arxiv: 2604.04976 · v1 · submitted 2026-04-04 · 💻 cs.IR

Recognition: 2 theorem links

· Lean Theorem

Tencent Advertising Algorithm Challenge 2025: All-Modality Generative Recommendation

Chao Zhou, Chengguo Yin, Fengye Xu, Haijie Gu, Haoran Xin, Huan Yu, Irwin King, Jie Jiang, Junbang Huo, Junwei Pan, Jun Zhang, Lei Xiao, Lunan Fan, Shudong Huang, Wei Xue, Wentao Ning, Xiaotian Li, Xing Zhou, Yanbo Wang, Yaozheng Wang, Yuliang Sun, Yurong Yang, Zhiyu Hu

Authors on Pith no claims yet

Pith reviewed 2026-05-13 16:57 UTC · model grok-4.3

classification 💻 cs.IR

keywords generative recommendationmulti-modal datasetsadvertising recommendationuser sequence modelingautoregressive modelsTencentGR-1MTencentGR-10Mindustrial benchmarks

0 comments

The pith

Two new datasets of real ad interaction sequences with multi-modal features enable training of generative recommender systems at industrial scale.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper presents the Tencent Advertising Algorithm Challenge 2025 and the two supporting datasets TencentGR-1M and TencentGR-10M. The datasets are built directly from de-identified Tencent Ads logs and supply user sequences that combine collaborative identifiers with multi-modal content representations. Each interaction carries exposure, click, and in the larger set conversion labels, allowing autoregressive models to generate behavior sequences rather than rank fixed candidates. A sympathetic reader cares because prior public resources for generative recommendation have lacked the combination of scale, realism, and full modality needed to test models in advertising settings. The paper also supplies a baseline model, an evaluation protocol that weights high-value conversions, and public releases of both data and code.

Core claim

The paper establishes that large-scale generative recommendation research can now proceed on realistic industrial data by releasing TencentGR-1M (1 million users, up to 100 items per sequence) and TencentGR-10M (10 million users), both containing collaborative IDs plus state-of-the-art multi-modal embeddings extracted from actual ad logs, with explicit distinction of click versus conversion events at sequence and target levels.

What carries the argument

The all-modality datasets TencentGR-1M and TencentGR-10M that map collaborative identifiers and multi-modal content into discrete token spaces so that user behavior can be modeled by autoregressive sequence models.

If this is right

Models can be trained and evaluated on explicit conversion events rather than clicks alone, with weighted scoring for high-value outcomes.
Research can directly compare generative sequence models against each other at 10 million user scale using a shared protocol and baseline.
Public release of the data and baseline code allows reproducible experiments on multi-modal tokenization for advertising recommendation.
Future work can explore whether autoregressive generation of entire interaction sequences improves personalization over ranking fixed candidate sets.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the datasets prove effective, similar construction pipelines could be applied to other recommendation domains that already collect rich multi-modal logs.
The emphasis on conversion-weighted evaluation may push the field toward metrics that better reflect business value rather than simple accuracy.
Limitations from de-identification could be tested by measuring how much predictive power is lost relative to non-public internal versions of the same logs.

Load-bearing premise

The de-identified logs and extracted multi-modal embeddings preserve enough genuine user behavior signal that generative models can learn realistic patterns without major distortion from privacy processing or embedding model choice.

What would settle it

A generative model trained on these datasets produces no measurable lift in click-through or conversion rates when its generated sequences are deployed against live ad traffic compared with strong non-generative baselines.

Figures

Figures reproduced from arXiv: 2604.04976 by Chao Zhou, Chengguo Yin, Fengye Xu, Haijie Gu, Haoran Xin, Huan Yu, Irwin King, Jie Jiang, Junbang Huo, Junwei Pan, Jun Zhang, Lei Xiao, Lunan Fan, Shudong Huang, Wei Xue, Wentao Ning, Xiaotian Li, Xing Zhou, Yanbo Wang, Yaozheng Wang, Yuliang Sun, Yurong Yang, Zhiyu Hu.

