Object Hallucination Probing on GQA Adversarial
82.73Accuracy
MESA
Updated 8d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.04 | 82.73 | 81.93 | |
2026.02 | 81.76 | 82.31 | |
2026.04 | 81.61 | 80.75 | |
2026.02 | 81 | 82.09 | |
2026.04 | 81 | 80.06 | |
2026.04 | 80.47 | 78.44 | |
2026.04 | 80.03 | 79.75 | |
2026.04 | 80.01 | 80.75 | |
2026.04 | 79.6 | 81.35 | |
2026.04 | 79.13 | 81.07 | |
2026.04 | 79.1 | 80.41 | |
2025.12 | 78.4 | 80.5 | |
2026.04 | 78.33 | 79.78 | |
2026.02 | 77.4 | 80.11 | |
2026.02 | 77.4 | 80.11 | |
2025.12 | 77.2 | 78.3 | |
2025.12 | 76.4 | 78.1 | |
2026.02 | 76.09 | 78.78 | |
2026.02 | 76.09 | 78.78 | |
2026.04 | 76.09 | 78.78 | |
2026.04 | 75.9 | 76.89 | |
2025.12 | 75.8 | 76.5 | |
2025.12 | 75.4 | 74.7 | |
2025.12 | 75.3 | 76.1 | |
2025.12 | 75.2 | 77.7 | |
2026.02 | 75.08 | 76.06 | |
2026.02 | 75.08 | 76.06 | |
2026.02 | 75 | 78.71 | |
2026.02 | 75 | 78.71 | |
2025.12 | 74.8 | 75.8 | |
2026.04 | 73.61 | 65.59 | |
2026.04 | 73.53 | 65.53 | |
2025.12 | 73.5 | 76 | |
2025.12 | 73.1 | 76.1 | |
2025.12 | 72.8 | 75.5 | |
2026.04 | 72.1 | 76.37 | |
2025.12 | 70.4 | 74.1 | |
2026.02 | 69.2 | 75.58 | |
2026.02 | 68.13 | 74.78 | |
2026.04 | 67.4 | 74.31 |