| title | Retrospective Harness Optimization: Improving LLM Agents via Self-Preference over Trajectory Rollouts |
|---|---|
| authors | Wenbo Pan, Shujie Liu, Chin-Yew Lin, Jingying Zeng, Xianfeng Tang, Xiangyang Zhou, Yan Lu, Xiaohua Jia |
| date | 2026-06-04 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.05922 |
| affiliation | Microsoft Research |
| upvotes | 36 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | AI agents rely on a harness of skills, tools, and workflows to solve complex problems. Continually improving this harness is essential for adapting to new tasks. However, existing optimization methods typically require ground-truth validation sets, yet such labeled data is difficult to acquire in practical deployment settings. To address this problem, we introduce Retrospective Harness Optimization (RHO), a self-supervised method that optimizes the agent harness using only past trajectories. Specifically, RHO selects a diverse coreset of challenging tasks from past trajectories and re-solves them in parallel. The agent analyzes these rollouts using self-validation and self-consistency, then generates candidate harness updates and selects the most effective one by its own pairwise self-preference. We evaluate RHO across three diverse domains, spanning software engineering, technical work, and knowledge work. Notably, a single optimization round improves the pass rate on SWE-Bench Pro from 59% to 78% without any external grading. Furthermore, our analysis demonstrates that RHO effectively targets prior failure modes. As a result, the optimized harness alters the agent's behavior patterns and sustains higher accuracy during long-horizon sessions. |
cite
@misc{pan2026retrospective,
title={Retrospective Harness Optimization: Improving LLM Agents via Self-Preference over Trajectory Rollouts},
author={Wenbo Pan and Shujie Liu and Chin-Yew Lin and Jingying Zeng and Xianfeng Tang and Xiangyang Zhou and Yan Lu and Xiaohua Jia},
year={2026},
eprint={2606.05922},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.05922}
}