Good SFT Optimizes for SFT, Better SFT Prepares for Reinforcement Learning Paper • 2602.01058 • Published 12 days ago • 39
