一作ACL 2026Findings· Findings of ACL
S hred B ench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLM s in Document Reconstruction
Zichun Guo · Yuling Shi · Wenhao Zeng · Chao Hu · Haotian Lin · Terry Yue Zhuo · Jiawei Chen · Xiaodong Gu · Wenping Ma
摘要
Multimodal Large Language Models (MLLMs) have achieved remarkable performance in Visually Rich Document Understanding (VRDU) tasks, but their capabilities are mainly evaluated on pristine, well-structured document images. We consider document reconstruction from shredded fragments, a challenging...