**Figure 2.** Figure 2: Illustration of the whole framework of the competition. The [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

read the original abstract

Generative recommender systems are rapidly emerging as a new paradigm for recommendation, where collaborative identifiers and/or multi-modal content are mapped into discrete token spaces and user behavior is modelled with autoregressive sequence models. Despite progress on multi-modal recommendation datasets, there is still a lack of public benchmarks that jointly offer large-scale, realistic and fully all-modality data designed specifically for generative recommendation (GR) in industrial advertising. To foster research in this direction, we organised the Tencent Advertising Algorithm Challenge 2025, a global competition built on top of two all-modality datasets for GR: TencentGR-1M and TencentGR-10M. Both datasets are constructed from real de-identified Tencent Ads logs and contain rich collaborative IDs and multi-modal representations extracted with state-of-the-art embedding models. The preliminary track (TencentGR-1M) provides 1 million user sequences with up to 100 interacted items each, where each interaction is labeled with exposure and click signals, while the final track (TencentGR-10M) scales this to 10 million users and explicitly distinguishes between click and conversion events at both the sequence and target level. This paper presents the task definition, data construction process, feature schema, baseline GR model, evaluation protocol, and key findings from top-ranked and award-winning solutions. Our datasets focus on multi-modal sequence generation in an advertising setting and introduce weighted evaluation for high-value conversion events. We release our datasets at https://huggingface.co/datasets/TAAC2025 and baseline implementations at https://github.com/TencentAdvertisingAlgorithmCompetition/baseline_2025 to enable future research on all-modality generative recommendation at an industrial scale. The official website is https://algo.qq.com/2025.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper releases two new public datasets for all-modality generative recommendation built from real advertising logs, which is its main contribution.

read the letter

The main thing to know is that this is a dataset and competition release paper for TencentGR-1M and TencentGR-10M. Both are constructed from de-identified Tencent Ads logs and combine collaborative IDs with multi-modal embeddings for generative sequence models, with labels covering exposure, clicks, and conversions plus a weighted evaluation that emphasizes high-value events. The preliminary track uses 1M users and the final track scales to 10M, with public releases on Hugging Face and baseline code on GitHub. The paper walks through the construction process, feature schema, a standard baseline autoregressive model, and the evaluation protocol. This setup is new in its explicit tailoring for all-modality generative recommendation in an industrial advertising context, where prior datasets often lacked the full combination of scale, conversion signals, and generative tokenization focus. The description supports reproducibility and gives enough detail on the pipeline to let others inspect statistics or run the baselines directly. Soft spots are minor and mostly tied to the paper type. The novelty sits primarily in the data release rather than new modeling techniques or theoretical results, and the embedding extraction and de-identification steps are described at a high level without deep analysis of possible artifacts. Nothing in the construction narrative shows internal inconsistencies or fitting issues. This is useful for researchers working on generative recommenders or industrial recsys who need large, realistic multi-modal sequence data for experiments. Readers hunting for major algorithmic advances will find less here. It deserves peer review because the datasets are new, documented, and publicly available at a scale that can support follow-on work.

Referee Report

1 major / 3 minor

Summary. The manuscript describes the organization of the Tencent Advertising Algorithm Challenge 2025 and the public release of two all-modality generative recommendation datasets, TencentGR-1M and TencentGR-10M, constructed from real de-identified Tencent Ads logs. It covers the task definition, data construction process, feature schema (collaborative IDs plus SOTA multi-modal embeddings), baseline GR model, evaluation protocol (including weighted metrics for high-value conversion events), and key findings from top-ranked competition solutions. The datasets are released on Hugging Face with baseline code on GitHub to support research on autoregressive sequence modeling for industrial advertising recommendation.

Significance. If the construction details hold, the work provides a valuable public benchmark for all-modality generative recommenders at industrial scale, filling a noted gap in realistic, large-scale datasets that jointly include collaborative signals, multi-modal content embeddings, and conversion labels. The explicit distinction between click and conversion events in the 10M track, combined with the released baseline and evaluation protocol, enables direct reproducibility and comparison of new GR methods on advertising data.

major comments (1)

Data Construction section: the description of multi-modal embedding extraction references 'state-of-the-art embedding models' without naming the specific models or versions used for each modality (text, image, etc.); this detail is load-bearing for the claim that the datasets are fully all-modality and replicable by the community.

minor comments (3)

Abstract and §4: the statement that TencentGR-1M contains 'up to 100 interacted items each' would benefit from a table or figure reporting the actual distribution of sequence lengths and item frequencies to substantiate the 'industrial scale' characterization.
Evaluation Protocol section: the weighted evaluation for conversion events is introduced but lacks the explicit weighting formula or pseudocode; adding this would improve clarity for readers implementing the metric.
The paper should include a summary table of key dataset statistics (number of unique items, modality coverage per item, label distributions) for both TencentGR-1M and TencentGR-10M to allow quick comparison.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive summary, significance assessment, and recommendation of minor revision. We address the single major comment point-by-point below.

read point-by-point responses

Referee: Data Construction section: the description of multi-modal embedding extraction references 'state-of-the-art embedding models' without naming the specific models or versions used for each modality (text, image, etc.); this detail is load-bearing for the claim that the datasets are fully all-modality and replicable by the community.

