arXiv cs.CL· Zhiyuan Shi, Xiang Li, Xingzhuo Guo, Shangchen Miao, Weiliang Chen, Qiyu Dai, Jinshan Ren, Jiangran Lyu, Fangfu Liu, Mingsheng Long, Jialong Wu·· 6 天前AI 评分31
S-Space：多模态语言模型中的空间工作区
S-Space: Exploring Spatial Workspace in Multimodal Language Models
AI 导读
研究者发现多模态语言模型内部存在 S-Space 线性表示子空间，物体 token 表示在其中编码水平、垂直和距离三个连续坐标轴，且该子空间是模型固有的，在非空间问题下仍可读取。S-Space 充当内部空间工作区：感知将物体位置编码其中，内部推理可变换视角，语言回答选择性地访问它。研究者还将其用作诊断工具，发现模型能形成空间表示，但在跨视角变换推理上存在瓶颈。
正文 · 原文
Authors:Zhiyuan Shi, Xiang Li, Xingzhuo Guo, Shangchen Miao, Weiliang Chen, Qiyu Dai, Jinshan Ren, Jiangran Lyu, Fangfu Liu, Mingsheng Long, Jialong Wu
Abstract:Multimodal language models (MLLMs) exhibit promising spatial reasoning capabilities, yet the internal mechanisms supporting them remain poorly understood. We uncover S-Space, a linear representation subspace in which object-token representations encode continuous coordinates along horizontal, vertical, and distance axes. Although we identify these axes by differentiating spatial-answer scores with respect to object-token activations, S-Space is intrinsic rather than induced by specific data or prompts, remaining readable under non-spatial questions. We show that S-Space functions as an internal spatial workspace of MLLMs. Perception encodes object locations into it, forming spatial representations that can be directly read out to support accurate spatial judgments. Internal reasoning transforms it for changing viewpoints. Verbal responses selectively access it, evidenced by causal interventions that alter spatial judgments while largely preserving non-spatial responses. Case studies further show that S-Space assigns locations to unseen entities and the observer, and connects physical directions with abstract linguistic meanings, suggesting both the capabilities and limitations of large-scale language training. Finally, we use S-Space as a diagnostic tool for spatial reasoning failures. Direction-word readouts reveal systematic coupling between cardinal and image-relative reference frames. More importantly, a perception-computation decomposition exposes a key bottleneck: MLLMs can form capable spatial representations, yet fall short in the spatial transformations required to reason reliably across viewpoint changes.
| Comments: | Project Page: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.13527 [cs.CV] |
| (or arXiv:2610.13527v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.13527 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jialong Wu [view email]
[v1]
Fri, 9 Oct 2026 12:48:41 UTC (13,229 KB)
来源：arXiv cs.CL · arxiv.org