How well do Large Language Models perform in Arithmetic tasks?

Assessing Arithmetic Competence in Contemporary Language Models Framing and aims At first glance this work positions a compact, targeted benchmark to probe a narrow but important capability: raw arithmetic inside language models. The authors assemble...