协同进化替代反向传播，资源受限环境下多轮工具型智能体的全参数后训练更省内存，性能不打折
arXiv:2608.02391v1 Announce Type: new Abstract: Tool-using large language model (LLM) agents produce long, multi-turn trajectories, making gradient-ba…
LLM智能体新范式：让模型权重与“缰绳”协同进化，提升适应性与性能。
arXiv:2607.22688v1 Announce Type: cross Abstract: Post-training agents for automated AI research requires optimizing not only model parameters, but al…
LLM通过技能自我博弈与协同进化突破能力天花板，实现模型能力的自主提升与前沿拓展。
arXiv:2607.22529v1 Announce Type: new Abstract: LLM training is shifting from manual design and annotation to interaction-driven self-evolution. Howev…
机器人扩散策略训练新思路，先验映射协同进化提升性能，ICML 2026 收录值得关注
arXiv:2512.02581v3 Announce Type: replace Abstract: Reinforcement learning (RL) faces a persistent tension: policies that are stable to optimize (e.g.…
传统奖励只看成败，这套方法用自适应评分标准给多步智能体每一步“打分讲理”，可解释强化学习的新思路。
arXiv:2606.21262v1 Announce Type: new Abstract: Reinforcement learning for multi-step LLM agents often relies on scalar rewards that indicate success …
LLM代理自我进化新突破：协同策略与环境共同进化，解决代理-环境错配难题
arXiv:2605.24426v1 Announce Type: new Abstract: Large Language Model (LLM) agents are increasingly improved through interaction, yet most self-evoluti…
用大语言模型自动化处理元模型与语法的共同演化，告别繁琐人工，挑战复杂场景下的规则局限。
arXiv:2605.21465v1 Announce Type: new Abstract: In model-driven engineering, metamodel evolution leads to the need to adapt corresponding grammars to …