Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

When users are LLM agents, recommendation collapses from personalization into structural pattern matching.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-13 07:13 UTC pith:O3EXJW6X

load-bearing objection First solid offline recsys map on a pure agent social platform; the structural-over-personalization ranking is real enough to take seriously, even if sparsity and missing significance tests keep the leap to "collapse" provisional. the 3 major comments →

arxiv 2606.29762 v2 pith:O3EXJW6X submitted 2026-06-29 cs.IR

Do Recommendation Algorithms Work When Users Are LLM Agents? A Case Study on Moltbook

Daming Li , Simeng Han , Jialu Zhang This is my paper
classification cs.IR
keywords recommender systemsMoltbookmulti-agent systemscollaborative filteringpersonalizationuser modelingLLM agentssocial networks
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper asks whether the recommender systems built for humans still work when the users are autonomous LLM agents that may not hold stable content preferences. On Moltbook, a large Reddit-style social platform used only by AI agents, the authors cast forum recommendation as the task of predicting which topic communities an agent will post or comment in next. Across simple popularity rules, matrix factorization, item-based collaborative filtering, graph models, sequential transformers, and content embeddings, the methods that win are those that exploit item co-occurrence and aggregate popularity rather than those that try to learn a personalized user embedding. Static agent persona descriptions, the closest available preference profiles, add little or nothing. The paper also shows that accuracy does not decay as the train–test time gap grows, unlike the familiar drift pattern on human platforms. The result is both a caution for platforms that mix human and agent traffic and a new empirical window on how agent societies actually consume content.

Core claim

On Moltbook, recommendation signal lives in platform- and item-level structure, not in user-specific personalization. ItemKNN and popularity heuristics outperform or match BPR-MF, hybrid content-MF, pure content matching, LightGCN, and SASRec; agent description embeddings fail to help; and performance is stationary across multi-week train–test gaps. Personalization therefore collapses into structural pattern matching for these agent users.

What carries the argument

Forum (submolt) recommendation framed as next-community prediction from binary or karma-weighted agent–submolt interaction matrices, evaluated by Recall, NDCG, Hit Rate, and MRR under temporal train–test splits that exclude already-visited communities.

Load-bearing premise

The authors treat an agent’s posts and comments as a reliable proxy for engagement preference, even though the platform never logs impressions or passive views.

What would settle it

If richer session or impression logs (or agents whose configurations explicitly store evolving content preferences) made user-embedding models such as BPR-MF or SASRec clearly beat ItemKNN and TopPopular, the claim that personalization collapses would be overturned.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper formulates a forum (submolt) recommendation task on Moltbook, a Reddit-style platform used only by OpenClaw LLM agents, and evaluates eight methods (Random, TopPopular, BPR-MF, HybridMF, ContentBased, ItemKNN, LightGCN, SASRec) plus description and karma ablations. Using a temporal train/test split on a public 10-week archive, the authors report that item-side collaborative filtering and popularity heuristics outperform methods that learn user embeddings; agent description embeddings add little or no value; karma weighting helps structural methods but not BPR-MF; and performance does not decay as the train–test gap grows. They interpret this as evidence that agent content consumption lacks persistent personal preferences, so recommendation collapses to structural pattern matching, with implications for mixed human–agent platforms.

Significance. This is, to my knowledge, the first large-scale empirical study of classical and sequential recommenders when users are autonomous LLM agents rather than humans. The setting is timely, the dataset is public, and the experimental design (temporal split, standard top-N metrics, karma and description ablations, temporal-gap stress test) is appropriate for a first baseline paper. The two-tier ranking and the flat temporal-decay curves are interesting and potentially important for both multi-agent systems and recommender design under agent pollution of human platforms. Strengths include transparent sparsity reporting, comparison to sparse human benchmarks (Amazon-Books), and explicit discussion of OpenClaw configuration constraints. The work opens a clear research direction even if some interpretive leaps need tightening.

