List your company
ResearchFormerly WizWand

Question Answering on DUDE

21.43ANLSM3DOCDEP
15.065216.717618.3720.0224Apr 14, 2026Apr 15, 2026Apr 16, 2026Apr 17, 2026
Updated 5mo ago

Evaluation results

MethodLinks
M3DOCDEPtop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0421.4318.2125.12
MultiDocFusiontop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0418.5916.9222.85
MultiDocFusiontop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0418.5916.9222.85
Structure-based chunkingtop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0417.5114.8919.21
Structure-based chunkingtop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0417.5114.8919.21
Perplexity chunkingtop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0416.5313.918.55
Perplexity chunkingtop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0416.5313.918.55
Length chunkingtop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0416.1114.4419.88
Length chunkingtop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0416.1114.4419.88
Semantic chunkingtop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0415.4812.6116.57
Semantic chunkingtop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0415.4812.6116.57
LumberChunkertop-k={1, 4}, Models=Llama-3.2-3B, Mistral-8B, Qwen-2.5-7B (averaged)2026.0415.3112.8417.52
LumberChunkertop-k={1, 2, 3, 4}, Readers=LLaVA-OneVision-1.5, InternVL-3.5, and Qwen2.5-VL, RAG protocol=shared corpus-level RAG protocol2026.0415.3112.8417.52