arXiv cs.AI· Yichen Feng, Zhangchen Xu, Fengqing Jiang, Yuetai Li, Bhaskar Ramasubramanian, Luyao Niu, Bill Yuchen Lin, Radha Poovendran·· 11 小时前AI 评分34
VisualSphinx：面向强化学习的大规模合成视觉逻辑谜题数据集
VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RL
AI 导读
VisualSphinx 是一个大规模合成视觉逻辑推理训练数据集，通过“规则到图像”合成流水线从种子题目提取并扩展谜题规则、生成可执行渲染程序来构建带答案的图像。实验用标准 GRPO 加二元正确性奖励验证，在该数据集上训练的 VLM 逻辑推理表现提升，并迁移到代数、算术和几何推理任务。
正文 · 原文
Abstract: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 dataset. To tackle the challenge of image synthesis with grounded answers, we propose a rule-to-image synthesis pipeline that extracts and expands puzzle rules from seed questions and generates executable rendering programs for puzzle assembly. Using standard GRPO with binary correctness rewards as a validation recipe, experiments show that VLMs trained on VisualSphinx benefit from the 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.
| Comments: | DMLR; Project page at this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2505.23977 [cs.CV] |
| (or arXiv:2505.23977v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2505.23977 arXiv-issued DOI via DataCite |
Submission history
From: Fengqing Jiang [view email]
[v1]
Thu, 29 May 2025 20:08:36 UTC (13,296 KB)
[v2]
Fri, 30 Oct 2026 17:01:52 UTC (5,296 KB)
来源：arXiv cs.AI · arxiv.org