-jev-opencode", "jev-webagent-bench empty stub", "Kiln-AI/jev_jsonschema noul_threshold 0.5", "NSStudent/JevSwiftSDK unofficial", "GLiNER2 native Apple path", "unofficial Swift/Core ML GLiNER 2.5-small", "entity spans + confidence", "not Choice/Score/Noul", "not TypeSafe", "label descriptions as schema", "on-device ANE economics", "honesty locks", "shershah1024/gliner-native-runtime ≠ Fastino", "≠ gliner25-compaction ≠ gliner2-ultrafast ≠ Eran-BA/Jev_from_GLiNER2 ≠ NSStudent/JevSwiftSDK ≠ jevmlx", "default threshold 0.1 still soft", "soft Noul ≠ hard safety", "Decision Graph Protocol frame→assess→commit", "app retains permissions/effects", "Jev-first assessor-neutral", "guarded commit / receipt/next frame", "assessment batching", "hard-gating DGP as safety theater", "numerous-com/dgp ≠ TypeSafe official", "jegrep calibrated path+range Nouls", "no embeddings/index/daemon", "~$0.01–0.03 typical", "agent --json", "can1357/jegrep ≠ Bentlybro/jevgrep ≠ uehaj/jev-semgrep", "Archer-arch fidelity", "kev family OOD 0.76–0.77 vs Jev 0.86", "block-causal isolation", "pointer/readout CE-trained", "/v1/systemone drop-in", "replica honesty", "cost-sensitive decision theory × System One probabilities → control flow", "thresholds derived from costs not hard-coded", "YES / NO / UNSURE from cost_false_yes / cost_false_no / cost_human", "auto-batching same-object questions", "Kungie/gut ≠ tpellet/hunch ≠ carldaws/hunch", "judgment vs generation", "deterministic execution after probabilistic judgment", "exactly one app-owned callback", "explicit uncertain branch", "Illusion47586/judge ≠ lexingtonhibiki/judgekit ≠ Ascurse/typed-judge-kit", "variable-N option scoring as the trainable object", "dynamic candidate bags not fixed label sets", "zwliJay/jev-forge ≠ NanoJev", "open replica economics / latency vs closed Jev", "NAR local drop-in", "wfzyx/von late-catch HIGH", "competing NAR claims / replica honesty", "typed judgments vs chat judges on guardrailing", "ishaannk/llm-vs-jev cross-note only", "deeper integrity fold is rh-guard", "nothing wins outright", "can be argued out of guarding"", "Jev IS the if-statement", "judgments/probabilities drive branches", "text model only writes prose", "interpreter owns variables/loops/budgets/replay", "otherwise maybe / confidence gate", "chaos samples after the gate", "southpolesteve/probably ≠ carldaws/hunch ≠ feelings ≠ Kungie/gut ≠ Illusion47586/judge ≠ tidymodels/probably", "133★ / forks 10 live", "build calibrated classifiers from human feedback", "retrieve by relevance not resemblance", "one calibrated yes/no per memory in one request", "pointer mode 17/18 19/20 *theirs*", "embedding resemblance misses the allergy", "samdotmak/jev-recall ≠ jev-search ≠ jev-sift ≠ carryforward ≠ chopratejas/invalidate", "memory leases ended by new evidence", "six Nouls then fixed rules in code", "0 of 157 false invalidations", "questions/plans/directives are not evidence", "unsure → review queue", "host keeps the store", "name↔body / comment truth / test-claims", "mizchi/jev-lint is mizchi/jevlint rename", "no shipped rule has severity error", "~1 in 5 findings wrong *theirs*", "mizchi/jev-lint ≠ huntedman/JevLint ≠ MichitoSugawara/jev-lint", "JSON Schema → typed JSON via Jev", "noul_threshold 0.5 decoder not a proof", "IncompatibleSchemaError lists every bad property", "on-device Laya CoreML ANE", "~5 ms P50 short decisions", "189/189 FP16 checkpoint parity", "10× not achieved", "mizorewww/laya-coreml ≠ gliner-native-runtime ≠ jevmlx ≠ NandhaKishorM/laya", "softmax over allowed tokens ≠ Noul", "question-first cache", "Micha0827/snapjudge ≠ githubnext/localjev ≠ jevmlx ≠ cendress/SnapJudge", "Jev-first Pi agent loop", "slow-LLM fallback", "explicit action menu / CandidateSource unimplemented", "62 tests wiring not quality", "direwolfiy/JevPi ≠ standardagents/jevpilot ≠ pi-jev-control", "resume-screening bias audit methodology", "name×resume factorial independent Nouls", "callback determined by resume quality", "mean-probability name gaps operationally negligible", "natemoo-re/bias-bench ≠ BBQ", "Plan/PRD panel → code-owned pass|review|block", "cheerleading out of scope", "austindixson/planalyzer ≠ single-goodness Noul", "cost-aware multi-model routing/escalation", "decide vs do", "successful-task cost", "cannacre8ive/switchboard-ai ≠ ha-switchboard ≠ hermes-switchyard", "frozen-protocol zero-shot bench", "TypeSafe Jev vs PrismNLI vs Laya", "contamination caveat", "elcronos/jev-vs-open-decision-models ≠ JevBench ≠ DMB", "context-window admission control", "VOI gate which tokens are worth the expensive model", "fail polarity per lens", "on small inputs lenses lose money", "cvsgireesh/jevusher ≠ jev-sift ≠ winnow", "typed decision control plane", "receipt ≠ authorization", "historical-v0 zero retained cases", "MokiMeow/jev-fabric ≠ jev-forge ≠ dgp", "live 15-dim typed rubric re-score per pause", "scoring economics exemplar", "OpenJev/Codiv ≠ TypeSafe hosted", "jose-troche/live-rubric ~$0.000004 desc / ~$0.000006 README", "adversarial pre-registered Jev eval", "28 predictions before data", "123,805 requests", "confidence does not track ignorance", "polite injection 65% / crude 0%", "willkelly/jev-evaluation ≠ jevals ≠ jev-baselines-eval", "provider-neutral Elixir/BEAM Noul/Choice/Score SDK", "class infrastructure", "nshkrdotcom/system_one_sdk ≠ typesafe_sdk ≠ dannote/jev", "question-linting of Jev questions themselves", "nine jaggedness rules, no API key, no labelled data", "static lint ≠ measured separation", "yodablocks/jevq ≠ tenbin ≠ JevLint ≠ commitjev", "open-weights Laya as class exemplar (binding)", "Nx/Bumblebee runtime", "host chooses backend", "ChristianAlexander/laya_ex ≠ system_one_sdk ≠ dannote/jev ≠ NandhaKishorM/laya", "on-chain/edge Laya deploy", "parity_verified stays false", "model output never grants Tx", "humandebri/IC-Laya ≠ laya_ex", "auditable weekend replica", "Jev outputs never used for training", "soft human-vote distributions", "unpaired 0.577 vs 0.727", "agilabs-ai/jev48 ≠ JevBench ≠ Mapika/decider", "adversarial dual-judge / framing attack surface", "comparative framing is the usable judgment", "prior injection crowds out evidence", "copyleftdev/ember ≠ ember.js", "Laya specialist fine-tune pipeline", "training still GPU-pending", "PIXELZX0/XERON ≠ convaiinnovations/laya", "Hub Laya replica drop", "daliborsb/laya ≠ convaiinnovations/laya ≠ NandhaKishorM/laya", "System One student distillation corpus", "gold is programmatic", "teacher is closed-API clone", "do not distill Jev as teacher of record", "MagaBitmex/jev-4b-distill-data ≠ missing student checkpoint", "non-LLM VIN System One", "planning depth not chat", "lewislululu/jevon ≠ douglance/jevon", "source-bound evidence checks", "local quote mismatch needs no API", "exit 0 ≠ claim truth", "WaynezProg/jev-kit ≠ jonathanavis96/jev-kit (Airlock) ≠ jev-use ≠ jev-mcp", "independent System One evidence catalog", "scores not one leaderboard", "no external record currently reproduced", "TokenTrim no-Jev matched hybrid 62.4%", "reachjalil/system-one-bench ≠ mallahyari/system-one-benchmark", "21 tasks · 134 items · 208 questions", "scenes from public GitHub contracts, not production logs", "SivletLabs/jev-eval ≠ willkelly/jev-evaluation ≠ 4esv/jev-eval ≠ xxkuboxx/jev-eval ≠ onlyoneaman/jev-eval ≠ dayhaysoos/jevals", "option isolation (sibling-blind)", "permutation-equivariant", "Hub OWNER not published", "nafisazizir/hev ≠ jaredpalmer/kev", "frozen local LLM logits, no trained decision head", "residual-head 9,222-param decreased 73/96→67/96", "confidence = 1−normalized entropy, not P(correct)", "yuki-oshio/mini-jev ≠ r-ms/mini-jev", "Jev classifier as autoregressive next-token predictor", "ChatJev-style soundness theater", "erik-dunteman/ChatJev ≠ dannote/jev ≠ jev-gpt", "calibrated decision head × AlphaProof value head", "implementation-layer isomorphism, semantic difference", "timeout = censoring", "do not launder Noul as proof", "parallel rank-prediction vs serial selection", "independent questions can conflict", "zzzzzec/jevsort ≠ keltokhy/jsort", "curated open System One ecosystem catalog", "rupeshpoojary9/awesome-open-system-one ≠ AnotiaWang/awesome-jev", "arXiv paper radar with Jev relevance scoring", "ranking ≠ calibration / 0.5 still soft", "fail-open failed evals not marked seen", "train calibrated ~27M from scratch", "typed Q→prob dist / one forward pass / no LLM decode", "hyusi2003/MiniSystemOne ≠ Colvin0315/MiniSystemOne", "description-only stub / size 5", "ESCI hard probe fails four of six", "jev_bool ECE 0.242 inversion 0.255", "do not re-fold §60 six-gates as new", "jobbyjev one-request-per-company from batch-size result", "find/design/evaluate TypeSafe Jev decision loops", "karanb192/jev-architect ≠ samtay32/jev-system-architect", "Jairik/jev-distiller size 1", "distill-Jev UI stub / do not distill Jev as teacher of record", "post-launch scored use-case map / Jev self-scores then human curation", "licensedsaucer9-web/jev-opportunities", "Jev-inize a use case into classifier/router", "gavinHuang/jevinize → simple-jev not TypeSafe", "featherless-ai/simple-jev", "compare saved decisions / same label can still change the branch", "VihaanAgarwal/jev-diff ≠ Saik0s/diffusiongemma-jev-macos", "not tested with a live Jev API key", "constrained logprob + temp/Platt ≠ Noul", "OpenJevPro pastes openjev-sglang JevBench as own", "zhangcy122/OpenJevPro ≠ IamBusy/OpenJev ≠ ekzhang/openjev-sglang", "PolyForm Noncommercial", "SmolLM-135M / sub-70ms / 0 output tokens", "demo P(True) 0.5052 / Choice conf 0.2872 / Score conf 0.0055", "README claims MIT / GitHub license null / no LICENSE file", "patelvishwa112/jev-system-one-rlcd ≠ arnabgho/rlcd-lite ≠ blackwood-rlcd", "source-backed Awesome Jev radar / 306+ commit-pinned", "logicrw/awesome-jev-projects ≠ AnotiaWang/awesome-jev ≠ yibie/awesome-jev ≠ cobanov/awesome-jev ≠ rupeshpoojary9/awesome-open-system-one", "auto GitHub sync / Issue-only submissions", "hashed n-gram encoder / rival-aware attention", "olanotolu/jevbetter vs jevlike starter", "synthetic hard menus top-1 0.916 vs 0.873 / ECE 0.0182 vs 0.0367 / 40 vs 4608 menus/sec", "shuffled-context control 0.335", "Turn any open LLM into System-One Jev", "uspraveen/Jevify ≠ Mintzs/jevify ≠ gulagala001/jevify", "Jevify-any-LLM architecture probe", "description-only stub / size 0", "Train encoder-only calibrated decision models from a task sentence", "Exu is a toolkit, not a method", "strictly proper scoring rule", "Pre-alpha", "Ruivalim/exu-base", "scratch-trained calibrated decision model", "typed Q → probability dists", "Colvin0315/MiniSystemOne ≠ hyusi2003/MiniSystemOne", "no published weights download URL", "90.5 seconds / 29.2% pipeline evidence", "p_i/p_j independent of other candidates", "Recipe for calibrated decision models — small model out", "init → synth → train → eval → serve", "91.1 % / ECE 0.022 *theirs*", "Jev zero-shot 75.1", "scienthoon/luce", "Put Jev's three headline claims on trial", "0.5B local GPU", "46x speedup / accuracy identical", "ECE 0.624 sentiment catastrophe", "bigger model worse calibration", "RichardoMrMu/jev-mini ≠ yuki-oshio/mini-jev ≠ r-ms/mini-jev", "System-1 decision engine for local LLMs", "structured choices only", "JSON parse of generated text ≠ Noul", "TypefAI JEV / Journal Entry Voucher", "tapsin/jev-local ≠ us/jev-local ≠ Argos1111/jev_local", "Jev 1.13 reward-model eval across 8 benchmark tracks", "40,940 examples / 0 API errors", "RewardBench v1 92.58%", "Precise IF 50.63%", "goya4140/jev-reward-model-evaluation", "Scaffolding in progress", "Jev vs LLM support-ticket routing", "static + live decision bench", "TypeSafe's own published benchmark", "illustrative simulations, not live API calls", "JevBench v1 — smart/cheap/fast/reliable", "I/C/S/K 25% geometric mean", "classifier.dev fast tier 84.8 is Jev behind its own API", "do not re-fold §78 v1.2 board as new", "Laya (421M) 70.1 now on board", "Zero-shot/few-shot LLM routing", "hard budget filter before Jev", "Jev never asked to perform budget arithmetic", "Jev judges the next state, XState enforces transitions", "simulation uses synthetic keyword fixtures", "catalog gravity", "v-modal/awesome-jev-tools", "★339 live REST", "curation is not endorsement", "crawler-maintained directory", "Daily GitHub + npm sweep, human-merged", "RadRebelSam/awesome-jev ≠ AnotiaWang ≠ yibie ≠ cobanov ≠ logicrw ≠ v-modal", "HF peft SPLADE/BGE reranker", "rdxtremity/jev-reranking ≠ carlaiau/jev-reranking", "query-side encoders, not a Jev replica", "ONNX System One Qwen3.5-4B scorer", "source:pngwn/system-one-qwen3.5-4b-scorer", "CC-BY-NC-4.0", "temperature 1.75", "transformers.js AutoModel cannot load this graph", "Consistency benchmark Space", "This Space contains no benchmark result yet", "12-case plumbing fixture", "Benchmark-driven Jev router and judge", "cheap alone is not success", "Jev does not write, sum prices, or claim accuracy %", "Sol 94.2 / Luna 83.9 / Jev path 89.7", "19.2% Sol / 62.3% cost save / 4.5pp miss of 2pp non-inferiority", "p50 latency worse than Sol due to routing overhead", "erendikmenn/jev-llm-router-benchmark ≠ jev-rag-benchmark ≠ ryantsai/jev-llm-router", "Express + node:sqlite", "mock and Jev decision engines", "previous_ticket_count >= 3 is code", "MIN_CONFIDENCE 0.6 still soft", "substring false positives", "aesaganda/jev-ticket-router ≠ SarathChandraBellam/jev-vs-llm-ticket-router", "Universal Figure & Diagram Router", "confidence ≥ 0.85 hard-gate is theater", "generative AI banned from scientific plots", "six visual branches", "hoangngochuong24947-gif/jev-figure-router", "human-labeled (state, question, label)", "166,054 rows / 22 configs", "soft_label for human uncertainty", "Praveenrajus/jev-bench ≠ fstandhartinger/jevbench", "ternary bonsai System One GGUF", "openjev's mechanism, Bonsai's weights", "Hub does not ship weights", "100/100 easy T/F is not Harbor", "label_mass ≠ correctness", "stock llama.cpp Q2_0 silently gibberish", "NicolaiMTLassen/open-bonzi-jev ≠ NicolaiLassen", "transformers.js DeBERTa ONNX", "source:com-kotobalabs/open-jev-deberta-v3-large", "temperature 1.05", "AutoModel from_pretrained works", "onnx-community/open-jev-deberta-v3-large-ONNX ≠ system-one-qwen3.5-4b-scorer-ONNX", "107★ densify", "GH 151M vs README 149.6M", "PR #1 now closed unmerged", "do not re-fold §71 claim-audit as a beat", "typed decisions, RLCD, confidence-gated routing", "structured ≠ correct", "mock not live API", "26 tests", "wjdjdakf17/jev-study ≠ baekenough/jev-study", "bonzi-27b-v2 / ternary-8b / 27b-v1 GGUF family densify", "WANLI-256 74.6% / 65.2% / 71.1% *theirs*", "Bonsai 1 27B Q1_0 runs on stock llama.cpp", "ternary still needs PrismML fork", "hf:heman10x/openJev-verdict-2.0 twin tokenizer-only", "OpenJev Vision image classification + uncertainty", "CLEVR-4 held-out joint 0%", "hfdataset:IamBusy/OpenJev-Vision-Research-v0.1 12,832", "294,912 derived targets not independent samples", "Laya multilingual ONNX WebGPU typed-decisions port", "63/63 selected answers / 5.1e-4 CPU / 1.2e-2 WebGPU", "UpHash-Network/mini-jev is yuki-oshio transfer", "jev-injection-bench 11,900 labelled prompts", "Jev best ranking / Haiku better ECE 0.021 vs 0.058", "0.5–0.9 band is where Jev's numbers do not mean what they say", "Prompt wording moves panic 28%", "manojlds/jev-dspy-bench ≠ dspachos/jev-dspy ≠ jmanhype/jev-dspy-lab", "Jev agreement is similarity, never ground truth", "no aggregate quality grade or merge gate", "AbstentionBench-on-Jev rank 1 of 20 vs 2025 field", "question-asymmetry", "forward-looking 0.465 never extreme", "openkev calibration layer not a runtime", "ECE vs coverage