arXiv 论文：多模态训练让纯文本推理产生"视觉式思维"
Emergent Vision-Shaped Thinking in Text-Only Reasoning through Multimodal Training
这篇论文做了很扎实的消融实验：在 Chameleon-7B 上关掉视觉-语言轴，纯文本空间推理直接掉了 27.6 个点。结论对做模型可解释性或评估的人很有用——探针能看到不等于模型真的在用，得做因果干预才能下结论。
一篇 arXiv 论文提出"涌现视觉式思维"（EVT）概念：多模态训练会改变语言模型纯文本推理的内部机制。研究覆盖 6 个模型家族的 10 种多模态配置，并用匹配的纯文本骨干模型做因果消融对照。
核心实验在 Chameleon-7B 上进行：在推理时消融隐藏状态空间中视觉-语言轴的投影后，BIG-Bench Hard（BBH）navigate 任务准确率下降 27.6 个百分点，而 boolean expressions 任务反而上升 5.6 个百分点，形成 33.2 个百分点的双重分离。该效应还延伸到心理旋转、折纸、视角转换等心理模拟任务，且不能用任务难度或推理深度解释。
机制层面，效应集中在 7-8B 规模，32B 配置中扩展到非空间视觉属性绑定。在 BAGEL、Janus-Pro 这类双通路统一模型中，因果作用点定位在生成通路上的一维载体，理解通路虽可解码出空间信息但并非因果关键。作者指出这解释了行为测量、探针分析和干预实验为何会得出分歧结论。
Emergent Vision-Shaped Thinking in Text-Only Reasoning through Multimodal Training
Does multimodal training shape how a language model thinks, even when the input is text only? We test this with targeted causal ablations across 10 multimodal configurations from 6 model families, with matched text-only backbones and cross-family transfer controls. We contrast spatial-imagery tasks that invite mental scene construction with symbol-manipulation tasks solvable by symbolic rules alone. We identify a vision-language axis in hidden-state space and ablate its projection at inference. In Chameleon-7B, this ablation drops accuracy on BIG-Bench Hard (BBH) navigate by 27.6 pp while leaving boolean expressions intact (+5.6 pp), a 33.2 pp double dissociation. The effect is direction-specific and layer-localized, extends to mental-simulation tasks such as mental rotation, paper folding, and perspective taking, and is not explained by task difficulty, reasoning depth, or sequential update load. Together, these results reveal emergent vision-shaped thinking (EVT): text-only reasoning that causally recruits representations introduced or strengthened by vision-language training, without assuming an image-like internal format. The measured effect is spatially concentrated at 7-8B and broadens to non-spatial visual attribute binding in the tested 32B configuration. In dual-path unified models (BAGEL, Janus-Pro), the causal handle localizes to the generation pathway as a compact one-dimensional carrier. The understanding pathway also encodes spatial information in a higher-rank, probe-decodable form, but is not the causal handle. To separate decodability from causal use, we analyze concentration, alignment, and semantic location. These diagnostics explain why behavior, probing, and intervention can diverge. Multimodal training thus reshapes the geometry of text-only reasoning, not just how visual input is processed.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：Chameleon-7B、BIG-Bench Hard、多模态、模型可解释性，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.AI」（@Qi Zhang, Bohan Wang, Qianru Sun, Hanwang Zhang），经 AITOP 多维评分管线结构化提炼。