Pith. sign in

REVIEW 4 major objections 6 minor 46 references

ConnectionMind: Leveraging Social Networks and Large Language Models for Personalized Recommendation at Meta

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read ConnectionMind claims that on social platforms, recommendation should be treated as LLM-guided path discovery over a typed user–item–social graph, and that this approach is production-viable and measurably improves engagement.

desk verdict A credible system paper with a coherent architecture and a real production deployment, but the A/B evidence doesn't actually show that LLM reasoning drives the gains, because the distilled student GNN serves most traffic. read the letter →

arxiv 2608.10187 v1 pith:X47PMLAY submitted 2026-08-10 cs.IR cs.SI

Haoyu Han , Yuming Liu , Lei Huang , Lizhu Zhang , Jiliang Tang , Xiangjun Fan This is my paper · ORCID
classification cs.IRcs.SI
keywords socialrecommendationheterogeneousgraphpathreasoninglargelanguagemodelsreinforcementlearningteacher-studentdistillationproductiondeploymentneuralnetworks
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

ConnectionMind sets out to prove that social evidence in recommendation does not need to be squeezed into a single learned embedding: it can be kept as explicit paths from a user to an item, such as user → friend → liked video or user → group → shared video. The paper constructs a heterogeneous graph of users, pages, groups, and items, and trains an LLM policy to walk that graph and surface candidates with their evidence chains. The policy is first taught to imitate reference paths from logged interactions, then sharpened by reinforcement learning with format, recommendation-quality, and step-wise rewards. On the Delicious and Foursquare benchmarks it beats GNN-based social recommenders and an LLM baseline, and in production it reports +88% relative Recall@10 offline and +0.43% watch time, +0.33% exposure, and +0.22% video sessions online. The paper is establishing that selective, interpretable social reasoning can be deployed at billion-user scale.

What carries the argument

The load-bearing mechanism is the typed temporal heterogeneous graph $G_T=(V,E,R)$, with $V=U\cup P\cup I$, together with an LLM exploration policy that consumes sampled subgraphs and outputs structured expansions and surfaced item–path pairs. The graph makes every social or semantic relationship an explicit, weighted, time-stamped edge, turning 'which friends or groups matter for this user' into a search problem rather than an aggregation problem. The two-stage training—SFT on reference shortest paths, then GRPO over rule-based rewards combining format validity, final-item F1, and step-wise shaping—supplies the policy with both structural validity and recommendation utility, while the teacher–student hybrid inference is what reconciles autoregressive reasoning with millisecond-level serving latency.

What would settle it

Re-run the online A/B test with results stratified by serving path, or measure teacher–student agreement on held-out subgraphs. If the watch-time lift is concentrated in the heavy-user segment served directly by the LLM and the standard-user segment served by the distilled GNN shows no significant lift, the claim that LLM-guided reasoning improves recommendation for the general population is not supported.

Watch

Extended reading notes

Core claim

The central claim is that recommendation can be reformulated as query-conditioned path exploration over a typed heterogeneous graph, with the path itself serving as both the scoring evidence and the explanation. ConnectionMind defines a graph $G_T=(V,E,R)$ whose nodes are users, pages/creators, and items, and whose typed edges encode friendships, follows, group membership, shares, posts, co-watches, and semantic similarity. An LLM policy expands partial paths one typed hop at a time and surfaces item–path pairs; supervised fine-tuning on logged engagement trajectories gives it graph-navigation competence, and end-to-end reinforcement learning with rule-based rewards optimizes the full rollout. The paper reports that this system outperforms GNN-based, diffusion-based, and LLM baselines on public datasets and, in the platform's production pipeline, an 88% relative offline Recall@10 gain over the production GNN baseline, with online A/B lifts of +0.33% exposure, +0.43% watch time, and +0.22% video sessions. The production version splits traffic: the full LLM serves only the top 5–10% of active users, while a distilled GNN emulates the teacher's path selection for everyone else.

