bayesian-optimization

The paper proposes AgenticTuner, an agentic pipeline using an LLM as the optimizer to calibrate cosmological simulations by leveraging human-readable priors like codebases and documentation, achieving competitive calibration performance at a fraction of the CPU cost of Bayesian optimization.

The paper proposes CAP (Candidate Acquisition Policy), a method that learns a feedback-conditioned query policy from sparse records of related black-box optimization tasks, achieving the best final-budget results on four synthetic families and two engineering benchmarks (BOPTEST, SU2).

提出 acquisition function-guided generation (AFGG) 方法，通过将采集函数倾斜的分布嵌入扩散桥，在离散空间上结合预训练生成模型与数据空间代理模型进行贝叶斯优化；在分子设计基准上，基于流的变体在多数场景中超越现有最优方法。

This paper evaluates whether an LLM can serve as a calibration agent for expensive engineering simulators, finding that physics-labeled prompts cut early calibration error 22-fold but lose their advantage after about ten runs, so a hybrid warm-start plus numerical optimizer workflow reaches target accuracy in 9 runs versus 17 for Nelder-Mead.

提出 FGSS-BO，一种基于反馈引导搜索状态（FGSS）的结构化假设管理框架，将 LLM 的语言引导与可观测证据和可执行参数干预相连接，用于贝叶斯优化，在 LLAMBO 基准上取得更高的搜索性能。

提出了 ELF-BO 算法，利用目标函数评估的等待时间提前从超参数后验中采样，仅在观测到目标值后重新加权，从而以标准 BO 的决策延迟实现完全贝叶斯优化。在合成函数和真实应用中，ELF-BO 匹配完全贝叶斯方法的性能且延迟更低。

该论文提出一种车辆集成的数字孪生框架，通过驱动感知（drive-by sensing）实现桥梁结构健康监测，结合傅里叶神经算子代理模型、贝叶斯优化的传感器设计以及无监督损伤评估，并在澳大利亚和日本的多站点现场试验中完成验证。

Google Cloud AI Research researchers introduced AIM (Agentic Idea Management), a framework that organizes research ideas into clusters, ranks their promise, and schedules explore/exploit experiments with a solution auditor and resource planner, reporting gains over ScientistOne on AutoLab tasks.

Google Cloud AI Research introduces AIM (Agentic Idea Manager), an autonomous framework for idea-driven automated research that uses an Agentic Surrogate and Agentic Acquisition (inspired by Bayesian optimization) to organize and select research ideas, plus a Solution Auditor and Resource Planner to maintain idea-solution alignment and allocate experimental budgets. AIM outperforms the strongest AutoLab baseline by up to 4.9 points and reaches baseline performance up to 3.1x faster.

This paper introduces a hierarchical Bayesian optimization framework for optimizing aircraft-based multi-agent systems-of-systems, applied to wildfire suppression, improving search efficiency and robustness in complex design spaces.

This paper introduces HACK GPs, a method that uses online learning with expert advice for kernel selection in Gaussian Processes, enhancing robustness in sequential decision-making tasks such as Bayesian optimization and active learning.

The paper proposes a fast Bayesian optimization method for de novo discovery by leveraging linear models in latent spaces, achieving over 100x speedup over existing methods while maintaining performance.

This paper presents methods to integrate rich auxiliary information into Bayesian Optimization using LLMs, showing superior performance over standard and existing LLM-based approaches in benchmarks and real-world tasks like nuclear fusion.

This technical note presents a decision-focused active learning approach for optimizing critical materials recovery, showing that adaptive policies can achieve maximum enrichment with fewer experiments than nonadaptive methods in autonomous laboratory workflows.

The paper proposes Finite-Library Input-Warped Bayesian Optimization (FLIWBO), which selects input warps from a finite library to adapt input geometry for improved optimization efficiency while maintaining high-probability convergence guarantees, outperforming standard GP-UCB in various benchmarks.

This paper proposes SBOCF, a scalable Bayesian optimization method for image-based inverse problems in materials characterization, demonstrating improved efficiency and accuracy in parameter estimation from electron microscopy data.

This paper proposes a calibration method for product-of-experts Gaussian process models to correct overestimated posterior variances, enhancing uncertainty quantification and scalability for large datasets.

This paper introduces GRAPE, a two-stage framework for query-efficient high-dimensional Bayesian optimization that refines gradient posteriors and selects update directions based on expected progress, demonstrating superior performance in adversarial attacks and large language model prompt optimization.

This paper introduces TTARO, an online deep-kernel Bayesian optimization framework that adapts circuit representations at test time using evaluated figure-of-merit labels. It improves sample efficiency for analog circuit topology search, reducing regret AUC by 15.2% over standard BO and 20.7% over fixed deep kernel learning.

This paper presents SABLE, an open-source agentic framework that uses natural-language orchestration with LLMs to guide synthetically constrained multi-objective hit-to-lead optimization in drug discovery, integrating reaction-templated enumeration, property prediction, and Bayesian optimization.