Implementation of FoldMark: Safeguarding Protein Structure Generative Models with Distributional and Evolutionary Watermarking
-
Updated
Apr 19, 2026 - Python
Implementation of FoldMark: Safeguarding Protein Structure Generative Models with Distributional and Evolutionary Watermarking
Safety concept for Global Health Engineering
Biological dual-use content classifier using Constitutional Classifiers methodology — biosafety constitution, synthetic data pipeline, DeBERTa-v3-base classifier
Mapping the distribution of high-containment laboratories and projected climate-related hazards
A deterministic, fail-closed cryptographic firewall for AI-controlled labs that prevents dangerous hallucinations and prompt injections by validating DNA, peptide, and chemical synthesis requests through Ed25519 signatures, PCA-chain authority narrowing, and rigorous biological invariant screening.
Empirical dual-use risk assessment of protein language models (ESM-2) and structure-based design tools (ProteinMPNN)
Public runtime library for microorganism risk-reference data derived from NITE/MRINDA.
30 machine-readable constitutional rules for biological dual-use content across 6 bio domains. JSON format for Constitutional Classifier pipeline integration.
Add a description, image, and links to the biosafety topic page so that developers can more easily learn about it.
To associate your repository with the biosafety topic, visit your repo's landing page and select "manage topics."