Visual Faithfulness Evaluation on SeeTRUE
82F1 ScoreBLPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BLPOJudge Model=Qwen2.5-VL-32B-instruct2026.02 | 82 | 82 | |
| TextGradJudge Model=Llama-4-Maverick-17B-128E-instruct2026.02 | 80 | 80 | |
| BLPOJudge Model=Llama-4-Maverick-17B-128E-instruct2026.02 | 80 | 80 | |
| OPROJudge Model=Llama-4-Maverick-17B-128E-instruct2026.02 | 79 | 79 | |
| OPROJudge Model=Qwen2.5-VL-32B-instruct2026.02 | 78 | 78 | |
| APO-imageJudge Model=Llama-4-Maverick-17B-128E-instruct2026.02 | 78 | 78 | |
| TextGradJudge Model=Llama-4-Scout-17B-16E-instruct2026.02 | 77 | 77 | |
| BLPOJudge Model=Llama-4-Scout-17B-16E-instruct2026.02 | 77 | 77 | |
| TextGradJudge Model=Qwen2.5-VL-32B-instruct2026.02 | 75 | 75 | |
| No Optim.Judge Model=Llama-4-Scout-17B-16E-instruct2026.02 | 75 | 75 | |
| APO-imageJudge Model=Llama-4-Scout-17B-16E-instruct2026.02 | 75 | 75 | |
| No Optim.Judge Model=Llama-4-Maverick-17B-128E-instruct2026.02 | 75 | 76 | |
| APO-imageJudge Model=Qwen2.5-VL-32B-instruct2026.02 | 74 | 74 | |
| OPROJudge Model=Llama-4-Scout-17B-16E-instruct2026.02 | 73 | 73 | |
| No Optim.Judge Model=Qwen2.5-VL-32B-instruct2026.02 | 71 | 75 |