Counting Without Running: Evaluating LLMs' Reasoning About Code Complexity

Open in new window