Authors: We agree that the current description is insufficient for full replicability. In the revised manuscript we will add an explicit subsection (or table) in Data Construction that names the precise embedding models and versions used for each modality (text, image, video, etc.), including any preprocessing steps. This directly addresses the referee's concern and strengthens the all-modality claim. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper is a competition announcement and dataset release describing construction of TencentGR-1M and TencentGR-10M from de-identified Tencent Ads logs, including schema, baseline, and evaluation protocol. No mathematical derivations, equations, fitted parameters, predictions, or first-principles claims exist that could reduce to inputs by construction. All content is descriptive of external data artifacts and standard industrial pipelines; no self-citation chains or ansatzes are load-bearing for any result. This matches the default non-circular case for data-release papers.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are introduced; the work relies on standard data processing from logs and existing embedding models.

pith-pipeline@v0.9.0 · 5693 in / 1049 out tokens · 34015 ms · 2026-05-13T16:57:05.038969+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean (washburn_uniqueness_aczel, Jcost) reality_from_one_distinction unclear
Generative recommender systems... autoregressive sequence models... InfoNCE loss... weighted evaluation for high-value conversion events

Reference graph

Works this paper leans on

78 extracted references · 78 canonical work pages · 9 internal anchors

[1]

Anirudhan Badrinath, Prabhat Agarwal, Laksh Bhasin, Jaewon Yang, Jiajing Xu, and Charles Rosenberg. 2025. PinRec: Outcome-Conditioned, Multi-Token Generative Retrieval for Industry-Scale Recommendation Systems.arXiv preprint arXiv:2504.10507(2025)

work page arXiv 2025
[2]

Simone Borg Bruun, Krisztian Balog, and Maria Maistro. 2024. Dataset and Models for Item Recommendation Using Multi-Modal User Interactions. InProceedings of SIGIR

work page 2024
[3]

Zheng Chai, Qin Ren, Xijun Xiao, Huizhi Yang, Bo Han, Sijun Zhang, Di Chen, Hui Lu, Wenlin Zhao, Lele Yu, Xionghang Xie, Shiru Ren, Xiang Sun, Yaocheng Tan, Peng Xu, Yuchao Zheng, and Di Wu. 2025. LONGER: Scaling Up Long Sequence Modeling in Industrial Recommenders. InProceedings of the Nineteenth ACM Conference on Recommender Systems. ACM. doi:10.48550/a...

work page doi:10.48550/arxiv.2505.04421 2025
[4]

Jianxin Chang, Chenbin Zhang, Zhiyi Fu, Xiaoxue Zang, Lin Guan, Jing Lu, Yiqun Hui, Dewei Leng, Yanan Niu, Yang Song, and Kun Gai. 2023. TWIN: TWo- stage Interest Network for Lifelong User Behavior Modeling in CTR Prediction at Kuaishou. InProceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM....

work page doi:10.1145/3539618 2023
[5]

Heng-Tze Cheng, Levent Koc, Jeremiah Harmsen, Tal Shaked, Tushar Chandra, Hrishi Aradhye, Glen Anderson, Greg Corrado, Wei Chai, Mustafa Ispir, Ro- han Anil, Zakaria Haque, Lichan Hong, Vihan Jain, Xiaobing Liu, and Hemal Shah. 2016. Wide & Deep Learning for Recommender Systems.arXiv preprint arXiv:1606.07792(2016). arXiv:1606.07792 [cs.LG]

work page arXiv 2016
[6]

Tri Dao, Dan Fu, Stefano Ermon, Atri Rudra, and Christopher Ré. 2022. Flashat- tention: Fast and memory-efficient exact attention with io-awareness.Advances in neural information processing systems35 (2022), 16344–16359

work page 2022
[7]

Matthijs Douze, Alexandr Guzhva, Chengqi Deng, Jeff Johnson, Gergely Szilvasy, Pierre-Emmanuel Mazaré, Maria Lomeli, Lucas Hosseini, and Hervé Jégou. 2024. The Faiss library. (2024). arXiv:2401.08281 [cs.LG]

work page internal anchor Pith review Pith/arXiv arXiv 2024
[8]

Ningya Feng, Junwei Pan, Jialong Wu, Baixu Chen, Ximei Wang, Qian Li, Xian Hu, Jie Jiang, and Mingsheng Long. 2024. Long-sequence recommendation models need decoupled embeddings.arXiv preprint arXiv:2410.02604(2024)

work page arXiv 2024
[9]

