Biological Reasoning on RPE1 (OOD)
79.18DE ScoreCoverage (Upper Bound)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Coverage (Upper Bound)Policy Model=RBIO12026.03 | 79.18 | 91.99 | 85.59 | |
| BoN w/ Full Set (Multi-Label)Policy Model=RBIO1, Strategy=Best-of-N, Reward Model Training=DC-W2S Full Set2026.03 | 78.88 | 62.3 | 70.59 | |
| BoN w/ CellProfiler (100k)Policy Model=RBIO1, Strategy=Best-of-N, Reward Model Training=DC-W2S CellProfiler (100k)2026.03 | 78.82 | 58.04 | 68.43 | |
| BoN w/ ESM (100k)Policy Model=RBIO1, Strategy=Best-of-N, Reward Model Training=DC-W2S ESM (100k)2026.03 | 78.81 | 60.65 | 69.73 | |
| Majority VotingPolicy Model=RBIO12026.03 | 78.51 | 30.69 | 54.6 | |
| Greedy DecodingPolicy Model=RBIO12026.03 | 78.33 | 34.89 | 56.61 |