1 paper
Xiangxi Zheng, Kuang He, Jiayi Hu +6
Chart-to-code generation demands strict visual precision and syntactic correctness from Vision-Language Models (VLMs). However, existing approaches are fundamentally constrained by…