Pith. sign in

REVIEW 3 major objections 4 minor 136 references

Learning Truthful Mechanisms without Discretization

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read TEDI represents truthful mechanisms as convex pricing rules and learns them directly on continuous outcome spaces, avoiding discretization.

desk verdict A genuinely new discretization-free truthful mechanism learner with solid architecture and experiments, but the truthfulness guarantee is certified only for exact inference while Algorithm 6 runs a finite heuristic; fixable and worth refereeing. read the letter →

arxiv 2506.22911 v1 pith:PJKT3Y4E submitted 2025-06-28 cs.GT cs.AIcs.LG

Yunxuan Ma , Siqiang Wang , Zhijian Duan , Yukun Cheng , Xiaotie Deng This is my paper
classification cs.GTcs.AIcs.LG
keywords automatedmechanismdesigntruthfullearningmenumechanismspricingrulespartialconvexityuniversalapproximationLangevinsamplingrevenuemaximization
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

This paper introduces TEDI, a way to learn truthful, revenue-maximizing mechanisms that never discretizes the outcome space. It argues that a mechanism can be represented by pricing rules—functions that assign a price to every possible outcome—and that if each pricing rule is convex in the outcome and charges nothing for the zero outcome, the resulting menu mechanism is equivalent to a truthful direct mechanism. The paper's central claim is that this function class is fully expressive: TEDI can approximate the optimal truthful mechanism to arbitrary precision, and it can do so with polynomial parameter counts where discretization-based methods need exponentially many candidates. On the experimental side, TEDI reaches about 98 percent of optimal revenue in a ten-good digital-goods example and matches or beats the compared baselines across moderate-scale settings. If correct, this removes the main barrier, the curse of dimensionality, that limited earlier differentiable economics methods.

What carries the argument

The central object is the pricing rule: a function $p_i(x_i; t_{-i})$ that names the price player $i$ pays for local outcome $x_i$ given reports of others. TEDI writes each pricing rule as $f(x_i,t_{-i};\theta) - f(0,t_{-i};\theta)$, which enforces no-buy-no-pay, and represents $f$ with a Partial GroupMax Network—a network whose affine layers (except the first) have nonnegative weights generated by a parameterized affine network and whose activations are group-wise maxes. This makes $f$ convex in $x_i$, since a positive sum and pointwise maximum of convex functions stay convex, and the paper shows these networks are universal approximators of functions convex in $x_i$ and continuous in $t_{-i}$. The second mechanism is the covariance trick: because the training distribution is an exponential (Gibbs) distribution over outcomes, the gradient of an expectation under it equals the expectation of the gradient plus a covariance with the weighted affine social welfare, which can be estimated unbiasedly from paired samples. Third, continuous sampling uses Langevin dynamics to keep samples approximately drawn from the current Gibbs distribution as $\theta$ changes. Together, the architecture guarantees the menu is truth-inducing, and the gradient estimator makes optimizing it a first-order optimization problem.

What would settle it

On a trained TEDI menu, replace Algorithm 6's SGD inference with an exact convex solver and compare the resulting direct mechanism: any type profile at which the exact best response differs from the SGD-chosen outcome shows the deployed mechanism is not the certified truthful one, and a misreport that gains positive utility would settle that the approximation breaks incentive compatibility.

Watch

Extended reading notes

Core claim

On the paper's own terms: in outcome-separable quasi-linear settings (digital goods, procurement, multi-good monopolist), learning an optimal truthful mechanism is equivalent to learning a menu of pricing rules. Each rule maps a player's local outcome to a price, and two conditions on the rule suffice: it is convex in the local outcome and it satisfies no-buy-no-pay, meaning the zero bundle costs nothing. The paper proves both directions: any menu mechanism with these two properties induces a truthful direct mechanism, and any truthful direct mechanism can be represented this way. Because Partial GroupMax Networks are universal approximators of exactly these partially convex functions, a parameterized family can represent the truthful-mechanism class, and the supreme expected utility achievable by TEDI menus equals that of all truthful mechanisms. Training uses a covariance trick that turns gradients through the sampling distribution into a computable expectation, together with continuous sampling via Langevin dynamics. The implemented system therefore learns pricing rules directly on continuous outcome spaces, and the Bernoulli digital-goods example has an $O(m)$-parameter representation where discretized menus require $\Omega(2^m)$ candidates.

Load-bearing premise

Truthfulness requires the deployed inference to return an exact global maximizer of each player's concave utility, while the algorithm's inference step performs 5,000 SGD iterations and reports residual accuracy rather than a certificate of global optimality.

Editorial extensions

If this is right

  • Any direct mechanism that is exactly equivalent to a TEDI menu mechanism is dominant-strategy incentive compatible and individually rational.
  • The class of mechanisms TEDI can represent achieves the same supreme expected utility as the class of all truthful mechanisms, so no truthful mechanism is out of reach by approximation.
  • TEDI removes outcome discretization: the optimal Bernoulli digital-goods mechanism is expressed with $O(m)$ parameters instead of $\Omega(2^m)$ menu entries.
  • Training only needs an unbiased gradient estimator and samples from the current distribution, so standard first-order optimizers such as Adam can be used.
  • In the studied reproducible-goods auctions, TEDI reaches about 98 percent of optimal revenue in the ten-good Bernoulli case and matches or exceeds all compared learning-based and classical baselines.