Load-bearing premise

In production, the reported gains mostly come from a lightweight student GNN that imitates the LLM teacher, because the full LLM is served only to the top 5–10% of active users; the whole result stands on the assumption that the student faithfully reproduces the teacher's path-finding behavior, which the paper never measures.

Editorial extensions

If this is right

  • Off-the-shelf LLMs cannot perform this kind of recommendation out of the box: the untrained backbone achieves near-zero recall, so task-specific post-training is a prerequisite.
  • Bigger reasoning capacity helps: the 8B policy consistently outperforms the 3B policy on both public benchmarks.
  • Semantic understanding alone is not enough: a vanilla 70B LLM ranker gains +39% offline Recall@10, while graph-reasoning ConnectionMind gains +88%, showing that structure adds signal beyond language.
  • The reasoning benefits can be scaled to latency-critical traffic by distilling the LLM teacher's chosen paths into a lightweight GNN, with the full LLM reserved for the most active users.
  • Recommendations come with explicit evidence paths, which makes social influence traceable for debugging and user-facing explanation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If path-based selectivity is the true driver of the gains, then teacher–student agreement—not LLM size—becomes the binding constraint on production quality; improving distillation fidelity is the most direct next step.
  • The same 'typed graph + LLM policy + distillation' recipe could transfer to other relational domains, such as job, commerce, or health recommendations, wherever a decision can be justified by an explicit chain of relationships.
  • The paper does not test whether surfacing the evidence paths to users changes trust, retention, or perceived relevance; an A/B test that shows or hides the path would isolate that effect.
  • Routing by user activity level may be a proxy; a router that serves direct LLM reasoning based on subgraph complexity or expected path value could allocate reasoning budget more precisely.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. ConnectionMind casts social recommendation as typed path discovery over a heterogeneous graph of users, pages, groups, and items, and uses an LLM policy trained by SFT followed by RL with rule-based rewards to select evidence paths. The paper reports strong public-dataset results on Delicious and Foursquare, and claims production deployment at Meta with a hybrid inference scheme in which the full LLM serves only the top 5–10% of active users and a distilled Student GNN serves the remaining traffic, reporting online A/B lifts in exposure, watch time, and video sessions. The manuscript is clearly written and the training pipeline is described in enough detail to be reproduced on the public datasets, but the central production claim is not yet fully supported because the online A/B test evaluates the hybrid system rather than the LLM reasoning policy in isolation, and the public experiments lack statistical reporting.

Significance. If the attribution gap is closed, this would be a notable systems contribution: it demonstrates a scalable way to inject relational, interpretable path evidence into industrial recommendation, with a clearly described two-stage SFT+RL training recipe and a plausible teacher–student deployment pattern. The graph schema and the path-discovery formulation are useful for follow-up work. However, as presented, the paper does not yet establish that LLM-guided graph reasoning, rather than the distilled student or the heavy-user serving path, is responsible for the measured production gains. The public experimental reporting also lacks variance estimates, so the claimed 'clear margin' over strong social-recommendation baselines is not yet rigorously established. The contribution at this stage is a well-specified framework with preliminary evidence, not a fully validated production result.

