1 paper
Aditya Thimmaiah, Jiyang Zhang, Jayanth Srinivasa +2
Recent work asks whether large language models (LLMs) condition their reasoning on explicit rules rather than statistical regularities from pretraining. Program execution provides…