Reading between the lines

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

  • The truthfulness certificate applies to the exact menu mechanism; a deployment that replaces exact global maximization in the inference step with approximate SGD is not covered by the certificate, so exact convex optimization or a certificate of optimality would be needed to guarantee incentive compatibility in practice.
  • The covariance-trick identity is not specific to auctions: any problem where players choose from a continuous set according to a log-concave Gibbs distribution and the designer's payoff is differentiable can be trained this way, which suggests applications beyond mechanism design, such as Stackelberg or contract settings.
  • The continuous-sampling step is the computational bottleneck; replacing Langevin re-sampling with a faster or amortized sampler for log-concave distributions, or finding a Partial GroupMax variant with $O(m^2)$ rather than $O(m^3)$ complexity, would improve scalability.
  • The paper limits TEDI to outcome-separable problems; extending the menu representation to unit-supply auctions with hard per-good capacity constraints would require additional feasibility handling beyond the penalties discussed in the paper.
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

3 major / 4 minor

Summary. The paper proposes TEDI, a discretization-free method for learning truthful mechanisms in outcome-separable quasi-linear domains. It represents a menu mechanism through pricing rules parameterized by a new Partial GroupMax Network, proves that such networks universally approximate partially convex continuous functions, derives an exact covariance-trick identity for gradients of the Gibbs-relaxed objective, and introduces a Langevin-based continuous sampling scheme. The authors claim truthfulness, full expressiveness, and dimension-insensitivity for TEDI, and support the claims with experiments in reproducible-goods auctions, reporting strong performance relative to discretization-based baselines such as MenuNet, GemNet, and LotteryAMA.

Significance. If the guarantees were fully realized by the implemented algorithm, this would be a notable advance: it would provide the first network-based universal approximator for partially convex functions, a parameter-efficient representation of optimal mechanisms in a class of multi-dimensional auction settings, and a genuinely discretization-free route through differentiable economics. The covariance identity and the menu-mechanism equivalence theorem are elegant and appear correct under the stated assumptions. The experimental comparison is informative and includes useful ablations. However, the central truthfulness and expressiveness claims are established only for an idealized exact-argmax version of the mechanism, while the deployed inference is a finite heuristic; this gap is load-bearing and currently prevents the paper's headline claims from being fully supported.

major comments (3)
  1. [§3.2, Algorithm 6, Corollary 3.7]
  2. [§3.3.5, Algorithm 1]
  3. [§3.4, Proposition 3.8 / Theorem B.10]
minor comments (4)
  1. [Abstract and §1.1]
  2. [Algorithm 6]
  3. [Appendix B.3.2, Figure 4]
  4. [Appendix D.1.2]

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: TEDI's truthfulness, full-expressiveness, and dimension-insensitivity are genuine theorems with self-contained proofs; the two co-author self-citations are not load-bearing.

full rationale

TEDI's derivation chain is self-contained. Corollary 3.7's truthfulness claim rests on Theorem 2.9, which is proved in Appendix C.2 by an explicit Fenchel-conjugate construction from an arbitrary DSIC mechanism to a partial-convex, no-buy-no-pay menu mechanism, together with a direct verification that any equivalent direct mechanism satisfies the textbook IC/IR inequalities of Definition 2.3; truthfulness is defined independently of the construction, so the guarantee is a theorem rather than a definitional restatement. Full expressiveness (Theorem B.10) is an approximation result chained from Theorem 2.9, Proposition B.5, and Lemmas B.8 and B.9; the only imported ingredients are external, classical results: PMA universality from Kim and Kim [67] and neural-network universality from Cybenko [30], neither involving the present authors. The covariance trick (Proposition 3.6) is a genuine log-derivative identity with a proof in Appendix C.1, and the O(m) parameterization of Example 1.1 (Example B.11) is an expressiveness statement about the known linear optimal rule, not a fitted-then-predicted quantity. The two self-citations ([35], [36]) appear only in related-work and future-work passages and are not load-bearing for any central claim. The main caveat, flagged in the paper at Appendix D.1.2, is that Corollary 3.7 certifies only direct mechanisms exactly equivalent to a TEDI menu (exact argmax per Definition 2.6), whereas Algorithm 6 finds the argmax by 5,000 gradient steps and reports residual utility accuracy 'less than 10^{-6}' rather than a global-optimality certificate; a sub-optimal inference point can create an incentive-compatibility gap equal to the utility shortfall. This is an implementation-correctness risk, not a circular reduction: no parameter is fitted to the target quantity and relabeled as a prediction, and no equation reduces to its own input by construction.