major comments (4)
  1. [§6.1, Table 2] The online A/B test evaluates the full hybrid teacher–student system, in which only the top 5–10% of active users receive direct LLM reasoning and the remaining majority is served by a Student GNN trained via offline distillation. The paper reports a single set of lifts for the combined deployment and does not provide teacher–student agreement statistics (e.g., path overlap or ranking agreement) or cohort-stratified A/B results. Because the student handles most traffic, the +0.43% watch-time gain could be driven by the student's graph features, by the heavy-user cohort, or by unrelated production changes, and the abstract and conclusion attribute the gain to ConnectionMind's reasoning policy. This is load-bearing for the paper's central claim; please report teacher–student agreement and A/B lifts separately for the two serving paths, or substantially soften the claim to a system-level result.
  2. [§5.2, Table 1] The public benchmark comparison reports a single number per method and metric, without standard deviations, confidence intervals, or significance tests, and no information is given about the number of random seeds or the tuning protocol for the baselines. The claim of a 'clear margin' over MHCN and SEPT cannot be assessed from point estimates alone, especially since the absolute Recall@5 values are very low (e.g., 0.0631 on Delicious). Please add repeated-run statistics, state the number of seeds, and describe the hyperparameter selection procedure for all baselines.
  3. [§6.2, Table 2] The offline production evaluation is underspecified: the 'Production GNN Baseline' is never named or described, and there is no definition of the offline evaluation set, the ground-truth labels, the candidate pool, or the way Recall@10 is computed in the production offline setting. The +88% relative lift is therefore not interpretable. Please name the baseline, describe the evaluation methodology, and report the underlying absolute metric values rather than only relative lifts.
  4. [§3] The three 'consistent patterns' from production-scale interaction logs are reported only qualitatively, with the authors explicitly stating that the findings are not shown due to privacy and platform policy constraints. While this constraint is understandable, these observations are used as the primary motivation for the graph and path-based design. Please provide sanitized summary statistics or an explicit internal reference, or clearly mark these observations as unverifiable motivating hypotheses rather than as established empirical findings.
minor comments (6)
  1. [Figure 3] The y-axis is labeled 'Recall' even though the figure plots Recall@5 for Delicious and Recall@10 for Foursquare in the same panel; the axis and caption should state this explicitly, and error bars should be added.
  2. [§5.1] Dataset statistics (number of users, items, relations, and density) and the exact train/validation/test split procedure are not reported; please add them, and clarify whether the split respects temporal order given that the graph edges are time-stamped.
  3. [§6.2, Table 2] The confidence intervals are reported without p-values, sample sizes, or a description of the statistical test, so the phrase 'statistically significant' is not substantiated; please specify the test and the randomization unit.
  4. [§4.3] The step-wise reward uses shortest paths to reachable positives in the sampled subgraph; this privileged information should be discussed as a potential bias, since the policy's behavior may then depend heavily on the subgraph sampler and on shortest-path coverage rather than on learned reasoning.
  5. [Abstract and Conclusion] The phrase 'deployed in Meta's large-scale recommendation pipeline' should be qualified in both the abstract and the conclusion, since the full LLM policy serves only a small slice of traffic; consider saying 'deployed in a hybrid teacher–student configuration.'
  6. [§2.2] The related-work section is adequate but does not discuss recent path-based or reasoning-based recommendation methods beyond metapath and GNN aggregation; adding a few recent references would sharpen the positioning of the contribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the derivation is standard supervised learning plus external A/B testing; production attribution is a validity caveat, not a circular reduction.

full rationale

The paper's claimed derivation chain is: construct a heterogeneous social-item graph, train an LLM policy via SFT on logged positive engagements followed by RL with rule-based rewards computed against the same kind of positive labels, and evaluate on held-out positives from public datasets and a production A/B test. This is ordinary supervised learning and reinforcement learning: the output items are not defined as the training labels by construction, and the held-out evaluation is external to the fitted values. The Section 3 production-log observations are qualitative and are used to motivate the graph and sampling design; the later offline/online results do not mathematically reduce those observations to the system's outputs, so they are at most a weak post-hoc consistency claim rather than a circular step. The teacher-student hybrid in Section 6.1 means the A/B lift is mostly produced by the distilled Student GNN, but this is an attribution/validity limitation, not a self-referential reduction: nothing in the paper's equations forces the measured lift to equal the teacher's fitted values. Self-citations (e.g., refs [11], [6], [25], [26]) are background literature and are not load-bearing. No equation quoted in the paper reduces a prediction to a fitted input, and no uniqueness or ansatz is imported from the authors' prior work. Therefore no circularity under the quoted-reduction standard.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central result rests on hand-set hyperparameters (M, k, D_max, reward weights), qualitative production-log observations, and the unverified assumption that the student GNN faithfully transfers the teacher's reasoning to the majority of traffic. No new physical or conceptual entities are introduced.