independent", "select_threshold returns inf", "escalation catches uncertainty not ignorance", "misakaikato/openkev ≠ jaredpalmer/kev", "pdf-race Docling→Jev vs Gemini", "parser owns the wall clock", "12/12 tie is a tie", "titles selected not generated", "flopcheck 16 calibrated tweet judgments", "mechanical tells in code", "ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas", "Laya calibration lab Gradio MCP", "T never changes argmax", "confidence ≠ top-label p", "easy probe set refused", "40–48 rows too small to ship T", "Gemma-4 26B-A4B jevify classification+calibration", "LoRA adapter twin not independent eval", "Gemma-4 E4B jevify", "E4B LoRA stub card", "kushalpatil/jevify-gemma4 ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify", "GH kushalpatil07/jevify 404", "PAWS 0.580/ece 0.288 is the weak cell", "smaller E4B slightly better OOD ECE than 26B-A4B", "Hub jevify merged LoRA ships weights", "bonzi Bonsai-8B v1 GGUF densify", "Bonsai-1.7B v1", "Bonsai-4B v1", "WANLI-256 64.5% / 60.2% / 52.0% *theirs*", "rank #4 / #5 / #6 of 6", "JulesHuisman/jev-eval scaffolding / README SHA c356a584 (was empty e69de29b)", "JulesHuisman/jev-eval ≠ SivletLabs/jev-eval ≠ willkelly/jev-evaluation ≠ 4esv ≠ xxkuboxx ≠ onlyoneaman ≠ dayhaysoos/jevals", "7 bands 6/10 vs 40 bands 0/10", "source receipts + confidence slider re-policy without re-inference", "32/32 synthetic is smoke not production", "classify HF datasets across typed semantic dimensions", "roadus2 watch misspelling; lock roadius2/ultra_laya", "ultra_laya REVIEW defects", "default branch claude/laya-jev-review-gg5ppo", "XNLI EN 88.3% ECE 0.032 → RU 77.3% ECE 0.096", "Δ −11.0 pp [−14.2,−7.8]; ECE +0.063", "MASSIVE no detectable difference at n=600", "confidence is function of p_max (r=1.000)", "pointer-not-generator 400 human-authored responses", "proposed ≠ authorized", "FewRel 160: Jev 85.0% vs lexical 13.125%", "gated 100% (95/95) coverage 59.375%", "J++ composable semantic computation language", "judge-jev 0.5 still soft", "947 repos scored; A 273 / B 302 / C 372", "LLM rubric ≠ benches", "No benchmark winner is claimed", "phishing: naive 62.6% vs regex 91.8%; 5-atomic + LR 95.0% *theirs*", "AITuber tension ±15", "README npm global; repo is Rust", "git-confess code owns counting/blame/ratio", "httpx exhibit 11% (13/119) *theirs*", "90d trend +12.40% vs random +12.75% vs BH +41.71%", "5m win rate 25%", "Awesomejev 656 entries / 38,160 stars", "tracker likes 64 (+4) lastModified UNCHANGED", "Laya present; Blackwood ABSENT; Archer still promised_not_landed", "Blackwood tracker ABSENT; likes 2 gated manual", "r = c - p_a", "ECE 0.021; acc 0.807 vs warmup 0.746", "Independent primitive", "11.57s vs 54.10s · 4.67× · 120/128 *theirs*", "default path is pretrained Gemma probs not trained RLCD head", "GH Meanblock 404; lock leesk212/JEV-CPU", "softmax over letter slots ≠ Noul", "WANLI 0.741 vs openjev v2 0.77 *theirs*", "3-way NLI ≠ Noul", "priority 0.464 = majority floor", "banking77 contaminated", "raw margins not probabilities", "do not distill Jev as teacher of record (they distilled Haiku)", "“0.9 is not one number”", "ranking ≠ calibration", "banking77 0.8–0.9 stated 0.86 actual 0.73 over-confident *theirs*", "≠ Praveenrajus/jev-bench ≠ fstandhartinger/jevbench", "$0.0000153–$0.0000226 vs circulating $0.0004 (~20×)", "Score is 0..n-1 expectation not 0–1", "Noul has no confidence field", "TCP floor 198.8 ms", "type reliability is not a reason to choose Jev (json_schema 5/5)", "gateway tax not one number", "Function-only 5/8 vs hybrid 8/8", "4/8 without Jev", "8 designed cases not conversion lift", "200-row pilot Jev 86.5% 173/200 vs Gemini Flash-Lite 86.0% 172/200 vs Pro 87.0% 174/200 *theirs*", "not a ranking", "情緒測謊器", "8-example Jev vs GPT-5.6 Sol ~64× cost 5.4× latency *theirs*", "synthetic; no inference", "≠ JevBench v1.2 §78", "Judged 3317 / listed 2560", "Jev judges, code applies policy", "APA “microsecond policy / zero hallucination” overclaim", "Client-side quiz; pointer from held docs; scanned-PDF warn", "Jev judges / agent reasons / user decides", "selecting an option is not permission to implement", "pattern exact, judgement must clear floor", "no matching pattern → no model call", "not a correctness oracle", "Spec vs artifact remainder", "treating 0.85 as 85% / minProbability hard-gate as Harbor", "VERIFY acquires discriminating evidence, never same-pool confidence-only rescoring", "fast/full/max are ceilings not sizes", "Solar writes, Jev chooses NEXT ACTION", "do not reopen or amend PR #23 or #24 or #25 or #26 or #27", , "Calibration is not alpha", "NO CURRENT ALPHA CANDIDATE", "ΔR² approximately +0.00084", "Brier 0.2131387", "ECE 0.0421875", "Adding Jev probability to deterministic volatility improved Brier by only 1.4058e-05", "default 0.5 keeps zero non pinned", "keepResult median 0.14 to 0.17", "keepCall median 0.28 to 0.35", "usable range is about 0.10 to 0.25", "7.8% to 57.9%", "judges results it never sees", "task-finish eval not built yet", "$0.002 per compaction", "slavadubrov/sgr-judge-bench ≠ slavadubrov/jev-judge-bench", "Jev 108/120 $0.083 0.34 s", "Luna SGR 114/120", "paired Jev accuracy-difference intervals include zero", "not evidence of equivalence", "GLM SGR 26/120 93 format failures", "Terra-planned Jev hybrid 55/120", "rule-based by default, optionally Jev-backed", "empty README", "missing key cannot break the experience", "prefill plus exactly one decode", "softmax over A/B/C ≠ Noul", "BBQ 9,053/10,000 (90.53%)", "ECE 0.0890", "Mean confidence 0.9943", "overconfident", "score and noul not implemented", "DGUI 12 rows (was 6)", "INSTRUCT 119 rows likes 2", "encode the state once, decide everything in parallel", "0.740 accuracy against a 0.508 majority", "ECE 0.047", "fine-tune's advantage ends where its 384-token training data does", "jasonkneen/open-jev ≠ pngwn/open-jev", "same sha d41dc3cd", "Space does not call Jev", "recomputes routing from saved probabilities", "200-case Jev 97.0% / 100.0% / 95.0% / MAE 9.22", "synthetic repository benchmark", "Jev evaluations are advisory", "YehuiTang0316/jev-nlgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep", "default threshold 0.8 still soft", "40-line windows cannot prove whole function", "token-native sequential start/end Choice", "Gemini/Haiku stubs not configured yet", "handful of hand-written examples, not a benchmark", "Jev judged exactly what it was given", "laguagu/jev-skills ≠ laguagu/jev-evidence-lab ≠ Pleo2/awesome-jev-agent-skills", "contract_passed is not a claim of guaranteed factual truth", "Wilson lower bound 0.85 floor", "fixture mode no savings claim", "SemIf 2207★ (+21 vs §110 2186)", "jevlike 1043★ (+5 vs 1038)", "TypeAR 15★ (+1 vs 14)", "AnotiaWang 97★ (+1 vs 96)", "yibie/awesome-jev 506★ (+16 vs 490)", "Laya likes 822 (was 802)", "tracker likes 64 flat, lastModified UNCHANGED", "do not reopen or amend PR #23/#24/#25/#26/#27/#28", "Heman10x-NGU/Verdict-open-jev ≠ Heman10x-NGU/openJev-verdict-2.0", "TF-IDF + LogReg ECE 0.0207 vs Jev 0.1440", "Verdict-open-jev 48.07% vs Jev 90.80%", "abstention combined recall 10.00%", "p50 35.58 ms", "K=25 (maximum capacity) 72.00%", "0.85 coverage 84.60% selective risk 1.18%", "26.1× faster than standard Qwen JSON generation", "Jevify 90.0% / 167 ms CUDA graphs disabled", "Finding 1: Brier on stated confidence alone is a trap", "grpo_rlcr 0.78 / ECE 0.084", "reliability 0.007 but resolution 0.000", "27 900 schema-driven decisions", "13 600 / 13 600 questions", "candidate mass min 0.99999624", "22 configs · 166,054 rows · 4 calibration-gold", "sha a39eba3f", "Student B MAE 0.148 / Pearson 0.836 / 86.0%", "pngwn/open-jev-laya-bench README 404", "sha 9f69c742 likes 2", "HDFS 0.9933 (745/750) / retain 0.0084", "BGL ERROR/FATAL protection 1.0000", "2,479 / 2,500 HDFS uncertain", "cache hit 0.9648 (2412/2500)", "$0.153936 estimated", "E2 recomputes from saved probabilities", "Space sha eda59e0a", "MASSIVE English 0.783 / Khmer 0.033 / Hindi 0.133", "40–48 rows too small to ship T", "T never changes argmax", "siren2345/jev-single-decode-transformers ≠ siren2345/jev-single-decode", "Split Transformers experiment from llama.cpp runtime", "tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab", "Four experiments stress-testing TypeSafe's Jev: calibration, bundle bias, label bias, and ensembling", "second pass must be $0.00 from cache", "The pages never call Jev", "Gemma 4 31B 77.0% / Jev 1.13.0 61.4% / Laya 322M 0.0%", "restriction state 95.0% against 84.4%", "None of the systems are particularly good at knowing when to stop and ask", "They skip the question and call a tool directly", "100% schema pass", "six-field joint 48.8% vs 72.8%", "ywchiu/jev_benchmark ≠ Running-Dolphins/jev-bench ≠ Praveenrajus/jev-bench", "ACT / REVIEW / FALLBACK", "A provider failure, timeout, malformed output, or missing answer is **not** a policy outcome", "confidence is descriptive provider output, not a substitute for probability", "Quality denominators include only valid scored answers", "an exact halfway tie chooses the lower level", "aiwithenoch/Jev-Skill ≠ simplosophy/jev-skill ≠ laguagu/jev-skills", "The local path does not claim to turn a smaller checkpoint into Jev", "Low support becomes decision: \"review\"", "MIT-0 SPDX NOASSERTION", "current-llm", "结构兼容，不是 Jev 模型能力", "altryne/jevify ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify", "Find where Jev belongs. Design the questions. Measure the difference", "TypeAR-AI/TypeAR 301 → TypeLLM/TypeLLM", "TypeLLM/TypeLLM 16★", "SemIf 2241★ (+34 vs §111 2207)", "jevlike 1051★ (+8 vs 1043)", "AnotiaWang 98★ (+1 vs 97)", "yibie/awesome-jev 525★ (+19 vs 506)", "Laya likes 864 (was 822)", "tracker likes 67 (+3 vs 64)", "lastModified UNCHANGED `2026-09-20T04:29:16.000Z`", "do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#32", "hysteresis enter/exit / replay policy without inference", "calibration does not compose / hop-ECE permutation-invariant", "equal-width vs quantile ECE / ranking ≠ calibration", "Qwen2.5 ≠ Archer / Qwen 3.8 sparring ≠ Archer / Qwen/Qwen3.8-27B ≠ Archer", "Deferred Crispification / TCE / AMS", "g0runmezadam/what-is-jev IS tunahansahin897/what-is-jev", "pd.cut equal-width vs jeval quantile", "A hunch is a probability with a policy attached", "soundness theater / measurement theater / hourly 0843", , "Jev Capability Resolver / NiazMorshed2007/jcr", "one tool nested capability tree / returns context / does not execute", "skills vs capabilities / workflow+judgment vs operations", "format independent of Jev / proposed open standard", "JCR_BAND_RATIO 0.6 is application policy / soft scores ≠ hard gates", "routing ≠ permission / docs ≠ authority to run", "sol-vs-opus5-20 lookup+explain / n=1 / Not Harbor task-execution", "wall-time mixed / Sol slower with JCR in 19/20", "NiazMorshed2007/jcr ≠ skill-broker ≠ skillranker ≠ jev-sift ≠ jev-lens ≠ jevusher ≠ jev_select_capability", "do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34", "notes.md §116", "copy the SemIf/MLX installer?", "quote 5.21× as beating Jev?", "treat 0.845 as a TypeSafe replica?", "collapse SemIf into kw2828/zhihz/semif-rs/semif-serve", "softmax over options as a Noul", "llm prompt to jev primitives", "conversion assistant not equivalent behavior", "heuristic conversion ≠ calibrated Noul", "alexwestco/llm-to-jev ≠ altryne/jevify", "user-provided 0940 / notes.md §118", "judge ≠ actuator", "candidate_mass", "softmax over A–H ≠ Noul", "hourly 0947 / notes.md §119", "ggmlc GGUF is not llama.cpp", "serving substrate ≠ calibrated replica", "Qwen3.5-9B ≠ Archer", "planner writes JEV selects", "hourly 1049 / notes.md §120", "open recreation ≠ calibrated replica", "semantic lint is a sensor not a proof", "cutoff 0.8 still soft", "paired bootstrap CIs *theirs*", "Same accuracy, 35x faster *theirs*", "hourly 1143 / notes.md §121", or "cascade sign-flip / calibration theater": read `references/faq.md`, then `references/mental-models.md`, then `references/mixed-architecture.md`, then `references/judgment-class.md` before any mapping. Proof, @@ -287,5 +287,21 @@ Do not copy keys. Rebased onto `8f446c4` (merged #41) after `fb15455` (merged #42 v0.5.0). Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. + + +## Hourly 1143 HIGH (`notes.md` §121) + +open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. +cutoff 0.8 still soft. paired bootstrap CIs *theirs*. +Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. +wire-compat ≠ replica. permission ≠ confidence. catalog ≠ endorsement. +Qwen3.5-4B ≠ Archer. Do not copy keys. Fresh PR off `a61372f` (merged #43). +Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. +`invented_signal: false`. + Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/agent-self-assessment.md b/.agents/skills/augustus/references/agent-self-assessment.md index 96579219..36836c8b 100644 --- a/.agents/skills/augustus/references/agent-self-assessment.md +++ b/.agents/skills/augustus/references/agent-self-assessment.md @@ -945,3 +945,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/applied-mappings.md b/.agents/skills/augustus/references/applied-mappings.md index 52718558..8cc68729 100644 --- a/.agents/skills/augustus/references/applied-mappings.md +++ b/.agents/skills/augustus/references/applied-mappings.md @@ -2524,3 +2524,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/composition-algebra.md b/.agents/skills/augustus/references/composition-algebra.md index 8bf2a3fe..288868f7 100644 --- a/.agents/skills/augustus/references/composition-algebra.md +++ b/.agents/skills/augustus/references/composition-algebra.md @@ -2322,6 +2322,80 @@ catalog ≠ endorsement; pick_by_id vs pick_second. do not reopen or amend PR #23–#42. Soft Noul ≠ hard safety. + +369. **open recreation PRIMARY** (kshetrajna12/reflex): + position 1 (Operand). It is an open re-creation of Jev. Qwen3.5-4B. + WebGPU 0.8B less calibrated. open recreation ≠ calibrated replica. + Qwen3.5-4B ≠ Archer. Full cards: `judgment-class.md`, `faq.md`. +370. **pngwn / DavidHatley independent not replica**: + pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*. + This dataset and model are independent research artifacts, not reproductions of Jev or RLCD. + Full cards: `validation.md`, `judgment-class.md`. +371. **semantic lint sensor PRIMARY** (lakeday-org/perch): + position 8 (Metric) × 3 (Guard). perch 164★ MIT HEAD ba775a9940b6. + Semantic code linting with Jev. semantic lint is a sensor not a proof. + Full cards: `formal-methods.md`, `faq.md`. +372. **oxlint cutoff 0.8 still soft** (wobsoriano/oxlint-plugin-jev): + position 3 (Guard). reports an error when the yes-probability clears cutoff. + cutoff 0.8 still soft. soft scores ≠ hard gates. + Full cards: `mixed-architecture.md`, `validation.md`. +373. **jevgrep / patdown / jevvy namesakes** (nassim-arifette + tyler-dot-earth + PanAchy): + nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep. + patdown fuzzy linter. PanAchy/jevvy ≠ Atominac/jevvy. + Full cards: `faq.md`, `applied-mappings.md`. +374. **paired bootstrap CIs PRIMARY** (emretheus/jev-rag-benchmark): + position 8 (Metric). +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31. + +7.62 pts SciFact CI +4.88 to +10.38. paired bootstrap CIs *theirs*. + emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark. + Full cards: `validation.md`, `faq.md`. +375. **BANKING77 speed-test *theirs*** (MohtashamMurshid/jev-speed-test): + position 8. BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*. + Confidence was useful, not a guarantee. Full cards: `validation.md`. +376. **cascade missed parity** (saurabhkumar8112/jev-gpt5-routing-study): + 430/500 vs GPT-5 432/500. This is not demonstrated equal-quality savings. + Full cards: `validation.md`, `mixed-architecture.md`. +377. **invented-ticket gate** (abh2050/jev-test-confidence-gate): + 24 invented tickets. Routing errors caught by the gate 0 of 3. + sample too small to establish calibration. Full cards: `validation.md`. +378. **cartpole not TypeSafe Jev** (tinmanlab/cartpole-jev): + This is not TypeSafe Jev. No real API requests were made. + Controller owns force. Full cards: `formal-methods.md`, `faq.md`. +379. **laya-jev wire-compat** (KonghaYao/laya-jev): + 按官方接口写的客户端只改一个 base URL. wire-compat ≠ replica. + serving substrate ≠ calibrated replica. Full cards: `judgment-class.md`. +380. **gqgs laya-onnx densify** (gqgs/laya-onnx): + densify §120. 