Assumptions & free parameters 3 free parameters · 7 assumptions · 3 invented entities

The central claim rests on standard mechanism-design assumptions (quasi-linearity, compact separable domains), two mild regularity assumptions on u0 and F, and the untested assumption that the finite-step inference procedure finds the exact argmax. The free parameters are training and architecture hyperparameters rather than parameters fitted to match the target result, so the circularity burden is low. The invented entities are mathematical constructions and a neural architecture, not new physical phenomena.

free parameters (3)
  • temperature beta for soft argmax relaxation = beta_max = 512 or 4096
    Program (8) equals the original argmax program only as beta goes to infinity; finite beta is a hand-chosen approximation parameter affecting the objective.
  • Langevin step size eta and inner iterations S = eta_0 = 0.03 down to 0.01; S = 2 for n=1, S = 16 for n>1
    Continuous sampling produces approximate samples from the target distribution; the bias of the gradient estimator depends on these hyperparameters.
  • Partial GroupMax width and depth = h_x = 64*(m+3), G = 2*(m+3), 1-2 layers
    The expressiveness and dimension-insensitivity experiments rely on these chosen sizes; no theorem states the required width for a given accuracy.
assumptions (7)
  • domain assumption Players have quasi-linear additive utilities ui(xi,pi;ti)=<xi,ti>+ci(xi)-pi with ci(0)=0
    Section 2 and Appendix A; used throughout for the menu-mechanism equivalence and for the Fenchel-conjugate construction in Theorem C.2.
  • domain assumption Type and outcome spaces are compact convex sets, and the global outcome space is separable as X=×Xi
    Appendix A and Appendix E.1; needed for universal approximation on compact domains and for menu mechanisms to always produce feasible global outcomes.
  • domain assumption Designer utility u0 is non-decreasing in each payment (Assumption 1)
    Appendix B.5, Lemma B.4; used to normalize no-buy-no-pay* without reducing expected utility.
  • domain assumption The type distribution F is non-degenerate (Assumption 2)
    Appendix B.5, Lemma B.9; used to transfer l1 closeness of direct mechanisms into closeness of expected designer utility.
  • standard math Standard universal approximation theorems for continuous functions and PMA (Cybenko; Kim and Kim)
    Proof of Theorem B.2 relies on Lemma C.6 to embed PMA inside Partial GroupMax and on Cybenko for Parameterized Affine Network approximation.
  • standard math Interchange of differentiation and expectation/integration in the covariance trick
    Appendix C.1, Proposition 3.6; requires regularity conditions on u0 and the Gibbs distribution that are not stated explicitly.
  • ad hoc to paper Exact global maximization in the inference step
    Corollary 3.7 certifies only direct mechanisms exactly equivalent to the menu mechanism; Algorithm 6 uses finite-step SGD with no global-optimality certificate.
invented entities (3)
  • Pricing rule
    purpose: A continuous function mapping local outcomes to prices, replacing the finite menu of discretization-based mechanisms.
    A mathematical modeling device; its validity rests on Theorem 2.9 rather than on any external falsifiable prediction.
  • Menu mechanism represented by pricing rules
    purpose: A multi-player mechanism in which each player chooses the outcome maximizing utility under a continuous price function.
    A reformulation of standard menu mechanisms with continuous outcome spaces; no independent empirical handle is provided beyond the paper's experiments.
  • Partial GroupMax Network
    purpose: A neural architecture claimed to be a universal approximator of continuous functions that are convex in part of their inputs.
    A new architecture; its usefulness is supported by the paper's own experiments but it has no external falsifiable handle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning Truthful Mechanisms without Discretization." pith.science (2026). https://pith.science/paper/PJKT3Y4E

@misc{pith2026250622911,
  author       = {Pith},
  title        = {Pith review of: Learning Truthful Mechanisms without Discretization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PJKT3Y4E}},
  note         = {Machine review of arXiv:2506.22911}
}
read the original abstract

This paper introduces TEDI (Truthful, Expressive, and Dimension-Insensitive approach), a discretization-free algorithm to learn truthful and utility-maximizing mechanisms. Existing learning-based approaches often rely on discretization of outcome spaces to ensure truthfulness, which leads to inefficiency with increasing problem size. To address this limitation, we formalize the concept of pricing rules, defined as functions that map outcomes to prices. Based on this concept, we propose a novel menu mechanism, which can be equivalent to a truthful direct mechanism under specific conditions. The core idea of TEDI lies in its parameterization of pricing rules using Partial GroupMax Network, a new network architecture designed to universally approximate partial convex functions. To learn optimal pricing rules, we develop novel training techniques, including covariance trick and continuous sampling, to derive unbiased gradient estimators compatible with first-order optimization. Theoretical analysis establishes that TEDI guarantees truthfulness, full expressiveness, and dimension-insensitivity. Experimental evaluation in the studied auction setting demonstrates that TEDI achieves strong performance, competitive with or exceeding state-of-the-art methods. This work presents the first approaches to learn truthful mechanisms without outcome discretization, thereby enhancing algorithmic efficiency. The proposed concepts, network architecture, and learning techniques might offer potential value and provide new insights for automated mechanism design and differentiable economics.

Figures

Figures reproduced from arXiv: 2506.22911 by the authors.