free parameters (4)
  • M (max sampled neighbors per relation per node)
    Introduced in Section 3.2; no value or tuning details reported, controls subgraph size and hence the search space for the policy.
  • k (local subgraph hop limit) and D_max (max exploration depth)
    Introduced in Sections 3.2 and 4.1; not specified, bounds path discovery and rollout cost.
  • Reward weights alpha_rec, alpha_step, alpha_fmt = 0.5, 0.3, 0.2
    Set by hand in Section 4.3; no sensitivity analysis reported.
  • Step-wise penalty schedule eta(1) and eta(>=2) = 0.3 and 0.5
    Hand-chosen in Section 4.3; no sensitivity analysis reported.
assumptions (6)
  • domain assumption The heterogeneous graph relations (friend, similar, follow, group, share, post, co-watch, semantic) are sufficient to capture the social signals that drive engagement.
    Section 3.1 constructs the graph from these relation types; if important relations are omitted, the path reasoning space is incomplete.
  • ad hoc to paper The top-M weighted neighbor sampling preserves the high-signal paths needed for recommendation.
    Section 3.2 states sampling up to M neighbors per relation based on edge weights; no analysis that this sparsification is lossless.
  • domain assumption Shortest paths from user to engaged items in the sampled subgraph are reliable supervision trajectories for learning reasoning behavior.
    Section 4.2 uses these as gold actions; this assumes logged engagement implies a causally valid reasoning chain.
  • ad hoc to paper The LLM policy output is executable and relation-consistent after SFT plus RL.
    The whole framework depends on the serialized structured actions being parseable and graph-valid, enforced only by a format reward.
  • ad hoc to paper The student GNN distillation faithfully approximates the teacher policy for the majority of traffic.
    Section 6.1 routes 90-95% of users to the student; the online A/B gain is therefore attributed to the student's imitation quality, which is not directly measured.
  • domain assumption The three qualitative production-log observations (social signals essential, selective, temporally stable) hold as stated.
    Section 3 states these patterns are reported qualitatively due to privacy; no numbers are given, yet they motivate the entire design.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ConnectionMind: Leveraging Social Networks and Large Language Models for Personalized Recommendation at Meta." pith.science (2026). https://pith.science/paper/X47PMLAY

@misc{pith2026260810187,
  author       = {Pith},
  title        = {Pith review of: ConnectionMind: Leveraging Social Networks and Large Language Models for Personalized Recommendation at Meta},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X47PMLAY}},
  note         = {Machine review of arXiv:2608.10187}
}
read the original abstract

Modern recommendation systems on social media platforms such as Meta must model complex social relationships, including friendships, group memberships, and creator interactions, alongside massive and heterogeneous content such as text and video. Traditional recommendation models, however, often omit these signals or treat them independently, lacking the reasoning capability to integrate multi-relational context for fine-grained personalization. We present ConnectionMind, a production-ready recommendation framework that tightly integrates the social network structure with large language models (LLMs) to enable scalable, interpretable, and reasoning-aware personalization in Meta. ConnectionMind constructs a heterogeneous graph connecting users, items, friends, groups, and creator pages, and formulates recommendation as a graph reasoning problem: discovering personalized paths from users to candidate items. An LLM-based policy is employed to reason over these graph structures and guide recommendation decisions. To train the system at scale, ConnectionMind adopts a two-stage learning strategy. We first perform supervised fine-tuning (SFT) on large-scale user-item interaction trajectories to initialize the reasoning policy, followed by end-to-end reinforcement learning (RL) to refine the model's ability to reason over social graphs for personalized recommendation. Extensive experiments on multiple real-world datasets demonstrate the effectiveness of ConnectionMind compared to representative baselines. More importantly, ConnectionMind has been deployed in Meta's large-scale recommendation pipeline and has been evaluated through online A/B tests, achieving a 0.43% improvement in video watch time. These results demonstrate measurable real-world impact in a production recommendation system.

