Beyond Embeddings: The Promise of Visual Table in Visual Reasoning

Benchmark Model Rank Results
visual-question-answering-on-mm-vetLLaVA-VT (Vicuna-13B)#77GPT-4 score: 39.8
visual-question-answering-on-mm-vetLLaVA-VT (Vicuna-7B)#135GPT-4 score: 31.8