Figure 1
Figure 1. (a): A visualization of the optimal mechanism (denoted as OPT) and the mechanism [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. An illustration of procedure of TEDI. Best viewed in color (same for the remaining [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. The architecture of a k-layer Partial GroupMax Network and its sub-components. Given an input y, a Parameterized Affine Network produces parameters W(y) ∈ R d out×d in and b(y) ∈ R d out , which can perform an affine transformation on a vector z ∈ R d in as: W(y) · z + b(y). Sometimes W(y) is required to be positive, then we implement an element-wise softplus function3 to ensure the positivity. The architecture of a… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Using Partial GroupMax Network to express a PMA. [PITH_FULL_IMAGE:figures/full_fig_p043_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

136 extracted references · 69 canonical work pages

  1. [1]

    Towards data auctions with externalities

    Anish Agarwal et al. “Towards data auctions with externalities”. In: Games and Economic Behavior 148 (2024), pp. 323–356

    2024

  2. [2]

    Automated design of robust mecha- nisms

    Michael Albert, Vincent Conitzer, and Peter Stone. “Automated design of robust mecha- nisms”. In: Proceedings of the AAAI Conference on Artificial Intelligence . Vol. 31. 1. 2017

    2017

  3. [3]

    Backpropagation and stochastic gradient descent method

    Shun-ichi Amari. “Backpropagation and stochastic gradient descent method”. In: Neuro- computing 5.4-5 (1993), pp. 185–196

    1993

  4. [4]

    Input convex neural networks

    Brandon Amos, Lei Xu, and J Zico Kolter. “Input convex neural networks”. In: International conference on machine learning . PMLR. 2017, pp. 146–155

    2017

  5. [5]

    Near-optimal max-affine estimators for convex regression

    G´ abor Bal´ azs, Andr´ as Gy¨ orgy, and Csaba Szepesv´ ari. “Near-optimal max-affine estimators for convex regression”. In: Artificial Intelligence and Statistics . PMLR. 2015, pp. 56–64

    2015

  6. [6]

    Sample complexity of automated mechanism design

    Maria-Florina F Balcan, Tuomas Sandholm, and Ellen Vitercik. “Sample complexity of automated mechanism design”. In: Advances in Neural Information Processing Systems 29 (2016)

    2016

  7. [7]

    MAC advice for facility loca- tion mechanism design

    Zohar Barak, Anupam Gupta, and Inbal Talgam-Cohen. “MAC advice for facility loca- tion mechanism design”. In: Advances in Neural Information Processing Systems 37 (2024), pp. 129564–129604

    2024

  8. [8]

    Universal approximation bounds for superpositions of a sigmoidal func- tion

    Andrew R Barron. “Universal approximation bounds for superpositions of a sigmoidal func- tion”. In: IEEE Transactions on Information theory 39.3 (1993), pp. 930–945

    1993

Show all 136 references
  1. [9]

    Dynamic programming

    Richard Bellman. “Dynamic programming”. In: science 153.3731 (1966), pp. 34–37

    1966

  2. [10]

    The curse of highly variable functions for local kernel machines

    Yoshua Bengio, Olivier Delalleau, and Nicolas Roux. “The curse of highly variable functions for local kernel machines”. In: Advances in neural information processing systems 18 (2005)

    2005

  3. [11]

    Methodology for Designing Reasonably Expressive Mechanisms with Application to Ad Auctions

    Michael Benisch, Norman M Sadeh, and Tuomas Sandholm. “Methodology for Designing Reasonably Expressive Mechanisms with Application to Ad Auctions.” In: IJCAI. 2009, pp. 46–52

    2009

  4. [12]

    Welfare and profit maximization with production costs

    Avrim Blum et al. “Welfare and profit maximization with production costs”. In: 2011 IEEE 52nd Annual Symposium on Foundations of Computer Science . IEEE. 2011, pp. 77–86

    2011

  5. [13]

    Large-scale machine learning with stochastic gradient descent

    L´ eon Bottou. “Large-scale machine learning with stochastic gradient descent”. In: Pro- ceedings of COMPSTAT’2010: 19th International Conference on Computational Statistic- sParis France, August 22-27, 2010 Keynote, Invited and Contributed Papers. Springer. 2010, pp. 177–186

    2010

  6. [14]

    Convex optimization

    Stephen Boyd and Lieven Vandenberghe. Convex optimization. Cambridge university press, 2004

    2004

  7. [15]

    An Introduction to the Theory of Mechanism Design

    Tilman B¨ orgers, Daniel Kr¨ ahmer, and Roland Strausz. An Introduction to the Theory of Mechanism Design. Oxford University Press, July 2015. isbn: 9780199734023

    2015

  8. [16]

    Optimal multi-dimensional mechanism design: Reducing revenue to welfare maximization

    Yang Cai, Constantinos Daskalakis, and S Matthew Weinberg. “Optimal multi-dimensional mechanism design: Reducing revenue to welfare maximization”. In: 2012 IEEE 53rd Annual Symposium on Foundations of Computer Science . IEEE. 2012, pp. 130–139

    2012

  9. [17]

    Log-sum-exp neural net- works and posynomial models for convex and log-log-convex data

    Giuseppe C Calafiore, Stephane Gaubert, and Corrado Possieri. “Log-sum-exp neural net- works and posynomial models for convex and log-log-convex data”. In: IEEE transactions on neural networks and learning systems 31.3 (2019), pp. 827–838. 19

    2019

  10. [18]

