From 2ac8a4224e285e0b9ef73148d4722f387cf0730a Mon Sep 17 00:00:00 2001
From: huangruiteng <14976749+huangruiteng@users.noreply.github.com>
Date: Sat, 19 Sep 2026 01:25:05 +0800
Subject: [PATCH] fix(site): align LHTB publication contract

Signed-off-by: huangruiteng <14976749+huangruiteng@users.noreply.github.com>
---
 apps/presentation/site/src/lhtb-copy.json |  4 ++--
 benchmark/tests/test_publication_scope.py | 18 ++++++++++++------
 2 files changed, 14 insertions(+), 8 deletions(-)

diff --git a/apps/presentation/site/src/lhtb-copy.json b/apps/presentation/site/src/lhtb-copy.json
index 3009fa9fcb..d7a087abf9 100644
--- a/apps/presentation/site/src/lhtb-copy.json
+++ b/apps/presentation/site/src/lhtb-copy.json
@@ -15,7 +15,7 @@
       ["04", "Task evidence", "#scores"]
     ],
     "resultEyebrow": "00 / Core result",
-    "resultTitle": "Higher mean reward than both baselines; no more strict solves than Plain.",
+    "resultTitle": "Higher mean reward than Plain and native Goal.",
     "resultBody": "LoopX averages 0.4948, Plain 0.4218, and native Goal 0.4475. At reward ≥ 0.95, they solve 7, 7, and 4 tasks respectively. Mean reward captures partial progress; full acceptance remains a separate outcome.",
     "summaryColumns": ["Execution", "Mean reward", "Solved ≥ 0.95", "Token usage", "USD"],
     "armLabels": {
@@ -170,7 +170,7 @@
       ["04", "逐题核对", "#scores"]
     ],
     "resultEyebrow": "00 / 核心结果",
-    "resultTitle": "均分高于两种基线，完整通过数尚未超过 Plain。",
+    "resultTitle": "均分高于 Plain 和原生 Goal。",
     "resultBody": "LoopX 为 0.4948，Plain 为 0.4218，原生 Goal 为 0.4475。以 Reward ≥ 0.95 计通过，三者分别为 7、7、4 题。均分反映部分进展，完整验收还需单独看。",
     "summaryColumns": ["执行方式", "平均 Reward", "通过 ≥ 0.95", "Token 用量", "USD"],
     "armLabels": {
diff --git a/benchmark/tests/test_publication_scope.py b/benchmark/tests/test_publication_scope.py
index 67fed33749..af96fbb560 100644
--- a/benchmark/tests/test_publication_scope.py
+++ b/benchmark/tests/test_publication_scope.py
@@ -85,16 +85,17 @@ def test_published_data_and_bilingual_tables_share_scope(exporters):
 def test_lhtb_published_data_and_bilingual_copy_share_scope():
     study = STUDY.parent / "LHTB" / "studies" / "five-arm-gpt56sol-max"
     data = json.loads((study / "data.json").read_text())
-    arms = {
+    all_arms = {
         "plain",
         "native_goal",
         "ssh_goal",
         "legacy_heartbeat",
         "new_heartbeat",
     }
-    assert set(data["arms"]) == arms
+    main_comparison_arms = {"plain", "native_goal", "new_heartbeat"}
+    assert set(data["arms"]) == all_arms
     assert len(data["tasks"]) == len({row["task"] for row in data["tasks"]}) == 46
-    assert all(set(row) == arms | {"task"} for row in data["tasks"])
+    assert all(set(row) == all_arms | {"task"} for row in data["tasks"])
 
     for arm, summary in data["arms"].items():
         rewards = [row[arm] for row in data["tasks"]]
@@ -105,6 +106,11 @@ def test_lhtb_published_data_and_bilingual_copy_share_scope():
     localized_copy = json.loads((site / "lhtb-copy.json").read_text())
     assert set(localized_copy) == {"en", "zh"}
     for localized in localized_copy.values():
-        assert set(localized["armLabels"]) == arms
-        assert set(localized["armKinds"]) == arms
-        assert {row[0] for row in localized["mechanismRows"]} == arms
+        assert set(localized["armLabels"]) == all_arms
+        assert set(localized["armKinds"]) == all_arms
+        assert {row[0] for row in localized["mechanismRows"]} == main_comparison_arms
+
+    assert localized_copy["en"]["resultTitle"] == "Higher mean reward than Plain and native Goal."
+    assert "7, 7, and 4 tasks" in localized_copy["en"]["resultBody"]
+    assert localized_copy["zh"]["resultTitle"] == "均分高于 Plain 和原生 Goal。"
+    assert "7、7、4 题" in localized_copy["zh"]["resultBody"]