List your company
ResearchFormerly WizWand

Fine-grained Detection on MMFakeBench 1000 (val)

38.9F1 ScoreREFORM
3.85212.95122.0531.149Mar 2, 2026
Updated 8mo ago

Evaluation results

MethodLinks
REFORMLanguage Model=Florence2-0.3B, Prompt Method=Ours (Sec.3.2), Zero-shot=true2026.0338.934.648.957.3
LLaVA-1.6Language Model=Vicuna-13B, Prompt Method=MMD-Agent, Zero-shot=true2026.033844.54140.6
BLIP2Language Model=FlanT5-XXL, Prompt Method=MMD-Agent, Zero-shot=true2026.0331.639.832.234.4
InstructBLIPLanguage Model=Vicuna-13B, Prompt Method=MMD-Agent, Zero-shot=true2026.032633.330.129.5
VILALanguage Model=LLaMA2-13B, Prompt Method=MMD-Agent, Zero-shot=true2026.0322.727.324.428.7
LLaVA-1.6Language Model=Vicuna-7B, Prompt Method=Standard, Zero-shot=true2026.0317.414.825.730.8
BLIP2Language Model=FlanT5-XXL, Prompt Method=Standard, Zero-shot=true2026.0316.734.927.332.8
BLIP2Language Model=FlanT5-XL, Prompt Method=Standard, Zero-shot=true2026.0316.420.127.533
mPLUG-Owl2Language Model=LLaMA2-7B, Prompt Method=Standard, Zero-shot=true2026.0314.522.225.931.1
InstructBLIPLanguage Model=Vicuna-13B, Prompt Method=Standard, Zero-shot=true2026.0313.713.22428.8
LLaVA-1.6Language Model=Vicuna-13B, Prompt Method=Standard, Zero-shot=true2026.031222.52530
PandaGPTLanguage Model=Vicuna-7B, Prompt Method=Standard, Zero-shot=true2026.0311.89.82530
VILALanguage Model=LLaMA2-7B, Prompt Method=Standard, Zero-shot=true2026.0311.57.52530
VILALanguage Model=LLaMA2-13B, Prompt Method=Standard, Zero-shot=true2026.0311.57.52530
Qwen-VLLanguage Model=Qwen-7B, Prompt Method=Standard, Zero-shot=true2026.037.510.324.311
InstructBLIPLanguage Model=Vicuna-7B, Prompt Method=Standard, Zero-shot=true2026.037.17.96.57.8
Otter-ImageLanguage Model=MPT-7B, Prompt Method=Standard, Zero-shot=true2026.035.210.53.44.1
MiniGPT4Language Model=Vicuna-7B, Prompt Method=Standard, Zero-shot=true2026.035.25.221.29