    Truthful implementation and preference aggregation in restricted domains

    Juan Carlos Carbajal, Andrew McLennan, and Rabee Tourky. “Truthful implementation and preference aggregation in restricted domains”. In: Journal of Economic Theory 148.3 (2013), pp. 1074–1101

    2013

  11. [19]

    Mechanism Design for Facility Location Problem: A Survey

    Hau Chan et al. “Mechanism Design for Facility Location Problem: A Survey”. In: The 30th International Joint Conference on Artificial Intelligence (IJCAI 2021) . 2021, pp. 1–17

    2021

  12. [20]

    Optimal competitive auctions

    Ning Chen, Nick Gravin, and Pinyan Lu. “Optimal competitive auctions”. In: Proceedings of the forty-sixth annual ACM symposium on Theory of computing . 2014, pp. 253–262

    2014

  13. [21]

    The complexity of optimal multidimensional pricing

    Xi Chen et al. “The complexity of optimal multidimensional pricing”. In: Proceedings of the twenty-fifth annual ACM-SIAM symposium on Discrete algorithms . SIAM. 2014, pp. 1319– 1328

    2014

  14. [22]

    Strategy-proofness and “median voters

    Stephen Ching. “Strategy-proofness and “median voters””. In: International Journal of Game Theory 26 (1997), pp. 473–490

    1997

  15. [23]

    Multipart pricing of public goods

    Edward H Clarke. “Multipart pricing of public goods”. In: Public choice (1971), pp. 17–33

    1971

  16. [24]

    Incremental mechanism design

    Vincent Conitzer and Tuomas W Sandholm. “Incremental mechanism design”. In: (2007)

    2007

  17. [25]

    Differentiable economics for ran- domized affine maximizer auctions

    Michael Curry, Tuomas Sandholm, and John Dickerson. “Differentiable economics for ran- domized affine maximizer auctions”. In: Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence . 2023, pp. 2633–2641

    2023

  18. [26]

    Automated design of affine maximizer mechanisms in dynamic set- tings

    Michael Curry et al. “Automated design of affine maximizer mechanisms in dynamic set- tings”. In: Proceedings of the AAAI Conference on Artificial Intelligence . Vol. 38. 9. 2024, pp. 9626–9635

    2024

  19. [27]

    Certifying strategyproof auction networks

    Michael Curry et al. “Certifying strategyproof auction networks”. In: Advances in Neural Information Processing Systems 33 (2020), pp. 4987–4998

    2020

  20. [28]

    Optimal Automated Market Makers: Differentiable Economics and Strong Duality

    Michael J Curry, Zhou Fan, and David C Parkes. “Optimal Automated Market Makers: Differentiable Economics and Strong Duality”. In: arXiv preprint arXiv:2402.09129 (2024)

    2024 arXiv

  21. [29]

    Learning revenue-maximizing auctions with differentiable matching

    Michael J Curry et al. “Learning revenue-maximizing auctions with differentiable matching”. In: International Conference on Artificial Intelligence and Statistics. PMLR. 2022, pp. 6062– 6073

    2022

  22. [30]

    Approximation by superpositions of a sigmoidal function

    George Cybenko. “Approximation by superpositions of a sigmoidal function”. In: Mathe- matics of control, signals and systems 2.4 (1989), pp. 303–314

    1989

  23. [31]

    Strong duality for a multiple-good monopolist

    Constantinos Daskalakis, Alan Deckelbaum, and Christos Tzamos. “Strong duality for a multiple-good monopolist”. In: Proceedings of the Sixteenth ACM Conference on Economics and Computation . 2015, pp. 449–450

    2015

  24. [32]

    Identifying and attacking the saddle point problem in high-dimensional non-convex optimization

    Yann N Dauphin et al. “Identifying and attacking the saddle point problem in high-dimensional non-convex optimization”. In: Advances in neural information processing systems 27 (2014)

    2014

  25. [33]

    Continuity properties of Paretian utility

    Gerard Debreu. “Continuity properties of Paretian utility”. In: International Economic Re- view 5.3 (1964), pp. 285–293

    1964

  26. [34]

    Procurement Auctions via Approximately Optimal Submodular Opti- mization

    Yuan Deng et al. “Procurement Auctions via Approximately Optimal Submodular Opti- mization”. In: arXiv preprint arXiv:2411.13513 (2024)

    2024 arXiv

  27. [35]

    A context-integrated transformer-based neural network for auction design

    Zhijian Duan et al. “A context-integrated transformer-based neural network for auction design”. In: International Conference on Machine Learning . PMLR. 2022, pp. 5609–5626

    2022

  28. [36]

    A scalable neural network for DSIC affine maximizer auction design

    Zhijian Duan et al. “A scalable neural network for DSIC affine maximizer auction design”. In: Advances in Neural Information Processing Systems 36 (2024). 20

    2024

  29. [37]

    Mechanism design for large language models

    Paul Duetting et al. “Mechanism design for large language models”. In: Proceedings of the ACM on Web Conference 2024 . 2024, pp. 144–155

    2024

  30. [38]

    Deep reinforcement learning in large discrete action spaces

    Gabriel Dulac-Arnold et al. “Deep reinforcement learning in large discrete action spaces”. In: arXiv preprint arXiv:1512.07679 (2015)

