Research map / Probabilistic and scientific modeling / Bayesian optimization
GIBBON: General-purpose Information-Based Bayesian Optimisation
Henry B. Moss, David S. Leslie, Javier Gonzalez, Paul Rayson
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use exact Thompson-sampled max-values instead of Gumbel samples (ablation) | GIBBON · noisy Hartmann, B=5 | regret: small boost for batch, no effect for B=1 | Mixed |
| use exact Thompson-sampled max-values instead of Gumbel samples (ablation) | GIBBON · noisy Hartmann | computational overhead: five and three times the cost for B=1 and B=5 | Hurt |
| increase batch size beyond 25 (ablation) | GIBBON · noisy Hartmann (20 iterations) | regret | Hurt |
| increase batch size to 50 with local penalisation (ablation) | LPEI · noisy Hartmann (20 iterations) | regret | Improved |
| down-weight diversity term by factor B^2 (ablation) | GIBBON · noisy Hartmann, large batches | regret | Improved |
| use GIBBON instead of MES in noiseless sequential case | GIBBON · Shekel (d=4) | regret | Improved |
| use GIBBON instead of KG for batch BO | GIBBON · synthetic batch benchmarks | computational overhead: at least 10 times smaller | Improved |
| use GIBBON instead of MUMBO | GIBBON · multi-fidelity synthetic benchmarks | regret and overhead | Improved |
| use GIBBON instead of MFMES | GIBBON · synthetic benchmarks | final regret | Improved |
| use DPP-based batch heuristic DPPEI instead of GIBBON | DPPEI · synthetic batch benchmarks | regret | Hurt |
| increase number of BO steps | batch KG, MFMES, NEI vs GIBBON · Hartmann | computational overhead | Mixed |
Closest papers by meaning
- Approximation-Aware Bayesian Optimization (NeurIPS 2024 · similarity 0.90)
- A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences (NeurIPS 2024 · similarity 0.88)
- A General Framework for User-Guided Bayesian Optimization (ICLR 2024 · similarity 0.87)
- Return of the Latent Space COWBOYS: Re-thinking the use of VAEs for Bayesian Optimisation of Structured Spaces (ICML 2025 · similarity 0.86)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization (ICML 2025 · similarity 0.86)
- Optimizing Posterior Samples for Bayesian Optimization via Rootfinding (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