Figures

Figures reproduced from arXiv: 2608.10187 by the authors.

Figure 1
Figure 1. The graph serves as the structured search space for our [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. An overview of the proposed framework ConnectionMind. We construct a heterogeneous social–item graph from production data and use an LLM policy to perform multi-step graph exploration that surfaces recommended items with evidence paths. The policy is trained in two stages: (1) supervised fine-tuning with step-wise supervision, followed by (2) end-to-end reinforcement learning with rule-based rewards. Trajectory cons… view at source ↗
Figure 3
Figure 3. Performance comparison of different variants. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The Hybrid Inference Architecture. The system [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 20 canonical work pages

  1. [1]

    Keqin Bao, Jizhi Zhang, Wenjie Wang, Yang Zhang, Zhengyi Yang, Yancheng Luo, Chong Chen, Fuli Feng, and Qi Tian. 2023. A Bi-Step Grounding Para- digm for Large Language Models in Recommendation Systems.arXiv preprint arXiv:2308.08434(2023). https://arxiv.org/abs/2308.08434

    arXiv 2023

  2. [2]

    Iván Cantador, Peter Brusilovsky, and Tsvi Kuflik. 2011. 2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011). InProceedings of the 5th ACM conference on Recommender systems(Chicago, IL, USA)(RecSys 2011). ACM, New York, NY, USA

    work page 2011

  3. [3]

    Heng-Tze Cheng, Levent Koc, Jeremiah Harmsen, Tal Shaked, Tushar Chandra, Hrishi Aradhye, Glen Anderson, Greg Corrado, Wei Chai, Mustafa Ispir, et al

  4. [4]

    Yuxiao Dong, Nitesh V Chawla, and Ananthram Swami. 2017. metapath2vec: Scalable representation learning for heterogeneous networks. InProceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining. 135–144

    2017

  5. [5]

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. 2024. The llama 3 herd of models.arXiv e-prints(2024), arXiv–2407

    2024

  6. [6]

    Wenqi Fan, Yao Ma, Qing Li, Yuan He, Eric Zhao, Jiliang Tang, and Dawei Yin

  7. [7]

    Shijie Geng, Shuchang Liu, Zuohui Fu, Yingqiang Ge, and Yongfeng Zhang. 2022. Recommendation as language processing (rlp): A unified pretrain, personalized prompt & predict paradigm (p5). InProceedings of the 16th ACM conference on recommender systems. 299–315

    2022

  8. [8]

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al . 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning.arXiv preprint arXiv:2501.12948(2025)

    arXiv 2025

Show all 46 references
  1. [9]

    Guibing Guo, Jie Zhang, and Neil Yorke-Smith. 2015. Trustsvd: Collaborative filtering with both the explicit and implicit influence of user trust and of item ratings. InProceedings of the AAAI conference on artificial intelligence, Vol. 29

    2015

  2. [10]

    Hanze Guo, Yijun Ma, and Xiao Zhou. 2025. Sorex: Towards self-explainable social recommendation with relevant ego-path extraction.ACM Transactions on Information Systems44, 2 (2025), 1–27

    2025

  3. [11]

    Haoyu Han, Mengdi Zhang, Min Hou, Fuzheng Zhang, Zhongyuan Wang, Enhong Chen, Hongwei Wang, Jianhui Ma, and Qi Liu. 2020. STGCN: a spatial-temporal aware graph learning method for POI recommendation. In2020 IEEE International Conference on Data Mining (ICDM). IEEE, 1052–1057

    2020

  4. [12]

    Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. 2017. Neural collaborative filtering. InProceedings of the 26th international conference on world wide web. 173–182

    2017

  5. [13]

    Mohsen Jamali and Martin Ester. 2009. Trustwalker: a random walk model for combining trust-based and item-based recommendation. InProceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining. 397–406

    2009

  6. [14]

    Hyeyoung Ko, Suyeon Lee, Yoonseo Park, and Anna Choi. 2022. A survey of recommendation systems: recommendation models, techniques, and application fields.Electronics11, 1 (2022), 141

