Elamur Research⌕
☰
home / papers / 2026 / Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers

Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers

Paper · 2026 · Zhiyang Zhang, Junda She, Kuo Cai et al.

Summary

LLM-based generative recommenders must name items in a form the model can produce and that maps back to real catalog entries. Free-text identifiers such as titles often lead to hallucinated, non-existent items. Semantic IDs are numeric codes outside the LLM's vocabulary, so they require vocabulary expansion and alignment training, and they transfer poorly across domains. The authors, from Kuaishou, propose GRLM, built on Term IDs (TIDs): short, standardized keyword sequences drawn from the model's native vocabulary.

GRLM has three components:

  • Context-aware Term Generation shows an LLM the metadata of embedding-retrieved neighbor items. This keeps shared attributes consistent while still capturing what makes each item distinct.
  • Integrative Instruction Fine-tuning jointly trains item-to-TID mapping and sequential recommendation.
  • Elastic Identifier Grounding maps generated TIDs to items, first by exact match and then by a structural fallback.

On real-world in-domain and cross-domain benchmarks, GRLM beats the baselines, scales with backbone size and almost never hallucinates. This suggests native-vocabulary identifiers are a practical alternative to Semantic IDs.

Key points

  • Term generation uses Qwen3-4B-2507, and the fine-tuned backbones are Qwen3 models from 0.6B to 14B. Recall rises steadily with model size, consistent with scaling-law behavior.
  • In the hallucination analysis, both Valid Rate@10 and Direct Hit Rate@10 exceed 0.99 on all reported datasets: Beauty, Sports, Toys, Leisure and Technology.
  • In cross-domain experiments, GRLM outperforms HSTU, LLM4CDSR, GenCDR and TriCDR, with relative gains reaching about 113% on some metrics.
  • Ablations show that removing either Context-aware Term Generation or the Generative Term Internalization task lowers Recall and NDCG on every in-domain dataset.
  • The authors note two limitations: retrieval relies on a fixed external embedding model, and validation was done only on Qwen3. Code is available at github.com/ZY0025/GRLM.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🏢KuaishouBeijing · China19 works · 95 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksInformation Systems · 1,730 works

Cite

Cite as Zhiyang Zhang, Junda She, Kuo Cai, Bo Chen, Shiyao Wang, Xinchen Luo, Qiang Luo, Ruiming Tang, Han Li, Kun Gai and Guorui Zhou. 2026. Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.984

@inproceedings{zhang2026unleashing,
  title = {{Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers}},
  author = {Zhiyang Zhang, Junda She, Kuo Cai, Bo Chen, Shiyao Wang, Xinchen Luo, Qiang Luo, Ruiming Tang, Han Li, Kun Gai and Guorui Zhou},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.984},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.984}
}

