Learn the Ropes, Then Trust the Wins: Self-imitation with Progressive Exploration for Agentic Reinforcement Learning Paper • 2509.22601 • Published Sep 26 • 29