Pith. sign in

REVIEW 4 major objections 6 minor 55 references

Explainable CTR Prediction via LLM Reasoning

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read ExpCTR jointly trains LLM explanations with the CTR model and reports higher accuracy and interpretability on three real-world datasets.

desk verdict ExpCTR is a plausible framework for joint CTR and explanation generation, but the paper's dual accuracy-plus-interpretability claim is only half-supported: the accuracy gains need ablations and variance, and interpretability is never actually measured. read the letter →

arxiv 2412.02588 v1 pith:725I7X46 submitted 2024-12-03 cs.IR cs.AI

classification cs.IRcs.AI
keywords LargeLanguageModelsExplainableRecommendationCTRPredictionReinforcementLearningProximalPolicyOptimizationLoRAData-freeTraining
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

The paper argues that the standard post-hoc pipeline for explainable recommendation—train a black-box CTR model, then fit a separate explanation generator—can be replaced by a joint framework in which the explanations themselves are trained to serve the CTR task. ExpCTR uses reinforcement learning (PPO) with two rewards: one that rewards an explanation when a frozen LLM can predict the user's actual interaction from it, and one that rewards an explanation when adding it as a textual feature improves an ID-based CTR model's prediction. The authors claim that this joint training removes the need for curated explanation datasets while improving both classification accuracy (AUC) and explanation quality on BookCrossing, MovieLens-20M, and Amazon Books. If the claim holds, explanation generation stops being an afterthought and becomes a trainable component of the recommender itself, with direct implications for trustworthy recommendation.

What carries the argument

The load-bearing object is the two-part reward in the PPO loop, computed by two predictors that stay fixed while the generator is updated. The LC head (Eq. 9) scores an explanation by how much it moves a frozen LLM's binary Yes/No prediction toward the true label; the IC head (Eq. 13) scores it by how much adding its text embedding to an ID-based CTR model improves prediction relative to a zero-embedding counterfactual. These normalized rewards (Eq. 10) feed a three-stage iterative training process—LC alignment, CTR training with textual features, and IC alignment—that closes the feedback loop between explanation and recommender.

What would settle it

Replace each generated explanation in the test set with a random sentence sampled from the same review corpus and recompute ExpCTR-Aug's AUC; if the AUC drops by less than about one point, the explanation features are not carrying the signal and the reward was optimizing for correlation rather than explanation content.

Watch

Extended reading notes

Core claim

ExpCTR's central claim is that a single LLM, fine-tuned with LoRA under a PPO objective, can generate explanations that are simultaneously faithful to user intent and useful to the recommender itself. The LC alignment reward (Eq. 9) checks whether the explanation, fed to a frozen LLM as the user's thought, makes that LLM's like/dislike prediction match the ground-truth label; the IC alignment reward (Eq. 13) checks whether appending the explanation's mean-pooled embedding to an ID-based CTR model moves its prediction toward the label, compared with zeroing the embedding. The paper reports that ExpCTR-Aug, the CTR model trained with the generated explanations as features, outperforms the DeepFM baseline by 18.2%, 11.9%, and 17.8% in AUC on BookCrossing, ML-20M, and Amazon Books, and that the case-study explanations align with user reviews more closely than an in-context-learning baseline. This constitutes, the authors claim, the first data-free joint training of explanations and CTR prediction.

Load-bearing premise

The training assumes that 'predictive usefulness' is a faithful stand-in for 'explanation quality': the LC reward relies on the frozen LLM's Yes/No verdict reflecting the user's real reasoning, and the IC reward relies on the ID-based CTR model's score change measuring truthfulness rather than exploitability.

Editorial extensions

If this is right

  • CTR models can treat LLM-generated free text as a first-class prediction feature, with no curated explanation dataset.
  • LoRA keeps the tuning cheap: one base LLM serves as the generator, the LC reward model, and the frozen encoder for explanations.
  • The three-stage loop makes explanation quality and CTR quality co-dependent, so improving one can improve the other in later rounds.
  • The method's prompt template is domain-transferable to any setting where items have textual titles and explicit like/dislike signals, not only books.

