MEGA Hub

OneTrans-GSR: A Generative Slate Ranking Paradigm Unifying Reranking and Mix-Ranking with One Autoregressive Transformer

Authors

Do you know Feng Cheng?You can claim authorship or link another user.Do you know Yu-Hsuan Huang?You can claim authorship or link another user.Do you know Fengyi Wang?You can claim authorship or link another user.Do you know Chaofan Zhou?You can claim authorship or link another user.Do you know Zhennan Zhu?You can claim authorship or link another user.Do you know Huan He?You can claim authorship or link another user.Do you know Bo Zhang?You can claim authorship or link another user.Do you know Ruoqi Zhang?You can claim authorship or link another user.Do you know Zhanfeng Mo?You can claim authorship or link another user.Do you know Ziwei Liu?You can claim authorship or link another user.Do you know Lanyi Li?You can claim authorship or link another user.Do you know Yifeng Pan?You can claim authorship or link another user.Do you know Junhua Long?You can claim authorship or link another user.Do you know Ziqiu Zhou?You can claim authorship or link another user.Do you know Yue Lin?You can claim authorship or link another user.Do you know Shaowei Liu?You can claim authorship or link another user.

Abstract

Industrial recommender systems allocate substantial computation to pointwise candidate scoring, yet separate reranking and mix-ranking stages still assemble the final exposure slate using hand-crafted rules and specialized low-capacity models. Mix-ranking combines product cards, short videos, livestreams, and advertising cards supplied by distinct upstream business channels. Although these rules and models encode valuable domain knowledge, they are inherently limited in leveraging additional data and computation. They constrain the combinatorial space of candidate slates and determine objective trade-offs through manual tuning. To enable scalable and effective slate generation, we reformulate reranking and mix-ranking as a unified learning problem whose performance can improve with computation. Specifically, we instantiate this formulation across the reranking and mix-ranking stages and propose OneTrans-GSR, a single decoder-only policy that extends our deployed OneTrans backbone to generate the exposure slate end-to-end. During training, a reward model evaluates complete slates and guides reinforcement learning toward slate-level and long-term objectives. During decoding, business constraints are explicitly enforced. Offline evaluation shows that slate quality consistently improves as reward-model and policy compute increase. Deployed in a large-scale industrial recommendation system, OneTrans-GSR mixes heterogeneous content types within a unified slate. In online A/B tests, it improves GMV/u by 2.63%, advertiser value by 2.06%, and repurchase rate by 2.03%, while reducing end-to-end latency across the reranking and mix-ranking stages by 13%. These results mark a paradigm shift in slate generation: moving beyond human-engineered heuristics toward a general learned policy creates a scalable path for further algorithmic optimization and compute-driven improvement.

Community

00