arXiv cs.CL· Jiahao Hui, Lin Zhu, Ben Tan, Yishen Hu, Jingdong Shu, Zetai Jiang, Xining Ran, Yeshou Cai, Gong Chen, Haijie Gu·· 12 小时前AI 评分40
腾讯 KuaFu：将十亿级用户长行为序列压缩为可复用表征
KuaFu: Compressing Long User Behavior into Understanding at Billion Scale
AI 导读
腾讯提出统一用户偏好建模系统 KuaFu，将每个行为条目只编码一次并在用户与下游任务间共享，避免十亿用户规模下的重复序列编码。其双轴投影器把行为序列缩短最多 15 倍、单条目从 150 KB 压到 10 KB，缓存表征降至 0.5 KB；在六个 MRQA 和五个 RecBench 基准上分别最高提升 17.7 EM 和 1.90 AUC。
正文 · 原文
Abstract:Learning user preferences from long behavioral sequences is fundamental to large-scale advertising, recommender systems, and conversational agents. In industrial systems, preference modeling is typically task-wise, with different downstream tasks independently processing the same user history to capture different aspects of preference. This repeated processing creates substantial redundancy at billion-user scale, especially as preferences evolve and representations require frequent updates. We present KuaFu, a unified user preference modeling system that replaces repeated user-level sequence encoding with reusable item-level compressed representations. Each behavioral item is encoded once and shared across users and downstream tasks, while a user's history is built chronologically and extended with new item representations as they arrive, letting KuaFu capture evolving preferences without re-encoding the entire history. To reduce inference and storage overhead, KuaFu introduces a two-axis projector: token-axis compression shortens behavior sequences by up to 15x and reduces each item from 150 KB to 10 KB, while width-axis compression reduces the cached representation to 0.5 KB. We further develop a four-stage training pipeline consisting of reconstruction pre-training, preference-understanding post-training, joint compression-generation training, and hallucination-aware reinforcement learning, progressively improving representation fidelity and downstream utility. Across six MRQA and five RecBench benchmarks, KuaFu consistently outperforms existing compression methods at comparable ratios, with gains of up to 17.7 EM and 1.90 AUC points respectively. KuaFu has been deployed and evaluated on the Tencent advertising and recommendation platform for ten months. Over this period, it improves per-GPU throughput by up to 350%, saves 190 GPUs, and increases overall GMV by 1.37%.
| Comments: | 12 pages, 6 figures, 3 tables |
| Subjects: | Information Retrieval (cs.IR); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.31045 [cs.IR] |
| (or arXiv:2609.31045v2 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2609.31045 arXiv-issued DOI via DataCite |
Submission history
From: Lin Zhu [view email]
[v1]
Fri, 25 Sep 2026 09:39:39 UTC (1,236 KB)
[v2]
Tue, 20 Oct 2026 12:09:04 UTC (1,795 KB)
来源：arXiv cs.CL · arxiv.org