Reading between the lines

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

  • A testable boundary of the claim: on datasets where review text is only weakly correlated with interactions, the IC reward should plateau and the reported AUC gains should shrink, which would delimit when joint training helps.
  • The LC reward makes the frozen LLM's own judgment the arbiter of 'user intent', so the method inherits that base model's biases; subgroups on which the base LLM is miscalibrated could receive confidently wrong explanations.
  • Nothing in the two rewards prevents the generator from producing verbose, template-like text that is predictive but not human-understandable; a readability or human-preference term would be a natural extension.
  • The paper does not isolate how much of the gain comes from the RL loop versus the LLM prior; comparing against a small review-trained generator with a similar parameter budget would separate the two.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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. The paper proposes ExpCTR, a framework that integrates LLM-generated textual explanations into CTR prediction. A generator LLM is fine-tuned with LoRA and PPO using two rewards: LC alignment (Eq. 9), which rewards explanations that make a frozen LLM's binary like/dislike prediction agree with the ground-truth label, and IC alignment (Eq. 13), which rewards explanations that change the prediction of an ID-based CTR model (DeepFM) toward the label. The training proceeds in three stages: LC-alignment fine-tuning of the generator, training the CTR model with generated explanations as extra text features, and IC-alignment fine-tuning of the generator; the last two stages are repeated iteratively. The paper reports AUC/LogLoss/MAE/RMSE on BookCrossing, ML-20M, and Amazon Books, plus two qualitative case studies, and claims that ExpCTR significantly improves both recommendation accuracy and interpretability without manually constructed explanation datasets.

Significance. If the claims were fully supported, ExpCTR would be a notable step toward joint training of explanation generation and CTR prediction, and the proposed two-reward RL scheme is interesting. The paper is also to be credited for making the method concrete, using public datasets, and reporting multiple metrics. However, the central evidence is currently undermined by a circular reward loop in the IC alignment stage and by the absence of any quantitative validation of interpretability. The reported accuracy gains are therefore not yet established as gains in explanation quality, and the interpretability claim rests entirely on two case studies. With additional ablations and an independent evaluation of explanation faithfulness, the method could be made convincing, but as it stands the paper's dual claim is unverified.

major comments (4)
  1. [§4.3.2, Eq. (13), and Table 1] The IC alignment reward is circular. In Stage 2, the CTR model f is trained on explanations produced by the generator πRL_φ; in Stage 3, R_IC = 1 − |y − s^r| + |s^r − s̃^r| is computed with that same f and used to update the very generator that produced the training explanations. Because f is trained to exploit any textual pattern that predicts the label, the second term |s^r − s̃^r| rewards any text feature that changes f's prediction, including spurious n-gram or embedding correlations. The ExpCTR-Aug gains over DeepFM reported in Table 1 (18.2%, 11.9%, 17.8% AUC) are thus computed with the co-adapted f and conflate explanation value with mutual adaptation. To support the accuracy claim, the authors should compare ExpCTR-Aug against DeepFM augmented with non-reward text features (e.g., random item descriptions, template sentences, or frozen ICL explanations) using the same concatenation and training recipe, and should also report the Stage-2 CTR model's performance before any Stage-3 RL update.
  2. [§4.2.2, Eq. (9), and §5.2, ExpCTR-LLM] The reported ExpCTR-LLM AUC is exactly the AUC of s^u_{u,i} in Eq. (8), which is the score whose label agreement R_LC in Eq. (9) maximizes. The PPO objective therefore directly optimizes the evaluation metric, so the improvement over ICL (e.g., 9.1% on ML-20M) is partly by construction. This does not by itself invalidate the method, but it means the result cannot be read as evidence that the explanations reflect user intentions. An independent evaluation is needed, such as a human rating of explanation helpfulness, a faithfulness metric computed against the user's actual review text, or a control experiment in which the same R_LC reward is applied to a non-explanatory text input to see whether the AUC gain is attributable to the explanation content.
  3. [§5.4, Table 2] The interpretability claim is not quantitatively validated. The paper provides only two qualitative case studies in Table 2; there is no human evaluation, no agreement metric with user reviews, no explanation ranking metric, and no faithfulness/grounding measure. The abstract states that ExpCTR 'significantly enhances both recommendation accuracy and interpretability,' but interpretability is asserted rather than measured. The authors should add a human study or an automatic fidelity evaluation (e.g., entailment/overlap with the user review, or feature attribution agreement) to support the interpretability part of the central claim.
  4. [§2 and §5.1.2] The paper cites Yang et al. (2024) [40], an RL-based explainable recommendation method, as related work but does not include it in the compared methods of Section 5.1.2 or in Table 1. Since ExpCTR is also an RL-based explainable method and the paper claims state-of-the-art performance, omitting [40] leaves the comparison incomplete. Please add this baseline or justify its exclusion.
minor comments (6)
  1. [References] References [44] and [45] are the same survey (Yongfeng Zhang et al., 'Explainable recommendation: A survey and new perspectives'); one should be removed or the citations should be merged.
  2. [§5.1.2 vs. Table 1] The text lists DSSM among the ID-based methods, but DSSM does not appear in Table 1; conversely, xDeepFM appears in Table 1 but is not described in the compared-methods list. Please align the list of baselines with the table.
  3. [Eq. (8)] The notation in Eq. (8) is typeset incorrectly: the subscript appears as "𝑠𝑢𝑢,𝑖" and should be something like s^u_{u,i}; also, the superscripts for s^r and s̃^r in Eq. (13) should be defined clearly before first use.
  4. [§4.1 vs. Table 2] The prompt template in Section 4.1 instructs 'The answer should be within one sentence,' but the explanations shown in Table 2 are multiple sentences long. Please either update the template or the examples so they match.
  5. [Abstract and §1] The term 'data-free' is misleading: the method still requires user-item interaction data and item titles; it avoids only manually constructed explanation datasets. Rewording to 'explanation-data-free' would be more precise.
  6. [Table 1] No variance or statistical significance tests are reported. Given that several baseline AUC differences are small (e.g., 0.02–0.05), the claim of 'significantly enhances' should be supported by results over multiple seeds or significance tests.

