Pith. sign in

REVIEW 4 major objections 5 minor 3 cited by

Contextual Generative Auction with Permutation-level Externalities for Online Advertising

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that an auction which models the order of ads shown to a user—permutation-level externalities—can be made both revenue-optimal and incentive-compatible, and that a generative model plus a learned payment rule…

desk verdict CGA is a credible engineering contribution with a real online lift, but the Lemma 1 proof gap and circular IC metric mean the advertised theoretical and incentive guarantees outrun the evidence. read the letter →

arxiv 2412.11544 v1 pith:6K25XKTN submitted 2024-12-16 cs.GT

Ruitao Zhu , Yangsu Liu , Dagui Chen , Zhenjia Ma , Chufeng Shi , Zhenzhe Zheng , Jie Zhang , Jian Xu
show 2 more authors
Bo Zheng Fan Wu
This is my paper · ORCID
classification cs.GT MSC 91B2668T07
keywords learning-basedmechanismdesignpermutation-levelexternalitiesgenerativeauctionincentivecompatibilityex-postregretmulti-slotadCTRpredictiononlineadvertising
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

Online ad auctions usually assume each ad's click-through rate is independent of the ads around it, but real user behavior depends on the full ordered list. This paper argues that the classic Myerson optimal-auction result survives even when CTR depends on the entire permutation of ads, so the optimal allocation and the optimal payment rule can be designed separately. It then builds the Contextual Generative Auction (CGA), which generates ad sequences with an autoregressive model and learns payments by minimizing ex-post regret, and reports that CGA reaches about 95% of the enumerated optimum revenue with small incentive-compatibility regret in offline tests and a 3.2% revenue lift in an online A/B test. The claim matters because it offers a tractable way to handle a hard combinatorial externality structure in real-time ad systems.

What carries the argument

The central object is the permutation-aware CTR model Θ(𝒙ᵢ; 𝑨, 𝑿, 𝒖), which maps an ordered allocation of ads to each ad's click probability, and the allocation rule that maximizes expected virtual welfare Σ ᵢ φ̃(𝑏ᵢ, 𝐹ᵢ)Θ(𝒙ᵢ; 𝑨, 𝑿, 𝒖) using ironed virtual values. The paper proves that this rule is monotone despite externalities, so Myerson's lemma applies and the optimal payment is the integral of the allocated CTR over bid changes. The mechanism that carries the argument is the Generator-Evaluator architecture: a permutation-invariant encoder with a permutation-equivariant autoregressive decoder (the Generator) produces allocations, while the Evaluator estimates permutation-aware CTR to supply rewards for policy-gradient training; PaymentNet then learns payments by minimizing differentiable ex-post regret, decoupled from allocation optimization.

What would settle it

Retrain the Evaluator on allocations generated by CGA itself and then recompute the 'Optimal' benchmark, the revenue estimates, and the IC regret using this retrained model; if the reported regret rises substantially above 3.7% or the revenue gain over DNA disappears, the central claim that CGA approximates the optimal DSIC auction would be undermined by distribution shift in the CTR model.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the ironed-virtual-value Myerson auction remains the revenue-optimal DSIC mechanism when each ad's CTR is a function of the whole allocation permutation, not just its own slot. The key step is proving that an allocation rule maximizing virtual welfare is still monotone in each advertiser's bid under permutation-level externalities (Lemma 1), so Myerson's payment formula applies and the optimal mechanism decomposes into an allocation rule and a separate payment rule (Corollary 1). Because direct enumeration of all permutations is infeasible online, the paper replaces the allocation rule with an autoregressive generative model (the Generator) guided by a permutation-aware CTR model (the Evaluator), and learns the payment rule with a neural network (PaymentNet) trained to minimize ex-post regret. Experiments on Taobao data show CGA achieving near-optimal revenue (~95% of the enumerated optimum) with IC regret of 2.1% to 3.7% offline, and online A/B tests show a 3.2% RPM improvement over the deployed DNA auction.

Load-bearing premise

The Evaluator's permutation-aware CTR model, trained on historical click logs from the existing system, remains accurate under CGA's new allocation policy and under the counterfactual bid perturbations used to compute virtual welfare and ex-post regret.

