List your company
ResearchFormerly WizWand

Knowledge-Intensive Visual Question Answering on E-VQA (test)

52.6Accuracy (All)ProMSA
9.75220.8763243.124Feb 15, 2026Mar 8, 2026Mar 30, 2026Apr 21, 2026May 13, 2026Jun 4, 2026Jun 26, 2026
Updated 4mo ago

Evaluation results

MethodLinks
ProMSARetriever=BGE+EVA-CLIP, Model=Qwen3-VL-8B, Category=Retrieval-Augmented Models2026.0652.652.2——
ProMSARetriever=BGE+EVA-CLIP, Model=Qwen2.5-VL-7B, Category=Retrieval-Augmented Models2026.0649.750——
REALRetriever=EVA-CLIP-8B, Model=Qwen3-VL-8B2026.0241.445.5——
REALRetriever=EVA-CLIP-8B, Model=Qwen3-VL-8B, Category=Retrieval-Augmented Models2026.0641.445.5——
MaS-VQARetriever=EVA-CLIP-8B, Model=Qwen3-VL-8B, Category=Retrieval-Augmented Models2026.0641.342.2——
ProMSARetriever=BGE+EVA-CLIP, Model=Qwen3-VL-2B, Category=Retrieval-Augmented Models2026.0641.242.4——
MMSearch-R1*Retriever=BGE+EVA-CLIP, Model=Qwen2.5-VL-7B, Category=Search-Agent Models2026.0640.740.6——
DeepEyesV2*Retriever=BGE+EVA-CLIP, Model=Qwen2.5-VL-7B, Category=Search-Agent Models2026.0639.540.1——
VLM-PRFRetriever=EVA-CLIP-8B, Model=InternVL3-8B2026.0239.240.1——
REALRetriever=EVA-CLIP-8B, Model=InternVL3.5-8B2026.0239.243.9——
Qwen3-VL-8B + LoTBackbone=Qwen3-VL-8B, Model Scale=Medium, Method=LoT2026.0436.438.6——
mKG-RAGRetriever=CLIP-ViT-L, Model=LLaMA3-8B2026.0236.338.4——
CC-VQARetriever=EVA-CLIP-8B, Model=Qwen2.5-VL-7B, Category=Retrieval-Augmented Models2026.0636.141.4——
Qwen2.5-VL-32B + LoTBackbone=Qwen2.5-VL-32B, Model Scale=Large, Method=LoT2026.043637.7——
ReflectiVARetriever=EVA-CLIP-8B, Model=LLaMA3.1-8B2026.0235.535.5——
VLM-PRFRetriever=EVA-CLIP-8B, Model=LLaMA3.1-8B2026.0235.536.3——
REALRetriever=EVA-CLIP-8B, Model=Qwen3-VL-2B2026.0235.440——
EchoSight†Retriever=EVA-CLIP-8B, Model=Qwen3-VL-8B, Category=Retrieval-Augmented Models2026.0635.238.1——
Qwen3-VL-8BBackbone=Qwen3-VL-8B, Model Scale=Medium, Method=Zero-shot retrieval-augmented2026.043536.6——
Qwen3-VL-4B + LoTBackbone=Qwen3-VL-4B, Model Scale=Small, Method=LoT2026.0434.837.3——
Qwen2.5-VL-32BBackbone=Qwen2.5-VL-32B, Model Scale=Large, Method=Zero-shot retrieval-augmented2026.0433.735.3——
InternVL3.5-38B + LoTBackbone=InternVL3.5-38B, Model Scale=Large, Method=LoT2026.0433.535.8——
Qwen3-VL-4BBackbone=Qwen3-VL-4B, Model Scale=Small, Method=Zero-shot retrieval-augmented2026.0432.835——
InternVL3.5-38BBackbone=InternVL3.5-38B, Model Scale=Large, Method=Zero-shot retrieval-augmented2026.0431.633.8——
Qwen2.5-VL-7B + LoTBackbone=Qwen2.5-VL-7B, Model Scale=Medium, Method=LoT2026.0431.433.6——
REALRetriever=EVA-CLIP-8B, Model=LLaVA-1.5-7B2026.0230.933.5——
InternVL3.5-8B + LoTBackbone=InternVL3.5-8B, Model Scale=Medium, Method=LoT2026.0430.733.2——
Qwen2.5-VL-3B + LoTBackbone=Qwen2.5-VL-3B, Model Scale=Small, Method=LoT2026.0430.432.9——
Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B, Model Scale=Medium, Method=Zero-shot retrieval-augmented2026.0430.132.1——
InternVL3.5-8BBackbone=InternVL3.5-8B, Model Scale=Medium, Method=Zero-shot retrieval-augmented2026.042931.4——
