1 paper · 1 filter
Jonghyun Song, Sangjun Song, Minjae Oh +3
Large language models (LLMs) achieve strong performance on mathematical reasoning benchmarks, yet the mathematically meaningful skills underlying their reasoning remain underexplor…