major comments (3)
  1. [§5.1, Table 2] Table 2 and §5.1: The central claim that personalization collapses to structural pattern matching rests on the ordering ItemKNN / LightGCN / SASRec / TopPopular ≫ BPR-MF / HybridMF / ContentBased. Absolute metrics are very small (e.g., Recall@10 ≈ 0.023–0.024) on a 0.041%-dense matrix, and the paper reports no per-agent variance, bootstrap intervals, or paired significance tests. Under extreme sparsity, user-factor models are more variance-sensitive than pure item co-occurrence; the observed gaps could be estimation noise. Please add statistical checks (e.g., paired tests or bootstrap CIs over agents) that the ranking is stable, or qualify the qualitative leap accordingly.
  2. [§4.1, §6.3] §4.1 and §6.3: The evaluation treats posts and comments as the sole positive labels for submolt engagement, while impressions and passive views are unobserved. If agents post for scheduled tasks, owner instructions, or platform affordances rather than content preference, the ranking measures next-action predictability, not preference. This is acknowledged as a limitation but is load-bearing for the preference-collapse interpretation. Strengthen the discussion with (i) any available evidence that posting/commenting correlates with topical affinity, or (ii) a clearer reframing of the claim as “next-forum engagement prediction under observed actions” rather than preference per se.
  3. [§5.4, Figure 2] §5.4 / Figure 2: The no-temporal-decay result is important for the stationarity argument, but the evaluation population shrinks sharply with gap (4,677 → 1,895 agents). Although the authors check training interaction counts, residual selection bias (only persistently active agents remain) could flatten decay curves. Report metrics restricted to a fixed agent cohort present at all gaps, or show that the flat profile survives that restriction; otherwise the stationarity claim is only partially supported.
minor comments (5)
  1. [Abstract, §1, §4.5] Abstract says “nine recommendation methods”; Introduction and §4.5 say “eight.” Align the count (and list) throughout.
  2. [Table 1, §4.2] Table 1 vs. §4.2: full-data active agents 79,643 / submolts 5,406 vs. training matrix 79,596 × 5,359. A one-sentence note on why the training matrix is slightly smaller would help.
  3. [Figure 1] Figure 1 caption lists three panels but the text refers to four metrics; MRR is only in tables. Minor consistency polish.
  4. [§4.5.3] §4.5.3 HybridMF uses α=0.1 and d=16 without reporting the validation grid; a short appendix table of tuned ranges would aid reproducibility.
  5. [§6.2] Related work on bot detection and agent pollution of social platforms could cite a few more recent bot-detection / agent-identification papers when discussing §6.2 implications.

Circularity Check

0 steps flagged

No circularity: purely empirical algorithm ranking on temporal hold-out; no derivation reduces to its own inputs by construction.

full rationale

The paper formulates a standard top-N forum recommendation task on public Moltbook interaction logs, trains eight classical methods (Random, TopPopular, BPR-MF, HybridMF, ContentBased, ItemKNN, LightGCN, SASRec) plus ablations (karma weighting, description embeddings) under a fixed temporal split, and reports Recall/NDCG/HR/MRR. All claims (ItemKNN/TopPopular beat user-factor models; descriptions add no value; no temporal decay) are direct numerical comparisons of those metrics; none is obtained by fitting a parameter and then re-labeling a related quantity as a prediction, nor by defining a quantity in terms of the target result. Self-citations (dataset archive, OpenClaw configuration) supply background only and are not load-bearing uniqueness theorems or ansatzes that force the ranking. Absolute metrics are low because of sparsity, but that is a correctness/statistical concern, not circularity. The derivation chain is therefore self-contained and non-circular.

Axiom & Free-Parameter Ledger

3 free parameters · 3 axioms · 0 invented entities

The central comparative claim rests on standard recsys evaluation axioms plus domain assumptions about OpenClaw agent architecture that cannot be verified from the public logs. No new physical or mathematical entities are postulated; free parameters are ordinary model hyperparameters that do not alter the qualitative ranking.

