Pith. sign in

REVIEW 5 major objections 8 minor 1 cited by

A Collaborative Ensemble Framework for CTR Prediction

T0 review · 5 major / 8 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Entropy-weighted ensemble of two CTR models beats either model alone with smaller embeddings.

desk verdict A useful ensemble recipe with a load-bearing fusion component that its own deployment experiment undermines. read the letter →

arxiv 2411.13700 v1 pith:NPEW2CKJ submitted 2024-11-20 cs.IR cs.LG

Xiaolong Liu , Zhichen Zeng , Xiaoyi Liu , Siyang Yuan , Weinan Song , Mengyue Hang , Yiqun Liu , Chaofei Yang
show 8 more authors
Donghyun Kim Wen-Yen Chen Jiyan Yang Yiping Han Rong Jin Bo Long Hanghang Tong Philip S. Yu
This is my paper · ORCID
classification cs.IRcs.LG
keywords RecommendersystemsCTRpredictioncollaborativelearningknowledgedistillationensemblemulti-embeddingconfidence-basedfusion
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

CETNet is a training framework that ensembles two click-through-rate models—InterFormer, which learns sequential and heterogeneous feature interactions, and DHEN, which learns hierarchical high-order interactions—each with its own independent embedding table. The two models are trained jointly with a symmetric KL-divergence term that aligns their predictions, and their embeddings are combined by a confidence-based fusion that weights each model by the negation entropy of its predicted click probability. The paper claims this collaborative ensemble outperforms either model alone and state-of-the-art baselines on three public datasets, and that on two additional datasets it matches or beats the multi-embedding paradigm while using substantially smaller embedding tables. The wider claim is that model diversity plus aligned joint training is a more efficient path to accuracy than simply scaling one model's embedding size.

What carries the argument

The mechanism is a confidence-based fusion built on the negation entropy of each model's scalar click probability: $C = -H(\hat y)$ with $H(\hat y)=-\hat y\log\hat y-(1-\hat y)\log(1-\hat y)$, detached from the computation graph. Softmax over the two confidence scores gives weights $w_{IF}$ and $w_{DHEN}$, and the final embedding is the weighted concatenation $[w_{IF}\cdot e_{IF}\parallel w_{DHEN}\cdot e_{DHEN}]$, fed to a readout layer with a sigmoid for the fused prediction. Around this sit three binary cross-entropy losses (one per model plus one for the fused output) and a symmetric KL-divergence term $\mathcal{L}_{kl}=\frac12 KL(\hat y_{IF}\|\hat y_{DHEN})+\frac12 KL(\hat y_{DHEN}\|\hat y_{IF})$ that aligns the two models' predictions during joint training. The separate embedding tables provide the diversity; the KL term provides the collaboration; the entropy weighting provides per-instance dynamic balancing.

What would settle it

Measure, across the three public datasets, the per-instance correlation between each component model's negation entropy and whether that model's prediction is closer to the ground-truth label than the other model's; if the correlation is zero or negative, the confidence-based fusion cannot be carrying the reported gains and concatenation should perform equally, which the ablation contradicts.

Watch

Extended reading notes

Core claim

The paper shows that complementary feature-interaction models can be made to cooperate better than either can perform alone, and that this cooperation is what unlocks gains without expanding embedding capacity. Specifically, it shows that giving InterFormer and DHEN separate embedding tables and training them with symmetric KL divergence prevents one model from dominating, while softmax weighting over the negation entropy of each model's click probability lets the more confident model's embedding carry more weight in the fused prediction. In experiments across AmazonElectronics, TaobaoAds, and KuaiVideo, the InterFormer+DHEN ensemble achieves the highest gAUC, AUC, and lowest LogLoss among all listed methods, and ablation removes each component (multi-embedding, KL alignment, confidence fusion, gradient stopping) degrades performance. On Criteo and Avazu, the framework with simple sum or concatenation fusion beats single-embedding and multi-embedding baselines at 2x or 3x embedding size where the baselines need 10x, showing the efficiency claim. In the industrial deployment, concatenation proved more stable than the confidence weighting, yet still gave a normalized-entropy gain of -0.062% over the baseline.

Load-bearing premise

The load-bearing premise is that a model's certainty, measured as the negation entropy of its predicted click probability, reliably indicates when that model's embedding deserves more weight in the final prediction.

