List your company

Vision-Language Compositional Reasoning on Winoground standard (test)

75.5Text ScoreGPT-4o
58.662.987567.37571.7625Jan 23, 2025

Evaluation results

MethodLinks
GPT-4oParadigm=Componential Analysis (CA), Reasoning Engine=GPT-4o2025.0175.558.552
Gemini 2.0Paradigm=Componential Analysis (CA), Reasoning Engine=Gemini 2.02025.017148.7542
Llama3.3-70BParadigm=Componential Analysis (CA), Reasoning Engine=Llama3.3-70B2025.0168.2549.2541.75
Qwen2.5-32BParadigm=Componential Analysis (CA), Reasoning Engine=Qwen2.5-32B2025.016746.2540
Phi-4-14BParadigm=Componential Analysis (CA), Reasoning Engine=Phi-4-14B2025.0165.254637.75
MMICL + CoCoT2025.0164.2552.550.75
Qwen2.5-14BParadigm=Componential Analysis (CA), Reasoning Engine=Qwen2.5-14B2025.0159.2534.527.25