arXiv cs.AI· Nicolas Samuel Blumer, Julien Martinelli, Samuel Kaski·· 11 小时前AI 评分27
面向不可靠反馈的上下文内黑盒优化（FICBO）
In-Context Black-Box Optimization with Unreliable Feedback
AI 导读
研究者提出 FICBO，一种在上下文内黑盒优化中同时利用优化历史与廉价辅助反馈的预训练优化器。该方法引入结构化先验刻画辅助反馈与目标函数的关系，并据此预训练一个反馈感知 Transformer，覆盖从高信息量到误导性反馈的多种情形。测试时模型通过对比观测目标值与辅助信号在上下文中估计来源可靠性，在合成与真实任务上既能利用有效反馈，也对弱或误导性来源保持稳健。
正文 · 原文
Abstract:Black-box optimization in science and engineering often comes with side information: experts, simulators, pretrained predictors, or heuristics can suggest promising candidates. Such feedback can accelerate search, but can also be biased, input-dependent, or misleading. Feedback-aware Bayesian optimization methods typically handle tasks independently, limiting adaptation across feedback sources. In-context optimizers enable cross-task adaptation, but usually rely only on optimization history at test time. We study feedback-informed in-context black-box optimization (FICBO), where a pretrained optimizer conditions on both the observed history and cheap auxiliary feedback for the current candidate set. We introduce structured priors over how auxiliary feedback relates to the objective, and use them to pretrain a feedback-aware transformer across regimes ranging from highly informative to misleading. At test time, the model estimates source reliability in context by comparing observed objective values with auxiliary signals, improving query selection. On synthetic and real-world tasks, FICBO effectively exploits informative feedback while remaining robust to weak or misleading sources. Empirical investigations further illustrate how the model perceives test-time sources, offering insights into its interpretability and decision-making process.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.06187 [cs.LG] |
| (or arXiv:2605.06187v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.06187 arXiv-issued DOI via DataCite |
Submission history
From: Nicolas Blumer [view email]
[v1]
Thu, 7 May 2026 13:03:19 UTC (1,425 KB)
[v2]
Wed, 4 Nov 2026 11:56:06 UTC (1,977 KB)
来源：arXiv cs.AI · arxiv.org