Yufei Feng, Fuyu Lv, Weichen Shen, Menghan Wang, Fei Sun, Yu Zhu, and Keping Yang. 2019. Deep Session Interest Network for Click-Through Rate Prediction. InProceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence. IJCAI, 2301–2307. doi:10.24963/ijcai.2019/319

work page doi:10.24963/ijcai.2019/319 2019
[10]

Kairui Fu, Tao Zhang, Shuwen Xiao, Ziyang Wang, Xinming Zhang, Chenchi Zhang, Yuliang Yan, Junjun Zheng, Yu Li, Zhihong Chen, et al . 2025. Forge: Forming semantic identifiers for generative retrieval in industrial datasets.arXiv preprint arXiv:2509.20904(2025)

work page arXiv 2025
[11]

Chongming Gao, Shijun Li, Wenqiang Lei, Jiawei Chen, Biao Li, Peng Jiang, Xiangnan He, Jiaxin Mao, and Tat-Seng Chua. 2022. KuaiRec: A Fully-observed Dataset and Insights for Evaluating Recommender Systems. InProceedings of CIKM

work page 2022
[12]

Chongming Gao, Shijun Li, Yuan Zhang, Jiawei Chen, Biao Li, Wenqiang Lei, Peng Jiang, and Xiangnan He. 2022. KuaiRand: An Unbiased Sequential Recom- mendation Dataset with Randomly Exposed Videos. InProceedings of CIKM

work page 2022
[13]

Tiancheng Gu, Kaicheng Yang, Ziyong Feng, Xingjun Wang, Yanzhao Zhang, Dingkun Long, Yingda Chen, Weidong Cai, and Jiankang Deng. 2025. Breaking the Modality Barrier: Universal Embedding Learning with Multimodal LLMs. arXiv:2504.17432 [cs.CV] https://arxiv.org/abs/2504.17432

work page arXiv 2025
[14]

Huifeng Guo, Ruiming Tang, Yunming Ye, Zhenguo Li, and Xiuqiang He. 2017. DeepFM: A Factorization-Machine Based Neural Network for CTR Prediction. InProceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence. IJCAI, 1725–1731. doi:10.24963/ijcai.2017/239

work page doi:10.24963/ijcai.2017/239 2017
[15]

Ting Guo, Zhaoyang Yang, Qinsong Zeng, and Ming Chen. 2025. Context-Aware Lifelong Sequential Modeling for Online Click-Through Rate Prediction.arXiv preprint arXiv:2502.12634(2025). arXiv:2502.12634 [cs.IR]

work page arXiv 2025
[16]

Xingzhuo Guo, Junwei Pan, Ximei Wang, Baixu Chen, Jie Jiang, and Mingsheng Long. 2024. On the Embedding Collapse when Scaling up Recommendation Models. InProceedings of the 41st International Conference on Machine Learning. PMLR, 16891–16909

work page 2024
[17]

Xiangnan He and Tat-Seng Chua. 2017. Neural Factorization Machines for Sparse Predictive Analytics. InProceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM, 355–364. doi:10.1145/3077136.3080777

work page doi:10.1145/3077136.3080777 2017
[18]

Min Hou, Le Wu, Yuxin Liao, Yonghui Yang, Zhen Zhang, Changlong Zheng, Han Wu, and Richang Hong. 2025. A Survey on Generative Recommendation: Data, Model, and Tasks.arXiv preprint arXiv:2510.27157(2025)

work page internal anchor Pith review Pith/arXiv arXiv 2025
[19]

Ruijie Hou, Zhaoyang Yang, Yu Ming, Hongyu Lu, Zhuobin Zheng, Yu Chen, Qinsong Zeng, and Ming Chen. 2024. Cross-Domain LifeLong Sequential Model- ing for Online Click-Through Rate Prediction. InProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. ACM, 5116–5125. doi:10.1145/3637528.3671601

work page doi:10.1145/3637528.3671601 2024
[20]

Yupeng Hou, Jiacheng Li, Ashley Shin, Jinsung Jeon, Abhishek Santhanam, Wei Shao, Kaveh Hassani, Ning Yao, and Julian McAuley. 2025. Generating Long Semantic IDs in Parallel for Recommendation. InProceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining

work page 2025
[21]

Xian Hu, Ming Yue, Zhixiang Feng, Junwei Pan, Junjie Zhai, Ximei Wang, Xinrui Miao, Qian Li, Xun Liu, Shangyu Zhang, Letian Wang, Hua Lu, Zijian Zeng, Chen Cai, Wei Wang, Fei Xiong, Pengfei Xiong, Jintao Zhang, Zhiyuan Wu, Chunhui The Tencent Advertising Algorithm Challenge 2025: All-Modality Generative Recommendation , , Zhang, Anan Liu, Jiulong You, Cha...

work page arXiv 2025
[22]

Tongwen Huang, Zhiqi Zhang, and Junlin Zhang. 2019. FiBiNET: Combining Feature Importance and Bilinear Feature Interaction for Click-Through Rate Prediction. InProceedings of the 13th ACM Conference on Recommender Systems. ACM, 169–177. doi:10.1145/3298689.3347043

work page doi:10.1145/3298689.3347043 2019
[23]

Yanhua Huang, Yuqi Chen, Xiong Cao, Rui Yang, Mingliang Qi, Yinghao Zhu, Qingchang Han, Yaowei Liu, Zhaoyu Liu, Xuefeng Yao, et al . 2025. Towards Large-scale Generative Ranking.arXiv preprint arXiv:2505.04180(2025)

work page arXiv 2025
[24]

Keller Jordan, Yuchen Jin, Vlado Boza, You Jiacheng, Franz Cecista, Laker New- house, and Jeremy Bernstein. [n. d.]. Muon: An optimizer for hidden layers in neural networks, 2024.URL https://kellerjordan. github. io/posts/muon6 ([n. d.])

work page 2024
[25]

Wang-Cheng Kang and Julian McAuley. 2018. Self-Attentive Sequential Rec- ommendation. InProceedings of the 2018 IEEE International Conference on Data Mining

work page 2018
[26]

Jared Kaplan, Sam McCandlish, Tom Henighan, Tom B Brown, Benjamin Chess, Rewon Child, Scott Gray, Alec Radford, Jeffrey Wu, and Dario Amodei. 2020. Scaling laws for neural language models.arXiv preprint arXiv:2001.08361(2020)

work page internal anchor Pith review Pith/arXiv arXiv 2020
[27]

Xiaoyu Kong, Leheng Sheng, Junfei Tan, Yuxin Chen, Jiancan Wu, An Zhang, Xiang Wang, and Xiangnan He. 2025. MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation.arXiv preprint arXiv:2510.24431(2025)

work page arXiv 2025
[28]

Seunghyun Lee et al. 2025. GRAM: Generative Recommendation via Semantic- aware Multi-granular Late Fusion. InProceedings of ACL

work page 2025
[29]

Fengxin Li, Yi Li, Yue Liu, Chao Zhou, Yuan Wang, Xiaoxiang Deng, Wei Xue, Dapeng Liu, Lei Xiao, Haijie Gu, Jie Jiang, Hongyan Liu, Biao Qin, and Jun He

work page
[30]

arXiv:2411.13789 [cs.IR] https://arxiv.org/ abs/2411.13789

LEADRE: Multi-Faceted Knowledge Enhanced LLM Empowered Display Advertisement Recommender System. arXiv:2411.13789 [cs.IR] https://arxiv.org/ abs/2411.13789

work page arXiv
[31]

Shiyu Li, Yang Tang, Shizhe Chen, and Xi Chen. 2024. Conan- embedding: General Text Embedding with More and Better Negative Samples. arXiv:2408.15710 [cs.CL] https://arxiv.org/abs/2408.15710

work page arXiv 2024
[32]

Zehan Li, Xin Zhang, Yanzhao Zhang, Dingkun Long, Pengjun Xie, and Meishan Zhang. 2023. Towards general text embeddings with multi-stage contrastive learning.arXiv preprint arXiv:2308.03281(2023)

work page internal anchor Pith review Pith/arXiv arXiv 2023
[33]

Jianxun Lian, Xiaohuan Zhou, Fuzheng Zhang, Zhongxia Chen, Xing Xie, and Guangzhong Sun. 2018. xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems. InProceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. ACM, 1754–1763. doi:10.1145/3219819.3220023

work page doi:10.1145/3219819.3220023 2018
[34]

Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Cheng- gang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, et al. 2024. Deepseek-v3 technical report.arXiv preprint arXiv:2412.19437(2024)

work page internal anchor Pith review Pith/arXiv arXiv 2024
[35]

Ilya Loshchilov and Frank Hutter. 2019. Decoupled Weight Decay Regularization. InInternational Conference on Learning Representations. https://arxiv.org/abs/ 1711.05101

work page internal anchor Pith review Pith/arXiv arXiv 2019
[36]

Xinchen Luo, Jiangxia Cao, Tianyu Sun, Jinkai Yu, Rui Huang, Wei Yuan, Hezheng Lin, Yichen Zheng, Shiyao Wang, Qigen Hu, et al . 2025. Qarm: Quantitative alignment multi-modal recommendation at kuaishou. InProceedings of the 34th ACM International Conference on Information and Knowledge Management. 5915– 5922

work page 2025
[37]

Lyu, and Irwin King

Hao Ma, Tom Chao Zhou, Michael R. Lyu, and Irwin King. 2011. Improving Recommender Systems by Incorporating Social Contextual Information.ACM Trans. Inf. Syst.29, 2, Article 9 (April 2011), 23 pages. doi:10.1145/1961209.1961212

work page doi:10.1145/1961209.1961212 2011
[38]

Kelong Mao, Jieming Zhu, Liangcai Su, Guohao Cai, Yuru Li, and Zhenhua Dong

work page
[39]

In Proceedings of the AAAI Conference on Artificial Intelligence, Vol

FinalMLP: An Enhanced Two-Stream MLP Model for CTR Prediction. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 37. 4552–4560. doi:10.1609/aaai.v37i4.25577

work page doi:10.1609/aaai.v37i4.25577
[40]

Azzolini, et al

Maxim Naumov, Dheevatsa Mudigere, Hao-Jun Michael Shi, Jianyu Huang, Narayanan Sundaram, Jongsoo Park, Xiaodong Wang, Udit Gupta, Carole-Jean Wu, Alisson G. Azzolini, et al. 2019. Deep Learning Recommendation Model for Personalization and Recommendation Systems.arXiv preprint arXiv:1906.00091 (2019)

work page arXiv 2019
[41]

Xiaonan Nie, Yi Liu, Fangcheng Fu, Jinbao Xue, Dian Jiao, Xupeng Miao, Yangyu Tao, and Bin Cui. 2023. Angel-ptm: A scalable and economical large-scale pre- training system in tencent.arXiv preprint arXiv:2303.02868(2023)

work page arXiv 2023
[42]

Aaron van den Oord, Yazhe Li, and Oriol Vinyals. 2018. Representation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748(2018)

work page internal anchor Pith review Pith/arXiv arXiv 2018
[43]

Florian Paischer et al. 2024. Preference Discerning with LLM-Enhanced Genera- tive Recommendation.arXiv preprint arXiv:2412.08604(2024)

work page arXiv 2024
[44]

Junwei Pan, Jian Xu, Alfonso Lobos Ruiz, Wenliang Zhao, Shengjun Pan, Yu Sun, and Quan Lu. 2018. Field-weighted Factorization Machines for Click-Through Rate Prediction in Display Advertising. InProceedings of the 2018 World Wide Web Conference. International World Wide Web Conferences Steering Committee, 1349–1357. doi:10.1145/3178876.3186040

work page doi:10.1145/3178876.3186040 2018
[45]

Junwei Pan, Wei Xue, Ximei Wang, Haibin Yu, Xun Liu, Shijie Quan, Xueming Qiu, Dapeng Liu, Lei Xiao, and Jie Jiang. 2024. Ads Recommendation in a Collapsed and Entangled World. InProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. ACM. doi:10.1145/3637528.3671607

work page doi:10.1145/3637528.3671607 2024
[46]

Ethan Perez, Florian Strub, Harm De Vries, Vincent Dumoulin, and Aaron Courville. 2018. Film: Visual reasoning with a general conditioning layer. In Proceedings of the AAAI conference on artificial intelligence, Vol. 32

work page 2018
[47]

Qi Pi, Guorui Zhou, Yujing Zhang, Zhe Wang, Xiaoqiang Zhu, Kun Gai, Peng Cui, and Wenwu Zhu. 2020. Search-based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Prediction. InProceedings of the 29th ACM International Conference on Information & Knowledge Management. ACM, 2685–2692. doi:10.1145/3340531.3412744

work page doi:10.1145/3340531.3412744 2020
[48]

Shashank Rajput, Nikhil Mehta, Anima Singh, Raghunandan Hulikal Keshavan, Trung Vu, Lukasz Heldt, Lichan Hong, Yi Tay, Vinh Tran, Jonah Samost, et al

work page
[49]

Recommender systems with generative retrieval.Advances in Neural Information Processing Systems36 (2023), 10299–10315

work page 2023
[50]

Steffen Rendle. 2010. Factorization Machines. In2010 IEEE International Confer- ence on Data Mining. IEEE, 995–1000. doi:10.1109/ICDM.2010.127

work page doi:10.1109/icdm.2010.127 2010
[51]

Spotify Research. 2025. Semantic IDs for Generative Search and Recommendation. Spotify Research Blog

work page 2025
[52]

Yu Shang, Chen Gao, Nian Li, and Yong Li. 2025. A Large-scale Dataset with Behavior, Attributes, and Content of Mobile Short-video Platform. InProceedings of the Web Conference Companion

work page 2025
[53]

Jianlin Su, Murtadha Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. 2024. Roformer: Enhanced transformer with rotary position embedding. Neurocomputing568 (2024), 127063

work page 2024
[54]

2021.𝐹 𝑀2: Field-matrixed Factorization Machines for Recommender Systems

Yang Sun, Junwei Pan, Alex Zhang, and Aaron Flores. 2021.𝐹 𝑀2: Field-matrixed Factorization Machines for Recommender Systems. InProceedings of the Web Conference 2021. ACM, 2828–2837. doi:10.1145/3442381.3449930

work page doi:10.1145/3442381.3449930 2021
[55]

Wenjie Wang et al. 2024. LETTER: Learnable Item Tokenization for Generative Recommendation.arXiv preprint arXiv:2405.07314(2024)

work page arXiv 2024
[56]

Fangzhao Wu et al. 2020. MIND: A Large-scale Dataset for News Recommenda- tion. InProceedings of ACL

work page 2020
[57]

Da Xu, Chuanwei Ruan, Evren Korpeoglu, Sushant Kumar, and Kannan Achan

work page
[58]

Inductive representation learning on temporal graphs.arXiv preprint arXiv:2002.07962(2020)

work page arXiv 2002
[59]

Yi Xu, Moyu Zhang, Chenxuan Li, Zhihao Liao, Haibo Xing, Hao Deng, Jinxin Hu, Yu Zhang, Xiaoyi Zeng, and Jing Zhang. 2025. MMQ: Multimodal Mixture- of-Quantization Tokenization for Semantic ID Generation and User Behavioral Adaptation.arXiv preprint arXiv:2508.15281(2025)

work page arXiv 2025
[60]

Ben Xue, Dan Liu, Lixiang Wang, Mingjie Sun, Peng Wang, Pengfei Zhang, Shaoyun Shi, Tianyu Xu, Yunhao Sha, Zhiqiang Liu, et al . 2026. Generative Recommendation for Large-Scale Advertising.arXiv preprint arXiv:2602.22732 (2026)

work page arXiv 2026
[61]

Youze Xue, Dian Li, and Gang Liu. 2025. Improve Multi-Modal Embedding Learning via Explicit Hard Negative Gradient Amplifying.arXiv preprint arXiv:2506.02020(2025)

work page arXiv 2025
[62]

An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. 2025. Qwen3 technical report.arXiv preprint arXiv:2505.09388(2025)

work page internal anchor Pith review Pith/arXiv arXiv 2025
[63]

Wei Ye et al. 2025. DAS: Dual-Aligned Semantic IDs Empowered Industrial Recommender System. InProceedings of KDD

work page 2025
[64]

Guanghu Yuan, Fajie Yuan, Yudong Li, Beibei Kong, Shujie Li, Lei Chen, Min Yang, Chenyun Yu, Bo Hu, Zang Li, Yu Xu, and Xiaohu Qie. 2022. Tenrec: A Large-scale Multipurpose Benchmark Dataset for Recommender Systems. In Advances in Neural Information Processing Systems

work page 2022
[65]

Jiaqi Zhai, Lucy Liao, Xing Liu, Yueming Wang, Rui Li, Xuan Cao, Leon Gao, Zhao- jie Gong, Fangda Gu, Michael He, et al. 2024. Actions speak louder than words: Trillion-parameter sequential transducers for generative recommendations.arXiv preprint arXiv:2402.17152(2024)

work page internal anchor Pith review arXiv 2024
[66]

Buyun Zhang, Liang Luo, Yuxin Chen, Jade Nie, Xi Liu, Shen Li, Yanli Zhao, Yuchen Hao, Yantao Yao, Ellie Dingqiao Wen, Jongsoo Park, Maxim Naumov, and Wenlin Chen. 2024. Wukong: Towards a Scaling Law for Large-Scale Rec- ommendation. InProceedings of the 41st International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 235)...

work page 2024
[67]