References

39 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774.
  2. Keqin Bao, Jizhi Zhang, Wenjie Wang, Yang Zhang, Zhengyi Yang, Yanchen Luo, Chong Chen, Fuli Feng, and Qi Tian. 2025. A bi-step grounding paradigm for large language models in recommendation systems. ACM Transactions on Recommender Systems, 3(4):1– 27. find
  3. Keqin Bao, Jizhi Zhang, Yang Zhang, Wenjie Wang, Fuli Feng, and Xiangnan He. 2023. Tallrec: An effective and efficient tuning framework to align large language model with recommendation. In Proceedings of the 17th ACM conference on recommender systems, pages 1007–1014.
  4. Tom B Brown. 2020. Language models are few-shot learners. arXiv preprint arXiv:2005.14165.
  5. Yuxin Chen, Junfei Tan, An Zhang, Zhengyi Yang, Leheng Sheng, Enzhi Zhang, Xiang Wang, and Tat-Seng Chua. 2024. On softmax direct preference optimization for recommendation. Advances in Neural Information Processing Systems, 37:27463–27489. find
  6. Zhixuan Chu, Hongyan Hao, Xin Ouyang, Simeng Wang, Yan Wang, Yue Shen, Jinjie Gu, Qing Cui, Longfei Li, Siqiao Xue, and 1 others. 2023. Leveraging large language models for pre-trained recommender systems. arXiv preprint arXiv:2308.10837. — [arXiv:2308.10837](https://arxiv.org/abs/2308.10837) find
  7. Paul Covington, Jay Adams, and Emre Sargin. 2016. Deep neural networks for youtube recommendations. In Proceedings of the 10th ACM conference on recommender systems, pages 191–198. find
  8. Yashar Deldjoo, Zhankui He, Julian McAuley, Anton Korikov, Scott Sanner, Arnau Ramisa, René Vidal, Maheswaran Sathiamoorthy, Atoosa Kasirzadeh, and Silvia Milano. 2024. A review of modern recommender systems using generative models (gen-recsys). In Proceedings of the 30th ACM SIGKDD conference on Kn… find
  9. Jiaxin Deng, Shiyao Wang, Kuo Cai, Lejian Ren, Qigen Hu, Weifeng Ding, Qiang Luo, and Guorui Zhou. 2025. Onerec: Unifying retrieve and rank with generative recommender and iterative preference alignment. arXiv preprint arXiv:2502.18965. — [arXiv:2502.18965](https://arxiv.org/abs/2502.18965) find
  10. Ruidong Han, Bin Yin, Shangyu Chen, He Jiang, Fei Jiang, Xiang Li, Chi Ma, Mincong Huang, Xiaoguang Li, Chunzhen Jing, and 1 others. 2025. Mtgr: Industrial-scale generative recommendation framework in meituan. arXiv preprint arXiv:2505.18654. — [arXiv:2505.18654](https://arxiv.org/abs/2505.18654) find
  11. Peiyu Hu, Wayne Lu, and Jia Wang. 2025. From ids to semantics: A generative framework for cross-domain recommendation with adaptive semantic tokenization. arXiv preprint arXiv:2511.08006. — [arXiv:2511.08006](https://arxiv.org/abs/2511.08006) find
  12. Yanhua Huang, Yuqi Chen, Xiong Cao, Rui Yang, Mingliang Qi, Yinghao Zhu, Qingchang Han, Yaowei Liu, Zhaoyu Liu, Xuefeng Yao, and 1 others. 2025. Towards large-scale generative ranking. arXiv preprint arXiv:2505.04180. — [arXiv:2505.04180](https://arxiv.org/abs/2505.04180) find
  13. Wang-Cheng Kang and Julian McAuley. 2018. Self-attentive sequential recommendation. In 2018 IEEE international conference on data mining (ICDM), pages 197–206. IEEE. find
  14. Saketh Reddy Karra and Theja Tulabandhula. 2024. Interarec: Interactive recommendations using multimodal large language models. In Pacific-Asia Conference on Knowledge Discovery and Data Mining, pages 32–43. Springer. find
  15. Doyup Lee, Chiheon Kim, Saehoon Kim, Minsu Cho, and Wook-Shin Han. 2022. Autoregressive image generation using residual quantization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11523–11532. find
  16. Jiayi Liao, Sihang Li, Zhengyi Yang, Jiancan Wu, Yancheng Yuan, Xiang Wang, and Xiangnan He. 2024. Llara: Large language-recommendation assistant. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 1785–1795. find
  17. Jingzhe Liu, Liam Collins, Jiliang Tang, Tong Zhao, Neil Shah, and Clark Mingxuan Ju. 2025a. Understanding generative recommendation with semantic ids from a model-scaling view. arXiv preprint arXiv:2509.25522. — [arXiv:2509.25522](https://arxiv.org/abs/2509.25522) find
  18. Qidong Liu, Xiangyu Zhao, Yejing Wang, Zijian Zhang, Howard Zhong, Chong Chen, Xiang Li, Wei Huang, and Feng Tian. 2025b. Bridge the domains: Large language models enhanced cross-domain sequential recommendation. In Proceedings of the 48th International ACM SIGIR Conference on Research and Developme… find
  19. Zhanyu Liu, Shiyao Wang, Xingmei Wang, Rongzhou Zhang, Jiaxin Deng, Honghui Bao, Jinghao Zhang, Wuchao Li, Pengfei Zheng, Xiangyu Wu, and 1 others. 2025c. Onerec-think: In-text reasoning for generative recommendation. arXiv preprint arXiv:2510.11639.
  20. Haokai Ma, Ruobing Xie, Lei Meng, Xin Chen, Xu Zhang, Leyu Lin, and Jie Zhou. 2024. Triple sequence learning for cross-domain recommendation. ACM Transactions on Information Systems, 42(4):1– 29. find
