List your company

Document Question Answering on DUDE (Accuracy Metrics)

80.7Accuracy (all)GPT-4o-200b-128
12.37230.11147.8565.589Oct 8, 2025

Evaluation results

MethodLinks
GPT-4o-200b-128Retrieval=Ground-Truth2025.1080.78363.3
GPT-4o-200b-128Retrieval=LAD-RAG2025.1072.574.654.5
GPT-4o-200b-128Retrieval=topk-adjusted2025.10727554.1
GPT-4o-200b-128Retrieval=retrieving@52025.1070.773.253.4
GPT-4o-200b-128Retrieval=retrieving@102025.1070.673.253.5
InternVL2-8BRetrieval=Ground-Truth2025.1064.169.757.4
InternVL2-8BRetrieval=LAD-RAG2025.106366.953.3
Pixtral-12B-2409Retrieval=Ground-Truth2025.1060.263.646.7
Phi-3.5-Vision-4BRetrieval=LAD-RAG2025.1059.665.646.7
InternVL2-8BRetrieval=topk-adjusted2025.1058.960.941.6
Phi-3.5-Vision-4BRetrieval=topk-adjusted2025.1058.265.646.6
Phi-3.5-Vision-4BRetrieval=Ground-Truth2025.1057.863.844.6
InternVL2-8BRetrieval=retrieving@102025.1057.659.135
Phi-3.5-Vision-4BRetrieval=retrieving@102025.1057.664.345
Phi-3.5-Vision-4BRetrieval=retrieving@52025.1056.261.542.2
InternVL2-8BRetrieval=retrieving@52025.105656.436.7
Pixtral-12B-2409Retrieval=LAD-RAG2025.1054.557.843.3
Pixtral-12B-2409Retrieval=topk-adjusted2025.1053.557.442.6
Pixtral-12B-2409Retrieval=retrieving@102025.1052.95641.7
Pixtral-12B-2409Retrieval=retrieving@52025.1051.451.241.7
MiniCPM-Llama3-V2.5-8BRetrieval=All-Pages2025.10192115
Idefics2-8BRetrieval=All-Pages2025.101720.513
mPLUG-DocOwl 1.5-8BRetrieval=All-Pages2025.101518.811.6