Editorial extensions

If this is right

  • If the central claim is correct, auction systems can model the full ordered context of displayed ads without enumerating permutations, making permutation-level externalities computationally feasible for online deployment.
  • Allocation and payment can be optimized separately in learned mechanisms, so the allocation model can focus purely on virtual-welfare maximization while the payment model handles incentive compatibility through regret minimization.
  • The ex-post-regret formulation provides a differentiable way to enforce DSIC in neural auction designs, which can be applied beyond the specific generative architecture used here.
  • The monotonicity result suggests that any allocation rule maximizing virtual welfare under permutation-dependent CTR is implementable, extending Myerson's theory to a broader class of sequential allocation problems.
  • In practice, the online A/B results imply that modeling permutation-level externalities can produce measurable revenue gains with only a few milliseconds of added latency.

Reading between the lines

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

  • The paper's decoupling result likely generalizes beyond advertising to other sequential allocation settings where the value of an allocation depends on the order of assigned items, provided a permutation-aware value model is available.
  • The authors' claim that CGA approximates the optimal auction is conditional on the Evaluator's CTR predictions being accurate under the new allocation policy; a natural extension would be to retrain the Evaluator on CGA's own allocations and re-measure regret and revenue.
  • The 2.1% to 3.7% IC regret reported is computed under a specific counterfactual bid-perturbation grid; in practice, strategic advertisers could exploit the learned payment rule in ways not captured by that grid, so the guarantee is only as strong as the regret measurement procedure.
  • Since the paper presets value distributions for evaluation rather than estimating them from bids, the real-world optimality claim would be tested by estimating advertiser value distributions from observed bidding behavior and re-running the comparison.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes Contextual Generative Auction (CGA), a learning-based multi-slot ad auction mechanism that models permutation-level externalities. The authors derive a Myerson-style optimal DSIC auction for this setting, then approximate it with an autoregressive Generator trained by policy gradient using a permutation-aware Evaluator as reward model, and a PaymentNet trained to minimize ex-post regret via an augmented Lagrangian method. Offline experiments on Taobao logs compare CGA against GSP, DNA, SW-VCG, EdgeNet, VCG, and an enumerated 'Optimal' baseline, reporting that CGA attains roughly 95% of the Optimal revenue with 2.1–3.7% IC regret; an online A/B test reports a 3.2% RPM lift over the deployed DNA system.

Significance. If the results hold, CGA is a practically relevant advance: it breaks the 'allocation-after-prediction' paradigm, provides a theoretically motivated decoupling of allocation and payment, and demonstrates strong empirical performance in a large-scale industrial setting. The paper is clearly written, includes ablations that isolate the main design choices, and provides a within-model optimal enumeration benchmark that is a useful reference. The main risks are the rigor of the theoretical monotonicity proof, the reliance on a learned Evaluator that may be biased under distribution shift, and the fact that the reported IC metric is the same objective used to train PaymentNet. These issues do not invalidate the engineering contribution but do affect the strength of the central optimality and incentive-compatibility claims.

