Molecular Property Optimization on high mw high qed
2.84Normalized RewardRL
Evaluation Results
| Method | Links | |
|---|---|---|
| RLRL training budget=20002026.04 | 2.84 | |
| RLRL training budget=10002026.04 | 1.26 | |
| π(·|y*)2026.04 | 0.7 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=642026.04 | 0.59 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=162026.04 | 0.58 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=322026.04 | 0.58 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=82026.04 | 0.55 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=42026.04 | 0.47 | |
| RLRL training budget=5002026.04 | 0.41 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=22026.04 | 0.34 |