Pinned Loading

1. emmy emmy Public

   Evaluation-invariant measurement for multi-agent AI systems.

   Python

2. llms-en-garde llms-en-garde Public

   Do LLMs misbehave less when they're en garde?

   Shell

3. pants-on-fire-eval pants-on-fire-eval Public

   Is the model lying or just wrong? Decomposing a deliberative alignment anti-scheming spec.

   Python

4. activation-tomography activation-tomography Public

   Natural language autoencoders as measurement instruments for AI safety applications. Research fork of kitft/natural_language_autoencoders.

   Python

5. paper-chase paper-chase Public

   multi-agent simulation of a scientific publishing ecosystem

   Python

6. awesome-agent-sandboxes awesome-agent-sandboxes Public

   Comprehensive list of sandboxing options for AI agents + detailed sandboxing guide + analysis of AI safety research specific concerns/solutions. nb: intermittent updates post-2026.05

   Python 3