DocAttriBench - VISA: leaderboard
Metric: Overall answer correctness (%; grounded answer generation on Paper-VISA and Wiki-VISA (2,779 test questions): an answer counts only when its text matches the reference and its evidence box is correct; zero-shot, with model-specific prompts and box parsers; element-level evidence boxes from the MAPPET attribution pipeline over Docling layout regions, a box counting when IoU >= 0.5). Source: arxiv.org. Saturation forecast: Around 2030. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 8B | 1.9 |
| 2 | Qwen 2.5 VL 32B Instruct | 0.6 |
| 3 | InternVL3-38B | 0.4 |
| 4 | InternVL3-8B | 0 |
| 5 | InternVL2.5-2B | 0 |
| 6 | Qwen 2.5 VL 7B Instruct | 0 |
Interactive version: theaggregate.ai/benchmark?slug=docattribench-visa · How It Works · Data refreshed daily, snapshot 2026-09-26.