    2022

  7. [15]

    Yehuda Koren, Robert Bell, and Chris Volinsky. 2009. Matrix factorization tech- niques for recommender systems.Computer42, 8 (2009), 30–37

    2009

  8. [16]

    Xiao Li, Li Sun, Mengjie Ling, and Yan Peng. 2023. A survey of graph neural network based recommendation in social networks.Neurocomputing549 (2023), 126441

    2023

  9. [17]

    Yang Li, Kangbo Liu, Ranjan Satapathy, Suhang Wang, and Erik Cambria. 2024. Recent developments in recommender systems: A survey.IEEE Computational Intelligence Magazine19, 2 (2024), 78–95

    2024

  10. [18]

    Zongwei Li, Lianghao Xia, and Chao Huang. 2024. Recdiff: Diffusion model for social recommendation. InProceedings of the 33rd ACM international conference on information and knowledge management. 1346–1355

    2024

  11. [19]

    Yuqing Liu, Yu Wang, Lichao Sun, and Philip S Yu. 2024. Rec-gpt4v: Mul- timodal recommendation with large vision-language models.arXiv preprint arXiv:2402.08670(2024)

    2024 arXiv

  12. [20]

    Yixin Liu, Kai Zhang, Yuan Li, Zhiling Yan, Chujie Gao, Ruoxi Chen, Zhengqing Yuan, Yue Huang, Hanchi Sun, Jianfeng Gao, et al . 2024. Sora: A review on background, technology, limitations, and opportunities of large vision models. arXiv preprint arXiv:2402.17177(2024)

    2024 arXiv

  13. [21]

    Sichun Luo, Bowei He, Haohan Zhao, Wei Shao, Yanlin Qi, Yinya Huang, Aojun Zhou, Yuxuan Yao, Zongpeng Li, Yuanzhang Xiao, et al. 2025. Recranker: Instruc- tion tuning large language model as ranker for top-k recommendation.ACM Transactions on Information Systems43, 5 (2025), 1–31

    2025

  14. [22]

    Hao Ma, Haixuan Yang, Michael R Lyu, and Irwin King. 2008. Sorec: social recommendation using probabilistic matrix factorization. InProceedings of the 17th ACM conference on Information and knowledge management. 931–940

    2008

  15. [23]

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

  16. [24]

    Chuan Shi, Binbin Hu, Wayne Xin Zhao, and Philip S Yu. 2018. Heterogeneous information network embedding for recommendation.IEEE transactions on knowledge and data engineering31, 2 (2018), 357–370

    2018

  17. [25]

    Jiliang Tang, Xia Hu, and Huan Liu. 2013. Social recommendation: a review. Social network analysis and mining3, 4 (2013), 1113–1133

    2013

  18. [26]

    Jiliang Tang, Suhang Wang, Xia Hu, Dawei Yin, Yingzhou Bi, Yi Chang, and Huan Liu. 2016. Recommendation with social dimensions. InProceedings of the AAAI conference on artificial intelligence, Vol. 30

    2016

  19. [27]

    Yunlong Tang, Jing Bi, Siting Xu, Luchuan Song, Susan Liang, Teng Wang, Daoan Zhang, Jie An, Jingyang Lin, Rongyi Zhu, et al. 2025. Video understanding with large language models: A survey.IEEE Transactions on Circuits and Systems for Video Technology(2025)

    2025

  20. [28]

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023. Llama: Open and efficient foundation language models.arXiv preprint arXiv:2302.13971(2023)

    2023 arXiv

  21. [29]

    Le Wu, Junwei Li, Peijie Sun, Richang Hong, Yong Ge, and Meng Wang. 2020. Diffnet++: A neural influence and interest diffusion network for social recom- mendation.IEEE Transactions on Knowledge and Data Engineering34, 10 (2020), 4753–4766

    2020

  22. [30]

