Chart Deep Research in LVLMs via Parallel Relative Policy Optimization

TL;DR

This paper introduces PRPO, a parallel reward optimization method, and MCDR-Bench, an evaluation framework, to enhance deep research capabilities in LVLMs for chart understanding, addressing training and evaluation limitations.

Contribution

The paper presents a novel parallel reward optimization technique and a new benchmark for objective evaluation, advancing deep research in LVLMs for chart analysis.

Findings

PRPO effectively disentangles multi-dimensional reward signals.

MCDR-Bench enables objective assessment of deep research capabilities.

Experimental results show improved model performance in deep chart understanding.

Abstract

With the rapid advancement of data science, charts have evolved from simple numerical presentation tools to essential instruments for insight discovery and decision-making support. However, current chart data intelligence exhibits significant limitations in deep research capabilities, with existing methods predominantly addressing shallow tasks such as visual recognition or factual question-answering, rather than the complex reasoning and high-level data analysis that deep research requires. This limitation stems from two primary technical bottlenecks: at the training level, existing post-training techniques exhibit deficiencies in handling multi-dimensional reward signal interference and heterogeneous data gradient conflicts, preventing models from achieving balanced development across multiple capability dimensions; at the evaluation level, current methods remain limited to factual retrieval and basic computation, failing to assess end-to-end analytic reasoning and other deep research capabilities. To address the training challenge, we propose PRPO, which performs parallel optimization across reward dimensions and capability partitioning across data types, effectively disentangling conflicts between heterogeneous data and multi-dimensional reward signals while ensuring optimization stability. For the evaluation challenge, we construct MCDR-Bench based on the ``error uniqueness principle," transforming subjective generation assessment into objective error identification through controllable error injection, enabling quantifiable evaluation of deep research capabilities. Experimental validation confirms that the proposed PRPO and MCDR-Bench jointly establish a unified framework that systematically advances chart deep research through enhanced collaborative training and objective evaluation.