Automated Discovery Has No Universally Superior Harness
- source_url: https://arxiv.org/abs/2607.18235
- source_type: paper
- platform: arxiv
- author: Akshat Gupta, Jermaine Lei, Alexander Lu, Gopala Anumanchipalli, Leshem Choshen
- original_date: 2026-07-20
- added_date: 2026-07-22
- arxiv_id: 2607.18235
- arxiv_categories: cs.CL, cs.AI
- pdf_url: https://arxiv.org/pdf/2607.18235v1
- category: agents
- tags: agent-harness, automated-discovery, openevolve, hyperparameter, llm-agents, arxiv
- quality_score: 5
摘要（中文）
系统分解 OpenEvolve/TTT-Discover 类自动发现 harness，在 12 组模型-问题对超 310 万次 LLM rollout 上评估 30 种预算对齐变体，发现没有固定 harness 能跨设置稳定领先，OpenEvolve 变体整体常弱于更简单方案；harness 应视为随问题与模型调整的超参，并用早期进度做在线自适应预算分配收录理由：直接挑战通用最优 harness叙事，为 agent 搜索脚手架选型提供可复用统计基线与工程结论
Summary (English)
Autonomous discovery systems such as OpenEvolve and TTT-Discover are often used as general-purpose harnesses. However, in practice these are composite systems combining several design choices about archives, parent selection, exploration, and budget allocation into a single recipe. Because discovery runs are expensive and inherently stochastic, existing harnesses are often compared using too few independent trials to distinguish key methodological improvements from run-to-run variance. We systematically decompose OpenEvolve-style evolutionary search and the TTT-Discover search harness into its constituent components and systematically evaluate 30 budget-matched harnesses across 12 model-problem pairs using more than 3.1 million LLM rollouts and repeated-trial statistical analysis. Our results show that discovery harnesses have a generalization problem: No fixed harness is reliably superior across the evaluated model-problem pairs, and variants of OpenEvolve generally underperform simpler alternatives. Thus, harness choice is better viewed as a hyperparameter rather than as a universal recipe, and should be tailored to the specific problem and underlying model. We also find that early discovery progress predicts final performance, and use this property to present a budget-matched adaptive-allocation experiment that starts multiple harnesses, prunes weak partial runs, and reallocates compute to stronger survivors, outperforming both commitment to a randomly sampled fixed harness and a non-adaptive harness ensemble. Together, these results motivate shifting from fixed harness selection to online adaptation guided by early performance. We release all run pools including baseline null distributions for every model-problem pair as reusable statistical infrastructure against for future harness proposals.
One-liner
系统分解 OpenEvolve/TTT-Discover 类自动发现 harness，在 12 组模型-问题对超 310 万次 LLM rollout 上评估 30 种预算对齐变体，发现没有固定 harness 能跨设置稳定领先，OpenEvolve 变体整体常弱于更简单方案.
原文 / 元数据抓取
Automated Discovery Has No Universally Superior Harness
作者: Akshat Gupta, Jermaine Lei, Alexander Lu, Gopala Anumanchipalli, Leshem Choshen
原文链接: https://arxiv.org/abs/2607.18235
PDF: https://arxiv.org/pdf/2607.18235v1
发布时间: 2026-07-20
更新时间: 2026-07-20
分类: cs.CL, cs.AI
Autonomous discovery systems such as OpenEvolve and TTT-Discover are often used as general-purpose harnesses. However, in practice these are composite systems combining several design choices about archives, parent selection, exploration, and budget allocation into a single recipe. Because discovery runs are expensive and inherently stochastic, existing harnesses are often compared using too few independent trials to distinguish key methodological improvements from run-to-run variance. We systematically decompose OpenEvolve-style evolutionary search and the TTT-Discover search harness into its constituent components and systematically evaluate 30 budget-matched harnesses across 12 model-problem pairs using more than 3.1 million LLM rollouts and repeated-trial statistical analysis. Our results show that discovery harnesses have a generalization problem: No fixed harness is reliably superior across the evaluated model-problem pairs, and variants of OpenEvolve generally underperform simpler alternatives. Thus, harness choice is better viewed as a hyperparameter rather than as a universal recipe, and should be tailored to the specific problem and underlying model. We also find that early discovery progress predicts final performance, and use this property to present a budget-matched adaptive-allocation experiment that starts multiple harnesses, prunes weak partial runs, and reallocates compute to stronger survivors, outperforming both commitment to a randomly sampled fixed harness and a non-adaptive harness ensemble. Together, these results motivate shifting from fixed harness selection to online adaptation guided by early performance. We release all run pools including baseline null distributions for every model-problem pair as reusable statistical infrastructure against for future harness proposals.
Obsidian intake evidence excerpt
该内容文件由 AAIF content-fetcher 根据 active/high-score entry 与 OpenCLI arXiv 元数据补齐。
- entry_id: 572f6d48
- title: Automated Discovery Has No Universally Superior Harness
- source: https://arxiv.org/abs/2607.18235
- existing_summary_zh: 系统分解 OpenEvolve/TTT-Discover 类自动发现 harness，在 12 组模型-问题对超 310 万次 LLM rollout 上评估 30 种预算对齐变体，发现没有固定 harness 能跨设置稳定领先，OpenEvolve 变体整体常弱于更简单方案；harness 应视为随问题与模型调整的超参，并用早期进度做在线自适应预算分配收录理由：直接挑战通用最优 harness叙事，为 agent 搜索脚手架选型提供可复用统计基线与工程结论