Editorial extensions

If this is right

  • Ensembling architecturally different models with separate embedding tables can improve CTR accuracy even when total embedding parameters are fewer than a single-model baseline.
  • The symmetric KL term prevents one model from dominating, and the in-ensemble versions of both InterFormer and DHEN outperform their independently trained versions.
  • Removing the confidence-based fusion, the KL alignment, the multi-embedding, or the gradient stop each degrades performance in the ablations.
  • With only one epoch of training, the framework still beats single- and multi-embedding baselines on Criteo at 2x or 3x embedding size, which matches industry training budgets.
  • On the large-scale industrial dataset, the concatenation variant of the fusion gives a normalized-entropy gain, showing the multi-embedding ensemble transfers to production-scale data.

Reading between the lines

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

  • Since the large-scale industrial deployment found confidence-based fusion unstable and simple concatenation superior, the entropy weighting may be dataset- or calibration-dependent; a direct test is to compare full fusion against concatenation on the three public datasets while measuring per-instance entropy-correctness correlation.
  • The paper's own results suggest the benefit depends on architectural distance between the paired models: InterFormer+DHEN outperforms InterFormer+InterFormer, so a systematic study of how much diversity (for example, interaction-order coverage or sequence-vs-hierarchy) drives the gain would sharpen the design rule.
  • A stronger confidence signal might come from the models' disagreement or from calibrated probabilities rather than the entropy of the raw sigmoid output, since uncalibrated probabilities conflate certainty with confidence.
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

5 major / 8 minor

Summary. This paper proposes CETNet, an ensemble framework for CTR prediction that trains two models (InterFormer and DHEN) with separate embedding tables, aligns their predictions with symmetric KL divergence, and fuses their output embeddings using softmax weights derived from negation entropy. The authors evaluate CETNet on AmazonElectronics, TaobaoAds, KuaiVideo, Criteo, and Avazu, report ablations, and describe a large-scale internal deployment at Meta. The paper's central claim is that CETNet outperforms individual models and state-of-the-art baselines, and that confidence-based fusion is a key component that lets the framework achieve this with smaller embedding sizes than multi-embedding baselines.

Significance. The question addressed, how to combine diverse CTR models without duplicating full embedding costs, is practically important, and the paper's comparisons against the multi-embedding paradigm on Criteo/Avazu are a useful addition to the literature. The broad public-benchmark evaluation, multiple ensemble combinations, and one-epoch experiments are strengths. However, the signature confidence-based fusion component is explicitly reported as unsuccessful in the internal deployment, and the public results are all single-seed runs with no variance estimates. These issues need to be resolved before the stated contributions can be considered established.

