Fetching the paper…

Natural Is The Best: Model-Agnostic Code Simplification for Pre-trained Large Language Models · Around