  21. Jiarui Qin, Jiachen Zhu, Bo Chen, Zhirong Liu, Weiwen Liu, Ruiming Tang, Rui Zhang, Yong Yu, and Weinan Zhang. 2022. Rankflow: Joint optimization of multi-stage cascade ranking systems as flows. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information… find
  22. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, and 1 others. 2019. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9. find
  23. Shashank Rajput, Nikhil Mehta, Anima Singh, Raghu-nandan Hulikal Keshavan, Trung Vu, Lukasz Heldt, Lichan Hong, Yi Tay, Vinh Tran, Jonah Samost, and 1 others. 2023. Recommender systems with generative retrieval. Advances in Neural Information Processing Systems, 36:10299–10315. find
  24. Fei Sun, Jun Liu, Jian Wu, Changhua Pei, Xiao Lin, Wenwu Ou, and Peng Jiang. 2019. Bert4rec: Sequential recommendation with bidirectional encoder representations from transformer. In Proceedings of the 28th ACM international conference on information and knowledge management, pages 1441–1450. find
  25. Juntao Tan, Shuyuan Xu, Wenyue Hua, Yingqiang Ge, Zelong Li, and Yongfeng Zhang. 2024. Idgenrec: Llm-recsys alignment with textual id learning. In Proceedings of the 47th international ACM SIGIR conference on research and development in information retrieval, pages 355–364. find
  26. Wenjie Wang, Honghui Bao, Xinyu Lin, Jizhi Zhang, Yongqi Li, Fuli Feng, See-Kiong Ng, and Tat-Seng Chua. 2024a. Learnable item tokenization for generative recommendation. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, pages 2400–2409. find
  27. Wenjie Wang, Xinyu Lin, Fuli Feng, Xiangnan He, and Tat-Seng Chua. 2023. Generative recommendation: Towards next-generation recommender paradigm. arXiv preprint arXiv:2304.03516. — [arXiv:2304.03516](https://arxiv.org/abs/2304.03516) find
  28. Ye Wang, Jiahao Xun, Minjie Hong, Jieming Zhu, Tao Jin, Wang Lin, Haoyuan Li, Linjun Li, Yan Xia, Zhou Zhao, and 1 others. 2024b. Eager: Two-stream generative recommender with behavior-semantic collaboration. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pa… find
  29. Zhipeng Wei, Kuo Cai, Junda She, Jie Chen, Minghao Chen, Yang Zeng, Qiang Luo, Wencong Zeng, Ruiming Tang, Kun Gai, and 1 others. 2025. Oneloc: Geo-aware generative recommender systems for local life service. arXiv preprint arXiv:2508.14646. — [arXiv:2508.14646](https://arxiv.org/abs/2508.14646) find
  30. Longtao Xiao, Haozhao Wang, Cheng Wang, Linfei Ji, Yifan Wang, Jieming Zhu, Zhenhua Dong, Rui Zhang, and Ruixuan Li. 2025. Unger: Generative recommendation with a unified code via semantic and collaborative integration. ACM Transactions on Information Systems. find
  31. Yuhao Yang, Zhi Ji, Zhaopeng Li, Yi Li, Zhonglin Mo, Yue Ding, Kai Chen, Zijian Zhang, Jie Li, Shuanglong Li, and 1 others. 2025. Sparse meets dense: Unified generative recommendations with cascaded sparse-dense representations. arXiv preprint arXiv:2503.02453. — [arXiv:2503.02453](https://arxiv.org/abs/2503.02453) find
  32. Xuesong Yin, Songcan Chen, and Enliang Hu. 2013. Regularized soft k-means for discriminant analysis. Neurocomputing, 103:29–42. find
  33. Jiaqi Zhai, Lucy Liao, Xing Liu, Yueming Wang, Rui Li, Xuan Cao, Leon Gao, Zhaojie Gong, Fangda Gu, Michael He, and 1 others. 2024. Actions speak louder than words: Trillion-parameter sequential transducers for generative recommendations. arXiv preprint arXiv:2402.17152. — [arXiv:2402.17152](https://arxiv.org/abs/2402.17152) find
  34. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems, 30.
  35. Wenlin Zhang, Chuhan Wu, Xiangyang Li, Yuhao Wang, Kuicai Dong, Yichao Wang, Xinyi Dai, Xiangyu Zhao, Huifeng Guo, and Ruiming Tang. 2025. Llmtreerec: Unleashing the power of large language models for cold-start recommendations. In Proceedings of the 31st International Conference on Computational Li… find
  36. Dongsheng Wang, Yuxi Huang, Shen Gao, Yifan Wang, Chengrui Huang, and Shuo Shang. 2025. Generative next poi recommendation with semantic id. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 2, pages 2904–2914. find
  37. Bowen Zheng, Yupeng Hou, Hongyu Lu, Yu Chen, Wayne Xin Zhao, Ming Chen, and Ji-Rong Wen. 2024. Adapting large language models by integrating collaborative semantics for recommendation. In 2024 IEEE 40th International Conference on Data Engineering (ICDE), pages 1435–1448. IEEE. find
  38. Guorui Zhou, Honghui Bao, Jiaming Huang, Jiaxin Deng, Jinghao Zhang, Junda She, Kuo Cai, Lejian Ren, Lu Ren, Qiang Luo, and 1 others. 2025a. Openonerec technical report. arXiv preprint arXiv:2512.24762. — [arXiv:2512.24762](https://arxiv.org/abs/2512.24762) find
  39. Guorui Zhou, Jiaxin Deng, Jinghao Zhang, Kuo Cai, Lejian Ren, Qiang Luo, Qianqian Wang, Qigen Hu, Rui Huang, Shiyao Wang, and 1 others. 2025b. Onerec technical report. arXiv preprint arXiv:2506.13695. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhang-2026-unleashing-native-recommendation-potential-llm/