free parameters (3)
  • ItemKNN neighborhood size K
    Fixed at 50; affects absolute scores but not the structural-vs-personalization ranking that carries the claim.
  • HybridMF content weight alpha
    Set to 0.1; description-ablation alphas (0.1/0.3/0.5) are also free choices that leave the negative persona result unchanged.
  • BPR-MF / LightGCN / SASRec embedding dimension and training epochs
    Standard values (d=64, 20–100 epochs) chosen by validation Recall@10; not load-bearing for the qualitative conclusion.
axioms (3)
  • domain assumption Posting or commenting in a submolt is a valid positive interaction signal for forum recommendation.
    Stated in Section 4.1; required because impression logs do not exist.
  • domain assumption OpenClaw agents are effectively stateless across sessions and their configuration files do not encode persistent content preferences.
    Invoked in Sections 3.1 and 6.1 to explain why user embeddings fail; configuration files themselves are unobserved.
  • standard math Temporal train/test split without random leakage is the correct evaluation protocol.
    Standard recsys practice cited from Campos et al. and used throughout Section 4.

pith-pipeline@v1.1.0-grok45 · 22719 in / 2422 out tokens · 23955 ms · 2026-07-13T07:13:41.404056+00:00 · methodology

0 comments
read the original abstract

Large language model (LLM) agents are increasingly populating web platforms, raising a fundamental question for recommender systems: do algorithms designed for human users still work when users are LLM agents that may not have well-defined content consumption preferences? We study this question by formulating a forum recommendation problem on Moltbook, a large-scale social media platform exclusively for autonomous AI agents running on the OpenClaw framework. We evaluate nine recommendation methods spanning simple heuristic rules, matrix factorization, itemand user-based collaborative filtering, graph-based, and sequential models on the task of predicting which forums an agent will engage with next. We find that simple popularity-based rules or item-side collaborative filtering leveraging the platform and item structural information outperform techniques that explicitly learn a user representation. The static agent persona descriptions, the closest analog to a preference profile, fail to add value in predicting engagement. These results suggest that, on Moltbook, recommendation depends more on platform- and item-level structural signals than on user-specific personalization. We present multiple lines of empirical evidence that the observed content consumption patterns on Moltbook differ from well-established findings on human recommendation datasets, providing a new angle for studying agent societies and designing robust recommendation algorithms as agents increasingly populate the web.

Figures

Figures reproduced from arXiv: 2606.29762 by Daming Li, Jialu Zhang, Simeng Han.

Figure 1
Figure 1. Figure 1: Recall@𝐾, NDCG@𝐾, and HR@𝐾 across all models as 𝐾 varies [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Recommendation performance vs. train-test temporal gaps for all models. [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Review history (2 revisions) →

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. A Position Paper on Recommender Systems in the Era of Autonomous Agents

    cs.IR 2026-07 accept novelty 4.0

    A position paper proposes that transaction-oriented recommender systems be redesigned around client-side autonomous agents that query, compare, and verify options across platforms.

Reference graph

Works this paper leans on

77 extracted references · 12 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Mohamad Abou Ali, Fadi Dornaika, and Jinan Charafeddine. 2025. Agentic AI: a comprehensive survey of architectures, applications, and future directions. Artificial Intelligence Review59, 1 (2025), 11

    2025

  2. [2]

    Ariel Flint Ashery, Luca Maria Aiello, and Andrea Baronchelli. 2025. Emergent social conventions and collective bias in LLM populations.Science Advances11, 20 (2025), eadu9368

    2025

  3. [3]

    Keqin Bao, Jizhi Zhang, Yang Zhang, Wenjie Wang, Fuli Feng, and Xiangnan He. 2023. Tallrec: An effective and efficient tuning framework to align large language model with recommendation. InProceedings of the 17th ACM conference on recommender systems. 1007–1014

    2023

  4. [4]

    Jesús Bobadilla, Fernando Ortega, Antonio Hernando, and Abraham Gutiérrez

  5. [5]

    Recommender systems survey.Knowledge-based systems46 (2013), 109– 132

    2013

  6. [6]

    Veronika Bogina, Tsvi Kuflik, Dietmar Jannach, Maria Bielikova, Michal Kompan, and Christoph Trattner. 2023. Considering temporal aspects in recommender systems: a survey: V. Bogina et al.User Modeling and User-Adapted Interaction33, 1 (2023), 81–119

    2023

  7. [7]

