Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.16742 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910026575118336 |
|---|---|
| author | Sun, Haoxiang Xu, Lizhen Zhao, Bing Yin, Wotao Wang, Wei Yang, Boyu Wang, Rui Wei, Hu |
| author_facet | Sun, Haoxiang Xu, Lizhen Zhao, Bing Yin, Wotao Wang, Wei Yang, Boyu Wang, Rui Wei, Hu |
| contents | Reinforcement Learning with Verifiable Rewards (RLVR) has been shown effective in enhancing the visual reflection and reasoning capabilities of Large Multimodal Models (LMMs). However, existing datasets are predominantly derived from either small-scale manual construction or recombination of prior resources, which limits data diversity and coverage, thereby constraining further gains in model performance. To this end, we introduce \textbf{DeepVision-103K}, a comprehensive dataset for RLVR training that covers diverse K12 mathematical topics, extensive knowledge points, and rich visual elements. Models trained on DeepVision achieve strong performance on multimodal mathematical benchmarks, and generalize effectively to general multimodal reasoning tasks. Further analysis reveals enhanced visual perception, reflection and reasoning capabilities in trained models, validating DeepVision's effectiveness for advancing multimodal reasoning. Data: \href{https://huggingface.co/datasets/skylenage/DeepVision-103K}{this url}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_16742 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DeepVision-103K: A Visually Diverse, Broad-Coverage, and Verifiable Mathematical Dataset for Multimodal Reasoning Sun, Haoxiang Xu, Lizhen Zhao, Bing Yin, Wotao Wang, Wei Yang, Boyu Wang, Rui Wei, Hu Machine Learning Artificial Intelligence Reinforcement Learning with Verifiable Rewards (RLVR) has been shown effective in enhancing the visual reflection and reasoning capabilities of Large Multimodal Models (LMMs). However, existing datasets are predominantly derived from either small-scale manual construction or recombination of prior resources, which limits data diversity and coverage, thereby constraining further gains in model performance. To this end, we introduce \textbf{DeepVision-103K}, a comprehensive dataset for RLVR training that covers diverse K12 mathematical topics, extensive knowledge points, and rich visual elements. Models trained on DeepVision achieve strong performance on multimodal mathematical benchmarks, and generalize effectively to general multimodal reasoning tasks. Further analysis reveals enhanced visual perception, reflection and reasoning capabilities in trained models, validating DeepVision's effectiveness for advancing multimodal reasoning. Data: \href{https://huggingface.co/datasets/skylenage/DeepVision-103K}{this url}. |
| title | DeepVision-103K: A Visually Diverse, Broad-Coverage, and Verifiable Mathematical Dataset for Multimodal Reasoning |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2602.16742 |