Prism: Policy Reuse via Interpretable Strategy Mapping in Reinforcement Learning

ArXi:2604.02353v1 Announce Type: cross We present PRISM (Policy Reuse via Interpretable Strategy Mapping), a framework that grounds reinforcement learning agents' decisions in discrete, causally validated concepts and uses those concepts as a zero-shot transfer interface between agents trained with different algorithms. PRISM clusters each agent's encoder features into $K$ concepts via K-means.