major comments (4)
  1. [Appendix A.1 (Lemma 1)] The proof of Lemma 1 shows that for a given pair of bids b_t < b'_t there exists some optimal allocation at the higher bid that preserves monotonicity of ad_t's CTR, but it does not construct a single tie-breaking rule for the argmax that is simultaneously monotone in every bidder's bid and for all bid profiles. In particular, the 'w.l.o.g.' choices in the ironed-virtual-value case (Case 2) and in the initial supposition 'suppose ad_t in A*' leave open the possibility that a random or arbitrary tie-breaking rule violates the monotonicity condition required by Myerson's Lemma. Since Corollary 1 rests on Lemma 1, the paper should either prove existence of a monotone tie-breaking rule explicitly (e.g., by a lexicographic selection rule) or weaken the optimality claim accordingly.
  2. [Section 5.2 (Table 1) and Section 3.2] The 'Optimal' enumeration baseline and the reported 95% revenue approximation are computed with the learned Evaluator Θ, which is trained on click logs generated by the existing DNA system. If Θ is biased on CGA-style allocations or on the counterfactual bid perturbations used in the Monte Carlo payment integral, then the 'Optimal' revenue is not a true upper bound and the CGA revenue ratio is measured against a misspecified model. The paper should provide evidence that Θ generalizes to CGA allocations, for example by reporting CTR calibration on a holdout of the online CGA arm or on allocations produced by the learned Generator, and by testing sensitivity of the Optimal benchmark to plausible perturbations of Θ.
  3. [Section 4.2 (Eq. 10) and Section 5.2 (Ψ metric)] The reported IC metric Ψ is the same normalized ex-post regret objective that PaymentNet is trained to minimize in Eq. (10). Reporting low Ψ on the test set is therefore partly by construction and does not constitute an independent test of the DSIC property. The authors should evaluate incentive compatibility with a distinct measure, for instance misreport perturbations that were not used during training (different α grids or non-multiplicative deviations), or a metric based on the payment rule's Myerson integral, or a small-scale online experiment with bid perturbations.
  4. [Section 3.1 (Eq. 5) and Section 5.2] The theoretical optimality result in Corollary 1 applies to the exact argmax allocation rule, but the deployed Generator selects the highest-probability ad at each slot greedily during inference, which is not guaranteed to maximize virtual welfare. The 95% offline comparison partially addresses this gap empirically, but the paper should explicitly state that the DSIC guarantee applies to the exact argmax rule and not to the greedy approximation, and should discuss how the greedy error affects the IC regret figures (which are computed using the greedy Generator's allocations).
minor comments (5)
  1. [Section 5.1.1] The candidate set size is stated as approximately 30, while Section 2.2's Taobao example uses n≈50 and k≈5; offline experiments use k=3. Please clarify the relation between the example and the experimental setting.
  2. [Section 3.2] The expression for the calibration vector γ_A is written as 2σ(r(r([...]))); please define the output dimension of the MLP and clarify that the sigmoid is applied element-wise to produce values in (0,2).
  3. [Table 1] It would help to report confidence intervals or statistical significance tests for the revenue differences, given the large sample size (100,000 auctions).
  4. [Table 4 (online A/B test)] The online A/B test reports a 3.2% RPM lift but does not report the IC regret or the distance from the Optimal benchmark; please state explicitly in the text that the online test validates aggregate metrics only.
  5. [Section 5.1.3] The paper should cite the original data-driven IC metric [7] more fully and explain how Ψ relates to the ex-post regret defined in Eq. (2), especially the normalization by utility.

Circularity Check

1 steps flagged · score 3.0 of 10

The theoretical derivation and revenue benchmark are self-contained, but the reported IC regret is measured with the same ex-post regret objective used to train PaymentNet, making the low-regret claim partly by construction.

  1. fitted input called prediction [Section 4.2, Eq. (10); Section 5.1.3, Ψ metric; Tables 1 and 3.]
    "L_P = −1/|D| ∑_{s∈D} [∑_{i∈k} P_i(A_s)Θ_i(A_s) − ∑_{i∈k} λ_i c_rgt_i − ρ/2 ∑_{i∈k} (c_rgt_i)^2] ... IC metric: Ψ = 1/|D| ∑_{s∈D} ∑_{i∈k} c_rgt_i / u_i(v_i,b_s;X_s,u_s), where c_rgt_i is defined in Equation (2)."

    PaymentNet is fit by minimizing the Lagrangian L_P, whose IC-enforcement terms are exactly the empirical ex-post regrets c_rgt_i. The reported IC metric Ψ is a normalized sum of the same c_rgt_i values. Reporting Ψ = 2.1–3.7% as evidence of IC is therefore an evaluation against the same objective used to train the payment network, not an independent test. The held-out test set and non-convex optimization give the number some empirical content, but the low regret is partly forced by construction.

full rationale

The core theoretical result (Corollary 1) is a direct application of Myerson's Lemma and virtual-welfare maximization with a permutation-aware CTR model; it does not reduce to a self-citation or to the paper's own definitions. The value distributions in the offline experiments are preset to uniform/exponential rather than fitted from bids, so the revenue comparison against the enumerated 'Optimal' baseline is a fair within-model benchmark: both CGA and the Optimal baseline use the same learned Evaluator Θ to define virtual welfare and payments. The online A/B test provides external evidence for the RPM/CTR gains. The only identifiable circular element is the IC metric: the ex-post regret minimized in Eq. (10) is the same quantity normalized in Ψ, so the low reported regret is partly a measure of the training objective. The distribution-shift risk of the learned Evaluator under CGA allocations and counterfactual bid perturbations is a correctness/robustness concern, not a circularity. Overall score 3 reflects this one mild, partially constructive metric loop.

