解决表格转LaTeX时的奖励歧义问题，提升结构、样式、内容三方面的生成质量。
CSPO: Alleviating Reward Ambiguity for Structured Table-to-LaTeX Generation
- 将LaTeX表格的结构、样式、内容分开优化，各自分配专属奖励信号。
- 在多个评估指标上显著优于基线方法，尤其在保留原始表格格式方面表现突出。
- 适合需要高精度结构化文档生成的研究者或工程师使用。
表格包含丰富的结构化信息，但以图像形式存储时其内容被锁定在像素中。将表格图像转换为LaTeX代码可实现内容的精准数字化与复用，但当前多模态大模型常无法保持结构、样式或内容的一致性。传统强化学习微调依赖单一聚合奖励，导致奖励歧义，混淆多种行为特征，阻碍有效优化。我们提出组件特定策略优化（CSPO），一种将LaTeX表格组件——结构、样式、内容——解耦优化的强化学习框架。具体而言，CSPO为各组件分配独立奖励，并仅通过相关标记反向传播对应信号，缓解奖励歧义，实现组件级精准优化。为全面评估性能，我们引入一套分层评估指标。大量实验表明，CSPO有效提升了生成质量，凸显了组件特定优化在可靠结构化生成中的重要性。代码已开源：https://github.com/microsoft/CSPO。
原文摘要 · Abstract (English)
Tables contain rich structured information, yet when stored as images their contents remain "locked" within pixels. Converting table images into LaTeX code enables faithful digitization and reuse, but current multimodal large language models (MLLMs) often fail to preserve structural, style, or content fidelity. Conventional post-training with reinforcement learning (RL) typically relies on a single aggregated reward, leading to reward ambiguity that conflates multiple behavioral aspects and hinders effective optimization. We propose Component-Specific Policy Optimization (CSPO), an RL framework that disentangles optimization across LaTeX tables components-structure, style, and content. In particular, CSPO assigns component-specific rewards and backpropagates each signal only through the tokens relevant to its component, alleviating reward ambiguity and enabling targeted component-wise optimization. To comprehensively assess performance, we introduce a set of hierarchical evaluation metrics. Extensive experiments demonstrate the effectiveness of CSPO, underscoring the importance of component-specific optimization for reliable structured generation. Our code is available at https://github.com/microsoft/CSPO.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。