Overview

Generative recommendation studies how to produce personalized items or ranked lists directly from user histories and context, instead of scoring a fixed candidate set. Common approaches encode items as semantic IDs and train autoregressive Transformers or language models, while diffusion, flow-based generation, and continuous retrieval offer alternatives. Training combines next-item supervision with semantic or behavioral signals, and sometimes reinforcement learning; evaluation uses sequential benchmarks such as Amazon datasets alongside production A/B tests. A central challenge is balancing useful item representations and reachability—including for cold or long-tail items—with accuracy, fairness, and the latency demands of large-scale serving.

Latest papers

CardsList

122 papers21 papers in the last four weeks, up 62% on the four weeks before. 0.2% of all new papers.

  1. xGR: Efficient Generative Recommendation Serving at Scale

    Dec 12, 2025Qingxiao Sun, Tongxuan Liu, Shen Zhang +13LLM ServingLLM Inference Acceleration

  2. CoFiRec: Coarse-to-Fine Tokenization for Generative Recommendation

    Nov 27, 2025Tianxin Wei, Xuying Ning, Xuxing Chen +6Sequential RecommendationHierarchical Representation Learning