Assumptions & free parameters 4 free parameters · 8 assumptions · 0 invented entities

The paper introduces no new physical or economic entities; Generator, Evaluator, and PaymentNet are learned model components rather than postulated objects. The main loaded assumptions are the known value distributions, the accuracy of the learned CTR model, and the transfer of exact theoretical guarantees to a greedy learned approximation.

free parameters (4)
  • Learnable virtual-value scale w = learned, not reported
    Appears in Eq. (5) as e^w scaling the virtual value in allocation probabilities; trained on data and directly affects which ads are selected.
  • Generator, Evaluator, PaymentNet network weights = learned on 500,000 Taobao auctions
    All neural network parameters are fitted to the offline dataset; no values, seeds, or checkpoints are provided.
  • Lagrange multipliers lambda and penalty rho = updated during training, not specified
    Used in Eq. (10) to enforce the IC constraint; the values and update schedule are not reported.
  • Value distributions for virtual values = preset Uniform and Exponential
    Ironed virtual value functions are constructed from user-chosen distributions, not estimated from observed bids; this avoids estimation bias but is a manual modeling choice.
assumptions (8)
  • domain assumption Advertiser values are independent draws from known distributions f_i
    Section 2.1 defines the model; distributions are later preset to uniform and exponential for experiments.
  • domain assumption Single-parameter cost-per-click setting: advertiser utility is (v_i - p_i) * CTR_i
    Section 2.1 and footnote 3; the entire Myerson adaptation depends on this utility form.
  • domain assumption The permutation-aware CTR model Theta is known or accurately learned
    Section 2.1 and 3.2; the Evaluator is trained on historical logs and assumed to generalize to new allocations and bid perturbations.
  • standard math Myerson's lemma and revenue equivalence hold for single-parameter environments
    Theorems 1 and 2 are invoked directly.
  • standard math Ironed virtual values are monotone non-decreasing in the bid
    Definition 3 and the proof of Lemma 1 rely on this property from Myerson.
  • ad hoc to paper A tie-breaking rule can be chosen so that the argmax allocation preserves monotonicity
    The proof of Lemma 1 selects the original allocation A* in Case 2, but the paper does not define a deterministic tie-breaking rule that works for all bid profiles.
  • ad hoc to paper The learned Generator's greedy autoregressive allocation approximates the virtual-welfare-maximizing permutation
    Section 3.1 uses sampling in training and greedy selection at inference; theoretical optimality assumes exact argmax over permutations.
  • ad hoc to paper Augmented Lagrangian converges to a low-regret solution despite non-convexity
    Section 4.2 explicitly states the method is not guaranteed to converge to the global optimum.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Contextual Generative Auction with Permutation-level Externalities for Online Advertising." pith.science (2026). https://pith.science/paper/6K25XKTN

@misc{pith2026241211544,
  author       = {Pith},
  title        = {Pith review of: Contextual Generative Auction with Permutation-level Externalities for Online Advertising},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6K25XKTN}},
  note         = {Machine review of arXiv:2412.11544}
}
read the original abstract

Online advertising has become a core revenue driver for the internet industry, with ad auctions playing a crucial role in ensuring platform revenue and advertiser incentives. Traditional auction mechanisms, like GSP, rely on the independent CTR assumption and fail to account for the influence of other displayed items, termed externalities. Recent advancements in learning-based auctions have enhanced the encoding of high-dimensional contextual features. However, existing methods are constrained by the "allocation-after-prediction" design paradigm, which models set-level externalities within candidate ads and fails to consider the sequential context of the final allocation, leading to suboptimal results. This paper introduces the Contextual Generative Auction (CGA), a novel framework that incorporates permutation-level externalities in multi-slot ad auctions. Built on the structure of our theoretically derived optimal solution, CGA decouples the optimization of allocation and payment. We construct an autoregressive generative model for allocation and reformulate the incentive compatibility (IC) constraint into minimizing ex-post regret that supports gradient computation, enabling end-to-end learning of the optimal payment rule. Extensive offline and online experiments demonstrate that CGA significantly enhances platform revenue and CTR compared to existing methods, while effectively approximating the optimal auction with nearly maximal revenue and minimal regret.

Figures

Figures reproduced from arXiv: 2412.11544 by the authors.

Figure 1
Figure 1. The Permutation-aware ad CTR as a function of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The architecture of Contextual Generative Auction (CGA). The middle part shows the overall framework of CGA, [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. EGA-V1: Unifying Online Advertising with End-to-End Learning

    cs.IR 2025-05 conditional novelty 6.0 of 10

    EGA-V1 unifies advertising ranking and auction into a single non-autoregressive generative model with cluster attention, and is reported to beat multi-stage cascades on Meituan's ad traffic.