    Robin Burke. 2002. Hybrid recommender systems: Survey and experiments.User modeling and user-adapted interaction12, 4 (2002), 331–370

    2002

  8. [8]

    Pedro G Campos, Fernando Díez, and Iván Cantador. 2014. Time-aware recom- mender systems: a comprehensive survey and analysis of existing evaluation protocols.User Modeling and User-Adapted Interaction24, 1 (2014), 67–119

    2014

  9. [9]

    Micah Carroll, Dylan Hadfield-Menell, Stuart Russell, and Anca Dragan. 2021. Es- timating and penalizing preference shift in recommender systems. InProceedings of the 15th ACM Conference on Recommender Systems. 661–667

    2021

  10. [10]

    Lin Chen, Yunke Zhang, Jie Feng, Haoye Chai, Honglin Zhang, Bingbing Fan, Yibo Ma, Shiyuan Zhang, Nian Li, Tianhui Liu, et al. 2026. AI agent behavioral science.Humanities and Social Sciences Communications(2026)

    2026

  11. [11]

    Paolo Cremonesi, Yehuda Koren, and Roberto Turrin. 2010. Performance of recommender algorithms on top-n recommendation tasks. InProceedings of the fourth ACM conference on Recommender systems. 39–46

    2010

  12. [12]

    Stefano Cresci. 2020. A decade of social bot detection.Commun. ACM63, 10 (2020), 72–83

    2020

  13. [13]

    Jiaxin Deng, Shiyao Wang, Kuo Cai, Lejian Ren, Qigen Hu, Weifeng Ding, Qiang Luo, and Guorui Zhou. 2025. Onerec: Unifying retrieve and rank with generative recommender and iterative preference alignment.arXiv preprint arXiv:2502.18965 (2025)

    Pith/arXiv arXiv 2025

  14. [14]

    Mukund Deshpande and George Karypis. 2004. Item-based top-n recommenda- tion algorithms.ACM Transactions on Information Systems (TOIS)22, 1 (2004), 143–177

    2004

  15. [15]

    Zane Durante, Qiuyuan Huang, Naoki Wake, Ran Gong, Jae Sung Park, Bidipta Sarkar, Rohan Taori, Yusuke Noda, Demetri Terzopoulos, Yejin Choi, et al. 2024. Agent ai: Surveying the horizons of multimodal interaction.arXiv preprint arXiv:2401.03568(2024)

    Pith/arXiv arXiv 2024

  16. [16]

    Yunfan Gao, Tao Sheng, Youlin Xiang, Yun Xiong, Haofen Wang, and Jiawei Zhang. 2023. Chat-rec: Towards interactive and explainable llms-augmented recommender system.arXiv preprint arXiv:2303.14524(2023)

    Pith/arXiv arXiv 2023

  17. [17]

    Sushant Gautam, Annika W Olstad, Klas H Pettersen, and Michael A Riegler

  18. [18]

    The Moltbook Observatory Archive: an incremental dataset of agent-only social network activity.arXiv preprint arXiv:2605.13860(2026)

    Pith/arXiv arXiv 2026

  19. [19]

    David Goldberg, David Nichols, Brian M Oki, and Douglas Terry. 1992. Using collaborative filtering to weave an information tapestry.Commun. ACM35, 12 (1992), 61–70

    1992

  20. [20]

    Asela Gunawardana and Guy Shani. 2009. A survey of accuracy evaluation metrics of recommendation tasks.Journal of Machine Learning Research10, 12 (2009)

    2009

  21. [21]

    F Maxwell Harper and Joseph A Konstan. 2015. The movielens datasets: History and context.Acm transactions on interactive intelligent systems (tiis)5, 4 (2015), 1–19

    2015

  22. [22]

    Xiangnan He, Kuan Deng, Xiang Wang, Yan Li, Yongdong Zhang, and Meng Wang. 2020. Lightgcn: Simplifying and powering graph convolution network for recommendation. InProceedings of the 43rd International ACM SIGIR conference on research and development in Information Retrieval. 639–648

    2020

  23. [23]

    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

  24. [24]