    Le Wu, Peijie Sun, Yanjie Fu, Richang Hong, Xiting Wang, and Meng Wang. 2019. A neural influence diffusion model for social recommendation. InProceedings of the 42nd international ACM SIGIR conference on research and development in information retrieval. 235–244

    2019

  23. [31]

    Likang Wu, Zhi Zheng, Zhaopeng Qiu, Hao Wang, Hongchao Gu, Tingjia Shen, Chuan Qin, Chen Zhu, Hengshu Zhu, Qi Liu, et al . 2024. A survey on large language models for recommendation.World Wide Web27, 5 (2024), 60

    2024

  24. [32]

    Shuyuan Xu, Wenyue Hua, and Yongfeng Zhang. 2023. Openp5: Benchmarking foundation models for recommendation.arXiv preprint arXiv:2306.11134(2023)

    2023 arXiv

  25. [33]

    Bo Yang, Yu Lei, Jiming Liu, and Wenjie Li. 2016. Social collaborative filtering by trust.IEEE transactions on pattern analysis and machine intelligence39, 8 (2016), 1633–1647

    2016

  26. [34]

    Dingqi Yang, Bingqing Qu, Jie Yang, and Philippe Cudre-Mauroux. 2019. Revis- iting user mobility and social relationships in lbsns: a hypergraph embedding approach. InThe world wide web conference. 2147–2157

    2019

  27. [35]

    Dingqi Yang, Bingqing Qu, Jie Yang, and Philippe Cudré-Mauroux. 2020. Lbsn2vec++: Heterogeneous hypergraph embedding for location-based social networks.IEEE Transactions on Knowledge and Data Engineering34, 4 (2020), 1843–1855

    2020

  28. [36]

    Yonghui Yang, Le Wu, Yuxin Liao, Zhuangzhuang He, Pengyang Shao, Richang Hong, and Meng Wang. 2025. Invariance matters: Empowering social recom- mendation via graph invariant learning. InProceedings of the 48th International ACM SIGIR Conference on Research and Development in ...

    2025

  29. [37]

    Junliang Yu, Hongzhi Yin, Min Gao, Xin Xia, Xiangliang Zhang, and Nguyen Quoc Viet Hung. 2021. Socially-aware self-supervised tri-training for recommendation. InProceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining. 2084–2092

    2021

  30. [38]

    Junliang Yu, Hongzhi Yin, Jundong Li, Qinyong Wang, Nguyen Quoc Viet Hung, and Xiangliang Zhang. 2021. Self-supervised multi-channel hypergraph convolu- tional network for social recommendation. InProceedings of the web conference

    2021

  31. [39]

    Zhenrui Yue, Sara Rabhi, Gabriel de Souza Pereira Moreira, Dong Wang, and Even Oldridge. 2023. Llamarec: Two-stage recommendation using large language models for ranking.arXiv preprint arXiv:2311.02089(2023)

    2023 arXiv

  32. [40]

    Jingyi Zhang, Jiaxing Huang, Sheng Jin, and Shijian Lu. 2024. Vision-language models for vision tasks: A survey.IEEE transactions on pattern analysis and machine intelligence46, 8 (2024), 5625–5644

    2024

  33. [41]

    Junjie Zhang, Ruobing Xie, Yupeng Hou, Xin Zhao, Leyu Lin, and Ji-Rong Wen

  34. [42]

    Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al. 2023. A survey of large language models.arXiv preprint arXiv:2303.182231, 2 (2023)

    2023 arXiv

  35. [2016]

    InProceedings of the 1st workshop on deep learning for recommender systems

    Wide & deep learning for recommender systems. InProceedings of the 1st workshop on deep learning for recommender systems. 7–10

  36. [2019]

    InThe world wide web conference

    Graph neural networks for social recommendation. InThe world wide web conference. 417–426

  37. [2023]

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

    2023

  38. [2025]

    Recommendation as instruction following: A large language model em- powered recommendation approach.ACM Transactions on Information Systems 43, 5 (2025), 1–37

    2025

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.