论文翻译：ICLR-2023.DYVAL: DYNAMIC EVALUATION OF LARGE LANGUAGE MODELS FOR REASONING TASKS

DYVAL: DYNAMIC EVALUATION OF LARGE LANGUAGE MODELS FOR REASONING TASKS

https://openreview.net/forum?id=gjfOL9z5Xr

Dynamic evaluation to mitigate potential test data contamination: we integrated the dynamic evaluation framework DyVal [5], which generates evaluation samples on-the-fly with controlled complexity. 来自：https://github.com/microsoft/promptbench

b站讲解视频：https://www.bilibili.com/video/BV1iSUmYREub

DYVAL：大型语言模型推理任务的动态评估

摘要

大型语言模型（LLMs）在各种