DeepMath-103K-VisualSuitability is an extension of DeepMath-103K, enriched with visual suitability annotations that indicate whether a mathematical problem is appropriate for figure-guided (visual) reasoning.
All original problems, answers, and metadata are preserved from DeepMath-103K.
This dataset only adds new annotations and does not modify the original content.
The goal of this dataset is to enable research on when… See the full description on the dataset page: https://huggingface.co/datasets/chenmeiqi/DeepMath-103K-VisualSuitability.
