aiqbench analysis
This research introduces FICBO, a transformer-based framework that leverages auxiliary feedback to improve black-box optimization while dynamically estimating source reliability in the presence of misleading data. By structuring priors over feedback relationships, the method enables robust query selection across diverse regimes without relying solely on historical optimization data. The work offers significant insights for applications in science and engineering where integrating expert heuristics or simulators with automated search is critical.
In-Context Black-Box Optimization with Unreliable Feedback
arXiv:2605.06187v2 Announce Type: replace-cross 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.
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.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2605.06187 [cs.LG]
(or arXiv:2605.06187v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.06187 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 268 words
Related coverage
LLM-as-a-Judge Scores Are Unreliable Optimization Signals in Closed-Loop Table Recognition
arXiv · Jul 20, 2026
Removing nodal and support-mismatch pathologies in Variational Monte Carlo via blurred sampling
arXiv · Sep 28, 2026
Response Renormalization for Critical Deep Equilibrium Models
arXiv · Aug 26, 2026
Navigating Runtime Dynamism in Semantic Query Serving
arXiv · Nov 4, 2026