Google DeepMind 提出 Evo2 框架实现智能体 harness 与权重协同进化
Google DeepMind 提出 Evo2 框架，让智能体的 harness 与模型权重协同进化：一个策略同时担任 Proposer（改写自身 harness）和 Reflector（诊断每次修改的效果），这些修改及结果再作为训练数据，经 SFT 后接交替 GRPO 训练两个角色。
Banger paper from Google DeepMind on co-evolving an agent's harness and its weights.
(bookmark it)
If you already log your harness changes and results, those logs can become training data for the model running inside the harness.
This is a nice paper from Google showing how to do this co-evolution.
Specifically, they propose Evo2, a framework for weight-harness co-evolution.
How it works:
One policy plays two roles. As a Proposer, it writes edits to its own harness, and as a Reflector, it diagnoses what each edit changed. Those edits and their outcomes then become training data for both roles, using SFT followed by interleaved GRPO.
The result:
Evo2 reaches 48.3% on SWE-bench Verified and 82.1% on Polyglot, against 41.0% and 62.6% for the strongest baselines. The harness part alone moves a GPT-5-mini agent from 24.4% to 48.3% on SWE-bench Verified.
The evolved harnesses also transfer to other benchmarks without further evolution.
Paper: arxiv.org/abs/2610.13025
Chat with Paper: academy.dair.ai/papers/evo-2…
来源：omarsar0 · x.com