When Tables Go Crazy: Evaluating Multimodal Models on French Financial Documents

ArXi:2602.10384v3 Announce Type: replace Vision-language models (VLMs) perform well on many document understanding tasks, yet their reliability in specialized, non-English domains remains underexplored. This gap is especially critical in finance, where documents mix dense regulatory text, numerical tables, and visual charts, and where extraction errors can have real-world consequences. We