Reinforcement Learning from Verifiable Rewards (RLVR)