496.8 MiB. tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx. + Full cards: `judgment-class.md`, `faq.md`. +381. **laya_router 35x systems comparison** (glukicov/laya_router): + Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. + 0.600 / 184 ms vs 0.600 / 6,415 ms. Full cards: `validation.md`. +382. **catalogs namesake lock** (BeatAPI/awesome-jev + cousins): + All 125 projects. catalog ≠ endorsement. + BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub. + Full cards: `faq.md`, `applied-mappings.md`. +383. **super-jev densify permission ≠ confidence** (Kevthetech143/super-jev): + densify experimental V0.2.0. not affiliated with TypeSafe AI. + permission ≠ confidence. Full cards: `mixed-architecture.md`, `agent-self-assessment.md`. +384. **jev-lab namesake + openjev collision + GGUF densify** (Pasblinn + allay-team + mradermacher): + Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab. + Independent project. Not affiliated with TypeSafe. + allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev. + 2022 Mineflayer Jevalent collision. static quants of kushalpatil/jevify-gemma4-e4b. + Full cards: `faq.md`, `judgment-class.md`. + +Hourly 1143 items 369–384 (`notes.md` §121). Do **not** +re-fold §120 items 353–368 / §119 items 337–352 / §118 items 322–329 +/ §117 items 330–336 / §116 items 309–316 / §115 items 303–308 / +§114 items 289–302. +Skip Archer rewrite. +open recreation ≠ calibrated replica; semantic lint is a sensor not a proof; +cutoff 0.8 still soft; paired bootstrap CIs *theirs*; +Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; +permission ≠ confidence; catalog ≠ endorsement. +do not reopen or amend PR #23–#43. +Soft Noul ≠ hard safety. + + Hourly 0947 items 337–352 (`notes.md` §119). Do **not** re-fold §118 items 322–329 / §117 items 330–336 / §116 items 309–316 / §115 items 303–308 / §114 items 289–302 / §113 items 273–288. @@ -2418,3 +2492,7 @@ User-provided 0922 uniqueness lock: SemIf was formerly OpenJev; independent; not User-provided 0940 uniqueness lock: Turn decision-shaped LLM prompts into proposed Jev primitives; This is a conversion assistant, not an automatic guarantee of equivalent behavior; The compiler uses deterministic heuristics, not an LLM or evaluation model; It understands a deliberately small set of common prompt patterns; Generated instructions and criteria must be reviewed before production use; Score ranges such as 0 to 1 are translated into ordered Jev criteria; Prompts requiring open-ended prose are not a fit; suitability strong/partial/not_a_fit; compatibility full/partial/none; Writing new text stays with an LLM; Review the generated Score rubric; Jev scores ordered criteria, not an arbitrary 0-to-1 range; Everything runs locally in the browser; There is no framework, database, account, API, or server-side prompt processing; The key is read from the process environment and is never stored or printed; connect-src 'none'; alexwestco/llm-to-jev ≠ altryne/jevify ≠ ryana/jevify ≠ fidecastro/jevify ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify; HEAD 234058ab372d; README SHA 43cd94fb; LICENSE SHA 5f334006; compiler SHA fdf235d0; 2★; MIT; JavaScript; size 29; Pages https://alexwestco.github.io/llm-to-jev/; invented_signal false; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35; notes.md §118 Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/faq.md b/.agents/skills/augustus/references/faq.md index 8ae37ce4..0f33c7c7 100644 --- a/.agents/skills/augustus/references/faq.md +++ b/.agents/skills/augustus/references/faq.md @@ -3682,5 +3682,27 @@ No across the board. AUTO_ACT is not a Noul. closed-set fail-open stdlib-only. v No. catalog ≠ endorsement. jev-skill 109★ 90 scenarios. awesome-jev-live 673 entries 4★. planner writes JEV selects. minecraft-agent 214★ 131 JEV decisions 35 Astra calls. nether-final-08 8 minutes 43.300 seconds. RoboJEV structured simulator state not images. 策略未通过自己的回测门槛. 36 组参数全部净期望为负. no positive expectation under real costs. typed_evals NOT an official TypeSafe AI product. jev-as-judge is a sensor. CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*. accuracy is a trap. 9.0% base rate always-no 91.0%. 111-case benchmark *theirs*. Do not reopen or amend PR #23–#42. `invented_signal: false`. `notes.md` §120. +## Treat reflex / pngwn / a blueprint as TypeSafe Jev? Treat Qwen3.5-4B as Archer? + +No. It is an open re-creation of Jev. less calibrated. open recreation ≠ calibrated replica. pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*. This dataset and model are independent research artifacts, not reproductions of Jev or RLCD. Qwen3.5-4B ≠ Archer. A blueprint is not a replica. `notes.md` §121. + +## Treat perch / oxlint / patdown as a proof? Treat cutoff 0.8 as a hard gate? + +No. semantic lint is a sensor not a proof. the plugin reports an error when the yes-probability clears your cutoff. cutoff 0.8 still soft. patdown fuzzy linter. nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep. PanAchy/jevvy ≠ Atominac/jevvy. soft scores ≠ hard gates. `notes.md` §121. + +## Treat paired CIs / BANKING77 / 35x / 430/500 as Harbor? Treat 24 invented tickets as calibration? + +No. paired bootstrap CIs *theirs*. +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31. +7.62 pts SciFact CI +4.88 to +10.38. BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500. This is not demonstrated equal-quality savings. 24 invented tickets. Routing errors caught by the gate 0 of 3. sample too small to establish calibration. This is not TypeSafe Jev. No real API requests were made. No orders, no advice. `notes.md` §121. + +## Treat a Laya HTTP/ONNX/MCP server as a calibrated replica? Treat BeatAPI 125 as endorsement? Treat a permission plugin as confidence? + +No. wire-compat ≠ replica. KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL. gqgs/laya-onnx densify 496.8 MiB. serving substrate ≠ calibrated replica. All 125 projects. catalog ≠ endorsement. BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub. Independent project. Not affiliated with TypeSafe. permission ≠ confidence. Kevthetech143/super-jev densify experimental V0.2.0. allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev. Do not reopen or amend PR #23–#43. `invented_signal: false`. `notes.md` §121. + + + Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/formal-methods.md b/.agents/skills/augustus/references/formal-methods.md index 30870cec..3b73e358 100644 --- a/.agents/skills/augustus/references/formal-methods.md +++ b/.agents/skills/augustus/references/formal-methods.md @@ -1374,3 +1374,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/formal-semi-formal.md b/.agents/skills/augustus/references/formal-semi-formal.md index 108461d7..7747623e 100644 --- a/.agents/skills/augustus/references/formal-semi-formal.md +++ b/.agents/skills/augustus/references/formal-semi-formal.md @@ -87,3 +87,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/judgment-class.md b/.agents/skills/augustus/references/judgment-class.md index 8873b056..000fe5e1 100644 --- a/.agents/skills/augustus/references/judgment-class.md +++ b/.agents/skills/augustus/references/judgment-class.md @@ -1470,3 +1470,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/mappings.md b/.agents/skills/augustus/references/mappings.md index 2e7feeed..9994ecdc 100644 --- a/.agents/skills/augustus/references/mappings.md +++ b/.agents/skills/augustus/references/mappings.md @@ -2479,3 +2479,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/mental-models.md b/.agents/skills/augustus/references/mental-models.md index a7f7d03c..bcb5f900 100644 --- a/.agents/skills/augustus/references/mental-models.md +++ b/.agents/skills/augustus/references/mental-models.md @@ -2986,5 +2986,21 @@ catalog ≠ endorsement. Life analogue: shuffle the menu; if dinner changes, you ranked seats, not dishes. A new bottle does not recertify the medicine. Negative EV stays negative. Do not copy keys. + + +## Apply 1143 (`notes.md` §121) + +Open recreation ≠ calibrated replica. A cover band is not the studio master. +Semantic lint is a sensor not a proof. cutoff 0.8 still soft. +Third-party benches are *theirs*: paired bootstrap CIs, BANKING77, cascade +parity miss, invented-ticket n=24. Same accuracy, 35x faster is a systems +comparison ≠ semantic equivalence. Local Laya wire-compat is still substrate +≠ replica. permission ≠ confidence. catalog ≠ endorsement. +Qwen3.5-4B ≠ Archer. Do not copy keys. + Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/methods-catalog.md b/.agents/skills/augustus/references/methods-catalog.md index 60e0a039..5384bf2c 100644 --- a/.agents/skills/augustus/references/methods-catalog.md +++ b/.agents/skills/augustus/references/methods-catalog.md @@ -281,3 +281,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/mixed-architecture.md b/.agents/skills/augustus/references/mixed-architecture.md index d0fbb623..b3ef8418 100644 --- a/.agents/skills/augustus/references/mixed-architecture.md +++ b/.agents/skills/augustus/references/mixed-architecture.md @@ -1448,3 +1448,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/question-design.md b/.agents/skills/augustus/references/question-design.md index 8a87b195..f2e7208d 100644 --- a/.agents/skills/augustus/references/question-design.md +++ b/.agents/skills/augustus/references/question-design.md @@ -428,3 +428,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/toolbox-mapping.md b/.agents/skills/augustus/references/toolbox-mapping.md index 128c0d42..1dae9a11 100644 --- a/.agents/skills/augustus/references/toolbox-mapping.md +++ b/.agents/skills/augustus/references/toolbox-mapping.md @@ -357,3 +357,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/references/validation.md b/.agents/skills/augustus/references/validation.md index 47d0a0d1..90007d04 100644 --- a/.agents/skills/augustus/references/validation.md +++ b/.agents/skills/augustus/references/validation.md @@ -1161,3 +1161,7 @@ Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 s **Hourly 1049 HIGH (`notes.md` §120).** ggmlc GGUF is not llama.cpp. serving substrate ≠ calibrated replica. Qwen3.5-9B ≠ Archer. planner writes JEV selects. pick_by_id vs pick_second. soft scores ≠ hard gates. catalog ≠ endorsement. Do not reopen or amend PR #23–#42. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/.agents/skills/augustus/scripts/evaluate_decisions.py b/.agents/skills/augustus/scripts/evaluate_decisions.py index 945a3585..f572fb60 100755 --- a/.agents/skills/augustus/scripts/evaluate_decisions.py +++ b/.agents/skills/augustus/scripts/evaluate_decisions.py @@ -17,6 +17,11 @@ peaked ranking, not a Noul (mass outside the bag can be hidden) * pick_by_id vs pick_second: the same scores, shuffled option order; slot-two is ranking theater, not a Noul + * paired CI includes-zero: a bootstrap interval that covers 0 is not + evidence of equivalence (paired bootstrap CIs are *theirs*) + * same-accuracy speedup: tied accuracy plus a latency ratio is a + systems comparison, not semantic equivalence + * lint cutoff still soft: p >= cutoff fires a sensor, not a proof Select thresholds on one split, evaluate on another: run twice with different files. Missing labels or costs produce a stated limitation, not defaults. @@ -177,6 +182,29 @@ def _softmax(xs): return [e / total for e in exps] +def ci_includes_zero(lo, hi): + """Paired bootstrap CIs are *theirs*. Covering zero is not equivalence.""" + if lo > hi: + raise ValueError("need lo <= hi") + return lo <= 0.0 <= hi + + +def same_accuracy_speedup(acc_a, acc_b, lat_a, lat_b): + """Tied accuracy plus a latency ratio is systems comparison, not meaning. + + Same accuracy, 35x faster *theirs* is not semantic equivalence. + """ + if lat_a <= 0 or lat_b <= 0: + raise ValueError("latencies must be positive") + tied = abs(acc_a - acc_b) < 1e-12 + speedup = lat_b / lat_a + return tied, speedup + + +def lint_cutoff_fires(p, cutoff): + """cutoff 0.8 still soft. A sensor fire is not a proof.""" + return p >= cutoff + def pick_by_id(scores_by_id): """Max score wins regardless of presentation order.""" @@ -300,6 +328,15 @@ def self_test(): assert by_id == ["B", "B", "B", "B"], by_id assert second == ["B", "A", "B", "C"], second + # paired CI / same-accuracy speedup / lint cutoff still soft (1143). + assert not ci_includes_zero(0.35, 1.31) # emretheus XQuAD *theirs* + assert ci_includes_zero(-0.10, 0.20) # covering zero is not equivalence + tied, speedup = same_accuracy_speedup(0.600, 0.600, 184.0, 6415.0) + assert tied + assert 34.0 < speedup < 36.0, speedup + assert lint_cutoff_fires(0.81, 0.8) + assert not lint_cutoff_fires(0.79, 0.8) + print("self-test ok") diff --git a/.agents/skills/augustus/scripts/uniqueness_gate.py b/.agents/skills/augustus/scripts/uniqueness_gate.py index abc5c808..75e210ba 100644 --- a/.agents/skills/augustus/scripts/uniqueness_gate.py +++ b/.agents/skills/augustus/scripts/uniqueness_gate.py @@ -1,14 +1,15 @@ #!