ProSR: 面向可靠思维链的过程塑造空间推理方法
ProSR: Process-Shaped Spatial Reasoning for Reliable Chain-of-Thought in VLMs
AI总结：
针对视觉语言模型在空间推理中存在的虚假基础与尾部不稳定性问题，提出ProSR框架，通过反事实不变性惩罚和尾部漂移惩罚优化推理过程，提升答案准确率及轨迹稳定性与视觉依赖性。
摘要：
可靠的空间推理仍然是视觉语言模型（VLM）的核心瓶颈。现有的空间推理主流训练范式主要依赖于结果对齐或过程模仿，缺乏对推理过程的显式约束，因此难以确保真正的视觉依赖和稳定的推理轨迹。在本文中，我们构建了一个覆盖多种空间现象的高质量思维链数据集，并诊断了模型的推理过程，揭示了强化学习优化过程中两种典型的过程退化类型：虚假基础（绕过视觉证据）和尾部不稳定性（推理后期不确定性异常上升）。为了解决这些问题，我们提出了ProSR，一种用于空间推理的过程塑造优化框架。通过反事实不变性惩罚和尾部漂移惩罚，ProSR将优化目标从单一的答案正确性扩展到两个过程级维度：视觉依赖性和轨迹稳定性。在多个复杂和分布外的空间推理基准上的实验表明，ProSR在提高答案准确率的同时，生成的推理轨迹更加稳定且更依赖于视觉证据。
英文摘要：
Reliable spatial reasoning remains a core bottleneck for vision-language models (VLMs). Existing mainstream training paradigms for spatial reasoning largely rely on outcome alignment or process imitation, lacking explicit constraints on the reasoning process, and therefore struggle to ensure genuine visual dependence and stable reasoning trajectories. In this paper, we construct a high-quality CoT dataset covering diverse spatial phenomena and diagnose the model's reasoning process, revealing two typical types of process degradation during reinforcement learning optimization: Spurious Grounding, which bypasses visual evidence, and Tail Instability, where uncertainty abnormally rises in the later stage of reasoning. To address these issues, we propose ProSR, a process-shaping optimization framework for spatial reasoning. Through a Counterfactual Invariance Penalty and a Tail Drift Penalty, ProSR extends the optimization objective from single answer correctness to two process-level dimensions: visual dependence and trajectory stability. Experiments on multiple complex and out-of-distribution spatial reasoning benchmarks show that ProSR improves answer accuracy while generating reasoning trajectories that are more stable and more dependent on visual evidence.
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。
