1
ChartAnno: Evaluating MLLMs for Chart Annotation Generation
首个系统评估多模态大模型生成图表注释能力的基准，揭示模型在数据推理与视觉布局上的短板。
arXiv:2608.03464v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) have made significant progress in chart understanding, gene…