/usr/bin/env python3 """Uniqueness gate for merged 0843 (§114), merged 0915 NanoJev (§115), merged 0920 jcr (§116), merged 0922 SemIf (§117), merged 0940 -llm-to-jev (§118), hourly 0947 HIGH (§119), and hourly 1049 HIGH (§120). +llm-to-jev (§118), hourly 0947 HIGH (§119), hourly 1049 HIGH (§120), +and hourly 1143 HIGH (§121). Each lock must appear as one consecutive substring in every listed overlay. Fragments scattered across files do not count. -Also: YAML-parse SKILL.md frontmatter; notes.md owns §114–§120; -composition items 289–316, 322–329, 330–336, 337–352, and 353–368 exist; -findings batches #97–#103 exist. Items 317–321 stay unused. +Also: YAML-parse SKILL.md frontmatter; notes.md owns §114–§121; +composition items 289–316, 322–329, 330–336, 337–352, 353–368, and 369–384 exist; +findings batches #97–#104 exist. Items 317–321 stay unused. CHANGELOG.md must not hold uniqueness dump walls (dumps live in changelog-hourly.md). README.md must not hold the 0743 dump wall. Pages greps stay in docs/index.md and docs/_layouts/default.html. @@ -124,6 +125,10 @@ 'Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120' ) +UNIQ_1143 = ( + 'Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121' +) + OVERLAYS = [ "research/notes.md", "research/changelog-hourly.md", @@ -183,6 +188,8 @@ def main() -> int: failed.append(f"0947 lock missing as one substring: {rel}") if UNIQ_1049 not in body: failed.append(f"1049 lock missing as one substring: {rel}") + if UNIQ_1143 not in body: + failed.append(f"1143 lock missing as one substring: {rel}") notes = (ROOT / "research/notes.md").read_text(encoding="utf-8") if "## 114. Hourly 0843 HIGH" not in notes: failed.append("notes.md missing §114 heading") @@ -198,10 +205,12 @@ def main() -> int: failed.append("notes.md missing §119 heading") if "## 120. Hourly 1049 HIGH" not in notes: failed.append("notes.md missing §120 heading") + if "## 121. Hourly 1143 HIGH" not in notes: + failed.append("notes.md missing §121 heading") algebra = (ROOT / ".agents/skills/augustus/references/composition-algebra.md").read_text( encoding="utf-8" ) - for n in list(range(289, 317)) + list(range(322, 330)) + list(range(330, 337)) + list(range(337, 353)) + list(range(353, 369)): + for n in list(range(289, 317)) + list(range(322, 330)) + list(range(330, 337)) + list(range(337, 353)) + list(range(353, 369)) + list(range(369, 385)): needle = f"{n}. **" if needle not in algebra: failed.append(f"composition-algebra missing item {n}") @@ -218,6 +227,7 @@ def main() -> int: "## Batch #101", "## Batch #102", "## Batch #103", + "## Batch #104", ): if batch not in findings: failed.append(f"findings.md missing {batch}") @@ -266,6 +276,14 @@ def main() -> int: "Qwen3.5-9B ≠ Archer", "planner writes JEV selects", "pick_by_id vs pick_second", + "open recreation ≠ calibrated replica", + "semantic lint is a sensor not a proof", + "cutoff 0.8 still soft", + "paired bootstrap CIs *theirs*", + "Same accuracy, 35x faster *theirs*", + "This is not demonstrated equal-quality savings", + "permission ≠ confidence", + "wire-compat ≠ replica", ): if frag not in haystack: failed.append(f"SKILL.md missing fragment {frag!r}") @@ -292,6 +310,12 @@ def main() -> int: "Qwen3.5-9B ≠ Archer", "planner writes JEV selects", "hourly 1049 / notes.md §120", + "open recreation ≠ calibrated replica", + "semantic lint is a sensor not a proof", + "cutoff 0.8 still soft", + "paired bootstrap CIs *theirs*", + "Same accuracy, 35x faster *theirs*", + "hourly 1143 / notes.md §121", ): if frag not in proto_line: failed.append(f"SKILL.md protocol missing {frag!r}") @@ -304,6 +328,7 @@ def main() -> int: ("0940", UNIQ_0940), ("0947", UNIQ_0947), ("1049", UNIQ_1049), + ("1143", UNIQ_1143), ): if lock in changelog: failed.append( @@ -336,6 +361,7 @@ def main() -> int: f"jcr chars={len(UNIQ_JCR)} lock0922 chars={len(UNIQ_0922)} " f"0940 chars={len(UNIQ_0940)} 0947 chars={len(UNIQ_0947)} " f"1049 chars={len(UNIQ_1049)} " + f"1143 chars={len(UNIQ_1143)} " f"overlays={len(OVERLAYS)}" ) return 0 diff --git a/CHANGELOG.md b/CHANGELOG.md index 0decb9ae..05f889ec 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -16,6 +16,34 @@ folds: `research/notes.md`. ## [Unreleased] +Hourly 1143 HIGH (`research/notes.md` §121 / composition items +369–384 / findings batch #104). Does **not** bump the 0.5.0 pin. +Uniqueness dumps live in +[`research/changelog-hourly.md`](research/changelog-hourly.md). +Do not reopen or amend PR #23–#43. Do not amend released 0.5.0 +(#42). Merged #43 owns §120. + +### Added + +- **Hourly 1143 HIGH (`notes.md` §121).** open recreation / semantic + lint sensor / third-party benches *theirs* / Laya serving densify / + catalogs. open recreation ≠ calibrated replica. semantic lint is a + sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs + *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ + semantic equivalence. permission ≠ confidence. catalog ≠ endorsement. + Evaluator: paired-CI / same-accuracy speedup / lint-cutoff still + soft. uniqueness_gate.py now checks 0843 + 0915 + jcr + 0922 + 0940 + + 0947 + 1049 + 1143. Composition items 369–384 / batch #104. + **HARD RULE:** do not reopen or amend PR #23–#43. Does **not** bump + 0.5.0. + +- **Recipe (class, not Jev-only).** Without Augustus: treat a Qwen3.5-4B + recreation as TypeSafe Jev, fire a 0.8 lint cutoff as a proof, or + quote 35x as semantic equivalence. With Augustus: open recreation ≠ + calibrated replica; semantic lint is a sensor not a proof; paired + bootstrap CIs stay *theirs*; serving substrate ≠ calibrated replica. + Same split for any Choice/Score/Noul-style head, not only hosted Jev. + Hourly 1049 HIGH (`research/notes.md` §120 / composition items 353–368 / findings batch #103). Does **not** bump the 0.5.0 pin. Uniqueness dumps live in diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 4a1a37b3..fa62c5c2 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -43,8 +43,8 @@ re-opened as "new." Before folding: - Read `research/notes.md` and the uniqueness fragments in `.agents/skills/augustus/SKILL.md` - Do not re-fold an already-landed section as a new beat -- Do not reopen or amend a merged fold PR (#23–#40) -- uniqueness_gate.py checks 0843 + 0915 + jcr + 0922 + 0940 + 0947 consecutive locks. +- Do not reopen or amend a merged fold PR (#23–#43) +- uniqueness_gate.py checks 0843 + 0915 + jcr + 0922 + 0940 + 0947 + 1049 + 1143 consecutive locks. - Hourly uniqueness dump: `research/changelog-hourly.md` (archive, not release notes) diff --git a/README.md b/README.md index c321b8fc..b8409fd1 100644 --- a/README.md +++ b/README.md @@ -142,6 +142,11 @@ User-provided 0920 jcr uniqueness lock: NiazMorshed2007/jcr MIT; site https://jc User-provided 0922 uniqueness lock: SemIf was formerly OpenJev; independent; not affiliated with Jev or TypeSafe; homepage openjev.com; default master; MIT; HEAD ca3ba65f1429; Tolerate float roundoff in MLX evidence verification; pushed 2026-09-19; live REST 2282★ / 140 forks; size 9177; README SHA 74ab7f7f; LICENSE SHA ca562883; interface pattern reproduction with open models; does not reproduce Jev undisclosed model/training; Direct option logits; 0 output tokens; shared-state parallel; MLX backend for Apple Silicon (`--backend mlx`); Qwen3.5-4B 3090 direct 1.023s vs AR JSON 5.332s (**5.21×**); argmax agree 18/21; systems comparison ≠ semantic equivalence; Parallel suffixes 20.03 dec/s on 777 decisions; Browser ladder Qwen3.5-4B authored BA 0.813, pert 0.766, TypeSafe subset agreement 0.845 vs Published Jev 0.883 (102 across 20 cases); Softmax over options ≠ calibrated Noul; typed output does not guarantee semantic correctness; wire/agreement ≠ replica of TypeSafe; SemIf ≠ kw2828/OpenJev playground ≠ zhihz/openjev ≠ apiplant/semif-rs port ≠ dddanielliu/semif-serve; rename is densify not a second census; JevBench 74.6 is §78 not this ladder; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#38; do not push onto open #39/#40; notes.md §117 User-provided 0940 uniqueness lock: Turn decision-shaped LLM prompts into proposed Jev primitives; This is a conversion assistant, not an automatic guarantee of equivalent behavior; The compiler uses deterministic heuristics, not an LLM or evaluation model; It understands a deliberately small set of common prompt patterns; Generated instructions and criteria must be reviewed before production use; Score ranges such as 0 to 1 are translated into ordered Jev criteria; Prompts requiring open-ended prose are not a fit; suitability strong/partial/not_a_fit; compatibility full/partial/none; Writing new text stays with an LLM; Review the generated Score rubric; Jev scores ordered criteria, not an arbitrary 0-to-1 range; Everything runs locally in the browser; There is no framework, database, account, API, or server-side prompt processing; The key is read from the process environment and is never stored or printed; connect-src 'none'; alexwestco/llm-to-jev ≠ altryne/jevify ≠ ryana/jevify ≠ fidecastro/jevify ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify; HEAD 234058ab372d; README SHA 43cd94fb; LICENSE SHA 5f334006; compiler SHA fdf235d0; 2★; MIT; JavaScript; size 29; Pages https://alexwestco.github.io/llm-to-jev/; invented_signal false; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35; notes.md §118 Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 +- Hourly 1143 HIGH (`research/notes.md` §121 / items 369–384 / batch #104). open recreation / semantic lint sensor / third-party benches *theirs* / Laya serving densify. uniqueness_gate 0843+0915+jcr+0922+0940+0947+1049+1143. Does not bump 0.5.0. Merged #43 owns §120. - Hourly 1049 HIGH (`research/notes.md` §120 / items 353–368 / batch #103). ggmlc GGUF serving / option-order / exact-p Minesweeper / adapters / catalogs. uniqueness_gate 0843+0915+jcr+0922+0940+0947+1049. Does not bump 0.5.0. Merged #41 owns §119. Merged #42 is the 0.5.0 release. - Hourly 0947 HIGH (`research/notes.md` §119 / items 337–352 / batch #102). jev-as-judge / OneForward candidate_mass / catalogs / replay / RLCD heads. uniqueness_gate 0843+0915+jcr+0922+0940+0947. Does not bump 0.5.0. Merged #42 is the 0.5.0 release. Merged #40 owns §118. Merged #39 is §114 hygiene. Merged #37 owns §117. Merged #38 owns §116. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/docs/ecosystem.md b/docs/ecosystem.md index bb7d10ed..67d01952 100644 --- a/docs/ecosystem.md +++ b/docs/ecosystem.md @@ -1141,3 +1141,7 @@ User-provided 0922 uniqueness lock: SemIf was formerly OpenJev; independent; not User-provided 0940 uniqueness lock: Turn decision-shaped LLM prompts into proposed Jev primitives; This is a conversion assistant, not an automatic guarantee of equivalent behavior; The compiler uses deterministic heuristics, not an LLM or evaluation model; It understands a deliberately small set of common prompt patterns; Generated instructions and criteria must be reviewed before production use; Score ranges such as 0 to 1 are translated into ordered Jev criteria; Prompts requiring open-ended prose are not a fit; suitability strong/partial/not_a_fit; compatibility full/partial/none; Writing new text stays with an LLM; Review the generated Score rubric; Jev scores ordered criteria, not an arbitrary 0-to-1 range; Everything runs locally in the browser; There is no framework, database, account, API, or server-side prompt processing; The key is read from the process environment and is never stored or printed; connect-src 'none'; alexwestco/llm-to-jev ≠ altryne/jevify ≠ ryana/jevify ≠ fidecastro/jevify ≠ Mintzs/jevify ≠ gulagala001/jevify ≠ uspraveen/Jevify; HEAD 234058ab372d; README SHA 43cd94fb; LICENSE SHA 5f334006; compiler SHA fdf235d0; 2★; MIT; JavaScript; size 29; Pages https://alexwestco.github.io/llm-to-jev/; invented_signal false; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35; notes.md §118 Hourly 0947 uniqueness lock: Fast and cheap agent evals. jev as judge.; 18,041 skills from the 200 most-starred repos; Not a security scanner; 最简 Jev 调用演示器; confidence 不是正确率; q93304989-bit/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab; 75% cheaper and 18% faster withdrawn; jev @0.15 100% recall 87% savings; 33Audits/jev-auto ≠ gargpratyush/jev-router; no Typesafe key, no PI_API_BASE, zero deps; tool-emitted Score/Noul ≠ calibrated Noul; semantic_compatibility: false; candidate_mass; Qwen3.5-2B ≠ Archer; Jev evaluates decisions; it cannot run a coding-agent session; Status: no model yet; S1LV3RJ1NX/openjev ≠ TheoLeeCJ/openjev; 28 accepted decisions; 3 targets; score 800; health 100; arcade game not a flight trainer; A successful live TypeSafe call has not been verified for v0.1.0; abhibansal60/tidy ≠ MANISH007700/tidy; No model, Jev included, predicted which channels its owner keeps; seed 1 selected on a held-out 400-item validation split; Brier 0.342 → 0.378; more accurate and more overconfident; Qwen3.5-4B ≠ Archer; static quants of kushalpatil/jevify-gemma4-26b-a4b; The labels were corrected, and one earlier result was retracted; zero of 23,869 eligible rows; Do not compare cost without checking task success; Exit 1 is not a proof; kisshan13/typesafe-ai-go ≠ Nibir1/typesafe-go ≠ official; 38 tests that cannot fail in a 356-model warehouse; if a parser can answer it, Jev is never asked; 359 of them; Games & Simulation 82; Education & Learning 1; Ratings are heuristics; syedabbasshaheer-art/jev-atlas ≠ ZeroX-01/jev-atlas ≠ Zaious/jev-capability-atlas ≠ gorock007/jev-atlas; anandi1989/awesome-jev-usecases ≠ whyashthakker/awesome-jev-use-cases ≠ walidboulanouar/awesome-jev-use-cases ≠ vamsikrishna2421/jev-usecases; Every headline result above is self-reported; Archer Hume 84.6% MMLU-Pro is a third-party probe not landed Archer; catalog ≠ endorsement; judge ≠ actuator; softmax over A–H ≠ Noul; SemIf 2270★; jevlike 1054★; TypeLLM/TypeLLM 16★; AnotiaWang 98★; yibie/awesome-jev 538★; Laya likes 889; tracker likes 68 lastModified UNCHANGED; Blackwood likes 2 gated manual; Archer still promised_not_landed; Hub archerhume/4rcherhume HTTP 401; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#35/#36/#37/#38/#40; do not push onto open #39; notes.md §119 Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + + +**Hourly 1143 HIGH (`notes.md` §121).** open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. cutoff 0.8 still soft. paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. serving substrate ≠ calibrated replica. catalog ≠ endorsement. permission ≠ confidence. Do not reopen or amend PR #23–#43. Does not bump 0.5.0. Skip Archer. `invented_signal: false`. +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/research/archive/findings.md b/research/archive/findings.md index c1c249ae..b156b41a 100644 --- a/research/archive/findings.md +++ b/research/archive/findings.md @@ -1,6 +1,44 @@ # Deep-read findings (evidence for research/notes.md) +## Batch #104 (2026-09-20 ~11:43 Boise / ~17:43 UTC) - hourly 1143 HIGH + +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 + +Note: `research/notes.md` §121. Docs + evaluator, fresh PR off latest +`main` (`a61372f` / merged #43 1049). Merged #43 owns §120 / 353–368 / +#103. Merged #41 owns §119 / 337–352 / #102. Merged #42 is the 0.5.0 +release. This fold stays §121 / items 369–384 / batch #104. +**HARD RULE:** do not reopen or amend PR #23–#43. +Quote READMEs. Soft Noul ≠ hard safety. Augustus owns +placement. `invented_signal: false`. + +- **Open recreation PRIMARY.** kshetrajna12/reflex 85★. It is an open re-creation of Jev. Qwen3.5-4B. less calibrated. + pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*. DavidHatley independent not Jev/RLCD. + open recreation ≠ calibrated replica. Qwen3.5-4B ≠ Archer. +- **Semantic lint sensor.** perch 164★. oxlint-plugin-jev cutoff 0.8 still soft. patdown fuzzy linter. + nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep. + PanAchy/jevvy ≠ Atominac/jevvy. semantic lint is a sensor not a proof. +- **Third-party benches *theirs*.** emretheus +0.82 / +7.62 with paired bootstrap CIs. + BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*. + frozen cascade 430/500 vs GPT-5 432/500. This is not demonstrated equal-quality savings. + 24 invented tickets. Routing errors caught by the gate 0 of 3. sample too small to establish calibration. + This is not TypeSafe Jev. No real API requests were made. No orders, no advice. + Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. +- **Laya serving densify.** KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL. wire-compat ≠ replica. + gqgs/laya-onnx densify 496.8 MiB. tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx. + serving substrate ≠ calibrated replica. +- **Catalogs / harnesses.