通知：联邦节假日调休：哥伦布日 / 原住民日（Columbus Day / Indigenous Peoples' Day），今日（13日）无更新。

arXiv 2511.179102026-03-20cs.CL

L2V-CoT：通过潜在干预实现链式推理的跨模态转移

L2V-CoT: Cross-Modal Transfer of Chain-of-Thought Reasoning via Latent Intervention

Yuliang Zhan, Xinyu Tang, Han Wan, Jian Li, Ji-Rong Wen, Hao Sun

备注：: AAAI 2026 oral

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出L2V-CoT方法，通过潜在干预将链式推理从LLM转移到VLM，利用低频潜在表示提升多步推理能力，实验表明优于无训练基线和监督方法。

摘要：

近年来，链式推理（CoT）显著增强了大型语言模型（LLMs）的能力，但视觉-语言模型（VLMs）由于多模态推理数据有限，在多步推理任务中仍面临挑战。为弥合这一差距，研究者探索了将CoT推理从LLMs转移到VLMs的方法。然而，现有方法要么需要高训练成本，要么要求架构对齐。在本文中，我们使用线性人工断层扫描（LAT）实证表明，尽管架构不同，LLMs和VLMs在CoT推理的低频潜在表示上具有相似性。基于这一见解，我们提出L2V-CoT，一种无需训练的潜在干预方法，用于将CoT推理从LLMs转移到VLMs。L2V-CoT在频域中提取并重采样LLMs的低频CoT表示，实现维度匹配，并在推理过程中将潜在表示注入VLMs以增强推理能力。大量实验表明，我们的方法在无训练基线中表现一致，并且甚至超越了监督方法。

英文摘要：

Recently, Chain-of-Thought (CoT) reasoning has significantly enhanced the capabilities of large language models (LLMs), but Vision-Language Models (VLMs) still struggle with multi-step reasoning tasks due to limited multimodal reasoning data. To bridge this gap, researchers have explored methods to transfer CoT reasoning from LLMs to VLMs. However, existing approaches either need high training costs or require architectural alignment. In this paper, we use Linear Artificial Tomography (LAT) to empirically show that LLMs and VLMs share similar low-frequency latent representations of CoT reasoning despite architectural differences. Based on this insight, we propose L2V-CoT, a novel training-free latent intervention approach that transfers CoT reasoning from LLMs to VLMs. L2V-CoT extracts and resamples low-frequency CoT representations from LLMs in the frequency domain, enabling dimension matching and latent injection into VLMs during inference to enhance reasoning capabilities. Extensive experiments demonstrate that our approach consistently outperforms training-free baselines and even surpasses supervised methods.

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。

↑