CharTool: Tool-Integrated Visual Reasoning for Chart Understanding
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918426316898304 |
|---|---|
| author | Zhang, Situo Zhang, Yifan Zhu, Zichen Ma, Da Pan, Lei Zhang, Danyang Zhao, Zihan Chen, Lu Yu, Kai |
| author_facet | Zhang, Situo Zhang, Yifan Zhu, Zichen Ma, Da Pan, Lei Zhang, Danyang Zhao, Zihan Chen, Lu Yu, Kai |
| contents | Charts are ubiquitous in scientific and financial literature for presenting structured data. However, chart reasoning remains challenging for multimodal large language models (MLLMs) due to the lack of high-quality training data, as well as the need for fine-grained visual grounding and precise numerical computation. To address these challenges, we first propose DuoChart, a scalable dual-source data pipeline that combines synthesized charts with real-world charts to construct diverse, high-quality chart training data. We then introduce CharTool, which equips MLLMs with external tools, including image cropping for localized visual perception and code-based computation for accurate numerical reasoning. Through agentic reinforcement learning on DuoChart, CharTool learns tool-integrated reasoning grounded in chart content. Extensive experiments on six chart benchmarks show that our method consistently improves over strong MLLM baselines across model scales. Notably, CharTool-7B outperforms the base model by **+8.0%** on CharXiv (Reasoning) and **+9.78%** on ChartQAPro, while achieving competitive performance with substantially larger or proprietary models. Moreover, CharTool demonstrates positive generalization to out-of-domain visual math reasoning benchmarks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_02794 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CharTool: Tool-Integrated Visual Reasoning for Chart Understanding Zhang, Situo Zhang, Yifan Zhu, Zichen Ma, Da Pan, Lei Zhang, Danyang Zhao, Zihan Chen, Lu Yu, Kai Artificial Intelligence Charts are ubiquitous in scientific and financial literature for presenting structured data. However, chart reasoning remains challenging for multimodal large language models (MLLMs) due to the lack of high-quality training data, as well as the need for fine-grained visual grounding and precise numerical computation. To address these challenges, we first propose DuoChart, a scalable dual-source data pipeline that combines synthesized charts with real-world charts to construct diverse, high-quality chart training data. We then introduce CharTool, which equips MLLMs with external tools, including image cropping for localized visual perception and code-based computation for accurate numerical reasoning. Through agentic reinforcement learning on DuoChart, CharTool learns tool-integrated reasoning grounded in chart content. Extensive experiments on six chart benchmarks show that our method consistently improves over strong MLLM baselines across model scales. Notably, CharTool-7B outperforms the base model by **+8.0%** on CharXiv (Reasoning) and **+9.78%** on ChartQAPro, while achieving competitive performance with substantially larger or proprietary models. Moreover, CharTool demonstrates positive generalization to out-of-domain visual math reasoning benchmarks. |
| title | CharTool: Tool-Integrated Visual Reasoning for Chart Understanding |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.02794 |