Circularity Check

2 steps flagged · score 6.0 of 10

ExpCTR's interpretability claims reduce to its own training rewards: the IC loop co-trains the CTR model with the generator, and the LC reward optimizes the exact LLM scorer reported as ExpCTR-LLM.

  1. fitted input called prediction [Section 4.2.2, Eq. (9); Section 4.3.2, Stage 1; Section 5.1.2, ExpCTR-LLM description]
    "ExpCTR-LLM reflects the effectiveness of the generated explanations in capturing and articulating user preferences and rationales for future interactions, which results in better outcomes under an LLM scorer."

    ExpCTR-LLM is the LLM scorer s from Eq. (8), and the LC reward in Eq. (9) is R_LC = 1 - |y - s|, i.e., the label agreement of that scorer. Stage 1 (Eq. (14)) optimizes the explanation generator against this reward. The reported AUC/LogLoss of ExpCTR-LLM is therefore computed from the very scorer whose agreement is the training signal. The paper presents this improvement as evidence that explanations 'reflect user preferences,' but no independent explanation-quality or faithfulness metric is used; the claimed effectiveness is the optimized proxy itself.

  2. self definitional [Section 4.2.3, Eq. (13); Section 4.3.2, Stages 2-3; Section 5.1.2, ExpCTR-Aug description]
    "To evaluate the impact of semantic representations of LLM's explanations, we compare the performance of the CTR model with and without these explanations and quantify the differences in CTR predictions. A notable performance improvement when explanations are incorporated indicates that the introduced semantic features contribute positively. This implies a better-aligned explanation, justifying a higher reward. This evaluation is formalized as follows: R_IC(Z_ui) = 1 - |y_ui - s^r_ui| + |s^r_ui - s_tilde^r_ui|"

    The score s^r in Eq. (13) is produced by the CTR model f from Eq. (12). Stage 2 (Eq. (15)) trains that f on the explanations Z generated by the very policy pi^RL_phi that Stage 3 (Eq. (16)) then optimizes against R_IC. R_IC is thus the objective of a closed co-adaptation loop: f learns to exploit whatever text the generator emits, and the generator is rewarded for making f's prediction match the label and for changing f's output relative to a zero text vector. The reported ExpCTR-Aug AUC is f's AUC, exactly the quantity the Stage-2 loss and the |y - s^r| reward term optimize. Interpreting this AUC gain as 'explanation quality' or 'interpretability' is by construction rather than independent validation; no human evaluation or faithfulness metric is provided.

