[论文] Can machines reason non-verbally?

论文概要

研究领域: CV 作者: Shaheer U. Saeed, Zion Tse, Gregory Slabaugh, Daniel C. Alexander, Yipeng Hu 发布时间: 2026-10-09 arXiv: 2610.13925

中文摘要

现代机器能解方程、遵逻辑、用语言推理，但在需要视觉非语言推理的问题上仍举步维艰——即便 DeepSeek、ChatGPT、Gemini 等先进系统在抽象模式补全或模糊兴趣区域勾画等视觉任务上仍表现有限。语言推理中的显式中间步骤可用逻辑或自动测试验证，而视觉推理的中间步骤是隐式的，缺乏明确验证方法，因此建模困难，构成当前机器智能的根本局限。我们提出 GenVerOpt：将生成、验证与优化统一为推理流水线的通用框架。通过生成候选解、验证正确性、迭代改进，实现逐样本优化而非仅依赖训练时捕捉的数据集级规律。这一过程模拟人类解题：想象多种可能、评估、精炼直至正确答案浮现。GenVerOpt 在多领域超越现有系统：在抽象 IQ 谜题上超越人类，在前列腺癌分割上匹配临床医生，在肝脏血管重建上超越专家，在肿瘤边界对齐上优于最先进水平。这些结果表明机器推理可以超越语言，视觉非语言推理可成为更灵活、更通用智能的基础。

原文摘要

Modern machines can solve equations, follow logic and reason through language, yet they still struggle with problems that require visual non-verbal reasoning. This is evident even in advanced systems such as DeepSeek, ChatGPT and Gemini, which remain limited on visual tasks such as completing abstract patterns or delineating ambiguous regions-of-interest. Unlike explicit intermediate steps in verbal reasoning which can be verified through logic or automated tests, visual reasoning involves implicit intermediate steps with no clear methods for verification. Non-verbal reasoning is therefore challenging to model and, as a result, remains a fundamental limitation of current machine intelligence. Here we introduce GenVerOpt, a general framework that unifies generation, verification and optimis...

#论文 #arXiv #CV #小凯