物理推理的因果支架：面向视觉语言模型中因果启发的物理世界理解基准
Causal Scaffolding for Physical Reasoning: A Benchmark for Causally-Informed Physical World Understanding in VLMs
AI总结：
提出CausalPhys基准（含3000+视频/图像问题及因果图），并设计因果图度量评估VLM推理，进一步提出因果理性微调（CRFT）提升推理准确性与可解释性。
摘要：
理解和推理物理世界是智能行为的基础，但最先进的视觉语言模型（VLM）在因果物理推理中仍会失败，常常产生看似合理但错误的答案。为解决这一问题，我们引入了CausalPhys，一个包含超过3000个精心策划的视频和图像问题的基准，涵盖四个领域：感知、预期、干预和目标导向。每个问题都配有一个专家注释的因果图，捕捉对象-属性-事件依赖关系，从而实现可解释且细粒度的因果理解评估。在此基础上，我们制定了一个因果图接地度量，定量衡量模型的思维链推理与正确因果关系的对齐程度，超越了仅基于答案的准确性，并能够系统诊断VLM的因果推理失败。使用该度量，我们对领先的VLM进行了全面分析，揭示了在捕捉因果依赖关系方面的系统性差距，并强调了因果感知学习的必要性。为解决这些局限性，我们进一步提出了因果理性微调（CRFT），明确将VLM推理与因果结构对齐。大量实验表明，CRFT在多个模型骨干上显著提升了推理准确性和可解释性。通过统一数据集整理、因果评估和因果感知学习，CausalPhys为推进现代VLM实现因果接地物理推理奠定了坚实基础。
英文摘要：
Understanding and reasoning about the physical world is the foundation of intelligent behavior, yet state-of-the-art vision-language models (VLMs) still fail at causal physical reasoning, often producing plausible but incorrect answers. To address this gap, we introduce CausalPhys, a benchmark of over 3,000 carefully curated video- and image-based questions spanning four domains: Perception, Anticipation, Intervention, and Goal Orientation. Each question is paired with an expert-annotated causal graph capturing object-attribute-event dependencies, enabling interpretable and fine-grained evaluation of causal understanding. Building on this, we formulate a causal-graph-grounded metric that quantitatively measures how well a model's chain-of-thought reasoning aligns with the correct causal relations, moving beyond answer-only accuracy and enabling systematic diagnosis of VLMs' causal reasoning failures. Using this metric, we conduct a comprehensive analysis of leading VLMs, revealing systematic gaps in capturing causal dependencies and underscoring the need for causality-aware learning. To address these limitations, we further propose Causal Rationale-informed Fine-Tuning (CRFT), which explicitly aligns VLM reasoning with causal structures. Extensive experiments demonstrate that CRFT substantially enhances both reasoning accuracy and interpretability across multiple model backbones. By unifying dataset curation, causal evaluation, and causality-informed learning, CausalPhys establishes a strong foundation for advancing modern VLMs toward causally grounded physical reasoning.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
