用一个智能体自主优化提示词、代码和机器学习流程，效果优于专用系统。
The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflows
- 智能体通过自我推理决定评估、诊断、修改和验证时机。
- 在14个任务中提升2%至40%，部分超越人类最佳结果。
- 统一框架适配多种任务，无需额外控制器设计。
近期优化提示词、程序和机器学习工作流的系统通常依赖显式的外层控制器，如进化搜索、多臂赌博机或文本梯度方法。我们提出一个根本性问题：这些搜索策略有多大比例可由单一工具使用型智能体内化？本文提出ReASearch，一种基于推理驱动优化的统一框架，使智能体能自主决定评估内容、故障诊断方式、修改操作及验证或重启时机。该智能体不仅作为受人工启发规则引导的生成器，更可分析结果、分配预算，并通过持久记忆在长周期中持续优化策略。借助共享智能体循环与领域特定工具，ReASearch以完全相同的架构优化提示词、程序和机器学习工作流。在14项不同任务中，其性能媲美且多数优于专用优化系统，在某些情况下实现较强基线2%至40%的提升，并发现改进先验人类最优解的新方案。关键发现是：复杂搜索行为——传统上需显式控制器实现——可自然从智能体的推理过程中涌现。
原文摘要 · Abstract (English)
Recent systems for optimizing prompts, programs, and ML workflows typically rely on explicit outer-loop controllers such as evolutionary search, bandits, or textual-gradient methods. We ask a fundamentally different question: how much of this search policy can be internalized by a single tool-using agent? We present ReASearch, a unified framework for reasoning-driven optimization in which the agent autonomously decides what to evaluate, how to diagnose failures, which edits to make, and when to verify or restart. Rather than serving only as a proposal generator guided by hand-designed heuristics, the agent actively analyzes outcomes, allocates budget, and refines its strategy over long horizons through persistent memory. With a shared agent loop and domain-specific tools, ReASearch instantiates the exact same scaffold to optimize prompts, programs, and ML workflows. Across 14 diverse tasks, it is competitive with and mostly better than specialized optimization systems, achieving gains of 2% to 40% over strong domain-specific baselines, and in some cases discovering solutions that improve on prior human best-known results. Crucially, we observe that complex search behaviors, which are typically implemented by explicit controllers, emerge naturally from the agent's reasoning process.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。