VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RL
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909628242067456 |
|---|---|
| author | Feng, Yichen Xu, Zhangchen Jiang, Fengqing Li, Yuetai Ramasubramanian, Bhaskar Niu, Luyao Lin, Bill Yuchen Poovendran, Radha |
| author_facet | Feng, Yichen Xu, Zhangchen Jiang, Fengqing Li, Yuetai Ramasubramanian, Bhaskar Niu, Luyao Lin, Bill Yuchen Poovendran, Radha |
| contents | Vision language models (VLMs) are expected to perform effective multimodal reasoning and make logically coherent decisions, which is critical to tasks such as diagram understanding and spatial problem solving. However, current VLM reasoning lacks large-scale and well-structured training datasets. To bridge this gap, we propose VisualSphinx, a first-of-its-kind large-scale synthetic visual logical reasoning training data. To tackle the challenge of image synthesis with grounding answers, we propose a rule-to-image synthesis pipeline, which extracts and expands puzzle rules from seed questions and generates the code of grounding synthesis image synthesis for puzzle sample assembly. Experiments demonstrate that VLM trained using GRPO on VisualSphinx benefit from logical coherence and readability of our dataset and exhibit improved performance on logical reasoning tasks. The enhanced reasoning capabilities developed from VisualSphinx also benefit other reasoning tasks such as algebraic reasoning, arithmetic reasoning and geometry reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23977 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RL Feng, Yichen Xu, Zhangchen Jiang, Fengqing Li, Yuetai Ramasubramanian, Bhaskar Niu, Luyao Lin, Bill Yuchen Poovendran, Radha Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Vision language models (VLMs) are expected to perform effective multimodal reasoning and make logically coherent decisions, which is critical to tasks such as diagram understanding and spatial problem solving. However, current VLM reasoning lacks large-scale and well-structured training datasets. To bridge this gap, we propose VisualSphinx, a first-of-its-kind large-scale synthetic visual logical reasoning training data. To tackle the challenge of image synthesis with grounding answers, we propose a rule-to-image synthesis pipeline, which extracts and expands puzzle rules from seed questions and generates the code of grounding synthesis image synthesis for puzzle sample assembly. Experiments demonstrate that VLM trained using GRPO on VisualSphinx benefit from logical coherence and readability of our dataset and exhibit improved performance on logical reasoning tasks. The enhanced reasoning capabilities developed from VisualSphinx also benefit other reasoning tasks such as algebraic reasoning, arithmetic reasoning and geometry reasoning. |
| title | VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RL |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2505.23977 |