InternVL3.5-4B + LoTBackbone=InternVL3.5-4B, Model Scale=Small, Method=LoT2026.0428.731.7——
EchoSightRetriever=EVA-CLIP-8B, Model=LLaVA-1.5-7B2026.0228.531.1——
GPT-4VRetriever=None, Model=None, Category=Zero-shot MLLMs2026.0628.126.9——
Qwen2.5-VL-3BBackbone=Qwen2.5-VL-3B, Model Scale=Small, Method=Zero-shot retrieval-augmented2026.0427.830.1——
Wiki-LLAVARetriever=CLIP-ViT-L, Model=LLaVA-1.5-7B2026.0227.125.7——
InternVL3.5-4BBackbone=InternVL3.5-4B, Model Scale=Small, Method=Zero-shot retrieval-augmented2026.0426.229.1——
Qwen2-VL-7B + LoTBackbone=Qwen2-VL-7B, Model Scale=Medium, Method=LoT2026.0425.628.7——
EchoSightRetriever=EVA-CLIP-8B, Model=Mistral-7B, Category=Retrieval-Augmented Models2026.0624.926.4——
Qwen3-VL-8BRetriever=None, Model=None, Category=Zero-shot MLLMs2026.0624.825.3——
NoteMRRetriever=EVA-CLIP-8B, Model=Qwen3-VL-8B2026.0223.625.6——
Qwen2-VL-7BBackbone=Qwen2-VL-7B, Model Scale=Medium, Method=Zero-shot retrieval-augmented2026.0422.925.6——
Qwen3-VL-2BRetriever=None, Model=None, Category=Zero-shot MLLMs2026.0621.922.2——
Qwen2.5-VL-7BRetriever=None, Model=None, Category=Zero-shot MLLMs2026.0620.622.1——
RORA-VLMRetriever=CLIP+Google Search, Model=LLaVA-1.5-7B2026.0220.3———
Qwen2-VL-2B + LoTBackbone=Qwen2-VL-2B, Model Scale=Small, Method=LoT2026.0416.218.2——
Qwen2-VL-2BBackbone=Qwen2-VL-2B, Model Scale=Small, Method=Zero-shot retrieval-augmented2026.041617.5——
BLIP-2Retriever=None, Model=Flan-T5, Category=Zero-shot MLLMs2026.0612.412.6——
PromptCapRetriever=EVA-CLIP-8B, Model=Flan-T5-11B2026.0211.410.8——
Prophet++Retriever=EVA-CLIP-8B, Model=Qwen3-VL-8B2026.0211.410.8——
BLIP-2Model=Flan-T5XL, Retrieval Mode=-, Category=Zero-shot MLLMs2026.02——12.612.4
DPR V+TModel=Multi-passage BERT, Retrieval Mode=CLIP ViT-B/32, Category=Retrieval-Augmented Models2026.02——29.1—
EchoSightModel=Mistral-7B/LLaMA-3-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——19.4—
GPT-4VModel=GPT-4V, Retrieval Mode=-, Category=Zero-shot MLLMs2026.02——26.928.1
InternVL3-8BModel=InternVL3-8B, Retrieval Mode=-, Category=Zero-shot MLLMs2026.02——15.715.9
MaS-VQAModel=InternVL3-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——40.840.1
MaS-VQAModel=Qwen3-VL-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——42.241.3
MMKB-RAGModel=Qwen2-7B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——39.735.9
NoteMRModel=Qwen3-VL-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——25.623.6
Prophet++Model=Qwen3-VL-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——22.520
Qwen2.5-VL-7BModel=Qwen2.5-VL-7B, Retrieval Mode=-, Category=Zero-shot MLLMs2026.02——16.316.3
Qwen3-VL-8BModel=Qwen3-VL-8B, Retrieval Mode=-, Category=Zero-shot MLLMs2026.02——19.519.5
ReflectiVAModel=LLaMA-3.1-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——2829.2
VLM-PRFModel=InternVL3-8B, Retrieval Mode=EVA-CLIP-8B, Category=Retrieval-Augmented Models2026.02——40.139.2
Wiki-LLaVAModel=Vicuna-7B, Retrieval Mode=CLIP ViT-L/14+Contriever, Category=Retrieval-Augmented Models2026.02——17.720.3