full rationale

The paper's accuracy numbers are held-out evaluations and are not themselves circular in the narrow statistical sense. However, the central dual claim of 'accuracy and interpretability' is partially self-referential. The LC alignment reward optimizes the exact LLM scorer whose test AUC is reported as ExpCTR-LLM, so that result is the optimized proxy rather than an independent measure of explanation quality. The IC alignment loop is more clearly circular: the CTR model f used in the reward is trained on the generator's own outputs, and then the generator is trained to maximize f's prediction agreement and f's sensitivity to the explanation text. Reporting f's AUC as evidence that explanations are faithful or interpretable therefore reduces to a co-adaptation objective. There is no external benchmark for explanation quality, no human study, and no faithfulness metric; the only evidence is the reward-based evaluation whose definition matches the training signal. The self-citations to [41,42] for the prompt template are not load-bearing, and no uniqueness theorem or imported-ansatz circularity appears. The accuracy component still has independent content on held-out data, which is why the score is 6 rather than higher.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper's central claims rest primarily on the validity of the two reward signals as proxies for explanation quality, and on the iterative co-adaptation between generator and CTR model converging to a useful solution. The free parameters are the threshold separating liked/disliked items, the KL penalty, the reward normalization bound, and the maximum history length. No new physical or abstract entities are introduced; the 'LC alignment' and 'IC alignment' are reward components, not separate entities.

free parameters (4)
  • Liked/disliked threshold = 5 (BookCrossing), 4 (ML-20M, Amazon Books)
    A threshold-based function classifies historical items into liked and disliked; this choice shapes the user profile and the resulting explanations, and is adjusted per dataset (Sections 4.1, 5.1.1).
  • KL penalty beta = 0.05
    Tuned via sensitivity analysis (Figure 4a) over {0.01,0.05,0.1,0.5}; affects the PPO objective (Eq. 7).
  • Reward normalization bound delta = 1.0
    Tuned via sensitivity analysis (Figure 4b) over {0.5,1.0,2.0,5.0}; clips the normalized rewards (Eq. 10).
  • Maximum history length = 10
    The prompt uses up to 10 chronologically preceding items; affects the user profile and explanation input (Section 5.1.1).
assumptions (5)
  • domain assumption The frozen LLM used in LC alignment is a valid proxy for user preferences: a higher 'Yes' probability when reading the generated explanation should correlate with the user actually liking the item.
    Stated in Section 4.2.2: 'A closer alignment between the CTR prediction and the ground-truth label indicates a more precise explanation.' This equates predictive accuracy on the binary label with explanation fidelity, which is an unproven assumption.
  • domain assumption The IC alignment reward treats the change in the ID-based CTR model's prediction caused by adding the explanation as a valid measure of explanation quality.
    Equations 11-13 in Section 4.2.3; this assumes the model's sensitivity to the text feature is a faithfulness measure, while the model itself is trained on those same explanations in Stage 2.
  • domain assumption The prompt template taken from prior work [41,42] adequately elicits reasoning and transfers across domains.
    Section 4.1 reuses a book recommendation prompt; the paper asserts it 'can be easily adapted to different recommendation scenarios with minor adjustments' without validation across domains.
  • domain assumption The probabilistic decomposition P(Z, y|D) = P(Z|D) P(y|Z,D) in Eq. 5 is a valid generative model of explanations and clicks.
    Section 3.2 introduces this factorization as the theoretical basis of the unified framework; it assumes the explanation is generated independently of the label given the data, which is a strong modeling assumption.
  • standard math PPO and LoRA training procedures follow standard implementations and converge as expected.
    Equation 7 uses the PPO objective from [30]; the paper relies on standard RLHF practice without proof of convergence or stability.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Explainable CTR Prediction via LLM Reasoning." pith.science (2026). https://pith.science/paper/725I7X46

@misc{pith2026241202588,
  author       = {Pith},
  title        = {Pith review of: Explainable CTR Prediction via LLM Reasoning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/725I7X46}},
  note         = {Machine review of arXiv:2412.02588}
}
read the original abstract

