Molecular Property Optimization on antibacterial like
90Normalized Rewardπ(·|y*)
Evaluation results
| Method | Links | |
|---|---|---|
| π(·|y*)2026.04 | 90 | |
| RLRL training budget=20002026.04 | 84 | |
| RLRL training budget=10002026.04 | 79 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=642026.04 | 69 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=162026.04 | 67 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=322026.04 | 66 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=82026.04 | 60 | |
| RLRL training budget=5002026.04 | 58 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=42026.04 | 48 | |
| Reward Weighted Classifier-Free Guidance (RCFG)Inference-time sampling budget |YS|=22026.04 | 30 |