1 paper · 2 filters
Yang Chen, Yunwen Li, Yufan Shen +6
Recent advancements in LVLMs necessitate robust benchmarks for complex, visually grounded reasoning. A critical limitation is identified in many document understanding benchmarks:…