Recommendation Systems have become integral to modern user experiences, but lack transparency in their decision-making processes. Existing explainable recommendation methods are hindered by reliance on a post-hoc paradigm, wherein explanation generators are trained independently of the underlying recommender models. This paradigm necessitates substantial human effort in data construction and raises concerns about explanation reliability. In this paper, we present ExpCTR, a novel framework that integrates large language model based explanation generation directly into the CTR prediction process. Inspired by recent advances in reinforcement learning, we employ two carefully designed reward mechanisms, LC alignment, which ensures explanations reflect user intentions, and IC alignment, which maintains consistency with traditional ID-based CTR models. Our approach incorporates an efficient training paradigm with LoRA and a three-stage iterative process. ExpCTR circumvents the need for extensive explanation datasets while fostering synergy between CTR prediction and explanation generation. Experimental results demonstrate that ExpCTR significantly enhances both recommendation accuracy and interpretability across three real-world datasets.

Figures

Figures reproduced from arXiv: 2412.02588 by the authors.

Figure 1
Figure 1. Comparison of current post-hoc paradigm methods [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall architecture of ExpCTR. are represented by their titles, while users are characterized by the titles of items they have interacted with. To further refine user profiles, we categorize these historical items into liked and disliked categories, using interaction signals such as ratings as indicators. A threshold-based function is employed to classify each item in a user’s interaction sequence. Items rated … view at source ↗
Figure 3
Figure 3. Trends of LC alignment and IC alignment re [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Hyperparameter sensitivity of ExpCTR in KL [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 15 canonical work pages

  1. [40]

    Jiaxi Tang, Ke Wang, Liqiang Zhang, Shuai Li, Jiajie Yan, and Zheng Zhang. 2016. Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery & Data Mining . 427–436

  2. [1]

    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

  3. [2]

    Zhongxia Chen, Xiting Wang, Xing Xie, Tong Wu, Guoqing Bu, Yining Wang, and Enhong Chen. 2019. Co-attentive multi-task learning for explainable recom- mendation.. In IJCAI, Vol. 2019. 2137–2143

  4. [3]

    Sunhao Dai, Ninglu Shao, Haiyuan Zhao, Weijie Yu, Zihua Si, Chen Xu, Zhongx- iang Sun, Xiao Zhang, and Jun Xu. 2023. Uncovering chatgpt’s capabilities in recommender systems. In Proceedings of the 17th ACM Conference on Recom- mender Systems. 1126–1132

  5. [4]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding.arXiv preprint arXiv:1810.04805 (2018)

  6. [5]

    Maurizio Ferrari Dacrema, Paolo Cremonesi, and Dietmar Jannach. 2019. Are we really making much progress? A worrying analysis of recent neural recommen- dation approaches. In Proceedings of the 13th ACM conference on recommender systems. 101–109

  7. [6]

    Jingyue Gao, Xiting Wang, Yasha Wang, and Xing Xie. 2019. Explainable rec- ommendation through attentive multi-view learning. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 33. 3622–3629

  8. [7]

    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)

Show all 55 references
  1. [8]

    Shuyu Guo, Shuo Zhang, Weiwei Sun, Pengjie Ren, Zhumin Chen, and Zhaochun Ren. 2023. Towards explainable conversational recommender systems. InProceed- ings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2786–2795

  2. [9]

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

  3. [10]

    Balázs Hidasi, Alexandros Karatzoglou, Linas Baltrunas, and Domonkos Tikk

  4. [11]

    Neil Houlsby, Andrei Giurgiu, Stanislaw Jastrzebski, Bruna Morrone, Quentin De Laroussilhe, Andrea Gesmundo, Mona Attariyan, and Sylvain Gelly. 2019. Parameter-efficient transfer learning for NLP. In International conference on machine learning. PMLR, 2790–2799

  5. [12]

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685 (2021)

  6. [13]

    Po-Sen Huang, Xiaodong He, Jianfeng Gao, Li Deng, Alex Acero, and Larry Heck. 2013. Learning deep structured semantic models for web search using clickthrough data. In Proceedings of the 22nd ACM international conference on Information & Knowledge Management . 2333–2338

