Show notes
🤗 Upvotes: 31 | cs.CL, cs.AI
作者:
Zhepei Wei, Wenlin Yao, Baolin Peng, Danqing Wang, Qianhui Wu, Hao Cheng, Yu Meng, Jianfeng Gao
标题:
Self-Evolving Coding Agents via Harness-Policy Optimization
Arxiv:
http://arxiv.org/abs/2610.22791v1
摘要:
Large language model (LLM) agents have shown promising potential in coding tasks, and their performance is jointly determined by the policy model and the harness that mediates their interaction with the environment. Indeed, harness and policy are mutually reinforcing and evolving only one component is suboptimal---a harness tuned for one checkpoint…
Website: share.transistor.fm

