Learning to Reflect and Correct: Towards Better Decoding Trajectories for Large-Scale Generative Recommendation

Re-assign community