    Xinran He, Junfeng Pan, Ou Jin, Tianbing Xu, Bo Liu, Tao Xu, Yanxin Shi, Antoine Atallah, Ralf Herbrich, Stuart Bowers, et al. 2014. Practical lessons from predicting clicks on ads at facebook. InProceedings of the eighth international workshop on data mining for online advertising. 1–9

    2014

  25. [25]

    Jonathan L Herlocker, Joseph A Konstan, Loren G Terveen, and John T Riedl

  26. [26]

    Evaluating collaborative filtering recommender systems.ACM Transactions on Information Systems (TOIS)22, 1 (2004), 5–53

    2004

  27. [27]

    Yifan Hu, Yehuda Koren, and Chris Volinsky. 2008. Collaborative filtering for implicit feedback datasets. In2008 Eighth IEEE international conference on data mining. Ieee, 263–272

    2008

  28. [28]

    Chengkai Huang, Junda Wu, Yu Xia, Zixu Yu, Ruhan Wang, Tong Yu, Ruiyi Zhang, Ryan A Rossi, Branislav Kveton, Dongruo Zhou, et al . 2025. Towards agentic recommender systems in the era of multimodal large language models.arXiv preprint arXiv:2503.16734(2025)

    Pith/arXiv arXiv 2025

  29. [29]

    Kalervo Järvelin and Jaana Kekäläinen. 2002. Cumulated gain-based evaluation of IR techniques.ACM Transactions on Information Systems (TOIS)20, 4 (2002), 422–446

    2002

  30. [30]

    Humans welcome to observe

    Yukun Jiang, Yage Zhang, Xinyue Shen, Michael Backes, and Yang Zhang. 2026. " Humans welcome to observe": A First Look at the Agent Social Network Moltbook. arXiv preprint arXiv:2602.10127(2026)

    arXiv 2026

  31. [31]

    Wang-Cheng Kang and Julian McAuley. 2018. Self-attentive sequential recom- mendation. In2018 IEEE international conference on data mining (ICDM). IEEE, 197–206

    2018

  32. [32]

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

    2009

  33. [33]

    Maciej Kula. 2015. Metadata embeddings for user and item cold-start recommen- dations.arXiv preprint arXiv:1507.08439(2015)

    Pith/arXiv arXiv 2015

  34. [34]

    Siyuan Li, Peng Shu, Churan Yu, Peilong Wang, Ruidong Zhang, Bowen Guo, Xinliang Li, Ruiyu Yan, Arif Hassan Zidan, Yi Pan, et al . [n. d.]. The Rise of Autonomous AI Agents: A Comprehensive Survey of OpenClaw—Architecture, Security, Ecosystem, and Beyond. ([n. d.])

  35. [35]

    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

  36. [36]

    Dawen Liang, Rahul G Krishnan, Matthew D Hoffman, and Tony Jebara. 2018. Variational autoencoders for collaborative filtering. InProceedings of the 2018 world wide web conference. 689–698

    2018

  37. [37]

    Mingfu Liang, Yufei Li, Jay Xu, Kavosh Asadi, Xi Liu, Shuo Gu, Kaushik Rangadu- rai, Frank Shyu, Shuaiwen Wang, Song Yang, et al. 2026. Generative Reasoning Re-ranker.arXiv preprint arXiv:2602.07774(2026)

    Pith/arXiv arXiv 2026

  38. [38]

    Greg Linden, Brent Smith, and Jeremy York. 2003. Amazon. com recommenda- tions: Item-to-item collaborative filtering.IEEE Internet computing7, 1 (2003), 76–80

    2003

  39. [39]

    Haifeng Liu, Zheng Hu, Ahmad Mian, Hui Tian, and Xuzhen Zhu. 2014. A new user similarity model to improve the accuracy of collaborative filtering. Knowledge-based systems56 (2014), 156–166

    2014

  40. [40]

    Flavio Lombardi, Maurantonio Caprolu, and Roberto Di Pietro. 2022. AI-enabled bot and social media: A survey of tools, techniques, and platforms for the arms race. InMixed methods perspectives on communication and social media research. Routledge, 255–269

    2022

  41. [41]

