PulseAugur
中
实时 15:28:12
English(EN) Self-Evolving Coding Agents via Harness-Policy Optimization

Self-HPO 框架通过优化策略和 Harness 递归地改进编码代理

研究人员推出了一种新颖的框架 Self-HPO,旨在通过交替优化代理的策略及其交互 Harness 来递归地改进编码代理。该方法允许代理通过强化学习识别并针对其自身不断变化的故障模式。在三个编码基准上的实验表明,Self-HPO 显著提升了性能,将 Qwen3.5-35B-A3B 的解决率从 57.8% 提高到 64.8%,并且在分布外设置中也显示出泛化收益。当两个组件都得到优化时,该框架的有效性进一步凸显,它能够实现复合收益,并且即使使用更强的外部模型作为 Harness 提议者也能受益。 AI

影响 该框架通过实现自我改进,可能导致更强大、更可靠的编码代理,从而加速软件开发。

排序理由 该集群描述了一个新的研究框架及其在编码基准上的性能,详细介绍在一篇论文和发布的代码/模型中。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

Self-HPO 框架通过优化策略和 Harness 递归地改进编码代理

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群描述了一个新的研究框架及其在编码基准上的性能,详细介绍在一篇论文和发布的代码/模型中。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, paper, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
18 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    通过 Harness-策略优化实现自演化编码代理

    Self-HPO is a recursive framework that alternates harness and policy optimization for coding agents, improving performance by targeting evolving failure modes through reinforcement learning.