Trinity-RFT: A General-Purpose and Unified Framework for Reinforcement Fine-Tuning of Large Language Models Paper • 2505.17826 • Published May 23 • 9