** BeatAPI All 125 projects. catalog ≠ endorsement. + Pasblinn/jev-lab Independent project. Not affiliated with TypeSafe. + Kevthetech143/super-jev densify experimental V0.2.0. permission ≠ confidence. + allay-team/openjev 2022 Mineflayer Jevalent collision. + +Pulse: Archer still NOT landed. Hub archerhume/4rcherhume HTTP **401**. +perch **164★**; reflex **85★**; oxlint-plugin-jev **36★**; SmartMoney-Cub **25★**. +`invented_signal: false`. + + + ## Batch #103 (2026-09-20 ~10:49 Boise / ~16:49 UTC) — hourly 1049 HIGH Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 diff --git a/research/archive/hourly/2026-09-20T17/novel_high_this_run.json b/research/archive/hourly/2026-09-20T17/novel_high_this_run.json new file mode 100644 index 00000000..fe47fc18 --- /dev/null +++ b/research/archive/hourly/2026-09-20T17/novel_high_this_run.json @@ -0,0 +1,1127 @@ +[ + { + "id": "lakeday-org/perch", + "source": "github", + "stars": 164, + "desc": "Semantic code linting with Jev", + "created": "2026-09-16T04:55:38Z", + "pushed": "2026-09-20T17:36:34Z", + "url": "https://github.com/lakeday-org/perch" + }, + { + "id": "kshetrajna12/reflex", + "source": "github", + "stars": 85, + "desc": "A small open decision model: state + typed questions -> calibrated probabilities. A Jev / System One re-creation on Qwen3.5.", + "created": "2026-09-17T23:03:08Z", + "pushed": "2026-09-20T17:28:58Z", + "url": "https://github.com/kshetrajna12/reflex" + }, + { + "id": "wobsoriano/oxlint-plugin-jev", + "source": "github", + "stars": 36, + "desc": "", + "created": "2026-09-19T05:44:11Z", + "pushed": "2026-09-20T17:40:26Z", + "url": "https://github.com/wobsoriano/oxlint-plugin-jev" + }, + { + "id": "myc0576/SmartMoney-Cub", + "source": "github", + "stars": 25, + "desc": "Read-only trading journal and review harness: Jev typed judgments, agent integration, and a reproducible finance benchmark. No orders, no advice.", + "created": "2026-06-25T05:56:49Z", + "pushed": "2026-09-20T15:37:19Z", + "url": "https://github.com/myc0576/SmartMoney-Cub" + }, + { + "id": "nassim-arifette/jevgrep", + "source": "github", + "stars": 12, + "desc": "Jev-powered semantic code search for coding agents \u2014 find behavior across repositories via CLI or MCP, with exact source excerpts and line numbers.", + "created": "2026-09-20T00:27:29Z", + "pushed": "2026-09-20T17:01:42Z", + "url": "https://github.com/nassim-arifette/jevgrep" + }, + { + "id": "tyler-dot-earth/patdown", + "source": "github", + "stars": 10, + "desc": "typesafe's jev as a \"fuzzy linter\". give your code an ocular patdown.", + "created": "2026-09-18T16:00:08Z", + "pushed": "2026-09-20T17:43:08Z", + "url": "https://github.com/tyler-dot-earth/patdown" + }, + { + "id": "PanAchy/jevvy", + "source": "github", + "stars": 9, + "desc": "Jev-powered plugins for coding agents", + "created": "2026-09-18T21:41:30Z", + "pushed": "2026-09-20T17:16:50Z", + "url": "https://github.com/PanAchy/jevvy" + }, + { + "id": "Kevthetech143/super-jev", + "source": "github", + "stars": 6, + "desc": "A small, extensible decision-to-action harness for TypeSafe Jev", + "created": "2026-09-17T03:56:36Z", + "pushed": "2026-09-20T17:42:11Z", + "url": "https://github.com/Kevthetech143/super-jev" + }, + { + "id": "BeatAPI/awesome-jev", + "source": "github", + "stars": 5, + "desc": "We only curate source-reviewed JEV-related projects with 100+ GitHub stars \u2014 integrations, tools, open models, and experiments. Live gallery: beatapi.io/awesome-jev", + "created": "2026-09-20T04:52:51Z", + "pushed": "2026-09-20T16:11:39Z", + "url": "https://github.com/BeatAPI/awesome-jev" + }, + { + "id": "KonghaYao/laya-jev", + "source": "github", + "stars": 1, + "desc": "", + "created": "2026-09-19T13:30:46Z", + "pushed": "2026-09-19T14:11:24Z", + "url": "https://github.com/KonghaYao/laya-jev" + }, + { + "id": "Pasblinn/jev-lab", + "source": "github", + "stars": 1, + "desc": "Open lab: Jev (TypeSafe System One) routing in front of Claude Code - measured bugs, patch, and a hard fallback with alerts", + "created": "2026-09-20T17:26:55Z", + "pushed": "2026-09-20T17:40:55Z", + "url": "https://github.com/Pasblinn/jev-lab" + }, + { + "id": "gqgs/laya-onnx", + "source": "github", + "stars": 1, + "desc": "Laya ONNX for browsers", + "created": "2026-09-19T06:24:24Z", + "pushed": "2026-09-19T06:24:51Z", + "url": "https://github.com/gqgs/laya-onnx" + }, + { + "id": "hfspace:pngwn/system-one-v2-retrain-trackio", + "source": "hf_space", + "stars": 1, + "desc": "", + "created": "2026-09-16T23:50:02.000Z", + "pushed": "2026-09-16T23:50:04.000Z", + "url": "https://huggingface.co/spaces/pngwn/system-one-v2-retrain-trackio" + }, + { + "id": "sah1l/jev-battleship", + "source": "github", + "stars": 1, + "desc": "Battleship against Jev, a model that answers in probabilities instead of text. Web game plus a CLI arena that plays it against general-purpose LLMs on identical fleets.", + "created": "2026-09-20T08:27:07Z", + "pushed": "2026-09-20T17:17:13Z", + "url": "https://github.com/sah1l/jev-battleship" + }, + { + "id": "4OH4/jev-compare", + "source": "github", + "stars": 0, + "desc": "Evaluating Jev for zero-shot evaluation", + "created": "2026-09-20T17:11:43Z", + "pushed": "2026-09-20T17:41:14Z", + "url": "https://github.com/4OH4/jev-compare" + }, + { + "id": "4esv/jev-joust", + "source": "github", + "stars": 0, + "desc": "TypeSafe Jev vs Jev in NES Joust, bring your own ROM", + "created": "2026-09-18T19:37:50Z", + "pushed": "2026-09-20T17:20:50Z", + "url": "https://github.com/4esv/jev-joust" + }, + { + "id": "99hansling/awesome-jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:16:46Z", + "pushed": "2026-09-20T16:43:54Z", + "url": "https://github.com/99hansling/awesome-jev" + }, + { + "id": "99hansling/bb-jev-browser", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:44:30Z", + "pushed": "2026-09-20T16:44:35Z", + "url": "https://github.com/99hansling/bb-jev-browser" + }, + { + "id": "AndreRochaDev/innovation-poc-jev", + "source": "github", + "stars": 0, + "desc": "Poc with jev", + "created": "2026-09-20T17:28:47Z", + "pushed": "2026-09-20T17:45:01Z", + "url": "https://github.com/AndreRochaDev/innovation-poc-jev" + }, + { + "id": "AparnaDhinakaran/jev-as-judge", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:39:25Z", + "pushed": "2026-09-20T17:40:45Z", + "url": "https://github.com/AparnaDhinakaran/jev-as-judge" + }, + { + "id": "ArronHC/windows2text", + "source": "github", + "stars": 0, + "desc": "Windows UI Automation screen-to-text, MCP server, and Jev-gated computer use", + "created": "2026-09-20T16:52:15Z", + "pushed": "2026-09-20T17:11:21Z", + "url": "https://github.com/ArronHC/windows2text" + }, + { + "id": "Atominac/jevvy", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:44:58Z", + "pushed": "2026-09-20T17:44:59Z", + "url": "https://github.com/Atominac/jevvy" + }, + { + "id": "CodeIA-Academy/jev-mcp", + "source": "github", + "stars": 0, + "desc": "MCP local que expone Jev (TypeSafe) como herramienta para Claude Code, Codex, Hermes y cualquier agente: ask_jev y list_jev_models, sin dependencias", + "created": "2026-09-20T17:24:10Z", + "pushed": "2026-09-20T17:34:43Z", + "url": "https://github.com/CodeIA-Academy/jev-mcp" + }, + { + "id": "Coditan/system-one-axi", + "source": "github", + "stars": 0, + "desc": "Unofficial AXI-style CLI for TypeSafe System One models, including Jev.", + "created": "2026-09-20T17:36:39Z", + "pushed": "2026-09-20T17:39:49Z", + "url": "https://github.com/Coditan/system-one-axi" + }, + { + "id": "Cyberesia/jev-traps", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T07:37:03Z", + "pushed": "2026-09-20T16:43:39Z", + "url": "https://github.com/Cyberesia/jev-traps" + }, + { + "id": "Hinstein/jevhub", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:47:32Z", + "pushed": "2026-09-20T17:41:16Z", + "url": "https://github.com/Hinstein/jevhub" + }, + { + "id": "Hugo-DDT/JevTape", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T14:28:24Z", + "pushed": "2026-09-20T17:32:29Z", + "url": "https://github.com/Hugo-DDT/JevTape" + }, + { + "id": "Je1zzz/Jev-Codex-accelerator", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T10:46:01Z", + "pushed": "2026-09-20T17:30:16Z", + "url": "https://github.com/Je1zzz/Jev-Codex-accelerator" + }, + { + "id": "JohnsonRan/pi-jev", + "source": "github", + "stars": 0, + "desc": "Claude-style auto-mode classifier for Pi, powered by TypeSafe Jev", + "created": "2026-09-20T14:51:02Z", + "pushed": "2026-09-20T17:28:24Z", + "url": "https://github.com/JohnsonRan/pi-jev" + }, + { + "id": "Kaos599/jev-writer", + "source": "github", + "stars": 0, + "desc": "Find out which qualities of your writing actually predict engagement. Analyses (never generates) your published posts with Jev, TypeSafe AI's System One model, using pre-registered rubrics and statist", + "created": "2026-09-20T17:32:54Z", + "pushed": "2026-09-20T17:39:01Z", + "url": "https://github.com/Kaos599/jev-writer" + }, + { + "id": "KaranSiddhu/jev-flappy-bird", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:18:14Z", + "pushed": "2026-09-20T17:21:16Z", + "url": "https://github.com/KaranSiddhu/jev-flappy-bird" + }, + { + "id": "Lancer59/SystemOneModels", + "source": "github", + "stars": 0, + "desc": "Testing out System One Models", + "created": "2026-09-20T15:45:21Z", + "pushed": "2026-09-20T16:51:01Z", + "url": "https://github.com/Lancer59/SystemOneModels" + }, + { + "id": "LouisUltra/jev-deep-dive", + "source": "github", + "stars": 0, + "desc": "An evidence-graded deep dive into Jev (TypeSafe AI's System One model): interface, verification, reproduction comparison, and field guide. Bilingual (EN/\u4e2d\u6587), with runnable probe tools.", + "created": "2026-09-20T17:19:10Z", + "pushed": "2026-09-20T17:23:34Z", + "url": "https://github.com/LouisUltra/jev-deep-dive" + }, + { + "id": "Mawfyy/jevflow", + "source": "github", + "stars": 0, + "desc": "Probabilistic AI decisions as composable backend programming primitives.", + "created": "2026-09-20T17:43:53Z", + "pushed": "2026-09-20T17:44:01Z", + "url": "https://github.com/Mawfyy/jevflow" + }, + { + "id": "MohtashamMurshid/jev-speed-test", + "source": "github", + "stars": 0, + "desc": "Reproducible Jev vs fast LLM experiment: BANKING77 data, raw responses, confidence evaluation, and analysis", + "created": "2026-09-20T17:43:40Z", + "pushed": "2026-09-20T17:44:10Z", + "url": "https://github.com/MohtashamMurshid/jev-speed-test" + }, + { + "id": "OsirianLegacy/JevSimulation", + "source": "github", + "stars": 0, + "desc": "Building a Simulated world using Typesafe Ai's Jev to run entity choices, C++ 20, JSON for Jev Interactions, Clay for TUI, and Raylib 6 for Rendering", + "created": "2026-09-20T17:02:59Z", + "pushed": "2026-09-20T17:04:02Z", + "url": "https://github.com/OsirianLegacy/JevSimulation" + }, + { + "id": "PavelLizunov/jev-minecraft-factory-sorter", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:48:13Z", + "pushed": "2026-09-20T16:48:22Z", + "url": "https://github.com/PavelLizunov/jev-minecraft-factory-sorter" + }, + { + "id": "PavelLizunov/jev-sentinel", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T12:35:08Z", + "pushed": "2026-09-20T16:59:51Z", + "url": "https://github.com/PavelLizunov/jev-sentinel" + }, + { + "id": "Pidbid/laya-deploy", + "source": "github", + "stars": 0, + "desc": "Deploy the Laya System 1 decision engine as a local judgment backend for omp (oh-my-pi): TypeSafe-compatible /v1/systemone bridge, MCP tools, systemd unit, installer, verification probes and the pitfa", + "created": "2026-09-20T17:10:40Z", + "pushed": "2026-09-20T17:17:02Z", + "url": "https://github.com/Pidbid/laya-deploy" + }, + { + "id": "Sahilll15/jobfit", + "source": "github", + "stars": 0, + "desc": "Check your CV against a job posting, one requirement at a time. Built on TypeSafe Jev.", + "created": "2026-09-20T15:48:26Z", + "pushed": "2026-09-20T17:40:57Z", + "url": "https://github.com/Sahilll15/jobfit" + }, + { + "id": "Santiago-j-s/jev-proxy", + "source": "github", + "stars": 0, + "desc": "Local observability proxy and playground for TypeSafe Jev", + "created": "2026-09-20T17:29:43Z", + "pushed": "2026-09-20T17:29:48Z", + "url": "https://github.com/Santiago-j-s/jev-proxy" + }, + { + "id": "Sskift/jev-sts2-agent", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T13:18:49Z", + "pushed": "2026-09-20T17:38:02Z", + "url": "https://github.com/Sskift/jev-sts2-agent" + }, + { + "id": "SwiftFaze/Jev-Studio", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T15:15:22Z", + "pushed": "2026-09-20T17:42:27Z", + "url": "https://github.com/SwiftFaze/Jev-Studio" + }, + { + "id": "Todmy/jevqa", + "source": "github", + "stars": 0, + "desc": "Jev-guided monkey tester: $0.35 and 5 minutes to find missing features, validation gaps and dead controls before QA", + "created": "2026-09-20T10:27:40Z", + "pushed": "2026-09-20T17:40:29Z", + "url": "https://github.com/Todmy/jevqa" + }, + { + "id": "TonyP-MR/jev-curation-engine", + "source": "github", + "stars": 0, + "desc": "Read-only TypeSafe Jev feasibility test rig for comparing structured Curation Engine classification decisions with existing LLM audit results.", + "created": "2026-09-18T12:14:14Z", + "pushed": "2026-09-20T17:42:49Z", + "url": "https://github.com/TonyP-MR/jev-curation-engine" + }, + { + "id": "Vazool/jev-escape-room", + "source": "github", + "stars": 0, + "desc": "An experiment using Jev System One as a probabilistic semantic layer within a deterministic Python game world.", + "created": "2026-09-20T17:17:56Z", + "pushed": "2026-09-20T17:19:42Z", + "url": "https://github.com/Vazool/jev-escape-room" + }, + { + "id": "VirtualMachinist/omapi-overlay", + "source": "github", + "stars": 0, + "desc": "Stock omp pin + overlay flake (skills, thin omapi wrap, Jev/cursor-agent gates). Not an omp source fork.", + "created": "2026-09-20T13:59:27Z", + "pushed": "2026-09-20T17:44:43Z", + "url": "https://github.com/VirtualMachinist/omapi-overlay" + }, + { + "id": "Vishnurr2k01/awesome-jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:18:11Z", + "pushed": "2026-09-20T16:52:28Z", + "url": "https://github.com/Vishnurr2k01/awesome-jev" + }, + { + "id": "a3ro-dev/jev-clone", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:53:07Z", + "pushed": "2026-09-20T17:25:30Z", + "url": "https://github.com/a3ro-dev/jev-clone" + }, + { + "id": "abdelrahmanmagdii/jevci", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T19:50:38Z", + "pushed": "2026-09-20T17:09:46Z", + "url": "https://github.com/abdelrahmanmagdii/jevci" + }, + { + "id": "abh2050/jev-test-confidence-gate", + "source": "github", + "stars": 0, + "desc": "One graph, one policy, two judgment engines. This experiment tests TypeSafe\u2019s published claims about Jev in a support-triage workflow, with a confidence gate that sends uncertain cases to a human.", + "created": "2026-09-20T17:01:34Z", + "pushed": "2026-09-20T17:32:11Z", + "url": "https://github.com/abh2050/jev-test-confidence-gate" + }, + { + "id": "agentik-os/jev-radar", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-18T19:01:25Z", + "pushed": "2026-09-20T17:38:58Z", + "url": "https://github.com/agentik-os/jev-radar" + }, + { + "id": "ahm-ge18/jev_replica", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:38:49Z", + "pushed": "2026-09-20T17:41:49Z", + "url": "https://github.com/ahm-ge18/jev_replica" + }, + { + "id": "allay-team/openjev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2022-07-11T19:11:20Z", + "pushed": "2022-07-12T19:20:08Z", + "url": "https://github.com/allay-team/openjev" + }, + { + "id": "api-evangelist/typesafe-ai", + "source": "github", + "stars": 0, + "desc": "TypeSafe AI is a San Francisco AI lab building System One models \u2014 a class of model trained to return typed, calibrated decisions for software instead of generated text. Its flagship and first System ", + "created": "2026-09-19T15:09:23Z", + "pushed": "2026-09-20T16:51:20Z", + "url": "https://github.com/api-evangelist/typesafe-ai" + }, + { + "id": "apurv101/jev-browser-use", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:47:26Z", + "pushed": "2026-09-20T16:47:26Z", + "url": "https://github.com/apurv101/jev-browser-use" + }, + { + "id": "aryamangoenka/traceassert", + "source": "github", + "stars": 0, + "desc": "Semantic assertions for coding agents \u2014 test what your agent actually did, powered by Jev", + "created": "2026-09-20T17:07:42Z", + "pushed": "2026-09-20T17:07:47Z", + "url": "https://github.com/aryamangoenka/traceassert" + }, + { + "id": "ayshrosine/Jev_Typesafe.Ai", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:45:05Z", + "pushed": "2026-09-20T17:45:06Z", + "url": "https://github.com/ayshrosine/Jev_Typesafe.Ai" + }, + { + "id": "ba8eman/trying-jev", + "source": "github", + "stars": 