Last released Oct 16, 2023
Tools for applying circuits-style interpretability techniques to RL agents.
Supported by