Context Understanding on DocVQA
0.994AccuracyTwC (Ours) - Img & Txt
Evaluation Results
| Method | Links | |
|---|---|---|
| TwC (Ours) - Img & TxtCategory=Think with Comic, Note=G-t-R2026.02 | 0.994 | |
| Gemini-3-ProCategory=MLLM, Note=direct2026.02 | 0.945 | |
| TwC (Ours) - Only ImageCategory=Think with Comic, Note=direct2026.02 | 0.928 | |
| Claude-Sonnet 4.5Category=MLLM, Note=direct2026.02 | 0.926 | |
| GPT-5.2Category=MLLM, Note=direct2026.02 | 0.728 | |
| TWI-1-Generated PhotoCategory=Think with Image, Note=G-t-R2026.02 | 0.675 | |
| DREAMLLMCategory=Think with Image, Note=G-t-R2026.02 | 0.655 | |
| Sora 2Category=Think with Video, Note=V-o-T, Evaluation Details=50 sampled instances2026.02 | 0.505 |