0, + "desc": "a small repo to try out jev and get a feel for it", + "created": "2026-09-20T17:04:11Z", + "pushed": "2026-09-20T17:21:15Z", + "url": "https://github.com/ba8eman/trying-jev" + }, + { + "id": "bahramzada/jev-canvas", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:15:45Z", + "pushed": "2026-09-20T17:36:45Z", + "url": "https://github.com/bahramzada/jev-canvas" + }, + { + "id": "bavovna/jev-like-srv", + "source": "github", + "stars": 0, + "desc": "jev test backend compatible with https://docs.typesafe.ai/api", + "created": "2026-09-20T16:56:28Z", + "pushed": "2026-09-20T17:37:20Z", + "url": "https://github.com/bavovna/jev-like-srv" + }, + { + "id": "blck-snwmn/playground-jev", + "source": "github", + "stars": 0, + "desc": "Small apps for experimenting with Jev.", + "created": "2026-09-19T05:56:15Z", + "pushed": "2026-09-20T17:29:57Z", + "url": "https://github.com/blck-snwmn/playground-jev" + }, + { + "id": "bogusweb/ship-game-with-jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T21:38:31Z", + "pushed": "2026-09-20T17:03:02Z", + "url": "https://github.com/bogusweb/ship-game-with-jev" + }, + { + "id": "chaseai-yt/jev-audit", + "source": "github", + "stars": 0, + "desc": "Find where TypeSafe Jev fits your real workflows. An evidence-based audit skill for Claude Code, Codex, and compatible agents.", + "created": "2026-09-20T17:18:11Z", + "pushed": "2026-09-20T17:18:14Z", + "url": "https://github.com/chaseai-yt/jev-audit" + }, + { + "id": "chr-kelly/jev-cookbook", + "source": "github", + "stars": 0, + "desc": "Runnable question sets for TypeSafe's Jev, an eval harness with measured CLINC150 results, and a linter for the request shapes the API silently mis-reads.", + "created": "2026-09-20T17:21:07Z", + "pushed": "2026-09-20T17:21:13Z", + "url": "https://github.com/chr-kelly/jev-cookbook" + }, + { + "id": "cleiton1231/Jev-mine", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:07:48Z", + "pushed": "2026-09-20T17:40:50Z", + "url": "https://github.com/cleiton1231/Jev-mine" + }, + { + "id": "crman/jev-ai-output-judge", + "source": "github", + "stars": 0, + "desc": "Benchmarking Jev and LLMs for RAG answer evaluation and hallucination detection", + "created": "2026-09-20T16:58:55Z", + "pushed": "2026-09-20T17:33:20Z", + "url": "https://github.com/crman/jev-ai-output-judge" + }, + { + "id": "curajorge/judgment-map", + "source": "github", + "stars": 0, + "desc": "A typed, reviewable AI judgment map for contract analysis with TypeSafe.", + "created": "2026-09-20T17:39:07Z", + "pushed": "2026-09-20T17:40:03Z", + "url": "https://github.com/curajorge/judgment-map" + }, + { + "id": "davidmasp/url-jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:52:12Z", + "pushed": "2026-09-20T16:54:46Z", + "url": "https://github.com/davidmasp/url-jev" + }, + { + "id": "ddwolfer/jev_play_guildrun", + "source": "github", + "stars": 0, + "desc": "jev_play_guildrun", + "created": "2026-09-20T15:33:43Z", + "pushed": "2026-09-20T17:20:46Z", + "url": "https://github.com/ddwolfer/jev_play_guildrun" + }, + { + "id": "emretheus/jev-rag-benchmark", + "source": "github", + "stars": 0, + "desc": "Free English RAG benchmark for TypeSafe Jev 1.13 \u2014 frozen candidate pools, calibration, paired bootstrap CIs, $0 runs. Compared with OpenJev and NVIDIA cross-encoders.", + "created": "2026-09-20T17:07:21Z", + "pushed": "2026-09-20T17:37:47Z", + "url": "https://github.com/emretheus/jev-rag-benchmark" + }, + { + "id": "expanso-io/demo-expanso-jev", + "source": "github", + "stars": 0, + "desc": "Expanso x Jev demos: live edge-pipeline dashboard, pipeline configs, and flow UI.", + "created": "2026-09-19T05:58:07Z", + "pushed": "2026-09-20T17:26:39Z", + "url": "https://github.com/expanso-io/demo-expanso-jev" + }, + { + "id": "glukicov/laya_router", + "source": "github", + "stars": 0, + "desc": "Model routing is a System 1 job: a 421M local decision engine (Laya) against a GPT-5 nano reasoning router. Same accuracy, 35x faster, free.", + "created": "2026-09-20T17:20:59Z", + "pushed": "2026-09-20T17:44:33Z", + "url": "https://github.com/glukicov/laya_router" + }, + { + "id": "gregjonesio/jev-nfl", + "source": "github", + "stars": 0, + "desc": "Before every NFL snap, a decision-only AI model (TypeSafe Jev) calls run or pass and go/punt/kick on fourth down, graded live against the coach.", + "created": "2026-09-20T04:19:01Z", + "pushed": "2026-09-20T17:19:19Z", + "url": "https://github.com/gregjonesio/jev-nfl" + }, + { + "id": "hf:mradermacher/jevify-gemma4-e4b-GGUF", + "source": "hf_model", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:04:35.000Z", + "pushed": "2026-09-20T17:41:16.000Z", + "url": "https://huggingface.co/mradermacher/jevify-gemma4-e4b-GGUF" + }, + { + "id": "hfdataset:DavidHatley/system-one-mini-data", + "source": "hf_dataset", + "stars": 0, + "desc": "", + "created": "2026-09-16T17:02:50.000Z", + "pushed": "2026-09-16T17:30:11.000Z", + "url": "https://huggingface.co/datasets/DavidHatley/system-one-mini-data" + }, + { + "id": "hfdataset:emretheus/jev-rag-benchmark", + "source": "hf_dataset", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:34:41.000Z", + "pushed": "2026-09-20T17:35:01.000Z", + "url": "https://huggingface.co/datasets/emretheus/jev-rag-benchmark" + }, + { + "id": "hfdataset:pngwn/system-one-decisions", + "source": "hf_dataset", + "stars": 0, + "desc": "", + "created": "2026-09-16T13:58:47.000Z", + "pushed": "2026-09-16T14:07:53.000Z", + "url": "https://huggingface.co/datasets/pngwn/system-one-decisions" + }, + { + "id": "hfspace:emretheus/jev-rag-benchmark-leaderboard", + "source": "hf_space", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:35:10.000Z", + "pushed": "2026-09-20T17:35:12.000Z", + "url": "https://huggingface.co/spaces/emretheus/jev-rag-benchmark-leaderboard" + }, + { + "id": "hfspace:pngwn/system-one-demo", + "source": "hf_space", + "stars": 0, + "desc": "", + "created": "2026-09-16T19:03:01.000Z", + "pushed": "2026-09-16T19:48:13.000Z", + "url": "https://huggingface.co/spaces/pngwn/system-one-demo" + }, + { + "id": "hfspace:pngwn/system-one-qwen3-5-4b-scorer-trackio", + "source": "hf_space", + "stars": 0, + "desc": "", + "created": "2026-09-16T14:21:41.000Z", + "pushed": "2026-09-16T14:21:43.000Z", + "url": "https://huggingface.co/spaces/pngwn/system-one-qwen3-5-4b-scorer-trackio" + }, + { + "id": "hobbs/jev-turn-analysis", + "source": "github", + "stars": 0, + "desc": "Jev Turn Analysis is a Rust CLI for analyzing completed Claude Code and Codex sessions.", + "created": "2026-09-20T17:12:32Z", + "pushed": "2026-09-20T17:40:20Z", + "url": "https://github.com/hobbs/jev-turn-analysis" + }, + { + "id": "i-priyanshuverma/laravel-jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:26:30Z", + "pushed": "2026-09-20T16:43:20Z", + "url": "https://github.com/i-priyanshuverma/laravel-jev" + }, + { + "id": "jammaru/jev-affected", + "source": "github", + "stars": 0, + "desc": "Semantic task routing for software development, powered by Jev.", + "created": "2026-09-20T17:08:51Z", + "pushed": "2026-09-20T17:25:21Z", + "url": "https://github.com/jammaru/jev-affected" + }, + { + "id": "jaysonsantos/sudoku-jev", + "source": "github", + "stars": 0, + "desc": "Sudoku game played by the TypeSafe Jev decision model through OpenRouter", + "created": "2026-09-19T09:12:20Z", + "pushed": "2026-09-20T17:10:40Z", + "url": "https://github.com/jaysonsantos/sudoku-jev" + }, + { + "id": "jdubpark/jevcode", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T22:05:37Z", + "pushed": "2026-09-20T17:30:03Z", + "url": "https://github.com/jdubpark/jevcode" + }, + { + "id": "jpollard-cs/jev-observatory", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T16:13:43Z", + "pushed": "2026-09-20T17:20:44Z", + "url": "https://github.com/jpollard-cs/jev-observatory" + }, + { + "id": "just-be-dev/jev-sat", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T13:54:20Z", + "pushed": "2026-09-20T17:11:02Z", + "url": "https://github.com/just-be-dev/jev-sat" + }, + { + "id": "jyothepro/jev-games", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:38:04Z", + "pushed": "2026-09-20T17:44:52Z", + "url": "https://github.com/jyothepro/jev-games" + }, + { + "id": "kaiyes/fast-jev-compaction-laya", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T14:39:01Z", + "pushed": "2026-09-20T14:50:47Z", + "url": "https://github.com/kaiyes/fast-jev-compaction-laya" + }, + { + "id": "kiroclawai/system-one-blueprint", + "source": "github", + "stars": 0, + "desc": "Jev-class System One model blueprint \u2014 built from open components (50M encoder + parallel decision heads, Needle3 distillation, RLCD path)", + "created": "2026-09-20T17:15:21Z", + "pushed": "2026-09-20T17:15:49Z", + "url": "https://github.com/kiroclawai/system-one-blueprint" + }, + { + "id": "klukacin/JEV_Claude_plugin", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:43:38Z", + "pushed": "2026-09-20T17:45:08Z", + "url": "https://github.com/klukacin/JEV_Claude_plugin" + }, + { + "id": "kuromoka/garmin-jev-buddy", + "source": "github", + "stars": 0, + "desc": "A playful Garmin Connect IQ character that reacts to running data using Jev, with Japanese and English display", + "created": "2026-09-20T16:39:04Z", + "pushed": "2026-09-20T17:38:37Z", + "url": "https://github.com/kuromoka/garmin-jev-buddy" + }, + { + "id": "leepokai/jev-adrank", + "source": "github", + "stars": 0, + "desc": "Real-time ad ranking and creative review on a typed evaluation model \u2014 one call prices a whole auction, no trained CTR model, no logged clicks.", + "created": "2026-09-20T17:05:17Z", + "pushed": "2026-09-20T17:28:12Z", + "url": "https://github.com/leepokai/jev-adrank" + }, + { + "id": "marcosnovaesq/laya-playground", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T00:58:17Z", + "pushed": "2026-09-19T01:07:40Z", + "url": "https://github.com/marcosnovaesq/laya-playground" + }, + { + "id": "michelbrigante46-art/Twitter-keyword-shield", + "source": "github", + "stars": 0, + "desc": "A high-performance content filtering userscript for X (Twitter), combining instant local rules with Jev (TypeSafe System One) semantic decisions to block spam and hidden promotions with sub-100ms late", + "created": "2026-09-20T17:33:37Z", + "pushed": "2026-09-20T17:38:42Z", + "url": "https://github.com/michelbrigante46-art/Twitter-keyword-shield" + }, + { + "id": "mmastrac/jevenator2", + "source": "github", + "stars": 0, + "desc": "Jevenator 2: Judgement Day - region-scan localisation on a diffusion LLM", + "created": "2026-09-20T16:29:22Z", + "pushed": "2026-09-20T17:14:58Z", + "url": "https://github.com/mmastrac/jevenator2" + }, + { + "id": "morre95/Laya-GuardRails-Harness", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T15:22:53Z", + "pushed": "2026-09-20T16:30:54Z", + "url": "https://github.com/morre95/Laya-GuardRails-Harness" + }, + { + "id": "muratcanberber/JEV-TheFishGame", + "source": "github", + "stars": 0, + "desc": "\ud83d\udc20 A multiplayer fish game where every AI decision is a TypeSafe Jev (System One) call \u2014 flee, hunt, roam, with live confidence bars. Node + WebSocket + Three.js.", + "created": "2026-09-20T00:16:07Z", + "pushed": "2026-09-20T17:44:26Z", + "url": "https://github.com/muratcanberber/JEV-TheFishGame" + }, + { + "id": "nayan-kunwar/jev-customer-support-triage", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T08:02:22Z", + "pushed": "2026-09-20T17:17:48Z", + "url": "https://github.com/nayan-kunwar/jev-customer-support-triage" + }, + { + "id": "ngallodev-software/agent-workflow-typesafe-ai", + "source": "github", + "stars": 0, + "desc": "Optional TypeSafe AI/Jev advisory decision evidence for Agent-Workflow, without transferring lifecycle authority from deterministic orchestration.", + "created": "2026-09-18T20:12:54Z", + "pushed": "2026-09-20T17:33:23Z", + "url": "https://github.com/ngallodev-software/agent-workflow-typesafe-ai" + }, + { + "id": "nsillik/jevvin-off", + "source": "github", + "stars": 0, + "desc": "Prototyping against TypeSafe's Jev System One API: a one-ticket quickstart and a Bluesky Jetstream firehose demo.", + "created": "2026-09-17T16:12:56Z", + "pushed": "2026-09-20T17:20:46Z", + "url": "https://github.com/nsillik/jevvin-off" + }, + { + "id": "nvkudva/laya-web", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T15:53:02Z", + "pushed": "2026-09-20T17:19:44Z", + "url": "https://github.com/nvkudva/laya-web" + }, + { + "id": "omerfeyzioglu/JevOps", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T21:59:11Z", + "pushed": "2026-09-20T17:42:02Z", + "url": "https://github.com/omerfeyzioglu/JevOps" + }, + { + "id": "piyush-infocusp/openjev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-17T12:36:44Z", + "pushed": "2026-09-17T12:36:45Z", + "url": "https://github.com/piyush-infocusp/openjev" + }, + { + "id": "rahulthakore16/n8n-nodes-jev", + "source": "github", + "stars": 0, + "desc": "Jev by TypeSafe AI for n8n: typed decisions, probabilities, and confidence-aware workflows", + "created": "2026-09-20T17:03:41Z", + "pushed": "2026-09-20T17:40:05Z", + "url": "https://github.com/rahulthakore16/n8n-nodes-jev" + }, + { + "id": "remotehostai/jg", + "source": "github", + "stars": 0, + "desc": "Proprietary jevgrep CLI preview: semantic code search for coding agents. Hosted search is not yet available.", + "created": "2026-09-20T03:29:39Z", + "pushed": "2026-09-20T17:35:45Z", + "url": "https://github.com/remotehostai/jg" + }, + { + "id": "renatobardi/jev-o-matic", + "source": "github", + "stars": 0, + "desc": "Lab of Jev", + "created": "2026-09-19T22:20:05Z", + "pushed": "2026-09-20T17:37:20Z", + "url": "https://github.com/renatobardi/jev-o-matic" + }, + { + "id": "robokrunch/awesome-jev", + "source": "github", + "stars": 0, + "desc": "A curated list of resources for Jev \u2014 TypeSafe AI's System One decision model. Maintained by RoboKrunch.", + "created": "2026-09-19T23:11:42Z", + "pushed": "2026-09-20T17:26:00Z", + "url": "https://github.com/robokrunch/awesome-jev" + }, + { + "id": "rudy2steiner/awesome-jev-hub", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T15:33:24Z", + "pushed": "2026-09-20T16:59:09Z", + "url": "https://github.com/rudy2steiner/awesome-jev-hub" + }, + { + "id": "rustybladerunner/jev-skills", + "source": "github", + "stars": 0, + "desc": "Practical Jev integration and experiment workflows, with runnable evidence-check examples.", + "created": "2026-09-20T17:03:48Z", + "pushed": "2026-09-20T17:03:51Z", + "url": "https://github.com/rustybladerunner/jev-skills" + }, + { + "id": "saurabhkumar8112/jev-gpt5-routing-study", + "source": "github", + "stars": 0, + "desc": "Reproducible offline Jev and GPT-5 routing study: cost, accuracy, and failed held-out parity", + "created": "2026-09-20T15:09:20Z", + "pushed": "2026-09-20T16:52:06Z", + "url": "https://github.com/saurabhkumar8112/jev-gpt5-routing-study" + }, + { + "id": "shaduf-labs/jev-catalog", + "source": "github", + "stars": 0, + "desc": "Jev use cases, open-source alternatives, real products and current access prices.", + "created": "2026-09-20T16:59:48Z", + "pushed": "2026-09-20T16:59:49Z", + "url": "https://github.com/shaduf-labs/jev-catalog" + }, + { + "id": "shinjiyu/jev-debug", + "source": "github", + "stars": 0, + "desc": "Locate a bug file, or flag a diff, by cross-checking small source batches with