  2. NGA: Non-autoregressive Generative Auction with Global Externalities for Advertising Systems

    cs.IR 2025-06 conditional novelty 5.0 of 10

    NGA is a non-autoregressive generative auction that models effects of adjacent organic content and computes rewards and payments in parallel, reporting gains in RPM, CTR, CVR, and latency over CGA.

  3. EGA-V2: An End-to-end Generative Framework for Industrial Advertising

    cs.IR 2025-05 conditional novelty 5.0 of 10

    EGA-V2 unifies ad ranking, creative selection, allocation, and payment into one generative transformer, and reports offline revenue and CTR improvements over cascaded and generative baselines on Meituan data.

Reference graph

Works this paper leans on

48 extracted references · 25 canonical work pages · cited by 3 Pith papers

  1. [1]

    Qingyao Ai, Keping Bi, Jiafeng Guo, and W Bruce Croft. 2018. Learning a deep listwise context model for ranking refinement. In The 41st international ACM SIGIR conference on research & development in information retrieval . 135–144

    2018

  2. [2]

    Yoram Bachrach, Sofia Ceppi, Ian A Kash, Peter Key, and David Kurokawa. 2014. Optimising trade-offs among stakeholders in ad auctions. In Proceedings of the fifteenth ACM conference on Economics and computation . 75–92

    2014

  3. [3]

    Santiago R Balseiro, Yuan Deng, Jieming Mao, Vahab S Mirrokni, and Song Zuo

    work page

  4. [4]

    Lucas Beyer, Bo Wan, Gagan Madan, Filip Pavetic, Andreas Steiner, Alexander Kolesnikov, André Susano Pinto, Emanuele Bugliarello, Xiao Wang, Qihang Yu, et al. 2023. A study of autoregressive decoders for multi-tasking in computer vision. arXiv preprint arXiv:2303.17376 (2023)

    arXiv 2023

  5. [5]

    Chi Chen, Hui Chen, Kangzhi Zhao, Junsheng Zhou, Li He, Hongbo Deng, Jian Xu, Bo Zheng, Yong Zhang, and Chunxiao Xing. 2022. Extr: click-through rate prediction with externalities in e-commerce sponsored search. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 2732–2740. KDD ’25, August 3–7, 2025, Toronto, Canada R...

    work page 2022

  6. [6]

    Yeshwanth Cherapanamjeri, Constantinos Daskalakis, Andrew Ilyas, and Manolis Zampetakis. 2022. Estimation of standard auction models. In Proceedings of the 23rd ACM Conference on Economics and Computation . 602–603

    work page 2022

  7. [7]

    Yuan Deng, Sébastien Lahaie, Vahab Mirrokni, and Song Zuo. 2020. A data- driven metric of incentive compatibility. In Proceedings of The Web Conference

    2020

  8. [8]

    Yuan Deng, Jieming Mao, Vahab Mirrokni, and Song Zuo. 2021. Towards efficient auctions in an auto-bidding world. In Proceedings of the Web Conference 2021 . 3965–3973

    work page 2021

Show all 48 references
  1. [9]

    Rahul Dey and Fathi M Salem. 2017. Gate-variants of gated recurrent unit (GRU) neural networks. In 2017 IEEE 60th international midwest symposium on circuits and systems (MWSCAS). IEEE, 1597–1600

    2017

  2. [10]

    Zhijian Duan, Haoran Sun, Yurong Chen, and Xiaotie Deng. 2024. A scalable neural network for dsic affine maximizer auction design. Advances in Neural Information Processing Systems 36 (2024)

    2024

  3. [11]

    Paul Dütting, Zhe Feng, Harikrishna Narasimhan, David Parkes, and Sai Srivatsa Ravindranath. 2019. Optimal auctions through deep learning. In International Conference on Machine Learning . PMLR, 1706–1715

    2019

  4. [12]

    Benjamin Edelman, Michael Ostrovsky, and Michael Schwarz. 2007. Internet advertising and the generalized second-price auction: Selling billions of dollars worth of keywords. American economic review 97, 1 (2007), 242–259

    2007

  5. [13]

