Precise Debugging Benchmark: Is Your Model Debugging or Regenerating?

Wang Bill Zhu, Miaosen Chai, Shangshang Wang +5

Unlike code completion, debugging requires localizing faults and applying targeted edits. We observe that frontier LLMs often regenerate correct but over-edited solutions during de…