AI RESEARCH
Beyond Facts: Benchmarking Distributional Reading Comprehension in Large Language Models
arXiv CS.CL
•
ArXi:2604.06201v1 Announce Type: new While most reading comprehension benchmarks for LLMs focus on factual information that can be answered by localizing specific textual evidence, many real-world tasks require understanding distributional information, such as population-level trends and preferences expressed across collections of text. We