CoVA-SFT：用于视觉抽象链的大规模数据集

摘要

思维链（Chain-of-thought, CoT）推理通过允许大语言模型（LLMs）将问题分解为中间步骤，显著提升了其性能。尽管CoT在语言任务中广泛有效，但仅文本CoT迫使模型将视觉问题序列化为笨拙的文本描述。虽然已有处理视觉输入的架构性解决方案，但社区仍缺乏一个大规模、多步骤、自我修正的数据集，用以教导模型在解决纯文本推理问题时如何构建和维持内部视觉工作空间。为解决这一局限，我们提出了CoVA-SFT，这是一个高度结构化的语料库，包含51.9K个样本、超过222K个多模态推理步骤，覆盖5种不同的布局家族和17项复杂任务；同时我们还提出了配套基准CoVA-Bench，包含覆盖相同任务的1,700个保留测试样本，用于可复现的评估。通过提供显式的推理依据表述、智能体渲染和验证循环，CoVA-SFT教导多模态语言模型交错使用文本与视觉抽象。我们验证了该数据集，结果表明在CoVA-SFT上微调的模型在CoVA-Bench上的平均性能超过所有交错CoT基线2倍以上，尽管它们仍不及强大的仅文本CoT基线，这凸显了未来工作面临的开放性挑战。

English

Chain-of-thought (CoT) reasoning has dramatically improved large language models (LLMs) by allowing them to decompose problems into intermediate steps. While CoT is widely effective for linguistic tasks, text-only CoT forces models to serialize visual problems into awkward prose. Although architectural solutions exist to process visual inputs, the community lacks a massive, multi-step, self-corrected dataset to teach models how to build and maintain internal visual workspaces when solving purely textual reasoning problems. To address this limitation, we introduce CoVA-SFT, a highly structured corpus of 51.9K samples containing over 222K multimodal reasoning steps across 5 distinct layout families and 17 complex tasks, and CoVA-Bench, a companion benchmark of 1,700 held-out test samples spanning the same tasks for reproducible evaluation. By providing explicit rationale formulations, agentic renderings, and verification loops, CoVA-SFT teaches multimodal language models to interleave text and visual abstractions. We validate the dataset by demonstrating that models fine-tuned on CoVA-SFT outperform all interleaved CoT baselines by more than 2x on average on CoVA-Bench, though they still fall short of strong text-only CoT baselines, highlighting open challenges for future work.