  7. [14]

    Dietmar Jannach, Markus Zanker, Alexander Felfernig, and Gerhard Friedrich

  8. [15]

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

  9. [16]

    Yuxuan Lei, Jianxun Lian, Jing Yao, Xu Huang, Defu Lian, and Xing Xie. 2023. Recexplainer: Aligning large language models for recommendation model inter- pretability. arXiv preprint arXiv:2311.10947 (2023)

  10. [17]

    Lei Li, Yongfeng Zhang, and Li Chen. 2021. Extra: Explanation ranking datasets for explainable recommendation. In Proceedings of the 44th International ACM SIGIR conference on Research and Development in Information Retrieval . 2463– 2469

  11. [18]

    Lei Li, Yongfeng Zhang, and Li Chen. 2021. Personalized transformer for explain- able recommendation. arXiv preprint arXiv:2105.11601 (2021)

  12. [19]

    Piji Li, Zihao Wang, Zhaochun Ren, Lidong Bing, and Wai Lam. 2017. Neural rating regression with abstractive tips generation for recommendation. InProceed- ings of the 40th International ACM SIGIR conference on Research and Development in Information Retrieval. 345–354

  13. [20]

    Xiang Lisa Li and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. arXiv preprint arXiv:2101.00190 (2021)

  14. [21]

    Zekun Li, Zeyu Cui, Shu Wu, Xiaoyu Zhang, and Liang Wang. 2019. Fi-gnn: Modeling feature interactions via graph neural networks for ctr prediction. In Proceedings of the 28th ACM international conference on information and knowledge management. 539–548

  15. [22]

    Jianxun Lian, Xiaohuan Zhou, Fuzheng Zhang, Zhongxia Chen, Xing Xie, and Guangzhong Sun. 2018. xdeepfm: Combining explicit and implicit feature in- teractions for recommender systems. In Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data ...

  16. [23]

    Hanmeng Liu, Ruoxi Ning, Zhiyang Teng, Jian Liu, Qiji Zhou, and Yue Zhang

  17. [24]

    Junling Liu, Chao Liu, Peilin Zhou, Renjie Lv, Kang Zhou, and Yan Zhang

  18. [25]

    Andriy Mnih and Russ R Salakhutdinov. 2007. Probabilistic matrix factorization. Advances in neural information processing systems 20 (2007)

  19. [26]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems 35...

  20. [27]

    Baolin Peng, Chunyuan Li, Pengcheng He, Michel Galley, and Jianfeng Gao. 2023. Instruction tuning with gpt-4. arXiv preprint arXiv:2304.03277 (2023)

  21. [28]

    arXiv preprint arXiv:2304.10149 (2023)

    Is chatgpt a good recommender? a preliminary study. arXiv preprint arXiv:2304.10149 (2023)

  22. [29]

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

  23. [30]

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov

  24. [31]

    Weiping Song, Chence Shi, Zhiping Xiao, Zhijian Duan, Yewen Xu, Ming Zhang, and Jian Tang. 2019. Autoint: Automatic feature interaction learning via self- attentive neural networks. InProceedings of the 28th ACM international conference on information and knowledge management ...

  25. [32]

    Yanru Qu, Han Cai, Kan Ren, Weinan Zhang, Yong Yu, Ying Wen, and Jun Wang

  26. [33]

    In 2016 IEEE 16th international conference on data mining (ICDM)

    Product-based neural networks for user response prediction. In 2016 IEEE 16th international conference on data mining (ICDM) . IEEE, 1149–1154

  27. [34]

    Nava Tintarev. 2007. Explanations of recommendations. In Proceedings of the 2007 ACM conference on Recommender systems . 203–206

  28. [35]

    Nan Wang, Hongning Wang, Yiling Jia, and Yue Yin. 2018. Explainable recommen- dation via multi-task learning in opinionated text data. In The 41st international ACM SIGIR conference on research & development in information retrieval. 165–174

  29. [36]

    Ruoxi Wang, Bin Fu, Gang Fu, and Mingliang Wang. 2017. Deep & cross network for ad click predictions. In Proceedings of the ADKDD’17 . 1–7

  30. [37]