JEV.", + "created": "2026-09-20T17:17:44Z", + "pushed": "2026-09-20T17:17:52Z", + "url": "https://github.com/shinjiyu/jev-debug" + }, + { + "id": "smithclay/dbt_jev", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T16:39:40Z", + "pushed": "2026-09-20T16:40:06Z", + "url": "https://github.com/smithclay/dbt_jev" + }, + { + "id": "stevenke1981/jev-codex-harness", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-20T17:05:45Z", + "pushed": "2026-09-20T17:30:44Z", + "url": "https://github.com/stevenke1981/jev-codex-harness" + }, + { + "id": "sumanmichael/jevlang", + "source": "github", + "stars": 0, + "desc": "Python with a smart if statement, backed by TypeSafe's Jev model", + "created": "2026-09-20T17:27:50Z", + "pushed": "2026-09-20T17:44:34Z", + "url": "https://github.com/sumanmichael/jevlang" + }, + { + "id": "taodav/jev_deep_rl", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T17:13:05Z", + "pushed": "2026-09-20T17:31:56Z", + "url": "https://github.com/taodav/jev_deep_rl" + }, + { + "id": "tinmanlab/cartpole-jev", + "source": "github", + "stars": 0, + "desc": "CartPole lab for calibrated decision supervision: Jev-inspired local heads, optional TypeSafe Jev, deterministic control, and evidence-first robotics research.", + "created": "2026-09-20T17:09:00Z", + "pushed": "2026-09-20T17:09:01Z", + "url": "https://github.com/tinmanlab/cartpole-jev" + }, + { + "id": "toorop/pi-jev-router", + "source": "github", + "stars": 0, + "desc": "", + "created": "2026-09-19T15:19:10Z", + "pushed": "2026-09-20T17:02:13Z", + "url": "https://github.com/toorop/pi-jev-router" + }, + { + "id": "wsargent/laya-mcp", + "source": "github", + "stars": 0, + "desc": "Local MCP server exposing laya-mlx typed-decision inference on Apple Silicon via fastmcp", + "created": "2026-09-20T16:58:30Z", + "pushed": "2026-09-20T17:00:23Z", + "url": "https://github.com/wsargent/laya-mcp" + }, + { + "id": "xiaohu0x/jevaimodel", + "source": "github", + "stars": 0, + "desc": "Source for jevaimodel.app", + "created": "2026-09-20T04:13:09Z", + "pushed": "2026-09-20T17:22:02Z", + "url": "https://github.com/xiaohu0x/jevaimodel" + }, + { + "id": "xiaohu0x/jevhunt", + "source": "github", + "stars": 0, + "desc": "jevhunt project", + "created": "2026-09-20T07:17:25Z", + "pushed": "2026-09-20T17:39:50Z", + "url": "https://github.com/xiaohu0x/jevhunt" + }, + { + "id": "zebedelu/chess-vs-jev", + "source": "github", + "stars": 0, + "desc": "A pygame chess game with Jev as the AI opponent. Play Human vs Human, Human vs Jev, or watch Jev vs Jev.", + "created": "2026-09-20T17:25:47Z", + "pushed": "2026-09-20T17:28:18Z", + "url": "https://github.com/zebedelu/chess-vs-jev" + }, + { + "id": "zurfyx/jev-browser-skill", + "source": "github", + "stars": 0, + "desc": "Let Jev, TypeSafe's ~100ms decision model, drive your browser. A plug-and-play skill for Claude Code and Codex.", + "created": "2026-09-19T22:05:25Z", + "pushed": "2026-09-20T17:40:21Z", + "url": "https://github.com/zurfyx/jev-browser-skill" + } +] \ No newline at end of file diff --git a/research/archive/hourly/2026-09-20T17/run_digest.json b/research/archive/hourly/2026-09-20T17/run_digest.json new file mode 100644 index 00000000..6d023855 --- /dev/null +++ b/research/archive/hourly/2026-09-20T17/run_digest.json @@ -0,0 +1,13 @@ +{ + "label": "1143", + "utc": "2026-09-20T17:43Z", + "notes": "\u00a7121", + "composition": "369-384", + "findings": "#104", + "invented_signal": false, + "primary": [ + "kshetrajna12/reflex", + "lakeday-org/perch", + "emretheus/jev-rag-benchmark" + ] +} diff --git a/research/changelog-hourly.md b/research/changelog-hourly.md index 494f168d..5696d6c5 100644 --- a/research/changelog-hourly.md +++ b/research/changelog-hourly.md @@ -13,6 +13,18 @@ This is the uniqueness-lock archive after hourly folds (#2–#40 / notes --- +## Hourly 1143 HIGH (notes.md §121 / items 369–384 / batch #104) + +- Fresh PR off latest `main` after merged #43 (1049 / §120). **HARD RULE:** + do not reopen or amend PR #23–#43. Skip Archer. Quote *theirs*. No wrappers. + `invented_signal: false`. +- open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. + cutoff 0.8 still soft. paired bootstrap CIs *theirs*. + Same accuracy, 35x faster *theirs*. systems comparison ≠ semantic equivalence. +- Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 + +--- + ## Hourly 1049 HIGH (notes.md §120 / items 353–368 / batch #103) - Fresh PR off latest `main` after merged #41 (0947 / §119) and diff --git a/research/notes.md b/research/notes.md index 4ec2d399..4143a6e8 100644 --- a/research/notes.md +++ b/research/notes.md @@ -29987,3 +29987,315 @@ Parent merge only after **CLEAN** adversarial review 0947 + 1049). No live Jev key. No wrappers. Hourly 1049 uniqueness lock: ggmlc GGUF is not llama.cpp; Loading them in llama.cpp will fail; one encoder pass; hf:mys/laya-GGUF sha 713ae6f6e39f likes 0 apache-2.0; hf:mys/laya-multilingual-GGUF sha 3b645ae54281; hf:mys/laya-typed-decisions-GGUF sha 1e9e8ba1f527; hf:tozp/laya-onnx sha 0862aeba1e65 Opset 14 FP32 and INT8; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; docker-laya MIT HEAD 1b8239a51ddd README SHA 9cb7bdc3; laya.cpp RTX ggml CUDA HEAD 8590937c79a2 README SHA cdd429b9; serving substrate ≠ calibrated replica; Softmax over options ≠ calibrated Noul; Qwen3.5-4B ≠ Archer; Qwen3.5-9B ≠ Archer; jev-position-test n=6 HEAD 7a56ca1c2698 README SHA 23f194c9; jevmlx slots 5 of 6; hosted Jev 0 of 6; prior_correction made it worse; jevSweeper mean Spearman ρ −0.274; picked exact-optimal 1/25 (4%); 31 of 36 still logically decidable; 86% of the time we should not have been asking; game success ≠ calibrated Noul; LLM2Jev 64★ Apache-2.0 HEAD 924618721277 README SHA da35fe61; not affiliated with or endorsed by Jev or TypeSafe; No answer tokens are generated; OpenSourceJev llama.cpp Qwen3-1.7B HEAD 3c41fba3681d; JEV-MLX Qwen3.5-9B HEAD dec24cd929ea; decision-head-rlcd Qwen3.5-4B 4.9M LoRA; AUTO_ACT is not a Noul; closed-set fail-open stdlib-only; verified=False; soft scores ≠ hard gates; 22 to 40% cheaper *theirs*; first version 70% more expensive; 111-case benchmark *theirs*; CDC 5,000 Jev 72.2% AUC 0.7725 *theirs*; accuracy is a trap; 9.0% base rate always-no 91.0%; catalog ≠ endorsement; jev-skill 109★ 90 scenarios HEAD 4f6e899a24d4; awesome-jev-live 673 entries 4★; minecraft-agent 214★ 131 JEV decisions 35 Astra calls; nether-final-08 8 minutes 43.300 seconds; planner writes JEV selects; RoboJEV structured simulator state not images; ashare-trader 策略未通过自己的回测门槛; 36 组参数全部净期望为负; no positive expectation under real costs; typed_evals NOT an official TypeSafe AI product; jev-as-judge is a sensor; third-person-audit 40% & 60% watermarks still soft; The included experience uses a handwritten demo provider; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42; notes.md §120 + +## 121. Hourly 1143 HIGH (2026-09-20 ~11:43 Boise / 2026-09-20T17:43Z) + +Measurement / class-member fold on a **fresh PR off latest `main`** +(`cursor/fold-hourly-1143-high-a434`) after `a61372f` (merged #43 hourly +1049, `notes.md` §120 / items 353–368 / batch #103). **HARD RULE:** do +not reopen or amend PR #23–#43. Do **not** push onto merged 1049 / 0947 +tracks. This fold's IDs: `notes.md` §121 / composition 369–384 / +findings batch #104. + +Never reopen merged #7–**#43**. Do **not** re-fold §120 1049 / §119 0947 +/ §118 llm-to-jev / §117 SemIf / §116 jcr / §115 NanoJev / §114 0843 *as +a second census*. Densify `gqgs/laya-onnx` (496.8 MiB browser int8, +already named in the 1049 lock), `Kevthetech143/super-jev` (experimental +V0.2.0), and `mradermacher/jevify-gemma4-e4b-GGUF` (static quants of +§119 `kushalpatil/jevify-gemma4-e4b`). Skip Archer rewrite. Quote +READMEs. Mark *theirs*. No wrappers, keys, `npm` / `pip` / `uv` / +`docker` install recipes. `invented_signal: false`. Hunches labeled. + +Lane is Augustus: **mathematical / logical / algorithmic mental models** +for Jev-class categorization/scoring across AI / SWE / **business / +knowledge work / life**, not SWE-only. PRIMARY this hour is **open +recreation ≠ calibrated replica**: kshetrajna12/reflex is an open +re-creation of Jev on Qwen3.5-4B; the WebGPU 0.8B path is "less +calibrated"; Qwen3.5-4B ≠ Archer. Semantic lint is a sensor not a +proof. Third-party benches (paired bootstrap CIs, BANKING77, cascade +parity miss, invented-ticket gate) are *theirs*. Local Laya serving is +still substrate ≠ replica. Catalogs are indexes. Soft scores ≠ hard +gates. Archer still **promised_not_landed**. + +Unique consecutive fragments (this hour) must appear as **one +substring** in overlays (see uniqueness gate): +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 + +### How-to-apply (five placements / measurement lenses) + +These are *class* lenses, not vendor tutorials. Same discipline as +§120 (serving substrate ≠ calibrated replica) and §119 (judge ≠ +actuator). Formal methods **compose** with scoring: a Noul is a SENSOR; +a lint plugin is a sensor; a recreation is a class member; policy / +parser / replay / cost table are exact work. + +1. **Open recreation ≠ calibrated replica** + (*theirs*, kshetrajna12/reflex PRIMARY + pngwn demo + DavidHatley + + kiroclawai blueprint + ahm-ge18/jev_replica). Quote *theirs*: It + is an open re-creation of Jev. Built on Qwen3.5-4B. WebGPU 0.8B is + less calibrated. pngwn demo accuracy 0.705 ECE 0.046 ~112 ms. + This dataset and model are independent research artifacts, not + reproductions of Jev or RLCD. Qwen3.5-4B ≠ Archer. Life analogue: + a cover band can play the set list; it is not the studio master. +2. **Third-party benches are *theirs*** + (emretheus RAG CIs + BANKING77 speed-test + gpt5 routing + + confidence-gate + cartpole + SmartMoney-Cub). paired bootstrap CIs + *theirs*. +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31. +7.62 pts + SciFact CI +4.88 to +10.38. BANKING77 500 Jev 81.0% GPT-OSS 82.8% + Mercury 73.2% Gemini 85.4% *theirs*. frozen cascade missed its + evaluation accuracy target 430/500 vs GPT-5 432/500. This is not + demonstrated equal-quality savings. 24 invented tickets. Routing + errors caught by the gate 0 of 3. sample too small to establish + calibration. This is not TypeSafe Jev. No real API requests were + made. No orders, no advice. Same accuracy, 35x faster *theirs* is + a systems comparison ≠ semantic equivalence. Evaluator: CI that + includes zero is not equivalence; cutoff fire is not a proof. +3. **Semantic lint is a sensor not a proof** + (*theirs*, perch 164★ PRIMARY + oxlint-plugin-jev + patdown + + nassim-arifette/jevgrep + PanAchy/jevvy). Semantic code linting + with Jev. The plugin reports an error when the yes-probability + clears your cutoff. cutoff 0.8 still soft. patdown fuzzy linter. + nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ + uehaj/jev-semgrep. PanAchy/jevvy ≠ Atominac/jevvy. A sensor can + flag; code owns the gate. Life analogue: a smoke alarm is not a + fire marshal. +4. **Local Laya serving densify, still substrate ≠ replica** + (KonghaYao/laya-jev + gqgs/laya-onnx densify + glukicov/laya_router + + Pidbid/laya-deploy + wsargent/laya-mcp). Quote *theirs*: + 按官方接口写的客户端只改一个 base URL. wire-compat ≠ replica. + gqgs/laya-onnx densify 496.8 MiB (520,921,501 bytes). + tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx. serving + substrate ≠ calibrated replica. Same accuracy, 35x faster *theirs* + (Laya 0.600 / 184 ms vs GPT-5 nano 0.600 / 6,415 ms on 180 labelled + requests) is systems comparison ≠ semantic equivalence. Do **not** + copy `uv` / `npm` / `docker`. +5. **Catalogs, harnesses, and games are indexes** + (BeatAPI/awesome-jev + cousins + super-jev densify + Pasblinn/jev-lab + + allay-team/openjev collision). All 125 projects. catalog ≠ + endorsement. BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ + Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ + rudy2steiner/awesome-jev-hub. Pasblinn/jev-lab ≠ tanay / dairui1 / + BrendanH18 / yibie / q93304989-bit. Independent project. Not + affiliated with TypeSafe. Kevthetech143/super-jev densify + experimental V0.2.0. permission ≠ confidence. allay-team/openjev ≠ + piyush-infocusp/openjev ≠ TheoLeeCJ/openjev. 2022 Mineflayer + Jevalent collision. Skip-thin: 4OH4/jev-compare Nothing to run yet. + +### HIGH + +1. **[kshetrajna12/reflex](https://github.com/kshetrajna12/reflex)** + - NEW HIGH PRIMARY recreation (Python MIT; **85★**; HEAD + `19586a1374dc`; README SHA `8f39fc38`). Quote *theirs*: It is an + open re-creation of Jev. Qwen3.5-4B by default. WebGPU 0.8B is + the same request format, just smaller and less calibrated. + open recreation ≠ calibrated replica. Qwen3.5-4B ≠ Archer. Do + **not** copy `pip` / keys. +2. **[lakeday-org/perch](https://github.com/lakeday-org/perch)** + - NEW HIGH PRIMARY sensor (JavaScript MIT; **164★**; HEAD + `ba775a9940b6`; README SHA `7ad0403b`). Quote *theirs*: Semantic + code linting with Jev. semantic lint is a sensor not a proof. + Do **not** copy `npm i @lakeday/perch`. +3. **[wobsoriano/oxlint-plugin-jev](https://github.com/wobsoriano/oxlint-plugin-jev)** + - NEW HIGH (TypeScript MIT; **36★**; HEAD `82703b841f93`; README + SHA `af703de3`). Quote *theirs*: the plugin reports an error when + the yes-probability clears your cutoff. Example cutoff 0.8. + cutoff 0.8 still soft. soft scores ≠ hard gates. Do **not** copy + `npm i` / `TYPESAFE_API_KEY`. +4. **[myc0576/SmartMoney-Cub](https://github.com/myc0576/SmartMoney-Cub)** + - NEW HIGH life/honesty (Python MIT; **25★**; HEAD `d93cf493853d`; + README SHA `678a9565`). GitHub description *theirs*: No orders, no + advice. Read-only journal. Human-in-the-loop required. Do **not** + paste finance scores as Harbor. +5. **[nassim-arifette/jevgrep](https://github.com/nassim-arifette/jevgrep)** + - NEW HIGH namesake (TypeScript MIT; **12★**; HEAD `52e2d895b69d`; + README SHA `82b93522`). Find code by what it does. nassim-arifette/jevgrep + ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep. Sensor, + not a proof. Do **not** copy `npm`. +6. **[tyler-dot-earth/patdown](https://github.com/tyler-dot-earth/patdown)** + - NEW HIGH (TypeScript NOASSERTION; **10★**; HEAD `ae0e277fdd64`; + README SHA `275f4b9c`). Quote *theirs*: fuzzy rules in one markdown + file. Swappable judge. cutoff still soft. Do **not** copy `npx`. +7. **[PanAchy/jevvy](https://github.com/PanAchy/jevvy)** + - NEW HIGH namesake (TypeScript MIT; **9★**; HEAD `f682078c0362`; + README SHA `f2a69efc`). Permission plugin for coding agents. + PanAchy/jevvy ≠ Atominac/jevvy. permission ≠ confidence. +8. **[Kevthetech143/super-jev](https://github.com/Kevthetech143/super-jev)** + - DENSIFY (TypeScript MIT; **6★**; HEAD `6710de64e7ad`; README SHA + `5754c76d`). Quote *theirs*: Status: experimental V0.2.0. independent + community project, not affiliated with TypeSafe AI. permission ≠ + confidence. Do **not** mint a second census. +9. **[BeatAPI/awesome-jev](https://github.com/BeatAPI/awesome-jev)** + - NEW HIGH catalog (MIT; **5★**; HEAD `32685ae0aab7`; README SHA + `8f821542`). Quote *theirs*: All 125 projects. 