# ChartCitor: Multi-Agent Framework for Fine-Grained Chart Visual Attribution

Source: [ChartCitor: Multi-Agent Framework for Fine-Grained Chart Visual Attribution](<https://www.sota2.com/research/paper/68fa76d11e209131cea6f43f>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T04:03:16.343Z

Authors: Kanika Goswami, Puneet Mathur, Ryan Rossi, Franck Dernoncourt

Published: 2025-02-03T02:00:51.000Z

## Abstract

Large Language Models (LLMs) can perform chart question-answering tasks but often generate unverified hallucinated responses. Existing answer attribution methods struggle to ground responses in source charts due to limited visual-semantic context, complex visual-text alignment requirements, and difficulties in bounding box prediction across complex layouts. We present ChartCitor, a multi-agent framework that provides fine-grained bounding box citations by identifying supporting evidence within chart images. The system orchestrates LLM agents to perform chart-to-table extraction, answer reformulation, table augmentation, evidence retrieval through pre-filtering and re-ranking, and table-to-chart mapping. ChartCitor outperforms existing baselines across different chart types. Qualitative user studies show that ChartCitor helps increase user trust in Generative AI by providing enhanced explainability for LLM-assisted chart QA and enables professionals to be more productive.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2502.00989>)

[PDF](<https://arxiv.org/pdf/2502.00989.pdf>)