major comments (5)
  1. [4.5 / Figure 3] Section 4.5 reports that confidence-based fusion 'does not result in significant improvements or stable learning' in the large-scale internal setting, whereas simple concatenation achieves a larger NE gain (-0.062% vs. -0.051%). Since confidence-based fusion is a central claimed contribution of the paper (abstract and Section 1) and Table 3's ablation says removing it degrades public-benchmark results, the paper needs to address this contradiction. At minimum, provide a multi-seed analysis on the public datasets to show whether the fusion component has a stable positive effect, and explicitly discuss in which settings it does and does not help.
  2. [4.2 / Table 1] The text states that the InterFormer+DHEN combination 'achieves the highest gAUC, AUC, and lowest LogLoss across all datasets.' Table 1 contradicts this for LogLoss on TaobaoAds: CETNet (InterFormer+DHEN) has LogLoss 0.1930, while CETNet (InterFormer+DCNv2) has 0.1921, CETNet (DIN+DIEN) has 0.1922, and InterFormer alone has 0.1924. The claim should be corrected to reflect the actual metric(s) for which the combination is best, and the authors should explain why LogLoss does not follow the same pattern.
  3. [3.4] The confidence measure used in the fusion module is the negation entropy of each model's scalar click probability. For a Bernoulli distribution, entropy is a deterministic function of p that is minimized as p approaches 0 or 1, irrespective of whether the prediction is correct; a confidently wrong model can therefore receive the largest fusion weight. The paper does not validate that exp(-H(P)) is a useful confidence signal for CTR predictions (e.g., by correlating it with per-instance error or showing calibration), so the theoretical justification for the fusion mechanism is currently weak.
  4. [4.1 / Tables 1 and 3] All experimental tables appear to be based on single runs with one random seed, and no standard deviations, confidence intervals, or significance tests are reported. Many of the claimed differences are small (e.g., Table 1, KuaiVideo AUC 0.7456 vs. 0.7453; Table 3, Amazon gAUC 0.8860 vs. 0.8851), so it is difficult to establish that the improvements are not run-to-run noise. Please provide multiple-seed results with summary statistics for the main claims in Tables 1 and 3.
  5. [4.4.2 / Table 4] The Criteo/Avazu experiments compare 'Ours-sum' and 'Ours-concat' for DNN, IPNN, DCNv2, and FinalMLP, but the manuscript never specifies how these variants relate to the CETNet architecture described in Section 3. For example, it is unclear whether symmetric KL alignment and confidence-based fusion are used in these configurations, or whether the reader should view them as simple concatenation/summation ensembles over the existing models. Without this specification, the parameter-efficiency claim relative to the multi-embedding baseline cannot be fully assessed.
minor comments (8)
  1. [Table 1] The table header row contains the typo 'InteFormer'; it should be 'InterFormer'.
  2. [4.3] The text says 'such as an increase in gAUC from 0.6637 to 0.6620,' but 0.6620 is a decrease; the sentence should be corrected.
  3. [4.1.2] The section says 'six state-of-the-art models' but lists seven model names, with four non-sequential models and three sequential models; the count and list should be aligned.
  4. [Figure 3] Figure 3 lacks axis labels and a clear definition of the NE gain metric; please label the axes and state the exact formula used for NE gain.
  5. [3.3.1] The KL divergence formula takes logarithms of predicted probabilities without an epsilon; a numerical stabilization note or clipping strategy should be added for cases where predictions are 0 or 1.
  6. [Abstract / 3.4.1] The abstract uses the term 'general softmax' while Section 3.4.1 describes standard softmax; the terminology should be made consistent.
  7. [Title page] The manuscript retains template placeholders such as 'Do Not Use This Code', 'Conference acronym XX', and the copyright block; these should be removed before submission.
  8. [4.4.1 / Figure 2] The labels 'Standard' and 'Ensemble' in Figure 2 are not defined in the caption; please clarify whether they refer to single-model predictions before and after collaborative training.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: CETNet is empirically validated against external benchmarks; the only self-citations supply component architectures, not the conclusion.

full rationale

I find no circular derivation. CETNet wraps two published models, InterFormer and DHEN, into an ensemble with separate embedding tables, collaborative KL-divergence training, and entropy-based fusion; its central claims are tested on external public datasets (AmazonElectronics, TaobaoAds, KuaiVideo, Criteo, Avazu) and against independent baselines. The fusion weights are explicitly defined from the negation entropy of each model's prediction, so the statement that lower-entropy models receive larger weights is true by construction, but the substantive claim that this weighting improves accuracy is settled empirically by the ablation study in Table 3, not by the definition. The KL-divergence term is an added training objective, and the multi-embedding comparison follows the external paradigm of [6]. The self-cited InterFormer and DHEN papers supply the component architectures, not the conclusion that the ensemble outperforms baselines; that conclusion rests on benchmark numbers. Section 4.5 is an honest limitation: it states that confidence-based fusion 'does not result in significant improvements or stable learning in our internal setting' and that simple concatenation achieved a better NE gain. This weakens the generality of the fusion contribution and is a correctness risk, but it is not circular. No equation is assumed equal to its target, and no fitted parameter is renamed as a prediction. The only mild issue is self-citation of the two component models, which is not load-bearing because the ensemble's value is judged against external data and baselines. I therefore assign a score of 1.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on modeling assumptions that separate embedding tables capture complementary patterns, that entropy-based confidence correlates with correctness, and that KL alignment preserves diversity. None of these is proven in the paper; the first is inherited from prior work, and the second is contradicted by the internal deployment. No new physical or conceptual entities are introduced.

