Research map / Learning theory and algorithms / Combinatorial and search optimization
Sample Efficiency Matters: A Benchmark for Practical Molecular Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SELFIES instead of SMILES strings | REINVENT · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 14.196 -> 13.471 | Hurt |
| use SELFIES instead of SMILES strings | LSTM HC · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 12.223 -> 10.246 | Hurt |
| use SELFIES-based STONED instead of SMILES GA | genetic algorithm · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 12.054 -> 13.024 | Helped |
| add GP surrogate model (GP BO) to Graph GA | Graph GA · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 13.751 -> 13.156 | Mixed |
| add active-learning surrogate (GFlowNet-AL) to GFlowNet | GFlowNet · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 9.131 -> 8.406 | Hurt |
| train a surrogate to prioritize molecules (MolPAL) vs random screening | Screening · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 8.635 -> 10.268 | Helped |
| use REINVENT instead of hill climbing LSTM | SMILES LSTM · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 12.223 -> 14.196 | Helped |
| retune hyperparameters (larger sigma) for limited oracle budget | REINVENT · PMO 23 oracles, 10K oracle calls | AUC Top-10 | Helped |
| single-start token-by-token or atom-by-atom assembly vs fragment/other | GA+D, MolDQN, Graph MCTS · PMO 23 oracles, 10K oracle calls | sum of mean AUC Top-10: 8.964, 5.620, 7.803 | Hurt |
Closest papers by meaning
- Beyond Drug Discovery: The Nanotechnology Molecular Optimization (NMO) Benchmark (NeurIPS 2026 · similarity 0.90)
- Multi-Objective Molecular Design Through Learning Latent Pareto Set (AAAI 2025 · similarity 0.87)
- On the Design Space of Discrete Diffusion Online Adaptation for Molecular Optimization (NeurIPS 2026 · similarity 0.86)
- Efficient Evolutionary Search Over Chemical Space with Large Language Models (ICLR 2025 · similarity 0.86)
- Gradual Optimization Learning for Conformational Energy Minimization (ICLR 2024 · similarity 0.86)
- Offline Model-based Optimization for Real-World Molecular Discovery (ICML 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.