UniPolicy is a multi-policy alignment framework for search advertising that jointly optimizes relevance, click propensity, and commercial value. It uses objective-specific prefix tokens, sparse MoE-LoRA routing, and residual FFNs to decouple parameters within a shared backbone, and builds pairwise preferences from multi-stage behavioral feedback to improve generation. In large-scale offline tests and a 7‑day online A/B test, UniPolicy achieved balanced gains across metrics, boosting CTR by 0.71%, RPS by 1.58%, and revenue by 1.32% while keeping latency stable.
arXiv:2512.23067v3 Announce Type: replace
Abstract: Personalized alignment from preference data has focused primarily on improving personal reward model (RM) accuracy, with the implicit assumption th...
By Fady Rezk, Yuangang Pan, Chuan-Sheng Foo, Xun Xu, Nancy Chen, Henry Gouk, Timothy Hospedales
arXiv:2609.08211v1 Announce Type: new
Abstract: Real-world Multi-Objective Reinforcement Learning (MORL) often suffers from sparse rewards, reward conflicts, and late-stage reward tug-of-war, causing...
By Shanwen Mao, Hao Zhang, Guangtao nie, Zhiheng Li, Huimu Wang, Sulong Xu, Gu Simiu
OneBid is a unified auto‑bidding foundation model that consolidates diverse cost‑per‑X (oCPX) advertising scenarios into a single framework. It builds on Decision Transformer by conditioning on two atomic signals—Return‑to‑Go for conversion value and Cost‑to‑Go for cost ratio—and incorporates value‑aware regularization. A sequence‑level Mixture‑of‑Experts architecture captures cross‑scenario knowledge while preserving low latency, and a Critic‑guided Relative Offline Policy optimization (CROP) aligns the backbone with scenario‑specific preferences without unsafe online exploration. In production at Kuaishou, OneBid achieved a 2.2% overall ADVV increase and up to 13.1% in the ROAS scenario.
By Yewen Li, Peng Jiang, Yitian Li, Pengfei Lv, Xialong Liu, Peng Jiang, Qingpeng Cai
The paper introduces GR4AD‑X, an enhanced generative recommendation framework for large‑scale advertising that tackles three key limitations of its predecessor: unreliable deterministic SID assignment, limited scalability of backbone models, and weak credit assignment in sequence‑level value optimization. GR4AD‑X improves SID reliability via Multi‑Path assignment, expands model capacity across complementary dimensions for better efficiency, and aligns generation with commercial value through Unified Value‑Guided Generation and Discrimination (UniVGD), which optimizes prefix‑level rewards, prunes beams, and reranks discriminatively. Online A/B tests on Kuaishou’s advertising platform show a 3.07% revenue increase over the baseline while meeting strict latency and throughput constraints.
By Yatong Xiao, Yibo Sheng, Lixiang Wang, Mingjie Sun, Yunhao Sha, Pengfei Zhang, Shengyu Wang, Peng Wang, Wenjin Wu, Peng Jiang
arXiv:2510. 01167v2 Announce Type: replace-cross Abstract: Aligning large language models to human preferences is inherently multidimensional, yet most pipelines collapse heterogeneous signals into a single objective.
By Yiran Shen, Yu Xia, Jonathan Chang, Prithviraj Ammanabrolu