| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Object Hallucination Evaluation | POPE | Accuracy94.42 | 2,056 | |
| Object Hallucination | POPE Popular | Accuracy93.33 | 406 | |
| Object Hallucination | POPE Adversarial | Accuracy90 | 367 | |
| Object Hallucination | POPE (Random) | F1 Score93.02 | 338 | |
| Hallucination Evaluation | POPE | Accuracy94.42 | 281 | |
| Object Hallucination Evaluation | POPE | Accuracy92 | 259 | |
| Object Hallucination Evaluation | POPE Adversarial | Accuracy89.8 | 174 | |
| Object Hallucination Evaluation | POPE Random | Accuracy94 | 152 | |
| Object Hallucination Evaluation | POPE | Accuracy (POPE)87.6 | 137 | |
| Visual Question Answering | POPE | Accuracy89.6 | 136 | |
| Object Hallucination Evaluation | POPE (test) | Accuracy90.6 | 123 | |
| Multimodal Understanding | POPE | POPE Score0.906 | 116 | |
| Object Hallucination Evaluation | POPE MSCOCO | F1 Score93.97 | 114 | |
| Object Hallucination Evaluation | POPE (popular) | Accuracy92 | 100 | |
| Object Hallucination Evaluation | POPE GQA | Accuracy91.03 | 86 | |
| Object Hallucination Evaluation | POPE Adversarial offline | F1 Score68.96 | 84 | |
| Object Hallucination Evaluation | POPE Popular offline | F1 Score84.43 | 84 | |
| Object Hallucination Evaluation | POPE Random offline | F1 Score73.6 | 84 | |
| Object Hallucination Evaluation | POPE A-OKVQA | Accuracy91.2 | 83 | |
| Object Hallucination | POPE | F1 Score91.9 | 79 | |
| Object Hallucination Evaluation | POPE GQA Popular | Accuracy89.4 | 70 | |
| Transfer Attack | POPE (test) | CAE0.2477 | 69 | |
| Object Hallucination Evaluation | POPE | Average Accuracy100.43 | 53 | |
| Object Probing | POPE Average | Accuracy87.84 | 52 | |
| Object Hallucination Evaluation | POPE Random, Popular, Adversarial v1.0 | Random Score94.27 | 51 |