Belief Prediction on EgoToM (test)
33.8True Accuracy
Qwen2.5-VL-7B
Updated 2mo ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.03 | 33.8 | 90.6 | 27.9 | |
2026.03 | 32.9 | 99.8 | 30.8 | |
2026.03 | 29 | 92.5 | 23.7 | |
2026.03 | 28.1 | 99.7 | 28.1 | |
2026.03 | 19.5 | 99.7 | 19.2 |
Qwen2.5-VL-7B
| Method | Links | |||
|---|---|---|---|---|
2026.03 | 33.8 | 90.6 | 27.9 | |
2026.03 | 32.9 | 99.8 | 30.8 | |
2026.03 | 29 | 92.5 | 23.7 | |
2026.03 | 28.1 | 99.7 | 28.1 | |
2026.03 | 19.5 | 99.7 | 19.2 |