List your company

Knowledge-Based Visual Question Answering on E-VQA All

45.8AccuracyDBAgent (SFT: E-VQA)
10.64819.77428.938.026Apr 8, 2026

Evaluation results

MethodLinks
DBAgent (SFT: E-VQA)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0445.8
DBAgent (SFT: Mixed)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0445.2
DBAgent (SFT: InfoSeek)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0444.3
VLM-PRFLLM=InternVL3-8B, Retriever=EVA-CLIP-8B2026.0439.2
ReflectiVALLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B2026.0436.8
VLM-PRFLLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B2026.0436
ReflectVALLM=LLaMA-3.1-8B, Retriever=EVA-CLIP-8B2026.0429.2
GPT-4VLLM=GPT-4V, Retriever=–2026.0428.1
DBAgent (w/o SFT)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0423.9
Qwen2.5-VL-7B (Base)LLM=Qwen2.5-VL-7B (Base), Retriever=–2026.0420.3
RORA-VLMLLM=LLaVA-v1.5-7B, Retriever=CLIP ViT-L/142026.0420.3
Wiki-LLaVALLM=LLaVA-v1.5-7B, Retriever=CLIP ViT-L/14 + Contriever2026.0420.3
LLaVA-v1.5LLM=Vicuna-7B, Retriever=–2026.0416.9
BLIP-2LLM=Flan-T5XL, Retriever=–2026.0412.4
InstructBLIPLLM=Flan-T5XL, Retriever=–2026.0412