VQA on VQA hard criterion
82ASRCoTTA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CoTTASurrogate=Ensemble2026.03 | 82 | 0.82 | |
| CoTTASurrogate=Ensemble2026.03 | 79 | 0.787 | |
| CoTTASurrogate=Ensemble2026.03 | 58 | 0.576 | |
| FOA-AttackSurrogate=Ensemble2026.03 | 53 | 0.492 | |
| M-AttackSurrogate=Ensemble2026.03 | 49 | 0.442 | |
| FOA-AttackSurrogate=Ensemble2026.03 | 42 | 0.418 | |
| M-AttackSurrogate=Ensemble2026.03 | 39 | 0.38 | |
| FOA-AttackSurrogate=Ensemble2026.03 | 12 | 0.119 | |
| M-AttackSurrogate=Ensemble2026.03 | 11 | 0.11 | |
| AttackVLMSurrogate=B/322026.03 | 3 | 0.022 | |
| Agent-AttackSurrogate=Ensemble2026.03 | 2 | 0.017 | |
| AnyAttackSurrogate=Ensemble2026.03 | 2 | 0.008 | |
| Agent-AttackSurrogate=Ensemble2026.03 | 2 | 0.01 | |
| CoTTASurrogate=Ensemble2026.03 | 2 | 0.02 | |
| AnyAttackSurrogate=Ensemble2026.03 | 1 | 0.003 | |
| AttackVLMSurrogate=B/162026.03 | 1 | 0.09 | |
| AttackVLMSurrogate=B/162026.03 | 0 | 0 | |
| AttackVLMSurrogate=B/322026.03 | 0 | 0 | |
| AttackVLMSurrogate=Laion2026.03 | 0 | 0 | |
| AttackVLMSurrogate=B/162026.03 | 0 | 0 | |
| AttackVLMSurrogate=B/322026.03 | 0 | 0 | |
| AttackVLMSurrogate=Laion2026.03 | 0 | 0 | |
| AnyAttackSurrogate=Ensemble2026.03 | 0 | 0 | |
| Agent-AttackSurrogate=Ensemble2026.03 | 0 | 0 | |
| AttackVLMSurrogate=Laion2026.03 | 0 | 0 | |
| AttackVLMSurrogate=B/162026.03 | 0 | 0 | |
| AttackVLMSurrogate=B/322026.03 | 0 | 0 | |
| AttackVLMSurrogate=Laion2026.03 | 0 | 0 | |
| AnyAttackSurrogate=Ensemble2026.03 | 0 | 0 | |
| M-AttackSurrogate=Ensemble2026.03 | 0 | 0 | |
| FOA-AttackSurrogate=Ensemble2026.03 | 0 | 0 | |
| Agent-AttackSurrogate=Ensemble2026.03 | 0 | 0 |