    2015 arXiv

  31. [39]

    Optimal auctions through deep learning

    Paul D¨ utting et al. “Optimal auctions through deep learning”. In: International Conference on Machine Learning . PMLR. 2019, pp. 1706–1715

    2019

  32. [40]

    Optimal auctions through deep learning: Advances in differentiable economics

    Paul D¨ utting et al. “Optimal auctions through deep learning: Advances in differentiable economics”. In: Journal of the ACM 71.1 (2024), pp. 1–53

    2024

  33. [41]

    Reverse Auction Relinquishing Broadcast Spectrum Rights

    Federal Communications Commission. Reverse Auction Relinquishing Broadcast Spectrum Rights. 2016. url: https://www.fcc.gov/auction/1001

    2016

  34. [42]

    College admissions and the stability of marriage

    David Gale and Lloyd S Shapley. “College admissions and the stability of marriage”. In: The American Mathematical Monthly 69.1 (1962), pp. 9–15

    1962

  35. [43]

    Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images

    Stuart Geman and Donald Geman. “Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images”. In: IEEE Transactions on pattern analysis and machine intelligence 6 (1984), pp. 721–741

    1984

  36. [44]

    Reverse auctions are different from auctions

    Matthias Gerstgrasser. “Reverse auctions are different from auctions”. In: Information Pro- cessing Letters 147 (2019), pp. 49–54. issn: 0020-0190. doi: https://doi.org/10.1016/j.ipl. 2019.03.006. url: https://www.sciencedirect.com/science/article/pii/S0020019019300560

    2019 doi

  37. [45]

    Duality and optimality of auctions for uniform distributions

    Yiannis Giannakopoulos and Elias Koutsoupias. “Duality and optimality of auctions for uniform distributions”. In: Proceedings of the fifteenth ACM conference on Economics and computation. 2014, pp. 259–276

    2014

  38. [46]

    Manipulation of voting schemes: a general result

    Allan Gibbard. “Manipulation of voting schemes: a general result”. In: Econometrica: journal of the Econometric Society (1973), pp. 587–601

    1973

  39. [47]

    Competitive auctions for multiple digital goods

    Andrew V Goldberg and Jason D Hartline. “Competitive auctions for multiple digital goods”. In: European Symposium on Algorithms. Springer. 2001, pp. 416–427

    2001

  40. [48]

    A lower bound on the competitive ratio of truthful auctions

    Andrew V Goldberg et al. “A lower bound on the competitive ratio of truthful auctions”. In: STACS 2004: 21st Annual Symposium on Theoretical Aspects of Computer Science, Montpellier, France, March 25-27, 2004. Proceedings 21 . Springer. 2004, pp. 644–655

    2004

  41. [49]

    Deep learning for multi- facility location mechanism design

    Noah Golowich, Harikrishna Narasimhan, and David C. Parkes. “Deep learning for multi- facility location mechanism design”. In: IJCAI’18. Stockholm, Sweden: AAAI Press, 2018. isbn: 9780999241127

    2018

  42. [50]

    Deep Learning

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep Learning. http://www.deeplearningbook. org. MIT Press, 2016

    2016

  43. [51]

    Approximation guarantees of Median Mechanism in Rd

    Nick Gravin and Jianhao Jia. “Approximation guarantees of Median Mechanism in Rd”. In: ACM Symposium on Theory of Computing (2025)

    2025

  44. [52]

    Incentives in teams

    Theodore Groves. “Incentives in teams”. In: Econometrica: Journal of the Econometric So- ciety (1973), pp. 617–631

    1973

  45. [53]

    Settling the sample complexity of single- parameter revenue maximization

    Chenghao Guo, Zhiyi Huang, and Xinzhi Zhang. “Settling the sample complexity of single- parameter revenue maximization”. In: Proceedings of the 51st Annual ACM SIGACT Sym- posium on Theory of Computing . 2019, pp. 662–673

    2019

  46. [54]

    Computationally feasible automated mechanism design: General approach and case studies

    Mingyu Guo and Vincent Conitzer. “Computationally feasible automated mechanism design: General approach and case studies”. In: Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 24. 1. 2010, pp. 1676–1679. 21

    2010

  47. [55]

    Optimizing affine maximizer auctions via linear programming: an application to revenue maximizing mechanism design for zero-day exploits markets

    Mingyu Guo, Hideaki Hata, and Ali Babar. “Optimizing affine maximizer auctions via linear programming: an application to revenue maximizing mechanism design for zero-day exploits markets”. In: PRIMA 2017: Principles and Practice of Multi-Agent Systems: 20th Interna- tional Con...

