AI RESEARCH
SEAL: Synergistic Co-Evolution of Agents and Learning Environments
arXiv CS.CL
•
ArXi:2605.24426v1 Announce Type: new Large Language Model (LLM) agents are increasingly improved through interaction, yet most self-evolution methods adapt either the policy or the learning environment in isolation. We identify this structural gap as \emph{Agent-Environment Misalignment}: the agent's capability frontier changes during