IFDECORATOR: Wrapping Instruction Following Reinforcement Learning with Verifiable Rewards Paper • 2508.04632 • Published 16 days ago • 2