Utility-Aware Generative Recommendation: Utility-Enriched Semantic Codes and Utility-Guided Decoding
cs.LG
Submitted: 2026-11-02
Updated: 2026-11-02
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
The gist: Generative recommendation has been widely adopted in large-scale industrial recommendation systems, where multimodal item embeddings are quantized into offline semantic identifiers (SIDs) and items…
Terminology
Abstract
Generative recommendation has been widely adopted in large-scale industrial recommendation systems, where multimodal item embeddings are quantized into offline semantic identifiers (SIDs) and items are retrieved by autoregressively generating their SIDs. However, conventional SID construction often underrepresents item utility reflected in post-exposure behaviors such as shares, long views, and video completion, making semantically similar items with different utility insufficiently distinguishable while struggling to accommodate rapidly changing utility signals. Moreover, generation likelihood may be misaligned with downstream utility, causing high-utility candidates to be prematurely pruned during beam search before post-generation reranking or training-time reward alignment can effectively compensate. To address these limitations, we propose UAGR, which aligns both SID construction and decoding with downstream multi-objective utility. Our Utility-Enriched Semantic Codes preserve the original semantic information while incorporating item-level utility without increasing SID length or vocabulary size, supporting incremental utility updates with no additional tokenization or sequence-decoding cost. Meanwhile, our Utility-Guided Decoding estimates candidate utility by combining predictions of multiple post-exposure outcomes and incorporates these estimates into beam scoring, reducing the risk of prematurely pruning high-utility paths while enabling flexible utility control without interfering with generative backbone training. Extensive offline experiments and online A/B tests under real-world traffic demonstrate the effectiveness and deployability of UAGR, which has been fully deployed on Kuaishou and increases video watch time by 0.206% online.
Sources
- Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
- OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search
- OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment
- RankGR: Rank-Enhanced Generative Retrieval with Listwise Direct Preference Optimization in Recommendation
- End-to-End Semantic ID Generation for Generative Advertisement Recommendation
- Generative Recommendation with Semantic IDs: A Practitioner's Handbook
- Breaking the Hourglass Phenomenon of Residual Quantization: Enhancing the Upper Bound of Generative Retrieval
- HiGR: Industrial-Scale Hierarchical Generative Slate Recommendation Framework in Tencent
- Generative Sequential Recommendation with GPTRec
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- History-Conditioned Joint-Prefix Alignment for Generative Recommendation
- Beyond Semantic IDs: Finite-Beam Reachability and Continuous-Query Generation in Generative Recommendation
- Understanding Semantic IDs: From Item Representation to Item Selection in Generative Recommendation
- OneLive: Dynamically Unified Generative Framework for Live-Streaming Recommendation
- Generative Recommendation for Large-Scale Advertising
- Unifying Generative and Dense Retrieval for Sequential Recommendation
- How Reliable Are Semantic-ID Tokenizer Comparisons in Generative Recommendation?
- Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models
- OpenOneRec Technical Report
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis