pith. sign in

arxiv: 2507.06121 · v2 · submitted 2025-07-08 · 💻 cs.IR

Brownian Bridge Diffusion for Sequential Recommendation

Yimeng Bai , Yang Zhang , Sihao Ding , Shaohui Ruan , Han Yao , Danhui Guan , Fuli Feng , Tat-Seng Chua This is my paper

Pith reviewed 2026-05-19 05:50 UTC · model grok-4.3

classification 💻 cs.IR
keywords sequential recommendationdiffusion modelsBrownian bridgeuser preferencesgenerative modelspersonalization
0
0 comments X

The pith

Brownian bridge diffusion builds direct paths from user history to target items instead of routing through noise.

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

The paper identifies that existing diffusion recommenders still force the model to reconstruct target items from Gaussian noise even when user history is provided as a condition. This creates an extra reconstruction task that can pull attention away from the actual preference patterns in the sequence. The authors replace that setup with a Brownian bridge process that defines a diffusion trajectory straight between the target item representation and the user's historical interactions. The result is BBDRec, which the experiments show beats both standard sequential models and prior diffusion recommenders on public datasets. A reader would care because the change suggests a more natural alignment between how diffusion works and what recommendation actually needs to predict.

Core claim

By adopting a preference-centric design that uses the Brownian bridge process to create a structured diffusion trajectory directly between target items and user historical representations, the model removes the distracting item-to-noise reconstruction step and thereby captures user-specific preference structures more effectively.

What carries the argument

The Brownian bridge process, which defines a direct stochastic path connecting a target item representation at one endpoint to the user's historical representations at the other endpoint.

If this is right

  • The diffusion process can be made to operate solely between observed history and the next item without an intermediate noise variable.
  • User history functions as the fixed endpoint rather than merely a conditioning signal.
  • Structured trajectories align the generative steps more closely with the sequential nature of user behavior.
  • Performance gains appear consistently across multiple public datasets compared with both sequential and diffusion baselines.

Where Pith is reading between the lines

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

  • The same bridging idea might transfer to other sequence-to-sequence generation settings where one endpoint is a history of states.
  • Training dynamics could become more stable because both ends of the diffusion path are deterministic user-derived vectors.
  • Similar direct-transition designs could be tested in session-based or next-basket recommendation without changing the overall diffusion framework.

Load-bearing premise

The extra reconstruction burden from Gaussian noise actually distracts the model from learning user preferences and the Brownian bridge removes that distraction.

What would settle it

An ablation that swaps the Brownian bridge trajectory for a standard Gaussian noise path conditioned on the same history and measures whether accuracy drops on the same test sets and metrics.

Figures

Figures reproduced from arXiv: 2507.06121 by Danhui Guan, Fuli Feng, Han Yao, Shaohui Ruan, Sihao Ding, Tat-Seng Chua, Yang Zhang, Yimeng Bai.

Figure 1
Figure 1. Figure 1: Comparison between the conditional and uncon [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: An overview of the proposed BBDRec framework. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Results of the performance of BBDRec across dif [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Results of the performance comparison across dif [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Results of the inference efficiency comparison. The [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
read the original abstract

Diffusion models, known for their strong generative capability derived from iterative noising and denoising processes, have recently emerged as a promising paradigm for sequential recommendation. To incorporate user history for personalization, existing methods typically follow a history-guided denoising paradigm inspired by text-guided image generation, where target item representations are reconstructed from Gaussian noise conditioned on user historical interactions. However, this design remains fundamentally anchored to an "item $\leftrightarrow$ noise" formulation, introducing an additional noise-reconstruction burden that may distract the model from capturing user-specific preference structures. Motivated by this limitation, we revisit diffusion-based sequential recommendation from a preference-centric perspective and adopt a preference bridging design that enables a direct "item $\leftrightarrow$ history" transition instead of relying on Gaussian noise. Based on this idea, we propose Brownian Bridge Diffusion Recommendation (BBDRec), which leverages the Brownian bridge process to construct a structured diffusion trajectory between target items and user historical representations, thereby better aligning diffusion modeling with the intrinsic nature of recommendation. Extensive experiments on multiple public datasets show that BBDRec consistently outperforms representative sequential and diffusion-based recommendation baselines. The implementation code is publicly available at https://github.com/baiyimeng/BBDRec.

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.

Referee Report

2 major / 2 minor

Summary. The paper proposes Brownian Bridge Diffusion Recommendation (BBDRec) for sequential recommendation. It argues that existing diffusion-based methods follow an item-to-noise reconstruction paradigm conditioned on user history, which introduces an unnecessary burden. Instead, BBDRec adopts a preference-centric design using the Brownian bridge stochastic process to construct a direct diffusion trajectory between target item representations and aggregated user historical representations, enabling better alignment with preference structures. Experiments on public datasets reportedly show consistent outperformance over sequential and diffusion-based baselines, with code released publicly.

Significance. If the central claims hold after addressing the issues below, the work could meaningfully advance diffusion models in recommendation systems by shifting from noise-centric to preference-bridging formulations. The explicit motivation from limitations in prior paradigms and the public code release are strengths that support reproducibility and potential follow-up work.

major comments (2)
  1. [§3 (Method) and §4 (Experiments)] The central claim that the Brownian bridge enables a direct item-to-history transition that removes the noise-reconstruction burden and better captures user preferences (motivation in abstract and §1) is load-bearing, yet the manuscript provides no controlled ablation isolating the bridge mechanics (linear interpolation and variance schedule X_t = (1-t)X_item + t X_history + sqrt(t(1-t)) noise) from a generic history-conditioning change. A standard conditional diffusion model injecting aggregated history at each reverse step could plausibly achieve similar alignment; without this comparison the attribution to the bridge specifically remains unverified.
  2. [§4 (Experiments)] Table 1 (or equivalent results table) and the abstract report consistent outperformance, but the experimental section lacks error bars, statistical significance tests (e.g., paired t-tests or Wilcoxon), and details on hyperparameter tuning or data splits. This undermines confidence in the superiority claim, especially given the low-confidence assessment of the setup.
minor comments (2)
  1. [§3] Notation for the Brownian bridge process and the preference bridging design could be clarified with an explicit equation for the forward/reverse transitions early in §3.
  2. [Abstract] The abstract would benefit from naming the specific public datasets and primary metrics (e.g., HR@10, NDCG@10) to give readers immediate context.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and suggestions. We address each major comment point by point below and will revise the manuscript to incorporate the recommended improvements where appropriate.

read point-by-point responses

  1. Referee: [§3 (Method) and §4 (Experiments)] The central claim that the Brownian bridge enables a direct item-to-history transition that removes the noise-reconstruction burden and better captures user preferences (motivation in abstract and §1) is load-bearing, yet the manuscript provides no controlled ablation isolating the bridge mechanics (linear interpolation and variance schedule X_t = (1-t)X_item + t X_history + sqrt(t(1-t)) noise) from a generic history-conditioning change. A standard conditional diffusion model injecting aggregated history at each reverse step could plausibly achieve similar alignment; without this comparison the attribution to the bridge specifically remains unverified.

    Authors: We agree that a controlled ablation isolating the Brownian bridge mechanics from generic history conditioning would strengthen attribution to the specific formulation. While the preference-centric motivation and the closed-form properties of the Brownian bridge (linear interpolation with the given variance schedule) provide theoretical grounding, we will add an explicit comparison to a standard conditional diffusion baseline that injects aggregated history at each reverse step, keeping the backbone and training identical for fairness. This ablation will be included in the revised experimental section. revision: yes

  2. Referee: [§4 (Experiments)] Table 1 (or equivalent results table) and the abstract report consistent outperformance, but the experimental section lacks error bars, statistical significance tests (e.g., paired t-tests or Wilcoxon), and details on hyperparameter tuning or data splits. This undermines confidence in the superiority claim, especially given the low-confidence assessment of the setup.

    Authors: We acknowledge that the current experimental reporting would benefit from greater statistical detail. In the revised manuscript we will add error bars (standard deviation over multiple random seeds) to all reported metrics in the main results table, include paired t-test or Wilcoxon signed-rank results with p-values to assess significance, and expand the experimental setup subsection with explicit descriptions of the hyperparameter search procedure and the train/validation/test split methodology used for each dataset. revision: yes

Circularity Check

0 steps flagged

No circularity: method applies known Brownian bridge to address stated limitation in prior diffusion paradigms

full rationale

The paper's derivation starts from an external critique of existing 'item ↔ noise' diffusion formulations in sequential recommendation, then adopts the standard Brownian bridge process (a well-established stochastic interpolation) to enable direct item-to-history transitions. No equations or steps reduce the proposed BBDRec outputs or performance claims to fitted parameters by construction, nor do self-citations supply the uniqueness or ansatz for the core bridge mechanics. The central design choice is motivated by and benchmarked against independent prior work on diffusion models and recommendation, remaining falsifiable via the reported experiments on public datasets.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are detailed beyond the high-level motivation and method name.

axioms (1)
  • domain assumption Brownian bridge process provides a suitable structured trajectory for modeling direct transitions between user history and target items in recommendation
    Invoked in the proposal of the preference bridging design

pith-pipeline@v0.9.0 · 5753 in / 1171 out tokens · 56951 ms · 2026-05-19T05:50:20.914687+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

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

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

63 extracted references · 63 canonical work pages · 2 internal anchors

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Floren- cia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 (2023)

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  2. [2]

    Donghoon Ahn, Jiwon Kang, Sanghyun Lee, Jaewon Min, Minjae Kim, Wooseok Jang, Hyoungwon Cho, Sayak Paul, SeonHwa Kim, Eunju Cha, et al . 2024. A Noise is Worth Diffusion Guidance. arXiv preprint arXiv:2412.03895 (2024)

    work page arXiv 2024

  3. [3]

    Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, et al. 2023. Qwen technical report. arXiv preprint arXiv:2309.16609 (2023)

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  4. [4]

    Yimeng Bai, Yang Zhang, Fuli Feng, Jing Lu, Xiaoxue Zang, Chenyi Lei, and Yang Song. 2024. GradCraft: Elevating Multi-task Recommendations through Holistic Gradient Crafting. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (Barcelona, Spain) (KDD ’24). Association for Computing Machinery, New York, NY, USA, 4774–4783

    work page 2024

  5. [5]

    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 Trans. Recomm. Syst. 3, 4, Article 53 (April 2025), 27 pages

    work page 2025

  6. [6]

    David Ben-Shimon, Alexander Tsikinovsky, Michael Friedmann, Bracha Shapira, Lior Rokach, and Johannes Hoerle. 2015. RecSys Challenge 2015 and the YOO- CHOOSE Dataset. In Proceedings of the 9th ACM Conference on Recommender Systems (Vienna, Austria) (RecSys ’15). Association for Computing Machinery, New York, NY, USA, 357–358

    work page 2015

  7. [7]

    Xingyan Bin, Jianfei Cui, Wujie Yan, Zhichen Zhao, Xintian Han, Chongyang Yan, Feng Zhang, Xun Zhou, Qi Wu, and Zuotao Liu. 2025. Real-time Indexing for Large-scale Recommendation by Streaming Vector Quantization Retriever. arXiv preprint arXiv:2501.08695 (2025)

    work page arXiv 2025

  8. [8]

    Hanqun Cao, Cheng Tan, Zhangyang Gao, Yilun Xu, Guangyong Chen, Pheng- Ann Heng, and Stan Z Li. 2024. A survey on generative diffusion models. IEEE Transactions on Knowledge and Data Engineering (2024)

    work page 2024

  9. [9]

    Yukuo Cen, Jianwei Zhang, Xu Zou, Chang Zhou, Hongxia Yang, and Jie Tang

    work page

  10. [10]

    InProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (Virtual Event, CA, USA) (KDD ’20)

    Controllable Multi-Interest Framework for Recommendation. InProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (Virtual Event, CA, USA) (KDD ’20). Association for Computing Machinery, New York, NY, USA, 2942–2951

    work page

  11. [11]

    Winston C Chow. 2009. Brownian bridge. Wiley interdisciplinary reviews: com- putational statistics 1, 3 (2009), 325–332

    work page 2009

  12. [12]

    Florinel-Alin Croitoru, Vlad Hondru, Radu Tudor Ionescu, and Mubarak Shah

    work page

  13. [13]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 9 (2023), 10850–10869

    Diffusion Models in Vision: A Survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 9 (2023), 10850–10869

    work page 2023

  14. [14]

    Ziqiang Cui, Haolun Wu, Bowei He, Ji Cheng, and Chen Ma. 2024. Context Matters: Enhancing Sequential Recommendation with Context-aware Diffusion- based Contrastive Learning. In Proceedings of the 33rd ACM International Con- ference on Information and Knowledge Management (Boise, ID, USA) (CIKM ’24). Association for Computing Machinery, New York, NY, USA, 404–414

    work page 2024

  15. [15]

    Boyi Deng, Wenjie Wang, Fengbin Zhu, Qifan Wang, and Fuli Feng. 2025. CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG. Proceedings of the AAAI Conference on Artificial Intelligence 39, 22 (Apr. 2025), 23760–23768

    work page 2025

  16. [16]

    Dave Epstein, Allan Jabri, Ben Poole, Alexei Efros, and Aleksander Holynski. 2023. Diffusion Self-Guidance for Controllable Image Generation. InAdvances in Neural Information Processing Systems , A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine (Eds.), Vol. 36. Curran Associates, Inc., 16222–16239

    work page 2023

  17. [17]

    Fiona Fui-Hoon Nah, Ruilin Zheng, Jingyuan Cai, Keng Siau, and Langtao Chen

    work page

  18. [18]

    , 277–304 pages

    Generative AI and ChatGPT: Applications, challenges, and AI-human collaboration. , 277–304 pages

    work page

  19. [19]

    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. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management (Atlanta, GA, USA) (CIKM ’22). 540–550

    work page 2022

  20. [20]

    Ruining He and Julian McAuley. 2016. Fusing Similarity Models with Markov Chains for Sparse Sequential Recommendation. In 2016 IEEE 16th International Conference on Data Mining (ICDM) . 191–200

    work page 2016

  21. [21]

    Balázs Hidasi and Alexandros Karatzoglou. 2018. Recurrent Neural Networks with Top-k Gains for Session-based Recommendations. In Proceedings of the 27th ACM International Conference on Information and Knowledge Management (Torino, Italy) (CIKM ’18). Association for Computing Machinery, New York, NY, USA, 843–852

    work page 2018

  22. [22]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. In Proceedings of the 34th International Conference on Neural Information Processing Systems (Vancouver, BC, Canada) (NIPS ’20). Curran Associates Inc., Red Hook, NY, USA, Article 574, 12 pages

    work page 2020

  23. [23]

    Jonathan Ho and Tim Salimans. 2021. Classifier-Free Diffusion Guidance. In NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications

    work page 2021

  24. [24]

    Guoqing Hu, Zhengyi Yang, Zhibo Cai, An Zhang, and Xiang Wang. 2024. Gen- erate and Instantiate What You Prefer: Text-Guided Diffusion for Sequential Recommendation. arXiv preprint arXiv:2410.13428 (2024)

    work page arXiv 2024

  25. [25]

    Wang-Cheng Kang and Julian McAuley. 2018. Self-Attentive Sequential Rec- ommendation. In 2018 IEEE International Conference on Data Mining (ICDM) . 197–206

    work page 2018

  26. [26]

    Bo Li, Kaitao Xue, Bin Liu, and Yu-Kun Lai. 2023. BBDM: Image-to-Image Translation with Brownian Bridge Diffusion Models. In2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . 1952–1961

    work page 2023

  27. [27]

    Lei Li, Yongfeng Zhang, Dugang Liu, and Li Chen. 2024. Large Language Models for Generative Recommendation: A Survey and Visionary Discussions. In Pro- ceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). ELRA and ICCL, Torino, Italia, 10146–10159

    work page 2024

  28. [28]

    Wuchao Li, Rui Huang, Haijun Zhao, Chi Liu, Kai Zheng, Qi Liu, Na Mou, Guorui Zhou, Defu Lian, Yang Song, Wentian Bao, Enyun Yu, and Wenwu Ou. 2025. DimeRec: A Unified Framework for Enhanced Sequential Recommendation via Generative Diffusion Models. In Proceedings of the Eighteenth ACM International Conference on Web Search and Data Mining (Hannover, Germ...

    work page 2025

  29. [29]

    Yongqi Li, Xinyu Lin, Wenjie Wang, Fuli Feng, Liang Pang, Wenjie Li, Liqiang Nie, Xiangnan He, and Tat-Seng Chua. 2024. A survey of generative search and recom- mendation in the era of large language models. arXiv preprint arXiv:2404.16924 (2024)

    work page arXiv 2024

  30. [30]

    Zihao Li, Aixin Sun, and Chenliang Li. 2023. DiffuRec: A Diffusion Model for Sequential Recommendation. ACM Trans. Inf. Syst. 42, 3, Article 66 (Dec. 2023), 28 pages

    work page 2023

  31. [31]

    Jianghao Lin, Jiaqi Liu, Jiachen Zhu, Yunjia Xi, Chengkai Liu, Yangtian Zhang, Yong Yu, and Weinan Zhang. 2024. A Survey on Diffusion Models for Recom- mender Systems. arXiv preprint arXiv:2409.05033 (2024)

    work page arXiv 2024

  32. [32]

    Qidong Liu, Fan Yan, Xiangyu Zhao, Zhaocheng Du, Huifeng Guo, Ruiming Tang, and Feng Tian. 2023. Diffusion Augmentation for Sequential Recommendation. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management (Birmingham, United Kingdom) (CIKM ’23). Association for Computing Machinery, New York, NY, USA, 1576–1586

    work page 2023

  33. [33]

    Ilya Loshchilov and Frank Hutter. 2019. Decoupled Weight Decay Regulariza- tion. In 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, May 6-9, 2019 . OpenReview.net

    work page 2019

  34. [34]

    Calvin Luo. 2022. Understanding diffusion models: A unified perspective. arXiv preprint arXiv:2208.11970 (2022)

    work page arXiv 2022

  35. [35]

    Haokai Ma, Ruobing Xie, Lei Meng, Xin Chen, Xu Zhang, Leyu Lin, and Zhanhui Kang. 2024. Plug-In Diffusion Model for Sequential Recommendation.Proceedings of the AAAI Conference on Artificial Intelligence 38, 8 (Mar. 2024), 8886–8894

    work page 2024

  36. [36]

    Yong Niu, Xing Xing, Zhichun Jia, Ruidi Liu, Mindong Xin, and Jianfu Cui. 2024. Diffusion Recommendation with Implicit Sequence Influence. In Companion Proceedings of the ACM Web Conference 2024 (Singapore, Singapore) (WWW ’24). Association for Computing Machinery, New York, NY, USA, 1719–1725

    work page 2024

  37. [37]

    Niket Patel, Luis Salamanca, and Luis Barba. 2023. Bridging the Gap: Addressing Discrepancies in Diffusion Model Training for Classifier-Free Guidance. arXiv preprint arXiv:2311.00938 (2023)

    work page arXiv 2023

  38. [38]

    Tran, Jonah Samost, Maciej Kula, Ed H

    Shashank Rajput, Nikhil Mehta, Anima Singh, Raghunandan Keshavan, Trung Vu, Lukasz Heidt, Lichan Hong, Yi Tay, Vinh Q. Tran, Jonah Samost, Maciej Kula, Ed H. Chi, and Maheswaran Sathiamoorthy. 2024. Recommender systems with generative retrieval. In Proceedings of the 37th International Conference on Neural Information Processing Systems (New Orleans, LA, ...

    work page 2024

  39. [39]

    Steffen Rendle, Christoph Freudenthaler, and Lars Schmidt-Thieme. 2010. Factor- izing personalized Markov chains for next-basket recommendation. InProceedings of the 19th International Conference on World Wide Web (Raleigh, North Carolina, USA) (WWW ’10). Association for Computing Machinery, New York, NY, USA, 811–820

    work page 2010

  40. [40]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-Resolution Image Synthesis With Latent Diffusion Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 10684–10695

    work page 2022

  41. [41]

    Dazhong Shen, Guanglu Song, Zeyue Xue, Fu-Yun Wang, and Yu Liu. 2024. Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidance. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 9370–9379

    work page 2024

  42. [42]

    Fei Sun, Jun Liu, Jian Wu, Changhua Pei, Xiao Lin, Wenwu Ou, and Peng Jiang

    work page

  43. [43]

    In Proceedings of the 28th ACM International Conference on Information and Knowledge Management (Beijing, China) (CIKM ’19)

    BERT4Rec: Sequential Recommendation with Bidirectional Encoder Rep- resentations from Transformer. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management (Beijing, China) (CIKM ’19). Association for Computing Machinery, New York, NY, USA, 1441–1450

    work page

  44. [44]

    Jiaxi Tang and Ke Wang. 2018. Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding. In Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining (Marina Del Rey, CA, USA) (WSDM ’18). Association for Computing Machinery, New York, NY, USA, 565–573. Conference acronym ’XX, June 03–05, 2018, Woodstock...

    work page 2018

  45. [45]

    Joojo Walker, Ting Zhong, Fengli Zhang, Qiang Gao, and Fan Zhou. 2022. Rec- ommendation via collaborative diffusion generative model. In International Con- ference on Knowledge Science, Engineering and Management . Springer, 593–605

    work page 2022

  46. [46]

    Chengbing Wang, Yang Zhang, Fengbin Zhu, Jizhi Zhang, Tianhao Shi, and Fuli Feng. 2025. Leveraging Memory Retrieval to Enhance LLM-based Generative Recommendation. In Companion Proceedings of the ACM on Web Conference 2025 . 1346–1350

    work page 2025

  47. [47]

    Peiyong Wang, Bohan Xiao, Qisheng He, Carri Glide-Hurst, and Ming Dong

    work page

  48. [48]

    In Proceedings of the 32nd ACM International Conference on Multimedia (Melbourne VIC, Australia) (MM ’24)

    Score-Based Image-to-Image Brownian Bridge. In Proceedings of the 32nd ACM International Conference on Multimedia (Melbourne VIC, Australia) (MM ’24). Association for Computing Machinery, New York, NY, USA, 10765–10773

    work page

  49. [49]

    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 (2023)

    work page arXiv 2023

  50. [50]

    Wenjie Wang, Yiyan Xu, Fuli Feng, Xinyu Lin, Xiangnan He, and Tat-Seng Chua

    work page

  51. [51]

    In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (Taipei, Taiwan) (SIGIR ’23)

    Diffusion Recommender Model. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (Taipei, Taiwan) (SIGIR ’23). Association for Computing Machinery, New York, NY, USA, 832–841

    work page

  52. [52]

    Yu Wang, Zhiwei Liu, Liangwei Yang, and Philip S. Yu. 2024. Conditional De- noising Diffusion for Sequential Recommendation. In Advances in Knowledge Discovery and Data Mining: 28th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2024, Taipei, Taiwan, May 7–10, 2024, Proceedings, Part V (Taipei, Taiwan). Springer-Verlag, Berlin, Heid...

    work page 2024

  53. [53]

    Jiancan Wu, Xiang Wang, Xingyu Gao, Jiawei Chen, Hongcheng Fu, and Tianyu Qiu. 2024. On the Effectiveness of Sampled Softmax Loss for Item Recommenda- tion. ACM Trans. Inf. Syst. 42, 4, Article 98 (March 2024), 26 pages

    work page 2024

  54. [54]

    Wenjia Xie, Rui Zhou, Hao Wang, Tingjia Shen, and Enhong Chen. 2024. Bridging User Dynamics: Transforming Sequential Recommendations with Schrödinger Bridge and Diffusion Models. In Proceedings of the 33rd ACM International Con- ference on Information and Knowledge Management (Boise, ID, USA) (CIKM ’24). Association for Computing Machinery, New York, NY, ...

    work page 2024

  55. [55]

    Ling Yang, Zhilong Zhang, Yang Song, Shenda Hong, Runsheng Xu, Yue Zhao, Wentao Zhang, Bin Cui, and Ming-Hsuan Yang. 2023. Diffusion Models: A Comprehensive Survey of Methods and Applications. ACM Comput. Surv. 56, 4, Article 105 (Nov. 2023), 39 pages

    work page 2023

  56. [56]

    Zhengyi Yang, Jiancan Wu, Zhicai Wang, Xiang Wang, Yancheng Yuan, and Xiangnan He. 2023. Generate What You Prefer: Reshaping Sequential Recom- mendation via Guided Diffusion. In Advances in Neural Information Processing Systems, A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine (Eds.), Vol. 36. Curran Associates, Inc., 24247–24261

    work page 2023

  57. [57]

    Jiaqi Zhai, Lucy Liao, Xing Liu, Yueming Wang, Rui Li, Xuan Cao, Leon Gao, Zhaojie Gong, Fangda Gu, Jiayuan He, Yinghai Lu, and Yu Shi. 2025. Actions speak louder than words: trillion-parameter sequential transducers for generative recommendations. In Proceedings of the 41st International Conference on Machine Learning (Vienna, Austria) (ICML’24). JMLR.or...

    work page 2025

  58. [58]

    Sheng, Jiajie Xu, Deqing Wang, Guanfeng Liu, and Xiaofang Zhou

    Tingting Zhang, Pengpeng Zhao, Yanchi Liu, Victor S. Sheng, Jiajie Xu, Deqing Wang, Guanfeng Liu, and Xiaofang Zhou. 2019. Feature-level deeper self-attention network for sequential recommendation. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (Macao, China) (IJCAI’19). AAAI Press, 4320–4326

    work page 2019

  59. [59]

    Yang Zhang, Wenxin Xu, Xiaoyan Zhao, Wenjie Wang, Fuli Feng, Xiangnan He, and Tat-Seng Chua. 2025. Reinforced Latent Reasoning for LLM-based Recommendation. arXiv preprint arXiv:2505.19092 (2025)

    work page arXiv 2025

  60. [60]

    Xiaoyan Zhao, Yang Deng, Wenjie Wang, Hong Cheng, Rui Zhang, See-Kiong Ng, Tat-Seng Chua, et al . 2025. Exploring the Impact of Personality Traits on Conversational Recommender Systems: A Simulation with Large Language Models. arXiv preprint arXiv:2504.12313 (2025)

    work page arXiv 2025

  61. [61]

    Xiaoyan Zhao, Juntao You, Yang Zhang, Wenjie Wang, Hong Cheng, Fuli Feng, See-Kiong Ng, and Tat-Seng Chua. 2025. NextQuill: Causal Preference Modeling for Enhancing LLM Personalization. arXiv preprint arXiv:2506.02368 (2025)

    work page arXiv 2025

  62. [62]

    Kun Zhou, Hui Wang, Wayne Xin Zhao, Yutao Zhu, Sirui Wang, Fuzheng Zhang, Zhongyuan Wang, and Ji-Rong Wen. 2020. S3-Rec: Self-Supervised Learning for Sequential Recommendation with Mutual Information Maximization. In Pro- ceedings of the 29th ACM International Conference on Information & Knowledge Management (Virtual Event, Ireland) (CIKM ’20). Associatio...

    work page 2020

  63. [63]

    Linqi Zhou, Aaron Lou, Samar Khanna, and Stefano Ermon. 2024. Denoising Diffusion Bridge Models. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024 . OpenReview.net. A Proofs and Derivations A.1 Deduction Details of Forward Process Initially, we provide the proof of Equation(8). Based on Equatio...

    work page 2024