    Ruoxi Wang, Rakesh Shivanna, Derek Cheng, Sagar Jain, Dong Lin, Lichan Hong, and Ed Chi. 2021. Dcn v2: Improved deep & cross network and practical lessons for web-scale learning to rank systems. In Proceedings of the web conference 2021 . 1785–1797

  31. [38]

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

  32. [39]

    Shitao Xiao, Zheng Liu, Peitian Zhang, and Niklas Muennighoff. 2023. C-Pack: Packaged Resources To Advance General Chinese Embedding. arXiv:2309.07597 [cs.CL]

  33. [41]

    Xiaohan Yu, Li Zhang, Xin Zhao, and Yue Wang. 2024. Break the ID-Language Barrier: An Adaption Framework for Sequential Recommendation. arXiv preprint arXiv:2411.18262 (2024)

  34. [42]

    Xiaohan Yu, Li Zhang, Xin Zhao, Yue Wang, and Zhongrui Ma. 2024. RA-Rec: An Efficient ID Representation Alignment Framework for LLM-based Recom- mendation. arXiv preprint arXiv:2402.04527 (2024)

  35. [43]

    Jingsen Zhang, Xu Chen, Jiakai Tang, Weiqi Shao, Quanyu Dai, Zhenhua Dong, and Rui Zhang. 2023. Recommendation with causality enhanced natural language explanations. In Proceedings of the ACM Web Conference 2023 . 876–886

  36. [45]

    Sean Wu, Michael Koo, Lesley Blum, Andy Black, Liyo Kao, Fabien Scalzo, and Ira Kurtz. 2023. A comparative study of open-source large language models, gpt-4 and claude 2: Multiple-choice test taking in nephrology. arXiv preprint arXiv:2308.04709 (2023)

  37. [46]

    Wayne Xin Zhao, Shanlei Mu, Yupeng Hou, Zihan Lin, Yushuo Chen, Xingyu Pan, Kaiyuan Li, Yujie Lu, Hui Wang, Changxin Tian, Yingqian Min, Zhichao Feng, Xinyan Fan, Xu Chen, Pengfei Wang, Wendi Ji, Yaliang Li, Xiaoling Wang, and Ji-Rong Wen. 2021. RecBole: Towards a Unified, Com...

  38. [47]

    Mengyuan Yang, Mengying Zhu, Yan Wang, Linxun Chen, Yilei Zhao, Xiuyuan Wang, Bing Han, Xiaolin Zheng, and Jianwei Yin. 2024. Fine-Tuning Large Language Model Based Explainable Recommendation with Explainable Quality WSDM, March 10–14, 2025, Hannover, Germany Xiaohan yu, Li Zh...

  39. [48]

    Guorui Zhou, Na Mou, Ying Fan, Qi Pi, Weijie Bian, Chang Zhou, Xiaoqiang Zhu, and Kun Gai. 2019. Deep interest evolution network for click-through rate prediction. In Proceedings of the AAAI conference on artificial intelligence , Vol. 33. 5941–5948

  40. [49]

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

  41. [52]

    Yongfeng Zhang, Xu Chen, et al. 2020. Explainable recommendation: A survey and new perspectives. Foundations and Trends ® in Information Retrieval 14, 1 (2020), 1–101

  42. [54]

    Rui Zheng, Shihan Dou, Songyang Gao, Yuan Hua, Wei Shen, Binghai Wang, Yan Liu, Senjie Jin, Qin Liu, Yuhao Zhou, et al. 2023. Secrets of rlhf in large language models part i: Ppo. arXiv preprint arXiv:2307.04964 (2023)

  43. [2010]

    Cambridge University Press

    Recommender systems: an introduction . Cambridge University Press

  44. [2016]

    In Proceedings of the 4th International Conference on Learning Representations

    Session-based Recommendations with Recurrent Neural Networks. In Proceedings of the 4th International Conference on Learning Representations

  45. [2017]

    arXiv preprint arXiv:1707.06347 (2017)

    Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347 (2017)

  46. [2019]

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

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

  47. [2023]

    arXiv preprint arXiv:2304.03439 (2023)

    Evaluating the logical reasoning ability of chatgpt and gpt-4. arXiv preprint arXiv:2304.03439 (2023)

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.