NARCBench: Detecting AI Agent Collusion

What it is

Free, open-source code and dataset for the paper "Detecting Multi-Agent Collusion Through Multi-Agent Interpretability." It tests whether AI agents secretly cooperating can be caught by reading the models' internal activations.

Why I recommend it

A research tool, not a beginner resource. Running it needs a powerful GPU and Python skills; the README and linked paper are free to read.

Topics

#ai ethics #ai safety #benchmark #interpretability #multi-agent systems #open source #research

Added Sep 23, 2026 · 0 opens