AI4Math/MathVerse

Dataset Card for MathVerse Dataset Description Paper Information Dataset Examples Leaderboard Citation Dataset Description The capabilities of Multi-modal Large Language Models (MLLMs) in visual math problem-solving remain insufficiently evaluated and understood. We investigate current benchmarks to incorporate excessive visual content within textual questions, which potentially assist MLLMs in deducing answers without truly interpreting the input diagrams.… See the full description on the dataset page: https://huggingface.co/datasets/AI4Math/MathVerse.