Somak Aditya

This study assesses the performance of language models on modified math problems.

Code prompts enhance reasoning performance in large language models like GPT 3.5.

MathSensei combines tools to improve solving math challenges effectively.