mlsanityData to 14 Oct 2026

Agentic High-Dimensional Bayesian Optimization with Hypothesis- and Evidence-Guided Search

arXiv 2609.34281 ↗First posted 28 Sep 2026cs.LGcited by 0

On the map: LLM agents and tool use → Web, search, and research agents

Abstract (arXiv)

High-dimensional Bayesian optimization (HDBO) seeks sample-efficient optimization when the number of variables is large relative to the evaluation budget. Recent LLM-based and agentic BO methods incorporate task knowledge and adapt search decisions during a run, but have primarily been evaluated on low- and moderate-dimensional problems. We ask whether this paradigm can transfer to the higher-dimensional regime. Our experiments show that these methods do not remain reliable in the high-dimensional regime, where the challenge is not only where to evaluate, but also which modeling assumption and search geometry to use when the objective's useful structure is unknown. We therefore introduce HERA, a Hypothesis- and Evidence-guided Research Agent that uses task context, optimization feedback, and structural diagnostics to revise search hypotheses, select and configure HDBO strategies, and determine their execution length. PRISM, its numerical optimization engine, generates and evaluates candidates sequentially within each search block, updating numerical models after each observation. HERA remains competitive with strong numerical HDBO baselines and outperforms the evaluated LLM-based and agentic methods on four metadata-free synthetic functions. Across eight real-world tasks, HERA achieves the best mean final objective among all evaluated systems on most benchmarks. Further analyses show that structural diagnostics change strategy use, metadata effects vary across tasks, and adaptive search blocks reduce inference cost.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use. Builds on 42 of its 50 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

42 of 50 references.

References · 42

  1. 2023Large Language Models as Optimizerscited by 864
  2. 2018A Tutorial on Bayesian Optimizationcited by 744
  3. 2019Scalable Global Optimization via Local Bayesian Optimizationcited by 374
  4. 2018GuacaMol: Benchmarking Models for De Novo Molecular Designcited by 297
  5. 2013Bayesian Optimization in a Billion Dimensions via Random Embeddingscited by 182
  6. 2019Benchmarking Model-Based Reinforcement Learningcited by 162
  7. 2021High-Dimensional Bayesian Optimization with Sparse Axis-Aligned Subspacescited by 141
  8. 2015High Dimensional Bayesian Optimisation and Bandits via Additive Modelscited by 139
  9. 2024Large Language Models to Enhance Bayesian Optimizationcited by 135
  10. 2017Batched Large-scale Bayesian Optimization in High-dimensional Spacescited by 117
  11. 2020Re-Examining Linear Embeddings for High-Dimensional Bayesian Optimizationcited by 92
  12. 2024Vanilla Bayesian Optimization Performs Great in High Dimensionscited by 92
30 more
  1. 2020Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Searchcited by 89
  2. 2024Large Language Model Agent for Hyper-Parameter Optimizationcited by 64
  3. 2018High-Dimensional Bayesian Optimization via Additive Models with Overlapping Groupscited by 59
  4. 2023Increasing the Scope as You Learn: Adaptive Bayesian Optimization in Nested Subspacescited by 52
  5. 2018High Dimensional Bayesian Optimization Using Dropoutcited by 40
  6. 2022Monte Carlo Tree Search based Variable Selection for High Dimensional Bayesian Optimizationcited by 37
  7. 2023Are Random Decompositions all we need in High Dimensional Bayesian Optimisation?cited by 36
  8. 2024Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimizationcited by 34
  9. 2023Looping in the Human Collaborative and Explainable Bayesian Optimizationcited by 27
  10. 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequencescited by 23
  11. 2020High-Dimensional Bayesian Optimization via Tree-Structured Additive Modelscited by 22
  12. 2024Position: Leverage Foundational Models for Black-Box Optimizationcited by 21
  13. 2025Understanding High-Dimensional Bayesian Optimizationcited by 20
  14. 2025Adaptive Kernel Design for Bayesian Optimization Is a Piece of CAKE with LLMscited by 17
  15. 2019Sequential Learning of Active Subspacescited by 16
  16. 2025Language-Based Bayesian Optimization Research Assistant (BORA)cited by 15
  17. 2026Can LLMs Beat Classical Hyperparameter Optimization Algorithms? A Study on autoresearchcited by 15
  18. 2025LLINBO: Trustworthy LLM-in-the-Loop Bayesian Optimizationcited by 14
  19. 2025Exploring Exploration in Bayesian Optimizationcited by 9
  20. 2026Adaptive Acquisition Selection for Bayesian Optimization with Large Language Modelscited by 9
  21. 2024Language Model Embeddings Can Be Sufficient for Bayesian Optimizationcited by 8
  22. 2025Reasoning BO: Enhancing Bayesian Optimization with Long-Context Reasoning Power of LLMscited by 8
  23. 2025Improving LLM-based Global Optimization with Search Space Partitioningcited by 6
  24. 2026LABO: LLM-Accelerated Bayesian Optimization through Broad Exploration and Selective Experimentationcited by 6
  25. 2026Agentic Bayesian Optimization through Surrogate-Augmented Autoresearchcited by 6
  26. 2025BBOPlace-Bench: Benchmarking Black-Box Optimization for Chip Placementcited by 5
  27. 2025High Dimensional Bayesian Optimization using Lasso Variable Selectioncited by 4
  28. 2025OptiMindTune: A Multi-Agent Framework for Intelligent Hyperparameter Optimizationcited by 3
  29. 2025LILO: Bayesian Optimization with Natural Language Feedbackcited by 3
  30. 2026Embedding by Elicitation: Dynamic Representations for Bayesian Optimization of System Promptscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Posterior Inference with Diffusion Models for High-dimensional Black-box Optimization
  2. 2026Unleashing LLMs in Bayesian Optimization: Preference-Guided Framework for Scientific Discovery
  3. 2026AgentHPOBench: A Benchmark For Evaluating LLM Agents as Sequential Hyperparameter Optimizers
  4. 2026TagBO: LLM-Driven Task-Aware Graph Bayesian Optimization for Scientific Discovery
  5. 2026Belief-Calibrated Optimization: An Explicit World Model for Agentic Optimization
  6. 2026Hypothesis-Driven Skill Optimization for LLM Agents
  7. 2026ASAP: Agent-System Co-Design for Wall-Clock-Centered Auto HPO Research for ML Experiments
  8. 2026Agentic-MatriBO: Capability-Compiled Agentic Bayesian Optimization for Materials and Chemical Design