M3Kang: A Multilingual Multimodal Mathematical Reasoning Dataset from Kangaroo Problems
Introduction
Despite state-of-the-art vision-language models (VLMs) have demonstrated strong reasoning capabilities, their performance in multilingual mathematical reasoning remains underexplored. To bridge this gap, we introduce M3Kang, the first massively multilingual, multimodal mathematical reasoning dataset for VLMs. It is derived from the Kangaroo Math Competition, the world’s… See the full description on the dataset page: https://huggingface.co/datasets/qualcomm/M3Kang.