free parameters (3)
  • alpha (KL divergence weight) = not reported
    Controls the strength of the symmetric KL collaboration term in the final loss; selected by tuning over {0.1, 0.5, 1.0} on the evaluation datasets, and the chosen value is not reported.
  • learning rate = not reported
    Tuned per the experimental setup but the final value is not reported, so the reported results depend on an unreported choice.
  • weight decay = not reported
    Tuned per the experimental setup but the final value is not reported, so the reported results depend on an unreported choice.
assumptions (4)
  • domain assumption A single embedding table limits a CTR model's ability to capture diverse feature interactions; separate embedding tables capture complementary patterns.
    Section 3.2 states this as motivation; it is a modeling assumption inherited from prior work [6] rather than a fact established in this paper.
  • domain assumption Lower entropy of a model's scalar click-probability prediction indicates higher confidence and should receive a larger fusion weight.
    Section 3.4 defines confidence as negation entropy; whether this correlates with correctness is not established and is contradicted by the internal deployment in Section 4.5.
  • domain assumption Symmetric KL divergence aligns the two models' predictions and prevents dominance without destroying their diversity.
    Section 3.3 asserts this; the paper provides no theoretical argument or direct measurement of diversity or dominance.
  • domain assumption InterFormer and DHEN, as described in references [29] and [31], capture sequential and hierarchical feature interactions respectively.
    The framework's complementarity claim inherits these architecture descriptions from the authors' own self-cited prior work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Collaborative Ensemble Framework for CTR Prediction." pith.science (2026). https://pith.science/paper/NPEW2CKJ

@misc{pith2026241113700,
  author       = {Pith},
  title        = {Pith review of: A Collaborative Ensemble Framework for CTR Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NPEW2CKJ}},
  note         = {Machine review of arXiv:2411.13700}
}
read the original abstract

Recent advances in foundation models have established scaling laws that enable the development of larger models to achieve enhanced performance, motivating extensive research into large-scale recommendation models. However, simply increasing the model size in recommendation systems, even with large amounts of data, does not always result in the expected performance improvements. In this paper, we propose a novel framework, Collaborative Ensemble Training Network (CETNet), to leverage multiple distinct models, each with its own embedding table, to capture unique feature interaction patterns. Unlike naive model scaling, our approach emphasizes diversity and collaboration through collaborative learning, where models iteratively refine their predictions. To dynamically balance contributions from each model, we introduce a confidence-based fusion mechanism using general softmax, where model confidence is computed via negation entropy. This design ensures that more confident models have a greater influence on the final prediction while benefiting from the complementary strengths of other models. We validate our framework on three public datasets (AmazonElectronics, TaobaoAds, and KuaiVideo) as well as a large-scale industrial dataset from Meta, demonstrating its superior performance over individual models and state-of-the-art baselines. Additionally, we conduct further experiments on the Criteo and Avazu datasets to compare our method with the multi-embedding paradigm. Our results show that our framework achieves comparable or better performance with smaller embedding sizes, offering a scalable and efficient solution for CTR prediction tasks.

Figures

Figures reproduced from arXiv: 2411.13700 by the authors.

Figure 1
Figure 1. The framework of CETNet. The left figure illustrates the overall framework of our ensemble with N models (for simplicity, we depict two models—InterFormer and DHEN—as used in our main experiments). Each model utilizes its own distinct embedding table. The output embeddings from these models are combined to leverage the unique strengths of each model. The right figure shows the confidence-based fusion module, where e… view at source ↗
Figure 2
Figure 2. Performance of Component Model on AmazonElec [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. NE learning curve on the Avazu dataset. Although the ME framework with DCNv2 achieves an AUC of 0.78975 at a 10x embedding size, our framework with Ours-concat achieves a higher AUC of 0.79033 with just a 2x embedding size. This result highlights that our ensemble design not only matches but surpasses the performance of ME with far fewer embedding parameters. In industry, models are often trained on large datasets w… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Learning Multi-Branch Cooperation for Enhanced Click-Through Rate Prediction at Taobao

    cs.IR 2024-11 conditional novelty 6.0 of 10

    MBCnet improves CTR prediction by having three feature-interaction branches selectively teach each other on disagreed samples while a regularizer keeps their representations moderately different.

Reference graph

Works this paper leans on

42 extracted references · 13 canonical work pages · cited by 1 Pith paper

  1. [1]

    Rohan Anil, Gabriel Pereyra, Alexandre Passos, Robert Ormandi, George E Dahl, and Geoffrey E Hinton. 2018. Large scale distributed neural network training through online distillation. arXiv preprint arXiv:1804.03235 (2018)

    arXiv 2018

  2. [2]

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

  3. [3]

    Xiuqi Deng, Lu Xu, Xiyao Li, Jinkai Yu, Erpeng Xue, Zhongyuan Wang, Di Zhang, Zhaojie Liu, Guorui Zhou, Yang Song, et al. 2024. End-to-end training of Multimodal Model and ranking Model. arXiv preprint arXiv:2404.06078 (2024)

    arXiv 2024

  4. [4]

