List your company

Vision-Language Compositionality on SugarCrepe

88.06AccuracyDINOv2
47.874458.307268.7479.1728Feb 7, 2026

Evaluation results

MethodLinks
DINOv2LLM=Qwen2-7B2026.0288.06
mFRESALLM=Qwen2-7B2026.0283.17
DINOv2LLM=Phi22026.0282.05
FT-DINOSAURLLM=Qwen2-7B2026.0281.24
VQDINOMLPLLM=Qwen2-7B2026.0280.16
StableLSDLLM=Qwen2-7B2026.0278.98
DINOSAURv2LLM=Qwen2-7B2026.0278.18
mFRESALLM=Phi22026.0277.27
DINOSAURv2LLM=Phi22026.0275.98
Slot DiffusionLLM=Qwen2-7B2026.0274.53
SPOTLLM=Qwen2-7B2026.0274.08
VQDINOMLPLLM=Phi22026.0273.05
StableLSDLLM=Phi22026.0272.92
DINOSAURLLM=Qwen2-7B2026.0272.45
SPOTLLM=Phi22026.0271.65
FT-DINOSAURLLM=Phi22026.0270.94
Slot DiffusionLLM=Phi22026.0270.39
DINOSAURLLM=Phi22026.0267.85
Blind-VLMLLM=Qwen2-7B2026.0253.29
Blind-VLMLLM=Phi22026.0249.42