Multi-page Document Question Answering on DUDE
56.1ANLSPDF-WuKong
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PDF-WuKong2024.10 | 56.1 | — | |
| GPT-4V (2024-06)2024.10 | 53.9 | — | |
| GRAM2024.10 | 53.4 | — | |
| CREAM2024.10 | 52.5 | — | |
| LayTokenLLM-8BBackbone=8B2025.03 | 52 | — | |
| LayTokenLLM-7BBackbone=7B2025.03 | 49.72 | — | |
| DocFormerv22024.10 | 48.4 | — | |
| DocOwl2-8BToken^V=3242024.09 | 46.77 | 0.94 | |
| mPLUG-DocOwl2-8BBackbone=8B2025.03 | 46.77 | — | |
| DocOwl22024.10 | 46.7 | — | |
| SV-RAG2024.10 | 45 | — | |
| Idefics3-8B2024.10 | 38.7 | — | |
| Idefics3-8BToken^V=~8382024.09 | 38.65 | 2.29 | |
| Idefics3-8BBackbone=8B2025.03 | 38.65 | — | |
| LongVA-7BToken^V=~2,0292024.09 | 38.37 | 2.26 | |
| LongVA-7BBackbone=7B2025.03 | 38.37 | — | |
| InternVL2-8BBackbone=8B2025.03 | 37 | — | |
| Hi-VT52024.10 | 35.7 | — | |
| LLaVA-next-interleave-7BToken^V=7292024.09 | 28.03 | 1.47 | |
| LLaVA-next-interleave-7BBackbone=7B2025.03 | 28.03 | — | |
| Longformer2024.10 | 27.1 | — | |
| BigBird2024.10 | 26.3 | — | |
| LayoutLMv32024.10 | 20.3 | — |