    Huifeng Guo, Ruiming Tang, Yunming Ye, Zhenguo Li, and Xiuqiang He. 2017. DeepFM: a factorization-machine based neural network for CTR prediction.arXiv preprint arXiv:1703.04247 (2017)

    arXiv 2017

  5. [5]

    Qiushan Guo, Xinjiang Wang, Yichao Wu, Zhipeng Yu, Ding Liang, Xiaolin Hu, and Ping Luo. 2020. Online knowledge distillation via collaborative learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 11020–11029

    work page 2020

  6. [6]

    Xingzhuo Guo, Junwei Pan, Ximei Wang, Baixu Chen, Jie Jiang, and Mingsheng Long. 2024. On the Embedding Collapse when Scaling up Recommendation Models. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024 . OpenReview.net

    work page 2024

  7. [7]

    Zongbo Han, Fan Yang, Junzhou Huang, Changqing Zhang, and Jianhua Yao

  8. [8]

    Ruining He and Julian McAuley. 2016. Ups and downs: Modeling the visual evolution of fashion trends with one-class collaborative filtering. In proceedings of the 25th international conference on world wide web . 507–517

    2016

Show all 42 references
  1. [9]

    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. In Proceedings of the eighth international workshop on data mining for online adver...

    2014

  2. [10]

    Geoffrey Hinton. 2015. Distilling the Knowledge in a Neural Network. arXiv preprint arXiv:1503.02531 (2015)

    2015 arXiv

  3. [11]

    Xuyang Hou, Zhe Wang, Qi Liu, Tan Qu, Jia Cheng, and Jun Lei. 2023. Deep Context Interest Network for Click-Through Rate Prediction. InProceedings of the 32nd ACM International Conference on Information and Knowledge Management . 3948–3952

    2023

  4. [12]

    Zhaoxin Huan, Ke Ding, Ang Li, Xiaolu Zhang, Xu Min, Yong He, Liang Zhang, Jun Zhou, Linjian Mo, Jinjie Gu, et al . 2024. Exploring Multi-Scenario Multi- Modal CTR Prediction with a Large Scale Dataset. In Proceedings of the 47th International ACM SIGIR Conference on Research ...

    2024

  5. [13]

    Yongqi Li, Meng Liu, Jianhua Yin, Chaoran Cui, Xin-Shun Xu, and Liqiang Nie

  6. [14]

    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 ...

    2018

  7. [15]

    Xiaolong Liu, Liangwei Yang, Zhiwei Liu, Xiaohan Li, Mingdai Yang, Chen Wang, and S Yu Philip. 2023. Group-Aware Interest Disentangled Dual-Training for Personalized Recommendation. In 2023 IEEE International Conference on Big Data (BigData). IEEE, 393–402

    2023

  8. [16]

    Maxim Naumov, Dheevatsa Mudigere, Hao-Jun Michael Shi, Jianyu Huang, Narayanan Sundaraman, Jongsoo Park, Xiaodong Wang, Udit Gupta, Carole- Jean Wu, Alisson G Azzolini, et al. 2019. Deep learning recommendation model for personalization and recommendation systems.arXiv preprin...

    2019 arXiv

  9. [17]

    Junwei Pan, Wei Xue, Ximei Wang, Haibin Yu, Xun Liu, Shijie Quan, Xueming Qiu, Dapeng Liu, Lei Xiao, and Jie Jiang. 2024. Ads recommendation in a collapsed and entangled world. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 5566–5577

    2024

  10. [18]

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

    2010

  11. [19]

    Matthew Richardson, Ewa Dominowska, and Robert Ragno. 2007. Predicting clicks: estimating the click-through rate for new ads. In Proceedings of the 16th international conference on World Wide Web. 521–530

    2007

  12. [20]

    J Ben Schafer, Joseph A Konstan, and John Riedl. 2001. E-commerce recommen- dation applications. Data mining and knowledge discovery 5 (2001), 115–153

    2001

  13. [21]

    Tianchi. [n. d.]. Ad display/click data on taobao.com, 2018. https://tianchi.aliyun. com/dataset/56

    2018

  14. [22]

    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

    2017

  15. [23]

    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

    2021

  16. [24]

    Xu Wang, Jiangxia Cao, Zhiyi Fu, Kun Gai, and Guorui Zhou. 2024. HoME: Hierarchy of Multi-Gate Experts for Multi-Task Learning at Kuaishou. arXiv preprint arXiv:2408.05430 (2024)

    2024 arXiv

  17. [25]

