Task Success on SciWorld
68.21Real
GPT-5
Updated 1mo ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2025.12 | 68.21 | 64.1 | 61.03 | 0.89 | |
2025.12 | 68.21 | 63.07 | 57.44 | 0.84 | |
2025.12 | 66 | 45.64 | 57.95 | 0.88 | |
2025.12 | 66 | 58.46 | 53.33 | 0.81 | |
2025.12 | 56 | 39.49 | 45.64 | 0.82 | |
2025.12 | 56 | 57.95 | 52.31 | 0.93 | |
2025.12 | 43.41 | 45.79 | 46.32 | 1.07 | |
2025.12 | 43.41 | 45.13 | 35.38 | 0.82 | |
2025.12 | 36.79 | 50 | 36.6 | 0.99 | |
2025.12 | 36.79 | 52.31 | 44.1 | 1.2 | |
2025.12 | 34.97 | 37.63 | 31.44 | 0.9 | |
2025.12 | 34.97 | 37.63 | 32.99 | 0.94 | |
2025.12 | 12.64 | 12.04 | 8.9 | 0.7 | |
2025.12 | 12.64 | 10.78 | 8.21 | 0.65 |