CSPO: Alleviating Reward Ambiguity for Structured Table-to-LaTeX Generation

• URL: http://arxiv.org/abs/2604.10918v1
• Date: Mon, 13 Apr 2026 02:34:30 GMT
• Title: CSPO: Alleviating Reward Ambiguity for Structured Table-to-LaTeX Generation
• Abstract summary: We propose Component-Specific Policy Optimization (CSPO), a framework that disentangles optimization across tables components-structure, style, and content.<n>CSPO assigns component-specific rewards and backpropagates each signal only through the tokens relevant to its component.
• Score: 45.77327540645401
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• StrucTab: A Structured Optimization Framework for Table Parsing [33.4550862402289]
  StrucTab is a table parsing model learned through intermediate structural supervision and reward decomposition.<n>At the modeling level, by decomposing the parsing process into human-inspired subtasks, StrucTab progressively unifies them through a sequential reasoning strategy.<n>At the optimization level, we introduce Uni-TabRL, a unified RL framework that leverages decomposed rewards to provide stable and informative optimization signals.
  arXiv  Detail & Related papers  (2026-06-29T07:41:39Z)
• Moving Beyond Diversity: Visual Token Pruning as Subspace Reconstruction for Efficient VLMs [7.371189496638082]
  We introduce SPARE, a subspace reconstruction method that reformulates token pruning as a column subset selection problem.<n>We show that SPARE consistently achieves state-of-the-art performance, with strong gains on tasks.
  arXiv  Detail & Related papers  (2026-06-17T04:45:10Z)
• PCA-Seg: Revisiting Cost Aggregation for Open-Vocabulary Semantic and Part Segmentation [58.1914505657064]
  This paper proposes a simple yet effective parallel cost aggregation (PCA-Seg) paradigm to alleviate the challenge of knowledge interference between class-level semantics and spatial context.<n>Experiments on eight benchmarks show that each parallel block in PCA-Seg adds merely 0.35M parameters while achieving state-of-the-art OSPS performance.
  arXiv  Detail & Related papers  (2026-03-18T09:26:43Z)
• BoundRL: Efficient Structured Text Segmentation through Reinforced Boundary Generation [26.825801831400003]
  BoundRL performs token-level text segmentation and label prediction for long structured texts.<n>Instead of generating complete contents for each segment, it generates only a sequence of starting tokens.<n>It reconstructs the complete contents by locating these tokens within the original texts.
  arXiv  Detail & Related papers  (2025-10-23T02:56:10Z)
• Table2LaTeX-RL: High-Fidelity LaTeX Code Generation from Table Images via Reinforced Multimodal Language Models [53.03670032402846]
  We address the task of table image to code generation, with the goal of automating the reconstruction of high-quality, publication-ready tables from visual inputs.<n>A central challenge of this task lies in accurately handling complex tables -- those with large sizes, deeply nested structures, and semantically rich or irregular cell content.<n>We propose a reinforced multimodal large language model (MLLM) framework, where a pre-trained MLLM is fine-tuned on a large-scale table-to-La dataset.
  arXiv  Detail & Related papers  (2025-09-22T11:13:48Z)
• CAL-RAG: Retrieval-Augmented Multi-Agent Generation for Content-Aware Layout Design [6.830055289299306]
  CAL-RAG is a retrieval-augmented, agentic framework for content-aware layout generation.<n>We implement our framework using LangGraph and evaluate it on a benchmark rich in semantic variability.<n>Results demonstrate that combining retrieval augmentation with agentic multi-step reasoning yields a scalable, interpretable, and high-fidelity solution.
  arXiv  Detail & Related papers  (2025-06-27T06:09:56Z)
• Learning Item Representations Directly from Multimodal Features for Effective Recommendation [51.49251689107541]
  multimodal recommender systems predominantly leverage Bayesian Personalized Ranking (BPR) optimization to learn item representations.<n>We propose a novel model (i.e., LIRDRec) that learns item representations directly from multimodal features to augment recommendation performance.
  arXiv  Detail & Related papers  (2025-05-08T05:42:22Z)
• LATex: Leveraging Attribute-based Text Knowledge for Aerial-Ground Person Re-Identification [78.73711446918814]
  We propose a novel framework named LATex for AG-ReID, which adopts prompt-tuning strategies to leverage attribute-based text knowledge.<n>Our framework can fully leverage attribute-based text knowledge to improve AGReID performance.
  arXiv  Detail & Related papers  (2025-03-31T04:47:05Z)
• HIP: Hierarchical Point Modeling and Pre-training for Visual Information Extraction [24.46493675079128]
  OCR-dependent methods rely on offline OCR engines, while OCR-free methods might produce outputs that lack interpretability or contain hallucinated content. We propose HIP, which models entities as HIerarchical Points to better conform to the hierarchical nature of the end-to-end VIE task. Specifically, such hierarchical points can be flexibly encoded and subsequently decoded into desired text transcripts, centers of various regions, and categories of entities.
  arXiv  Detail & Related papers  (2024-11-02T05:00:13Z)
• Contextualization Distillation from Large Language Model for Knowledge Graph Completion [51.126166442122546]
  We introduce the Contextualization Distillation strategy, a plug-in-and-play approach compatible with both discriminative and generative KGC frameworks. Our method begins by instructing large language models to transform compact, structural triplets into context-rich segments. Comprehensive evaluations across diverse datasets and KGC techniques highlight the efficacy and adaptability of our approach.
  arXiv  Detail & Related papers  (2024-01-28T08:56:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.