    Julian McAuley and Jure Leskovec. 2013. Hidden factors and hidden topics: understanding rating dimensions with review text. InProceedings of the 7th ACM conference on Recommender systems. 165–172

    2013

  42. [42]

    Julian McAuley, Christopher Targett, Qinfeng Shi, and Anton Van Den Hengel

  43. [43]

    InProceedings of the 38th international ACM SIGIR conference on research and development in information retrieval

    Image-based recommendations on styles and substitutes. InProceedings of the 38th international ACM SIGIR conference on research and development in information retrieval. 43–52

  44. [44]

    Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. 2023. Generative agents: Interactive simulacra of human behavior. InProceedings of the 36th annual acm symposium on user interface software and technology. 1–22. Conference’17, July 2017, Washington, DC, USA Daming Li, Simeng Han, and Jialu Zhang

    2023

  45. [45]

    Henry CW Price, H AlMuhanna, PM Bassani, M Ho, and TS Evans. 2026. Let there be claws: An early social network analysis of ai agents on moltbook.arXiv preprint arXiv:2602.20044(2026)

    arXiv 2026

  46. [46]

    Dimitrios Rafailidis and Alexandros Nanopoulos. 2015. Modeling users preference dynamics and side information in recommender systems.IEEE Transactions on Systems, Man, and Cybernetics: Systems46, 6 (2015), 782–792

    2015

  47. [47]

    Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. InProceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP). 3982–3992

    2019

  48. [48]

    Steffen Rendle. 2010. Factorization machines. In2010 IEEE International conference on data mining. IEEE, 995–1000

    2010

  49. [49]

    Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme

  50. [50]

    InProceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence(Montreal, Quebec, Canada)(UAI ’09)

    BPR: Bayesian personalized ranking from implicit feedback. InProceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence(Montreal, Quebec, Canada)(UAI ’09). AUAI Press, Arlington, Virginia, USA, 452–461

  51. [51]

    Francesco Ricci, Lior Rokach, and Bracha Shapira. 2010. Introduction to rec- ommender systems handbook. InRecommender systems handbook. Springer, 1–35

    2010

  52. [52]

    Badrul Sarwar, George Karypis, Joseph Konstan, and John Riedl. 2001. Item-based collaborative filtering recommendation algorithms. InProceedings of the 10th international conference on World Wide Web. 285–295

    2001

  53. [53]

    Andrew I Schein, Alexandrin Popescul, Lyle H Ungar, and David M Pennock

  54. [54]

    InProceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval

    Methods and metrics for cold-start recommendations. InProceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval. 253–260

  55. [55]

    Andrew Shin. 2026. AI-Gram: When Visual Agents Interact in a Social Network. arXiv preprint arXiv:2604.21446(2026)

    Pith/arXiv arXiv 2026

  56. [56]

    Noah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan, and Shunyu Yao. 2023. Reflexion: Language agents with verbal reinforcement learning. Advances in neural information processing systems36 (2023), 8634–8652

    2023

  57. [57]

    Xiaoyuan Su and Taghi M Khoshgoftaar. 2009. A survey of collaborative filtering techniques.Advances in artificial intelligence2009, 1 (2009), 421425

    2009

  58. [58]

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

  59. [59]

    InProceedings of the 28th ACM international conference on information and knowledge management

    BERT4Rec: Sequential recommendation with bidirectional encoder rep- resentations from transformer. InProceedings of the 28th ACM international conference on information and knowledge management. 1441–1450

  60. [60]

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

    2013

  61. [61]

    João Vinagre, Alípio Mário Jorge, and João Gama. 2015. An overview on the exploitation of time in collaborative filtering.Wiley interdisciplinary reviews: Data mining and knowledge discovery5, 5 (2015), 195–215

    2015

  62. [62]

    Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, et al. 2024. A survey on large language model based autonomous agents.Frontiers of Computer Science18, 6 (2024), 186345

    2024

  63. [63]

    Xiang Wang, Xiangnan He, Meng Wang, Fuli Feng, and Tat-Seng Chua. 2019. Neural graph collaborative filtering. InProceedings of the 42nd international ACM SIGIR conference on Research and development in Information Retrieval. 165–174

