Research map / Probabilistic and scientific modeling / Bayesian optimization
Local and high-dimensional Bayesian optimization
Make Bayesian optimization scale with local search, structure, and analyses of high-dimensional behavior.
90 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Bayesian optimization, keyword cluster “acquisition · bayesian · multi objective”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
15 papers in 2023–24 and 37 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 16.3% to 27.6%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Gaussian Process Planning with Lipschitz Continuous Reward Functions: Towards Unifying Bayesian Optimization, Active Learning, and Beyond (AAAI 2016 · 53 citations)
- 2017 · Max-value Entropy Search for Efficient Bayesian Optimization (ICML 2017 · 189 citations)
- 2018 · BOCK : Bayesian Optimization with Cylindrical Kernels (ICML 2018 · 73 citations)
- 2019 · BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning (NeurIPS 2019 · 294 citations)
- 2020 · BoTorch: A Framework for Efficient Monte-Carlo Bayesian Optimization (NeurIPS 2020 · 201 citations)
- 2021 · Interpretable Neural Architecture Search via Bayesian Optimisation with Weisfeiler-Lehman Kernels (ICLR 2021 · 16 citations)
- 2022 · Accelerating Bayesian Optimization for Biological Sequence Design with Denoising Autoencoders (ICML 2022 · 22 citations)
- 2023 · Bayesian Optimization-Based Combinatorial Assignment (AAAI 2023 · 11 citations)
- 2024 · Constrained Bayesian Optimization under Partial Observations: Balanced Improvements and Provable Convergence (AAAI 2024 · 15 citations)
- 2025 · Decomposed Quadratization: Efficient QUBO Formulation for Learning Bayesian Network (AAAI 2025 · 3 citations)
- 2026 · A Genetic Algorithm for Navigating Synthesizable Molecular Spaces (ICLR 2026 · 1 citation)
Earliest papers here
- Gaussian Process Planning with Lipschitz Continuous Reward Functions: Towards Unifying Bayesian Optimization, Active Learning, and Beyond (AAAI 2016 · 53 citations)
- Bayesian optimization for automated model selection (NeurIPS 2016 · 51 citations)
- BASC: Applying Bayesian Optimization to the Search for Global Minima on Potential Energy Surfaces (ICML 2016 · 23 citations)
What has been tried
64 papers here report 583 measured changes (a component added, removed or swapped, compared with a baseline): helped 65%, no clear effect 10%, hurt 17%, mixed 7%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Replace GPC constraint models with HLGP | S2 · BOV | Mixed | Constrained Bayesian Optimization under Partial Observations: Balanced Improvements and Provable Convergence (AAAI 2024 · 15 citations) | |
| Increase uncertainty of the lengthscale prior while keeping its mode fixed | SVM · best observed value | Hurt | Vanilla Bayesian Optimization Performs Great in High Dimensions (ICML 2024 · 8 citations) | |
| Optimize candidate parent subsets with the proposed reduction algorithm rather than IP only | Instances with larger parent set size · number of binary variables | Helped | Decomposed Quadratization: Efficient QUBO Formulation for Learning Bayesian Network (AAAI 2025 · 3 citations) | |
| Add large clonal families (use M greater than 1 instead of M=1) | In silico binding and stability optimization · best achieved objective | Helped | Bayesian Optimization of Antibodies Informed by a Generative Model of Evolving Sequences (ICLR 2025 · 3 citations) | |
| exclude 4 unobservable OD pairs | One-Way Corridor · NRMSE | 0.187 -> 0.094 | Helped | BO4Mob: Bayesian Optimization Benchmarks for High-Dimensional Urban Mobility Problem (NeurIPS 2025 · 2 citations) |
| Increase the sample-average control parameter from N_c=1 to N_c=10 | 2D Schwefel, 4D Rosenbrock, and 6D Ackley · outer-loop solution quality | Helped | Optimizing Posterior Samples for Bayesian Optimization via Rootfinding (ICLR 2025 · 1 citation) | |
| use folded-normal entropy approximation 1 rather than approximation 3 | DGSM estimation · DGSM RMSE | No clear effect | Active Learning for Derivative-Based Global Sensitivity Analysis with Gaussian Processes (NeurIPS 2024 · 1 citation) | |
| randomize dominant dimensions | Lasso-DNA (d=180) · average objective value | 0.315 -> 0.315 (± \num{4e-4}) | Hurt | Understanding High-Dimensional Bayesian Optimization (ICML 2025 · 0 citations) |
| Use random initialization instead of hybrid initialization | Gaussian (15, 5000, d=1.0) · BIC score | -101954.63 -> -101939.75 | Helped | Inexact Column Generation for Bayesian Network Structure Learning via Difference-of-Submodular Optimization (NeurIPS 2025 · 0 citations) |
| add transfer learning with LC mixup | PD1 · normalized regret (×100) | 15.2{ ±2.0} -> 11.7{ ±0.4} | Helped | Cost-Sensitive Freeze-thaw Bayesian Optimization for Efficient Hyperparameter Tuning (NeurIPS 2025 · 0 citations) |
| add the LMC second-batch component | MLP and PSF · mean absolute relative error | Helped | Kernelized Normalizing Constant Estimation: Bridging Bayesian Quadrature and Bayesian Optimization (AAAI 2024 · 0 citations) | |
| Set q=0.1 rather than q=0 | MOF storage capacity optimization · best objective value | Helped | Quality-Weighted Vendi Scores And Their Application To Diverse Experimental Design (ICML 2024 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 21 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- Mopta08 · 4 of 21 papers
- Lasso-DNA · 4 of 21 papers
- SVM · 4 of 21 papers
- Rover · 3 of 21 papers
- Levy · 2 of 21 papers
- Griewank · 2 of 21 papers
- Ackley · 2 of 21 papers
- Hartmann 6D · 2 of 21 papers
Baselines
- Random search · 4 of 21 papers
- Bounce · 3 of 21 papers
- GOBNILP · 2 of 21 papers
- DSP · 2 of 21 papers
- TuRBO · 2 of 21 papers
- SAASBO · 2 of 21 papers
- CMA-ES · 2 of 21 papers
- EI · 2 of 21 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 52 papers; from OpenAlex and the papers' affiliations).
- Stanford University · 5 papers
- Lund University · 5 papers
- Nagoya University · 2 papers
- University of Chicago · 2 papers
- Amazon · 2 papers
- Lund University / Lund Institute of Technology · 2 papers
- Meta · 2 papers
- Korea University · 2 papers
- Korea Advanced Institute of Science and Technology · 2 papers
Use it
Download BibTeX of 40 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Acquisition function design (97 papers)
- Multi-fidelity hyperparameter optimization (44 papers)
- Multi-objective Bayesian optimization (33 papers)
- Regret bounds for Bayesian optimization (43 papers)
Elsewhere in the topic
- Offline black-box optimization with surrogates (56 papers)
- High-dimensional Bayesian optimization (34 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).