    Yufei Feng, Binbin Hu, Yu Gong, Fei Sun, Qingwen Liu, and Wenwu Ou. 2021. GRN: Generative Rerank Network for Context-wise Recommendation. arXiv preprint arXiv:2104.00860 (2021)

    2021 arXiv

  6. [14]

    Nicola Gatti, Alessandro Lazaric, and Francesco Trovò. 2012. A truthful learning mechanism for contextual multi-slot sponsored search auctions with externalities. In Proceedings of the 13th ACM Conference on Electronic Commerce . 605–622

    2012

  7. [15]

    Arpita Ghosh and Mohammad Mahdian. 2008. Externalities in online advertising. In Proceedings of the 17th international conference on World Wide Web . 161–168

    2008

  8. [16]

    Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory.Neural computation 9, 8 (1997), 1735–1780

    1997

  9. [17]

    Patrick Hummel and R Preston McAfee. 2014. Position auctions with externalities. In Web and Internet Economics: 10th International Conference, WINE 2014, Beijing, China, December 14-17, 2014. Proceedings 10 . Springer, 417–422

    2014

  10. [18]

    Dmitry Ivanov, Iskander Safiulin, Igor Filippov, and Ksenia Balabaeva. 2022. Optimal-er auctions through attention. Advances in Neural Information Processing Systems 35 (2022), 34734–34747

    2022

  11. [19]

    Bernard J Jansen and Tracy Mullen. 2008. Sponsored search: an overview of the concept, history, and technology. International Journal of Electronic Business 6, 2 (2008), 114–131

    2008

  12. [20]

    Przemyslaw Jeziorski and Ilya Segal. 2015. What makes them click: Empirical analysis of consumer demand for search advertising. American Economic Journal: Microeconomics 7, 3 (2015), 24–53

    2015

  13. [21]

    Albert Xin Jiang and Kevin Leyton-Brown. 2005. Estimating bidders’ valuation distributions in online auctions. In Proceedings of IJCAI-05 workshop on game theoretic and decision theoretic agents . 98–107

    2005

  14. [22]

    Ningyuan Li, Yunxuan Ma, Yang Zhao, Zhijian Duan, Yurong Chen, Zhilin Zhang, Jian Xu, Bo Zheng, and Xiaotie Deng. 2023. Learning-Based Ad Auction Design with Externalities: The Framework and A Matching-Based Approach. In Pro- ceedings of the 29th ACM SIGKDD Conference on Knowl...

    2023

  15. [23]

    Guogang Liao, Xuejian Li, Ze Wang, Fan Yang, Muzhi Guan, Bingqi Zhu, Yongkang Wang, Xingxing Wang, and Dong Wang. 2022. NMA: Neural Multi-slot Auctions with Externalities for Online Advertising. arXiv preprint arXiv:2205.10018 (2022)

    2022 arXiv

  16. [24]

    Shuchang Liu, Qingpeng Cai, Zhankui He, Bowen Sun, Julian McAuley, Dong Zheng, Peng Jiang, and Kun Gai. 2023. Generative flow network for listwise rec- ommendation. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 1524–1534

    2023

  17. [25]

    Xiangyu Liu, Chuan Yu, Zhilin Zhang, Zhenzhe Zheng, Yu Rong, Hongtao Lv, Da Huo, Yiqing Wang, Dagui Chen, Jian Xu, Fan Wu, Guihai Chen, and Xiaoqiang Zhu. 2021. Neural auction: End-to-end learning of auction mechanisms for e- commerce advertising. In Proceedings of the 27th AC...

    2021

  18. [26]

    Roger B Myerson. 1981. Optimal auction design. Mathematics of operations research 6, 1 (1981), 58–73

    1981

  19. [27]

    Gali Noti and Vasilis Syrgkanis. 2021. Bid prediction in repeated auctions with learning. In Proceedings of the Web Conference 2021 . 3953–3964

    2021

  20. [28]

    Michael Ostrovsky and Michael Schwarz. 2023. Reserve prices in internet adver- tising auctions: A field experiment. Journal of Political Economy 131, 12 (2023), 3352–3376

    2023

  21. [29]

    Changhua Pei, Yi Zhang, Yongfeng Zhang, Fei Sun, Xiao Lin, Hanxiao Sun, Jian Wu, Peng Jiang, Junfeng Ge, Wenwu Ou, et al. 2019. Personalized re-ranking for recommendation. In Proceedings of the 13th ACM conference on recommender systems. 3–11

