Molecular Property Optimization on soluble permeable
0.6Normalized Rewardπ(·|y*)
Evaluation results
| Method | Links | |
|---|---|---|
| π(·|y*)2026.04 | 0.6 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=642026.04 | 0.39 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=322026.04 | 0.35 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=162026.04 | 0.25 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=82026.04 | 0.14 | |
| RLRL training budget=10002026.04 | 0.05 | |
| RLRL training budget=20002026.04 | 0.04 | |
| RLRL training budget=5002026.04 | 0.03 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=42026.04 | 0 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=22026.04 | -0.15 |