LoopRPT: Reinforcement Pre-Training for Looped Language Models Paper • 2603.19714 • Published 6 days ago • 12