Unleashing the Reasoning Potential of Pre-trained LLMs by Critique Fine-Tuning on One Problem Paper • 2506.03295 • Published 8 days ago • 17