From Diagnosis to Improvement: Probing Spatio-Physical Reasoning in Vision Language Models
This addresses the need for robust world models in AI, but it is incremental as it builds on existing methods with limited generalization.
The paper tackled the problem of spatio-physical reasoning in vision language models, revealing that current models perform inadequately due to biases and lack of deep reasoning, and improved Qwen2.5-VL-7B to surpass leading proprietary models.
Spatio-physical reasoning, a foundation capability for understanding the real physics world, is a critical step towards building robust world models. While recent vision language models (VLMs) have shown remarkable progress in specialized domains like multimodal mathematics and pure spatial understanding, their capability for spatio-physical reasoning remains largely unexplored. This paper provides a comprehensive diagnostic analysis of mainstream VLMs, revealing that current models perform inadequately on this crucial task. Further detailed analysis shows that this underperformance is largely attributable to biases caused by human-like prior and a lack of deep reasoning. To address these challenges, we apply supervised fine-tuning followed by rule-based reinforcement learning to Qwen2.5-VL-7B, resulting in significant improvements in spatio-physical reasoning capabilities and surpassing leading proprietary models. Nevertheless, despite this success, the model's generalization to new physics scenarios remains limited -- underscoring the pressing need for new approaches in spatio-physical reasoning.