Research map / Probabilistic and scientific modeling / Bayesian optimization
SnAKe: Bayesian Optimization with Pathwise Exploration
Jose Pablo Folch, Shiqiang Zhang, Robert Matthew Lee, Behrang Shafei, David Walz, Calvin Tsay, Mark van der Wilk, Ruth Misener
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 SnAKe instead of TrEI | Ypacarai Lake O1 optimization · Ypacarai Lake O1 | regret: 0.475 \pm 0.596 vs 0.074 \pm 0.063 | Improved |
| Use SnAKe instead of EIpu | Ypacarai Lake O2 optimization · Ypacarai Lake O2 | regret: 0.068 \pm 0.073 vs 0.109 \pm 0.186 | Hurt |
| Use SnAKe instead of EIpu | Ypacarai Lake O3 optimization · Ypacarai Lake O3 | regret: 0.071 \pm 0.049 vs 0.036 \pm 0.048 | Improved |
| Optimize all three objectives using a mixture of Thompson samples | SnAKe · Ypacarai simultaneous MO1 | regret: 13.636 \pm 35.842 (TrEI); 0.995 \pm 0.661 (EIpu); 0.996 \pm 2.743 (SnAKe) | Mixed |
| Optimize all three objectives using a mixture of Thompson samples | SnAKe · Ypacarai simultaneous MO2 | regret: 0.802 \pm 0.828 (TrEI); 19.308 \pm 55.901 (EIpu); 1.740 \pm 4.813 (SnAKe) | Mixed |
| Optimize all three objectives using a mixture of Thompson samples | SnAKe · Ypacarai simultaneous MO3 | regret: 0.162 \pm 0.108 (TrEI); 0.212 \pm 0.137 (EIpu); 0.064 \pm 0.053 (SnAKe) | Improved |
| Set the deletion distance to zero (ablation) | SnAKe · Synthetic benchmarks | input cost | Improved |
| Set the deletion distance to zero (ablation) | SnAKe · Synthetic benchmarks | regret | Hurt |
| Use nonzero deletion distance (ablation) | SnAKe · Synthetic benchmarks | robustness to deletion-distance choice | Improved |
| Remove point deletion and resample the batch at every iteration (ablation) | SnAKe · Local-optimum escape experiment | escape from local optimum | Hurt |
| Add ε-point deletion to the resampled batch (ablation) | SnAKe · Local-optimum escape experiment | exploration after escaping: pT; \hat{p} \approx 0.74 | Improved |
Closest papers by meaning
- A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences (NeurIPS 2024 · similarity 0.88)
- Transition Constrained Bayesian Optimization via Markov Decision Processes (NeurIPS 2024 · similarity 0.86)
- Approximation-Aware Bayesian Optimization (NeurIPS 2024 · similarity 0.86)
- A General Framework for User-Guided Bayesian Optimization (ICLR 2024 · similarity 0.86)
- Optimistic Games for Combinatorial Bayesian Optimization with Application to Protein Design (ICLR 2025 · similarity 0.85)
- Stopping Bayesian Optimization with Probabilistic Regret Bounds (NeurIPS 2024 · similarity 0.85)
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.