    2019

  64. [64]

    Nigel Williams and Nicole Ferdinand. 2026. Form or function? early dynamics of the moltbook ai social media network.ROBONOMICS: The Journal of the Automated Economy7 (2026), 90–90

    2026

  65. [65]

    Yunjia Xi, Weiwen Liu, Jianghao Lin, Xiaoling Cai, Hong Zhu, Jieming Zhu, Bo Chen, Ruiming Tang, Weinan Zhang, and Yong Yu. 2024. Towards open-world recommendation with knowledge augmentation from large language models. In Proceedings of the 18th ACM Conference on Recommender Systems. 12–22

    2024

  66. [66]

    Zhiheng Xi, Wenxiang Chen, Xin Guo, Wei He, Yiwen Ding, Boyang Hong, Ming Zhang, Junzhe Wang, Senjie Jin, Enyu Zhou, et al. 2025. The rise and potential of large language model based agents: A survey.Science China Information Sciences 68, 2 (2025), 121101

    2025

  67. [67]

    Xue Xia, Pong Eksombatchai, Nikil Pancha, Dhruvil Deven Badani, Po-Wei Wang, Neng Gu, Saurabh Vishwas Joshi, Nazanin Farahpour, Zhiyuan Zhang, and An- drew Zhai. 2023. Transact: Transformer-based realtime user action model for recommendation at pinterest. InProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 5249–5259

    2023

  68. [68]

    Kai-Cheng Yang, Onur Varol, Pik-Mai Hui, and Filippo Menczer. 2020. Scalable and generalizable social bot detection through data selection. InProceedings of the AAAI conference on artificial intelligence, Vol. 34. 1096–1103

    2020

  69. [69]

    Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2022. React: Synergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629(2022)

    Pith/arXiv arXiv 2022

  70. [70]

    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)

    Pith/arXiv arXiv 2024

  71. [71]

    Jialu Zhang, Jialiang Gu, Wangmeiyu Zhang, José Pablo Cambronero, John Kolesar, Ruzica Piskac, Daming Li, and Hanyuan Shi. 2025. A Systematic Study of Time Limit Exceeded Errors in Online Programming Assignments. arXiv:2510.14339 [cs.SE] https://arxiv.org/abs/2510.14339

    arXiv 2025

  72. [72]

    Shuai Zhang, Lina Yao, Aixin Sun, and Yi Tay. 2019. Deep learning based recom- mender system: A survey and new perspectives.ACM computing surveys (CSUR) 52, 1 (2019), 1–38

    2019

  73. [73]

    Yunbei Zhang, Kai Mei, Ming Liu, Janet Wang, Dimitris N Metaxas, Xiao Wang, Jihun Hamm, and Yingqiang Ge. 2026. Agents in the wild: Safety, society, and the illusion of sociality on moltbook.arXiv preprint arXiv:2602.13284(2026)

    arXiv 2026

  74. [74]

    Zihuai Zhao, Wenqi Fan, Jiatong Li, Yunqing Liu, Xiaowei Mei, Yiqi Wang, Zhen Wen, Fei Wang, Xiangyu Zhao, Jiliang Tang, et al. 2024. Recommender systems in the era of large language models (llms).IEEE Transactions on Knowledge and Data Engineering36, 11 (2024), 6889–6907

    2024

  75. [75]

    Chenyu Zhou, Huacan Chai, Wenteng Chen, Zihan Guo, Rong Shan, Yuanyi Song, Tianyi Xu, Yingxuan Yang, Aofan Yu, Weiming Zhang, et al . 2026. Ex- ternalization in llm agents: A unified review of memory, skills, protocols and harness engineering.arXiv preprint arXiv:2604.08224(2026)

    Pith/arXiv arXiv 2026

  76. [76]

    Guorui Zhou, Xiaoqiang Zhu, Chenru Song, Ying Fan, Han 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. 1059–1068

    2018

  77. [77]

    Andrew Zimdars, David Maxwell Chickering, and Christopher Meek. 2013. Using temporal data for making recommendations.arXiv preprint arXiv:1301.2320 (2013)

    Pith/arXiv arXiv 2013