Research map / Probabilistic and scientific modeling / Bayesian optimization
Combining Latent Space and Structured Kernels for Bayesian Optimization over Combinatorial Spaces
Aryan Deshwal, Jana Doppa
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 |
|---|---|---|---|
| add structure-coupled kernel to the latent-space GP surrogate | Naive LSBO · arithmetic expressions and chemical design | mean absolute error | Improved |
| use LADDER surrogate in place of Matern GP with same CMA-ES optimizer | Naive LSBO · arithmetic expressions and chemical design | best objective found | Improved |
| switch acquisition optimizer from CMA-ES to L-BFGS (ablation) | Naive LSBO · arithmetic expressions and chemical design | best objective found | No clear effect |
| add weighted retraining of the latent space model | Naive LSBO · arithmetic expressions and chemical design | best objective found | Improved |
| compare LADDER with Naive LSBO w/ retraining | LADDER · chemical design | best logP found: p-value = 0.0489 | No clear effect |
| compare LADDER with weighted retraining | LADDER · arithmetic expressions | best objective found | Improved |
| use latent space with structured kernel instead of BO directly on fingerprints | LADDER · chemical design, iteration 15 | best logP found: 2.46 +- 0.32 -> 2.78 +- 0.26 | Improved |
| use latent space with structured kernel instead of BO directly on fingerprints | LADDER · chemical design, iteration 100 | best logP found: 3.36 +- 0.23 -> 3.25 +- 0.18 | Hurt |
Closest papers by meaning
- Return of the Latent Space COWBOYS: Re-thinking the use of VAEs for Bayesian Optimisation of Structured Spaces (ICML 2025 · similarity 0.90)
- Joint Composite Latent Space Bayesian Optimization (ICML 2024 · similarity 0.89)
- A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences (NeurIPS 2024 · similarity 0.87)
- Omnipresent Yet Overlooked: Heat Kernels in Combinatorial Bayesian Optimization (NeurIPS 2025 · similarity 0.86)
- De Novo Molecular and Crystal Design with Latent Space Bayesian Optimization (AAAI 2025 · similarity 0.86)
- Inversion-based Latent Bayesian Optimization (NeurIPS 2024 · 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.