    Andrew M Webb, Charles Reynolds, Dan-Andrei Iliescu, Henry Reeve, Mikel Luján, and Gavin Brown. 2019. Joint training of neural network ensembles. stat 1050 (2019), 12

    2019

  18. [26]

    Zhibo Xiao, Luwei Yang, Wen Jiang, Yi Wei, Yi Hu, and Hao Wang. 2020. Deep multi-interest network for click-through rate prediction. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management . 2265–2268

    2020

  19. [27]

    Liangwei Yang, Zhiwei Liu, Chen Wang, Mingdai Yang, Xiaolong Liu, Jing Ma, and Philip S Yu. 2023. Graph-based alignment and uniformity for recommendation. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management. 4395–4399

    2023

  20. [28]

    Ibrahim Can Yilmaz and Said Aldemir. 2024. Mutual Learning for Finetuning Click-Through Rate Prediction Models. arXiv preprint arXiv:2406.12087 (2024)

    2024 arXiv

  21. [29]

    Yu, Hanghang Tong, and Jiyan Yang

    Zhichen Zeng, Xiaolong Liu, Mengyue Hang, Xiaoyi Liu, Qinghai Zhou, Chaofei Yang, Yiqun Liu, Yichen Ruan, Laming Chen, Yuxin Chen, Yujia Hao, Jiaqi Xu, Jade Nie, Xi Liu, Buyun Zhang, Wei Wen, Siyang Yuan, Kai Wang, Wen-Yen Chen, Yiping Han, Huayu Li, Chunzhi Yang, Bo Long, Phi...

    2024 arXiv

  22. [30]

    Buyun Zhang, Liang Luo, Yuxin Chen, Jade Nie, Xi Liu, Daifeng Guo, Yanli Zhao, Shen Li, Yuchen Hao, Yantao Yao, et al. 2024. Wukong: Towards a Scaling Law for Large-Scale Recommendation. arXiv preprint arXiv:2403.02545 (2024)

    2024 arXiv

  23. [31]

    Buyun Zhang, Liang Luo, Xi Liu, Jay Li, Zeliang Chen, Weilin Zhang, Xiaohan Wei, Yuchen Hao, Michael Tsang, Wenjun Wang, et al. 2022. DHEN: A deep and hierarchical ensemble network for large-scale click-through rate prediction. arXiv preprint arXiv:2203.11014 (2022)

    2022 arXiv

  24. [32]

    Xiaohui Zhang, Jaehong Yoon, Mohit Bansal, and Huaxiu Yao. 2024. Multimodal representation learning by alternating unimodal adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 27456–27466

    2024

  25. [33]

    Ying Zhang, Tao Xiang, Timothy M Hospedales, and Huchuan Lu. 2018. Deep mutual learning. In Proceedings of the IEEE conference on computer vision and pattern recognition. 4320–4328

    2018

  26. [34]

    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

    2019

  27. [35]

    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. In Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining ...

    2018

  28. [36]

    Haolin Zhou, Junwei Pan, Xinyi Zhou, Xihua Chen, Jie Jiang, Xiaofeng Gao, and Guihai Chen. 2024. Temporal Interest Network for User Response Prediction. In Companion Proceedings of the ACM on Web Conference 2024 . 413–422

    2024

  29. [37]

    Jieming Zhu, Jinyang Liu, Weiqi Li, Jincai Lai, Xiuqiang He, Liang Chen, and Zibin Zheng. 2020. Ensembled CTR prediction via knowledge distillation. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management. 2941–2958

    2020

  30. [38]

    Jieming Zhu, Jinyang Liu, Shuai Yang, Qi Zhang, and Xiuqiang He. 2021. Open benchmarking for click-through rate prediction. In Proceedings of the 30th ACM international conference on information & knowledge management . 2759–2769

    2021

  31. [39]

    Xiatian Zhu, Shaogang Gong, et al. 2018. Knowledge distillation by on-the-fly native ensemble. Advances in neural information processing systems 31 (2018). Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Trovato et al. Table 5: Model performance with one epoch training...

    2018

  32. [2016]

    In Proceedings of the 1st workshop on deep learning for recommender systems

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

  33. [2019]

    In Proceedings of the 27th ACM international conference on multimedia

    Routing micro-videos via a temporal graph-guided recommendation system. In Proceedings of the 27th ACM international conference on multimedia. 1464–1472

  34. [2022]

    In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Multimodal dynamics: Dynamical fusion for trustworthy multimodal classification. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 20707–20717

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.