1 paper
Jinyuan Feng, Dongmin Li, Yiqun Chen +4
Skill libraries improve the sample efficiency of agentic reinforcement learning (RL) by enabling large language model (LLM) agents to reuse procedural knowledge. Yet existing parad…