arXiv cs.CL· Hao Chen, Ziyang Ling, Ronald X. Xu, Mingzhai Sun·· 2 天前AI 评分22
FGSS-BO：面向 LLM 贝叶斯优化的反馈引导假设管理
FGSS-BO: Feedback-Guided Hypothesis Management for LLM-Based Bayesian Optimization
AI 导读
研究者提出 Feedback-Guided Search State（FGSS），一种将语言引导与观测证据、可执行参数干预及对比支持更新连接起来的结构化假设管理接口，由 LLM reflector 提出并修订假设、确定性 curator 验证并维护其身份与测试记录。
正文 · 原文
Abstract:Large language model (LLM)-guided Bayesian optimization (BO) uses past observations to guide search, but proposed directions and their revision can remain implicit. We propose Feedback-Guided Search State (FGSS), a structured hypothesis-management interface connecting language-based guidance to observed evidence, executable parameter interventions, and comparative support updates. An LLM reflector proposes and revises hypotheses; a deterministic curator validates them and maintains identities and test records. The evolving state guides candidate construction and priorities. Instantiated on LLAMBO, FGSS-BO retains its surrogate and default expected improvement selection, with a bounded testing policy. On two retrospective protocol/media tasks with nearest-neighbor response oracles, it attains higher sample mean final best and search AUC than the matched LLAMBO baseline and evaluated non-LLM baselines. Prose-reflection and EI-only controls show mixed component effects. Six direct-training hyperparameter-optimization tasks show task-dependent transfer, strongest on diabetes SVM; exact-grid and reaction controls examine other settings. FGSS exposes hypotheses, their execution, and cross-round revisions for inspection while supporting search gains in selected settings.
| Comments: | 21 pages, 8 figures |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.21246 [cs.CL] |
| (or arXiv:2610.21246v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.21246 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hao Chen [view email]
[v1]
Thu, 15 Oct 2026 04:42:42 UTC (224 KB)
来源：arXiv cs.CL · arxiv.org