List your company

Causal-order reasoning on VQA-Causal

99.17AccuracyHuman Estimate
39.723655.156870.5986.0232Jun 1, 2025

Evaluation results

MethodLinks
Human Estimate2025.0699.17
LLaVA 1.6Model Category=Text-Generation Models2025.0658.11
LLaVA 1.5Model Category=Text-Generation Models2025.0653.19
CLIP ViT B/32Model Category=Score-Based Models2025.0651.62
BLIP2 FEModel Category=Score-Based Models, Note=BLIP2 feature extractor model2025.0651.51
NegCLIPModel Category=Score-Based Models2025.0650.89
Vicuna 1.5Model Category=Text-Generation Models2025.0650.86
BLIP2 ITMModel Category=Score-Based Models2025.0650.76
CLIP ViT L/14Model Category=Score-Based Models2025.0650.74
RobustCLIPModel Category=Score-Based Models2025.0650.66
Qwen2.5-VLModel Category=Text-Generation Models2025.0650.66
Random2025.0650
BLIP ITM BaseModel Category=Score-Based Models2025.0648.94
BLIP ITM LargeModel Category=Score-Based Models2025.0648.68
FLAVAModel Category=Score-Based Models2025.0648.52
Qwen3-VLModel Category=Text-Generation Models2025.0642.01