DynamicBench: Evaluating Real-Time Report Generation in Large Language Models

Jingyao Li, Hao Sun, Zile Qiao +5

Traditional benchmarks for large language models (LLMs) typically rely on static evaluations through storytelling or opinion expression, which fail to capture the dynamic requireme…