    2017

  48. [56]

    Prior-Independent Auctions for Heterogeneous Bidders

    Guru Guruganesh et al. “Prior-Independent Auctions for Heterogeneous Bidders”. In: Pro- ceedings of the 2024 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA). SIAM. 2024, pp. 1–18

    2024

  49. [57]

    Automated on- line mechanism design and prophet inequalities

    Mohammad Taghi Hajiaghayi, Robert Kleinberg, and Tuomas Sandholm. “Automated on- line mechanism design and prophet inequalities”. In: AAAI. Vol. 7. 2007, pp. 58–65

    2007

  50. [58]

    Straightforward individual incentive compatibility in large economies

    Peter J Hammond. “Straightforward individual incentive compatibility in large economies”. In: The Review of Economic Studies 46.2 (1979), pp. 263–282

    1979

  51. [59]

    Universal approximation of symmetric and anti-symmetric functions

    Jiequn Han et al. “Universal approximation of symmetric and anti-symmetric functions”. In: Communications in Mathematical Sciences 20.5 (2022), pp. 1397–1408

    2022

  52. [60]

    Profit maximization in mechanism design

    Jason Hartline and Anna Karlin. “Profit maximization in mechanism design”. In: Algorith- mic Game Theory 331 (2007)

    2007

  53. [61]

    Gaussian Error Linear Units (GELUs)

    Dan Hendrycks and Kevin Gimpel. Gaussian Error Linear Units (GELUs) . 2023. arXiv: 1606.08415 [cs.LG]. url: https://arxiv.org/abs/1606.08415

    2023 arXiv

  54. [62]

    Welfare maximization with production costs: A primal dual approach

    Zhiyi Huang and Anthony Kim. “Welfare maximization with production costs: A primal dual approach”. In: Games and Economic Behavior 118 (2019), pp. 648–667

    2019

  55. [63]

    Optimal-er auctions through attention

    Dmitry Ivanov et al. “Optimal-er auctions through attention”. In: Advances in Neural In- formation Processing Systems 35 (2022), pp. 34734–34747

    2022

  56. [64]

    Posted Price Mechanisms for Online Allocation with Disec- onomies of Scale

    Hossein Nekouyan Jazi et al. “Posted Price Mechanisms for Online Allocation with Disec- onomies of Scale”. In: Proceedings of the ACM on Web Conference 2025 . 2025, pp. 2710– 2728

    2025

  57. [65]

    An Online Intelligent Task Pricing Mechanism Based on Reverse Auction in Mobile Crowdsensing Networks for the Internet of Things

    Bing Jia et al. “An Online Intelligent Task Pricing Mechanism Based on Reverse Auction in Mobile Crowdsensing Networks for the Internet of Things”. In: 2021 IEEE Symposium Series on Computational Intelligence (SSCI) . IEEE. 2021, pp. 1–6

    2021

  58. [66]

    Estimation of particle transmission by random sampling

    Herman Kahn and Theodore E Harris. “Estimation of particle transmission by random sampling”. In: National Bureau of Standards applied mathematics series 12 (1951), pp. 27– 30

    1951

  59. [67]

    Parameterized convex universal approximators for decision- making problems

    Jinrae Kim and Youdan Kim. “Parameterized convex universal approximators for decision- making problems”. In: IEEE Transactions on Neural Networks and Learning Systems 35.2 (2022), pp. 2448–2459

    2022

  60. [68]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A Method for Stochastic Optimization . 2017. arXiv: 1412.6980 [cs.LG]

    2017 arXiv

  61. [69]

    Bayesian estimates of equation system parameters: an application of integration by Monte Carlo

    Teun Kloek and Herman K Van Dijk. “Bayesian estimates of equation system parameters: an application of integration by Monte Carlo”. In: Econometrica: Journal of the Econometric Society (1978), pp. 1–19

    1978

  62. [70]

    Total-cost procurement auctions: Impact of suppliers’ cost adjustments on auction format choice

    Dimitris Kostamis, Damian R Beil, and Izak Duenyas. “Total-cost procurement auctions: Impact of suppliers’ cost adjustments on auction format choice”. In: Management Science 55.12 (2009), pp. 1985–1999. 22

    2009

  63. [71]

    Faster first-order methods for extensive-form game solving

    Christian Kroer et al. “Faster first-order methods for extensive-form game solving”. In: Pro- ceedings of the Sixteenth ACM Conference on Economics and Computation . 2015, pp. 817– 834

    2015

  64. [72]

    Towards a characterization of truthful combi- natorial auctions

    Ron Lavi, Ahuva Mu’Alem, and Noam Nisan. “Towards a characterization of truthful combi- natorial auctions”. In: 44th Annual IEEE Symposium on Foundations of Computer Science,

  65. [73]

    Two simplified proofs for Roberts’ theorem

    Ron Lavi, Ahuva Mu’alem, and Noam Nisan. “Two simplified proofs for Roberts’ theorem”. In: Social Choice and Welfare 32.3 (2009), pp. 407–423

    2009

  66. [74]

    Truthful and near-optimal mechanism design via linear programming

    Ron Lavi and Chaitanya Swamy. “Truthful and near-optimal mechanism design via linear programming”. In: Journal of the ACM (JACM) 58.6 (2011), pp. 1–24

    2011

  67. [75]

    Deep learning

    Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. “Deep learning”. In: nature 521.7553 (2015), pp. 436–444

    2015

  68. [76]

    Set transformer: A framework for attention-based permutation-invariant neural networks

    Juho Lee et al. “Set transformer: A framework for attention-based permutation-invariant neural networks”. In: International conference on machine learning. PMLR. 2019, pp. 3744– 3753

