TY  - RPRT
TI  - Learning to Reflect and Correct: Towards Better Decoding Trajectories for Large-Scale Generative Recommendation
AU  - Haibo Xing
AU  - Hao Deng
AU  - Lingyu Mu
AU  - Jinxin Hu
AU  - Yu Zhang
AU  - Xiaoyi Zeng
AU  - Jing Zhang
PY  - 2026
UR  - https://arxiv.org/abs/2602.23639
ID  - 2602.23639
ER  -