Buyun Zhang, Liang Luo, Xi Liu, Jay Li, Zeliang Chen, Weilin Zhang, Xiaohan Wei, Yuchen Hao, Michael Tsang, Wenjun Wang, Yang Liu, Huayu Li, Yasmine Badr, Jongsoo Park, Jiyan Yang, Dheevatsa Mudigere, and Ellie Wen. 2022. DHEN: A Deep and Hierarchical Ensemble Network for Large-Scale Click-Through Rate Prediction.arXiv preprint arXiv:2203.11014(2022). arX...

work page doi:10.48550/arxiv.2203.11014 2022
[68]

Taolin Zhang, Junwei Pan, Jinpeng Wang, Yaohua Zha, Tao Dai, Bin Chen, Ruisheng Luo, Xiaoxiang Deng, Yuan Wang, Ming Yue, et al. 2024. Towards scal- able semantic representation for recommendation.arXiv preprint arXiv:2410.09560 (2024)

work page arXiv 2024
[69]

Zhen Zhang et al. 2025. Semantic IDs for Joint Generative Search and Recom- mendation. InProceedings of CIKM

work page 2025
[70]

Pinxue Zhao, Hailin Zhang, Fangcheng Fu, Xiaonan Nie, Qibin Liu, Fang Yang, Yuanbo Peng, Dian Jiao, Shuaipeng Li, Jinbao Xue, et al. 2024. Efficiently training , , Author et al. 7b llm with 1 million sequence length on 8 gpus.arXiv e-prints(2024), arXiv–2407

work page 2024
[71]

Yuqiu Zhao, Chaohong Tan, Lei Shi, and Chen Ma. 2025. Generative Recom- mender Systems: A Comprehensive Survey on Model, Framework, and Applica- tion.Information Fusion127 (2025), 103919

work page 2025
[72]

Guorui Zhou, Jiaxin Deng, Jinghao Zhang, Kuo Cai, Lejian Ren, Qiang Luo, Qian- qian Wang, Qigen Hu, Rui Huang, Shiyao Wang, et al. 2025. OneRec Technical Report.arXiv preprint arXiv:2506.13695(2025)

work page arXiv 2025
[73]

Guorui Zhou, Hengrui Hu, Hongtao Cheng, Huanjie Wang, Jiaxin Deng, Jinghao Zhang, Kuo Cai, Lejian Ren, Lu Ren, Liao Yu, et al. 2025. Onerec-v2 technical report.arXiv preprint arXiv:2508.20900(2025)

work page arXiv 2025
[74]

Guorui Zhou, Chengru Song, Xiaoqiang Zhu, Xiao Ma, Yanghui Yan, Junqi Jin, Han Li, and Kun Gai. 2018. Deep Interest Network for Click-Through Rate Prediction. InProceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. ACM, 1059–1068. doi:10.1145/3219819. 3219823

work page doi:10.1145/3219819 2018
[75]

Guorui Zhou, Xiaoqiang Zhu, Chengru Song, Ying Fan, Han Zhu, Xiao Ma, Yanghui Yan, Junqi Jin, Han Li, and Kun Gai. 2019. Deep Interest Evolution Network for Click-Through Rate Prediction. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 33. 5941–5948. doi:10.1609/aaai.v33i01.33015941

work page doi:10.1609/aaai.v33i01.33015941 2019
[76]

Hao Zhou, Junwei Pan, Xun Liu, Wei Xue, Liqiang Nie, and Ji-Rong Wen. 2024. Temporal Interest Network for User Response Prediction. InProceedings of the ACM on Web Conference 2024. ACM, 2496–2507. doi:10.1145/3589335.3648340

work page doi:10.1145/3589335.3648340 2024
[77]

Tom Zhou, Hao Ma, Michael Lyu, and Irwin King. 2010. UserRec: A User Rec- ommendation Framework in Social Tagging Systems.Proceedings of the AAAI Conference on Artificial Intelligence24, 1 (Jul. 2010), 1486–1491. doi:10.1609/aaai. v24i1.7524

work page doi:10.1609/aaai 2010
[78]

Jie Zhu, Zhifang Fan, Xiaoxie Zhu, Yuchen Jiang, Hangyu Wang, Xintian Han, Haoran Ding, Xinmin Wang, Wenlin Zhao, Zhen Gong, Huizhi Yang, Zheng Chai, Zhe Chen, Yuchao Zheng, Qiwei Chen, Feng Zhang, Xun Zhou, Peng Xu, Xiao Yang, Di Wu, and Zuotao Liu. 2025. RankMixer: Scaling Up Ranking Models in Industrial Recommenders. InProceedings of the 34th ACM Inter...

work page arXiv 2025