100+ GitHub stars + filter. catalog ≠ endorsement. BeatAPI/awesome-jev ≠ 99hansling ≠ + Vishnurr2k01 ≠ robokrunch ≠ rudy2steiner/awesome-jev-hub. +10. **[KonghaYao/laya-jev](https://github.com/KonghaYao/laya-jev)** + - NEW HIGH serving (license **null**; **1★**; HEAD `4e940441bab0`; + README SHA `e7450b77`). Quote *theirs*: 按官方接口写的客户端只改一个 + base URL. wire-compat ≠ replica. serving substrate ≠ calibrated + replica. Do **not** copy `uv pip` / serve recipes. +11. **[Pasblinn/jev-lab](https://github.com/Pasblinn/jev-lab)** + - NEW HIGH namesake (MIT; **1★**; size **0** WITH CONTENTS; HEAD + `eafb858c0c65`; README SHA `f98704d2`). Quote *theirs*: Independent + project. Not affiliated with TypeSafe. Builds on gargpratyush/jev-router. + Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ + BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab. +12. **[gqgs/laya-onnx](https://github.com/gqgs/laya-onnx)** + - DENSIFY §120 namesake (license **null**; **1★**; HEAD + `7ae05e1f1c04`; README SHA `9aa8b9e6`). Quote *theirs*: + 520,921,501 bytes (496.8 MiB). Complete browser int8. tozp/laya-onnx + ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx. serving substrate ≠ + calibrated replica. Do **not** mint a second census. +13. **[emretheus/jev-rag-benchmark](https://github.com/emretheus/jev-rag-benchmark)** + - NEW HIGH measurement (Python MIT; **0★**; size **0** WITH + CONTENTS; HEAD `f02b0f1e44d5`; README SHA `1ee58b90`). Quote + *theirs*: +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31. +7.62 pts + SciFact CI +4.88 to +10.38. paired bootstrap CIs *theirs*. + emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark + (§89). Do **not** paste CIs as Harbor. +14. **[glukicov/laya_router](https://github.com/glukicov/laya_router)** + - NEW HIGH systems comparison (Python Apache-2.0; **0★**; HEAD + `c9b83958112a`; README SHA `9e17af89`). GitHub description + *theirs*: Same accuracy, 35x faster. Table: Laya 0.600 / 184 ms / + $0 vs GPT-5 nano 0.600 / 6,415 ms / $0.58 on 180 labelled + requests. They disagree on 84 of 180. systems comparison ≠ + semantic equivalence. Do **not** copy `uv`. +15. **[MohtashamMurshid/jev-speed-test](https://github.com/MohtashamMurshid/jev-speed-test)** + - NEW HIGH (license **null**; **0★**; HEAD `14b9632ea270`; README + SHA `78ea7973`). Quote *theirs*: BANKING77 500. Jev 81.0% 347 ms + $0.068. GPT-OSS 82.8%. Mercury 73.2%. Gemini 85.4%. Confidence + was useful, not a guarantee. *theirs*, not Harbor. +16. **[saurabhkumar8112/jev-gpt5-routing-study](https://github.com/saurabhkumar8112/jev-gpt5-routing-study)** + - NEW HIGH (NOASSERTION; **0★**; HEAD `6fca78deb7b0`; README SHA + `33126c29`). Quote *theirs*: The frozen cascade missed its + evaluation accuracy target: 430/500 vs GPT-5 432/500. This is + not demonstrated equal-quality savings. Do not compare cost + without task success (§119). +17. **[abh2050/jev-test-confidence-gate](https://github.com/abh2050/jev-test-confidence-gate)** + - NEW HIGH (license **null**; **0★**; HEAD `aee1451ed8dc`; README + SHA `d8b6783d`). Quote *theirs*: 24 invented tickets. Routing + errors caught by the gate 0 of 3. sample too small to establish + calibration. Evaluator: n=24 is not a calibration proof. +18. **[tinmanlab/cartpole-jev](https://github.com/tinmanlab/cartpole-jev)** + - NEW HIGH honesty (Python MIT; **0★**; HEAD `282d294d1fae`; + README SHA `803270e3`). Quote *theirs*: This is not TypeSafe Jev. + No real API requests were made for the committed results. + Controller owns force. Decision head is a supervisor. +19. **[hf:mradermacher/jevify-gemma4-e4b-GGUF](https://huggingface.co/mradermacher/jevify-gemma4-e4b-GGUF)** + - DENSIFY §119 (likes **0**; sha `174d1e3c9432`). Quote *theirs*: + static quants of kushalpatil/jevify-gemma4-e4b. Quant ≠ + calibration. Qwen3.5-4B ≠ Archer stays the class lock. +20. **[hfdataset:DavidHatley/system-one-mini-data](https://huggingface.co/datasets/DavidHatley/system-one-mini-data)** + - NEW HIGH honesty (likes **0**; sha `319c9329b417`). Quote + *theirs*: This dataset and model are independent research + artifacts, not reproductions of Jev or RLCD. +21. **[hfspace:pngwn/system-one-demo](https://huggingface.co/spaces/pngwn/system-one-demo)** + - NEW HIGH (likes **0**; sha `70450cb7d8e2`). Quote *theirs*: + accuracy 0.705, ECE 0.046, ~112 ms. Qwen3.5-4B-Base + LoRA. + Qwen3.5-4B ≠ Archer. Trackio cousins densify the same card. +22. **[kiroclawai/system-one-blueprint](https://github.com/kiroclawai/system-one-blueprint)** + - NEW HIGH (license **null**; **0★**; HEAD `c455d741bc50`; README + SHA `76ee0a23`). Quote *theirs*: Status: Blueprint / design + document. A blueprint is not a replica. +23. **[Pidbid/laya-deploy](https://github.com/Pidbid/laya-deploy)** + + **[wsargent/laya-mcp](https://github.com/wsargent/laya-mcp)** + - NEW HIGH serving (HEADs `9e0b9ac0228f` / `344d1b0a0513`). Local + omp / MCP Laya backends. serving substrate ≠ calibrated replica. + Do **not** copy install scripts. +24. **[TonyP-MR/jev-curation-engine](https://github.com/TonyP-MR/jev-curation-engine)** + - NEW HIGH (HEAD `caed0279bf27`; README SHA `980f4fcd`). Read-only + feasibility test rig. Not a production claim. +25. **[chr-kelly/jev-cookbook](https://github.com/chr-kelly/jev-cookbook)** + - NEW HIGH (MIT; HEAD `3727b8b4d2fb`; README SHA `25d83f1d`). + Runnable question sets. Catalog of shapes, not Harbor. + +### Remainder (short cards, same hour) + +sah1l/jev-battleship is a game measurement (web + CLI arena). +crman/jev-ai-output-judge is planned, not measured. ahm-ge18/jev_replica +is a Laya sentiment dashboard, not a Jev replica. 99hansling / +Vishnurr2k01 / robokrunch / rudy2steiner are catalog cousins of BeatAPI +(namesake lock). Games, proxies, n8n nodes, and thin apps +(jev-battleship, jev-joust, jev-flappy-bird, jobfit, jev-proxy, jevqa, +traceassert, jev-audit, jev-nfl, jev-affected, sudoku-jev, garmin-jev-buddy, +jev-adrank, Twitter-keyword-shield, jevenator2, TheFishGame, jev-sts2-agent, +innovation-poc-jev, trying-jev, playground-jev, and similar) are first +sightings or thin applications: catalog, do not elevate. + +### Skips (thin / collision / 404) + +- 4OH4/jev-compare README **610 B**: Nothing to run yet. skip-thin. +- Kaos599/jev-writer GitHub **404**. +- ayshrosine/Jev_Typesafe.Ai no HEAD. +- piyush-infocusp/openjev empty (size 0, no HEAD). +- shaduf-labs/jev-catalog empty. +- AparnaDhinakaran/jev-as-judge README **15 B**. +- jyothepro/jev-games README **11 B**. +- smithclay/dbt_jev README **10 B**. +- OsirianLegacy/JevSimulation, Vazool/jev-escape-room, jdubpark/jevcode, + nsillik/jevvin-off: no README. +- allay-team/openjev created **2022-07-11**: Mineflayer/Jevalent + collision, not TypeSafe Jev. Namesake lock only. +- Archer rewrite: **promised_not_landed**. Hub archerhume/4rcherhume + HTTP **401**. + +### Pulse (live REST this hour) + +Archer still promised_not_landed. Hub archerhume/4rcherhume HTTP **401**. +This hour does not re-census SemIf / Laya likes / tracker; those numbers +stay §119 until a dedicated pulse. perch **164★**. reflex **85★**. +oxlint-plugin-jev **36★**. SmartMoney-Cub **25★**. + +### Formal compose + anti-patterns + +Formal methods **compose** with scoring. A Noul is a SENSOR. A lint +plugin is a SENSOR. A recreation on Qwen3.5-4B is a class member, not +Archer. Serving (ONNX / local HTTP / MCP) is a substrate. Treating +reflex as TypeSafe Jev, cutoff 0.8 as a proof, paired CIs as Harbor, +35x as semantic equivalence, 430/500 as equal-quality savings, 24 +invented tickets as a calibration study, a catalog row as endorsement, +or permission-plugin confidence as authorization is soundness theater. +open recreation ≠ calibrated replica. semantic lint is a sensor not a +proof. systems comparison ≠ semantic equivalence. wire-compat ≠ replica. +permission ≠ confidence. catalog ≠ endorsement. + +### Adversarial review + testing hooks (Basit standing +order) + +Parent merge only after **CLEAN** adversarial review **AND** testing. +Hooks for the reviewer: + +- Uniqueness-gate: the consecutive `Hourly 1143 uniqueness lock:` + string must appear in every overlay listed below. Prior walls 0843 / + 0915 / jcr / 0922 / 0940 / 0947 / 1049 stay one substring each (do + not mutate them; do not reopen #23–#43). +- Namesake locks: nassim-arifette/jevgrep ≠ Bentlybro / can1357 / + uehaj; PanAchy/jevvy ≠ Atominac/jevvy; emretheus/jev-rag-benchmark ≠ + erendikmenn; BeatAPI/awesome-jev ≠ 99hansling ≠ Vishnurr2k01 ≠ + robokrunch ≠ rudy2steiner; Pasblinn/jev-lab ≠ tanay / dairui1 / + BrendanH18 / yibie / q93304989-bit; allay-team/openjev ≠ + piyush-infocusp ≠ TheoLeeCJ; tozp/laya-onnx ≠ Mattepiu ≠ gqgs. +- Densify vs new: gqgs/laya-onnx densifies §120; super-jev densifies + experimental V0.2.0; mradermacher GGUF densifies §119 e4b. reflex / + perch / oxlint / RAG CIs / BANKING77 / cascade miss are first + sightings this hour. +- Harbor-jevals: +0.82 / +7.62 / 81.0% / 430/500 / 0.705 / 35x are + *theirs*, not Harbor. 128/128 stays NanoJev §115. 0.779 JevBench is + *theirs* from §119. +- Anti-patterns to refuse: TypeSafe drop-in; Qwen3.5-4B as Archer; + cutoff 0.8 as a proof; paired CI as Harbor; 35x as semantic + equivalence; catalog as endorsement; permission as confidence; + copying keys / `npm` / `pip` / `uv`. +- Overlay set: SKILL.md body (not YAML surgery beyond protocol + fragments), mental-models Apply 1143, composition-algebra items + 369–384, faq, mixed-architecture, validation, toolbox-mapping, + methods-catalog, formal-methods, formal-semi-formal, + applied-mappings, judgment-class, question-design, + agent-self-assessment, mappings, CHANGELOG, README, docs/ecosystem, + findings batch #104, refresh-log, sources.json, changelog-hourly.md. +- Offline check: `evaluate_decisions.py --self-test` (now includes + paired-CI / same-accuracy speedup / lint-cutoff still soft) and + `uniqueness_gate.py` (0843 + 0915 + jcr + 0922 + 0940 + 0947 + 1049 + + 1143). No live Jev key. No wrappers. + +Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 diff --git a/research/refresh-log.md b/research/refresh-log.md index 3b59733c..85e4241e 100644 --- a/research/refresh-log.md +++ b/research/refresh-log.md @@ -1,4 +1,17 @@ +## 2026-09-20 ~17:43 UTC / ~11:43 Boise - Hourly 1143 HIGH +- Fresh PR off latest `main` after merged #43 (1049 / `notes.md` §120 / + items 353–368 / batch #103). Next free IDs: + `notes.md` §121 / composition 369–384 / findings batch #104. + **HARD RULE:** do not reopen or amend PR #23–#43. +- PRIMARY: open recreation ≠ calibrated replica. semantic lint is a sensor not a proof. + paired bootstrap CIs *theirs*. Same accuracy, 35x faster *theirs*. +- Evidence: `research/archive/hourly/2026-09-20T17/` (novel_high_this_run). +- uniqueness_gate 0843+0915+jcr+0922+0940+0947+1049+1143. Evaluator paired-CI / cutoff. + Quote *theirs*. No wrappers. `invented_signal: false`. +- Hourly 1143 uniqueness lock: open recreation ≠ calibrated replica; Qwen3.5-4B ≠ Archer; It is an open re-creation of Jev; less calibrated; perch 164★ MIT HEAD ba775a9940b6 README SHA 7ad0403b; semantic lint is a sensor not a proof; oxlint-plugin-jev cutoff 0.8 still soft; nassim-arifette/jevgrep ≠ Bentlybro/jevgrep ≠ can1357/jegrep ≠ uehaj/jev-semgrep; patdown fuzzy linter; PanAchy/jevvy ≠ Atominac/jevvy; No orders, no advice; SmartMoney-Cub 25★ HEAD d93cf493853d; paired bootstrap CIs *theirs*; emretheus/jev-rag-benchmark ≠ erendikmenn/jev-rag-benchmark; +0.82 pts XQuAD-EN 95% CI +0.35 to +1.31; +7.62 pts SciFact CI +4.88 to +10.38; Same accuracy, 35x faster *theirs*; systems comparison ≠ semantic equivalence; BANKING77 500 Jev 81.0% GPT-OSS 82.8% Mercury 73.2% Gemini 85.4% *theirs*; frozen cascade missed its evaluation accuracy target 430/500 vs GPT-5 432/500; This is not demonstrated equal-quality savings; 24 invented tickets; Routing errors caught by the gate 0 of 3; sample too small to establish calibration; This is not TypeSafe Jev; No real API requests were made; wire-compat ≠ replica; KonghaYao/laya-jev 按官方接口写的客户端只改一个 base URL; gqgs/laya-onnx densify 496.8 MiB; tozp/laya-onnx ≠ Mattepiu/laya-onnx ≠ gqgs/laya-onnx; serving substrate ≠ calibrated replica; BeatAPI/awesome-jev ≠ 99hansling/awesome-jev ≠ Vishnurr2k01/awesome-jev ≠ robokrunch/awesome-jev ≠ rudy2steiner/awesome-jev-hub; All 125 projects; catalog ≠ endorsement; Pasblinn/jev-lab ≠ tanayvasishtha/jev-lab ≠ dairui1/jev-lab ≠ BrendanH18/jev-lab ≠ yibie/laya-jev-lab ≠ q93304989-bit/jev-lab; Independent project. Not affiliated with TypeSafe; Kevthetech143/super-jev densify experimental V0.2.0; permission ≠ confidence; allay-team/openjev ≠ piyush-infocusp/openjev ≠ TheoLeeCJ/openjev; 2022 Mineflayer Jevalent collision; kushalpatil/jevify-gemma4-e4b GGUF densify; static quants; This dataset and model are independent research artifacts, not reproductions of Jev or RLCD; pngwn demo accuracy 0.705 ECE 0.046 ~112 ms *theirs*; cutoff 0.8 still soft; soft scores ≠ hard gates; do not reopen or amend PR #23/#24/#25/#26/#27/#28/#29/#30/#31/#32/#33/#34/#35/#36/#37/#38/#39/#40/#41/#42/#43; notes.md §121 + + ## 2026-09-20 ~16:49 UTC / ~10:49 Boise — Hourly 1049 HIGH - Fresh PR off latest `main` after merged #41 (0947 / `notes.md` §119 / items 337–352 / batch #102) and merged #42 v0.5.0. Next free IDs: diff --git a/research/sources.json b/research/sources.json index 96710594..f596829d 100644 --- a/research/sources.json +++ b/research/sources.json @@ -1,6 +1,6 @@ { "refresh_cadence": "hourly", - "retrieved": "2026-09-20T15:47Z", + "retrieved": "2026-09-20T17:43Z", "sources": [ { "kind": "docs", @@ -5233,6 +5233,78 @@ "title": "ghubnab99/jev-enterprise-decision-fabric", "url": "https://github.com/ghubnab99/jev-enterprise-decision-fabric", "note": "MIT HEAD 51dce51c359b 111-case benchmark *theirs*. notes.md §120." + }, + { + "kind": "github", + "title": "kshetrajna12/reflex", + "url": "https://github.com/kshetrajna12/reflex", + "note": "MIT 85★ HEAD 19586a1374dc open re-creation of Jev on Qwen3.5-4B; less calibrated. notes.md §121." + }, + { + "kind": "github", + "title": "lakeday-org/perch", + "url": "https://github.com/lakeday-org/perch", + "note": "MIT 164★ HEAD ba775a9940b6 semantic lint is a sensor not a proof. notes.md §121." + }, + { + "kind": "github", + "title": "wobsoriano/oxlint-plugin-jev", + "url": "https://github.com/wobsoriano/oxlint-plugin-jev", + "note": "MIT 36★ HEAD 82703b841f93 cutoff 0.8 still soft. notes.md §121." + }, + { + "kind": "github", + "title": "myc0576/SmartMoney-Cub", + "url": "https://github.com/myc0576/SmartMoney-Cub", + "note": "MIT 25★ HEAD d93cf493853d No orders, no advice. notes.md §121." + }, + { + "kind": "github", + "title": "emretheus/jev-rag-benchmark", + "url": "https://github.com/emretheus/jev-rag-benchmark", + "note": "MIT HEAD f02b0f1e44d5 paired bootstrap CIs *theirs* +0.82 / +7.62. notes.md §121." + }, + { + "kind": "github", + "title": "glukicov/laya_router", + "url": "https://github.com/glukicov/laya_router", + "note": "Apache-2.0 HEAD c9b83958112a Same accuracy, 35x faster *theirs*. notes.md §121." + }, + { + "kind": "github", + "title": "MohtashamMurshid/jev-speed-test", + "url": "https://github.com/MohtashamMurshid/jev-speed-test", + "note": "HEAD 14b9632ea270 BANKING77 500 Jev 81.0% *theirs*. notes.md §121." + }, + { + "kind": "github", + "title": "saurabhkumar8112/jev-gpt5-routing-study", + "url": "https://github.com/saurabhkumar8112/jev-gpt5-routing-study", + "note": "HEAD 6fca78deb7b0 430/500 vs 432/500 not demonstrated equal-quality savings. notes.md §121." + }, + { + "kind": "github", + "title": "KonghaYao/laya-jev", + "url": "https://github.com/KonghaYao/laya-jev", + "note": "HEAD 4e940441bab0 wire-compat ≠ replica. notes.md §121." + }, + { + "kind": "github", + "title": "BeatAPI/awesome-jev", + "url": "https://github.com/BeatAPI/awesome-jev", + "note": "MIT 5★ HEAD 32685ae0aab7 All 125 projects catalog ≠ endorsement. notes.md §121." + }, + { + "kind": "hf-space", + "title": "pngwn/system-one-demo", + "url": "https://huggingface.co/spaces/pngwn/system-one-demo", + "note": "sha 70450cb7d8e2 accuracy 0.705 ECE 0.046 ~112 ms *theirs*. notes.md §121." + }, + { + "kind": "hf-dataset", + "title": "DavidHatley/system-one-mini-data", + "url": "https://huggingface.co/datasets/DavidHatley/system-one-mini-data", + "note": "sha 319c9329b417 independent research artifacts, not reproductions of Jev or RLCD. notes.md §121." } ] }