arXiv cs.AI· Sudipta Paul, Pengrui Quan, Aounon Kumar, Nicholas Tzou, Vivek Kulkarni, Srinivas Chappidi·· 13 小时前AI 评分34
HGM-Dual：面向长时程规划的谱系感知引导式 Agent Harness 优化
HGM-Dual: Lineage-Aware Guided Agent Harness Optimization for Long-Horizon Planning
AI 导读
HGM-Dual 是一个面向规划智能体的诊断引导式 harness 进化框架，在案例准确率上比最强基线最高相对提升 14.45%，并以 2x-4x 更少的评估预算达到 HGM 变体的最终性能。
正文 · 原文
Abstract:Harness optimization, which improves an agent's prompts, tools, and workflows, is an increasingly effective way to enhance agents built on frozen or closed-source LLMs. Leading approaches such as the Darwin--Godel Machine and the Huxley--Godel Machine (HGM) rely on recursive self-improvement and were developed mainly for coding, where the competence needed to improve the agent coincides with the competence being improved. Planning tasks, in which an agent must produce a sequence of tool calls that jointly satisfies interacting constraints, break this alignment. There, open-ended self-modification repeatedly explores ineffective changes and consumes substantial evaluation budgets. We introduce HGM-Dual, a diagnostic-guided framework for budget-efficient harness evolution for planning agents. First, HGM-Dual represents agent evolution as lineage-aware context. This context records what each ancestor changed, which behaviors improved or regressed, and which modifications were ineffective, giving the editor explicit edit-outcome attribution instead of an unstructured archive of past agents. Second, HGM-Dual introduces error-conditioned dual-path child generation. For each selected parent, it generates a general child that preserves open-ended exploration and specialized children that target the harness components most likely responsible for the parent's highest-ranked constraint violations. HGM-Dual outperforms the strongest baseline by up to 14.45% relative improvement in case accuracy while matching HGM variants' final performance with 2x-4x fewer evaluation budget.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE) |
| Cite as: | arXiv:2610.30042 [cs.AI] |
| (or arXiv:2610.30042v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.30042 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Pengrui Quan [view email]
[v1]
Wed, 21 Oct 2026 17:02:28 UTC (208 KB)
来源：arXiv cs.AI · arxiv.org