    2019

  22. [30]

    Tian Qin, Fengxiang He, Dingfeng Shi, Wenbing Huang, and Dacheng Tao. 2022. Benefits of permutation-equivariance in auction mechanisms. Advances in Neural Information Processing Systems 35 (2022), 18131–18142

    2022

  23. [31]

    Jad Rahme, Samy Jelassi, Joan Bruna, and S Matthew Weinberg. 2021. A permutation-equivariant neural network architecture for auction design. In Pro- ceedings of the AAAI conference on artificial intelligence , Vol. 35. 5664–5672

    2021

  24. [32]

    Yuxin Ren, Qiya Yang, Yichun Wu, Wei Xu, Yalong Wang, and Zhiqiang Zhang

  25. [33]

    Guangyuan Shen, Shengjie Sun, Dehong Gao, Duanxiao Song, Libin Yang, Zhen Wang, Yongping Shi, and Wei Ning. 2023. EdgeNet: Encoder-decoder generative Network for Auction Design in E-commerce Online Advertising. In Proceed- ings of the 32nd ACM International Conference on Infor...

    2023

  26. [34]

    Xiaowen Shi, Fan Yang, Ze Wang, Xiaoxu Wu, Muzhi Guan, Guogang Liao, Wang Yongkang, Xingxing Wang, and Dong Wang. 2023. PIER: Permutation-Level Interest-Based End-to-End Re-ranking Framework in E-commerce. InProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery ...

    2023

  27. [35]

    Hal R Varian and Christopher Harris. 2014. The VCG auction in theory and practice. American Economic Review 104, 5 (2014), 442–445

    2014

  28. [36]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems 30 (2017)

    2017

  29. [37]

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

  30. [38]

    Yiqing Wang, Xiangyu Liu, Zhenzhe Zheng, Zhilin Zhang, Miao Xu, Chuan Yu, and Fan Wu. 2022. On designing a two-stage auction for online advertising. In Proceedings of the ACM Web Conference 2022 . 90–99

    2022

  31. [40]

    Yunjia Xi, Weiwen Liu, Xinyi Dai, Ruiming Tang, Weinan Zhang, Qing Liu, Xi- uqiang He, and Yong Yu. 2021. Context-aware reranking with utility maximization for recommendation. arXiv preprint arXiv:2110.09059 (2021)

    2021 arXiv

  32. [41]

    Yidan Xing, Zhilin Zhang, Zhenzhe Zheng, Chuan Yu, Jian Xu, Fan Wu, and Guihai Chen. 2023. Truthful auctions for automated bidding in online advertising. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence. 2915–2922

    2023

  33. [42]

    Manzil Zaheer, Satwik Kottur, Siamak Ravanbakhsh, Barnabas Poczos, Russ R Salakhutdinov, and Alexander J Smola. 2017. Deep sets. Advances in neural information processing systems 30 (2017)

    2017

  34. [43]

    Ruizhe Zhang, Xiaohui Xie, Jiaxin Mao, Yiqun Liu, Min Zhang, and Shaoping Ma

  35. [44]

    Zhanhao Zhang. 2021. A survey of online auction mechanism design using deep learning approaches. arXiv preprint arXiv:2110.06880 (2021)

    2021 arXiv

  36. [45]

    Zhilin Zhang, Xiangyu Liu, Zhenzhe Zheng, Chenrui Zhang, Miao Xu, Junwei Pan, Chuan Yu, Fan Wu, Jian Xu, and Kun Gai. 2021. Optimizing multiple performance metrics with deep GSP auctions for e-commerce advertising. InProceedings of the 14th ACM International Conference on Web ...

    2021

  37. [47]

    InProceedings of the Web Conference 2021

    Constructing a comparison-based click model for web search. InProceedings of the Web Conference 2021. 270–283

    2021

  38. [2021]

    In Proceedings of the 22nd ACM Conference on Economics and Computation

    The landscape of auto-bidding auctions: Value versus utility maximization. In Proceedings of the 22nd ACM Conference on Economics and Computation . 132– 133

  39. [2023]

    In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval

    Diffusion recommender model. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval . 832–841

  40. [2024]

    arXiv preprint arXiv:2402.06871 (2024)

    Non-autoregressive Generative Models for Reranking Recommendation. arXiv preprint arXiv:2402.06871 (2024)

    2024 arXiv

Pith tools

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