2 papers
cs.CL2026
EvoGuard: Mitigating Degradation in Online Skill Evolution
Yanli Wang, Zichen Bai, Tianyue Jiang +12
Skills let language agents reuse experience across tasks as natural-language documents. Recent work evolves skills from interaction, but existing methods have limitations when skil…
cs.CL2026
LiSAO: Lifting Light Single-Rollout Asynchronous RL Optimization via Group-Relative Actor-Critic Bootstrapping
Zichen Bai, Yanli Wang, Yang Liu +9
Asynchronous reinforcement learning (RL) is becoming a mainstream paradigm for post-training large language models (LLMs), as it improves training efficiency by updating the model…