Molecular Optimization on NMO
1.29AUCGenetic GFN
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Genetic GFNRepresentation=SMILES, Pre-training Dataset=ZINC2026.06 | 1.29 | 1.71 | 4.55 | 5 | |
| Genetic GFNRepresentation=SMILES, Pre-training Dataset=Synth. Dataset2026.06 | 0.62 | 0.85 | 5.15 | 5 | |
| molGARepresentation=SMILES2026.06 | 0.57 | 1.28 | 4.58 | 5 | |
| Genetic GFNRepresentation=GGS2026.06 | 0.55 | 0.79 | 4.17 | 5 | |
| Full method (ours)Representation=GGS2026.06 | 0.42 | 0.59 | 4.4 | 5 | |
| molGARepresentation=GGS, Genetic Operator=our GA2026.06 | 0.36 | 0.43 | 4.49 | 5 | |
| REINVENTRepresentation=SMILES, Pre-training Dataset=Synth. Dataset2026.06 | 0.14 | 0.26 | 4.54 | 5 | |
| REINVENTRepresentation=GGS2026.06 | 0.11 | 0.28 | 3.86 | 3 | |
| f-RAGRepresentation=SMILES2026.06 | 0.1 | 0.22 | 4.81 | 0 | |
| REINVENTRepresentation=SMILES, Pre-training Dataset=ZINC2026.06 | 0.01 | 0.02 | 3.11 | 1 | |
| GENMOLRepresentation=SMILES2026.06 | 0 | 0.01 | 4.1 | 0 |