arXiv cs.CL· Ziyi Zhu, Luka Smyth, Saki Shinoda, Jinghong Chen·· 12 小时前AI 评分38
SAGE：通过智能体引导探索实现随机提示词优化
SAGE: Stochastic Prompt Optimization via Agent-Guided Exploration
AI 导读
研究提出 SPO（随机提示词优化）框架，将自动提示词优化视为黑盒搜索，并对比误差引导随机搜索、遗传算法与多智能体流水线 SAGE 三种策略。在三个基准上无单一策略占优，效果取决于搜索空间结构与错误类型的交互。SAGE 在心理健康聊天机器人上连续优化八轮 A/B 测试，将单次噪声结果累积为次日留存率的统计显著提升。
正文 · 原文
Abstract:Context engineering has emerged as a primary lever for improving AI systems without parameter updates. Recent work showing that textual gradients do not function as real gradients motivates treating automatic prompt optimization (APO) as black-box search. We introduce SPO (Stochastic Prompt Optimization), a framework for stochastic search over prompt space, and compare three strategies of increasing sophistication: error-informed random search, a genetic algorithm with evolutionary operators, and SAGE (SPO via Agent-Guided Exploration), a multi-agent pipeline with diagnostic code execution. Across three benchmarks, no single strategy dominates; effectiveness depends on the interaction of landscape structure with error type. We further deploy SAGE on a mental-health chatbot under a continuous optimization paradigm, where it compounds eight cycles of individually-noisy A/B tests into a statistically robust gain in next-day retention. We argue that coupling qualitative diagnosis with quantitative validation is what makes agentic optimization effective for open-ended task-oriented dialogue.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.18902 [cs.CL] |
| (or arXiv:2606.18902v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.18902 arXiv-issued DOI via DataCite |
|
| Journal reference: | Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track, pages 3695-3711, Budapest, Hungary, October 2026 |
| Related DOI: | https://doi.org/10.18653/2026.emnlp-industry.221
DOI(s) linking to related resources |
Submission history
From: Ziyi Zhu [view email]
[v1]
Wed, 17 Jun 2026 10:25:25 UTC (739 KB)
[v2]
Tue, 28 Jul 2026 13:10:27 UTC (745 KB)
[v3]
Tue, 3 Nov 2026 21:32:56 UTC (749 KB)
来源：arXiv cs.CL · arxiv.org