List your company

Knowledge-Based Visual Question Answering on InfoSeek All

49.9AccuracyDBAgent (SFT: InfoSeek)
6.42817.7142940.286Apr 8, 2026

Evaluation results

MethodLinks
DBAgent (SFT: InfoSeek)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0449.9
DBAgent (SFT: Mixed)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0448.7
DBAgent (SFT: E-VQA)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0448.4
ReflectiVALLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B2026.0443.9
VLM-PRFLLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B2026.0442.8
VLM-PRFLLM=InternVL3-8B, Retriever=EVA-CLIP-8B2026.0442.5
ReflectVALLM=LLaMA-3.1-8B, Retriever=EVA-CLIP-8B2026.0440.1
Wiki-LLaVALLM=LLaVA-v1.5-7B, Retriever=CLIP ViT-L/14 + Contriever2026.0428.9
EchoSightLLM=Mistral-7B/LLaMA-3-8B, Retriever=EVA-CLIP-8B2026.0427.7
DBAgent (w/o SFT)LLM=Qwen2.5-VL-7B, Retriever=EVA-CLIP-8B + bge-m32026.0424.4
Qwen2.5-VL-7B (Base)LLM=Qwen2.5-VL-7B (Base), Retriever=–2026.0423.7
GPT-4VLLM=GPT-4V, Retriever=–2026.0414.6
BLIP-2LLM=Flan-T5XL, Retriever=–2026.0412.5
DPRLLM=Multi-passage BERT, Retriever=CLIP ViT-B/322026.0412.4
LLaVA-v1.5LLM=Vicuna-7B, Retriever=–2026.049.5
InstructBLIPLLM=Flan-T5XL, Retriever=–2026.048.1