    2019

  69. [77]

    Approximating revenue-maximizing combi- natorial auctions

    Alexander Likhodedov and Tuomas Sandholm. “Approximating revenue-maximizing combi- natorial auctions”. In: Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 5. 2005, pp. 267–274

    2005

  70. [78]

    Bundling as an optimal selling mechanism for a multiple-good monopolist

    Alejandro M Manelli and Daniel R Vincent. “Bundling as an optimal selling mechanism for a multiple-good monopolist”. In: Journal of Economic Theory 127.1 (2006), pp. 1–35

    2006

  71. [79]

    Microeconomic theory

    Andreu Mas-Colell, Michael Dennis Whinston, Jerry R Green, et al. Microeconomic theory. Vol. 1. Oxford University Press New York, 1995

    1995

  72. [80]

    The maximum numbers of faces of a convex polytope

    Peter McMullen. “The maximum numbers of faces of a convex polytope”. In: Mathematika 17.2 (1970), pp. 179–184

    1970

  73. [81]

    Equation of state calculations by fast computing machines

    Nicholas Metropolis et al. “Equation of state calculations by fast computing machines”. In: The journal of chemical physics 21.6 (1953), pp. 1087–1092

    1953

  74. [82]

    Envelope theorems for arbitrary choice sets

    Paul Milgrom and Ilya Segal. “Envelope theorems for arbitrary choice sets”. In: Economet- rica 70.2 (2002), pp. 583–601

    2002

  75. [83]

    A theory of auctions and competitive bidding

    Paul R Milgrom and Robert J Weber. “A theory of auctions and competitive bidding”. In: Econometrica: Journal of the Econometric Society (1982), pp. 1089–1122

    1982

  76. [84]

    On strategy-proofness and single peakedness

    Herv´ e Moulin. “On strategy-proofness and single peakedness”. In:Public Choice 35.4 (1980), pp. 437–455

    1980

  77. [85]

    Incentive compatibility and the bargaining problem

    Roger B Myerson. “Incentive compatibility and the bargaining problem”. In: Econometrica: journal of the Econometric Society (1979), pp. 61–73

    1979

  78. [86]

    Optimal auction design

    Roger B Myerson. “Optimal auction design”. In: Mathematics of operations research 6.1 (1981), pp. 58–73

    1981

  79. [87]

    Automated mech- anism design without money via machine learning

    Harikrishna Narasimhan, Shivani Brinda Agarwal, and David C Parkes. “Automated mech- anism design without money via machine learning”. In: Proceedings of the 25th International Joint Conference on Artificial Intelligence . 2016

    2016

  80. [88]

    Affine maximizers in domains with selfish valuations

    Swaprava Nath and Arunava Sen. “Affine maximizers in domains with selfish valuations”. In: ACM Transactions on Economics and Computation (TEAC) 3.4 (2015), pp. 1–19. 23

    2015

  81. [89]

    Sur une g´ en´ eralisation des int´ egrales de MJ Radon

    Otton Nikodym. “Sur une g´ en´ eralisation des int´ egrales de MJ Radon”. In: Fundamenta Mathematicae 15.1 (1930), pp. 131–179

    1930

  82. [90]

    Algorithmic Game Theory

    Noam Nisan et al. “Algorithmic Game Theory”. In: 2007. url: https://api.semanticscholar. org/CorpusID:239540

    2007

  83. [91]

    Optimal mechanism for selling two goods

    Gregory Pavlov. “Optimal mechanism for selling two goods”. In: The BE Journal of Theo- retical Economics 11.1 (2011), p. 0000102202193517041664

    2011

  84. [92]

    Preferencenet: Encoding human preferences in auction design with deep learning

    Neehar Peri et al. “Preferencenet: Encoding human preferences in auction design with deep learning”. In: Advances in Neural Information Processing Systems 34 (2021), pp. 17532– 17542

    2021

  85. [93]

    Why and when can deep-but not shallow-networks avoid the curse of dimensionality: a review

    Tomaso Poggio et al. “Why and when can deep-but not shallow-networks avoid the curse of dimensionality: a review”. In: International Journal of Automation and Computing 14.5 (2017), pp. 503–519

    2017

  86. [94]

    Benefits of permutation-equivariance in auction mechanisms

    Tian Qin et al. “Benefits of permutation-equivariance in auction mechanisms”. In: Advances in Neural Information Processing Systems 35 (2022), pp. 18131–18142

    2022

  87. [95]

    Auction learning as a two-player game

    Jad Rahme, Samy Jelassi, and S Matthew Weinberg. “Auction learning as a two-player game”. In: arXiv preprint arXiv:2006.05684 (2020)

    2020 arXiv

  88. [96]

    Auction Learning as a Two-Player Game

    Jad Rahme, Samy Jelassi, and S. Matthew Weinberg. “Auction Learning as a Two-Player Game”. In: International Conference on Learning Representations . 2021. url: https : / / openreview.net/forum?id=YHdeAO61l6T

    2021

  89. [97]

    A permutation-equivariant neural network architecture for auction de- sign

    Jad Rahme et al. “A permutation-equivariant neural network architecture for auction de- sign”. In: Proceedings of the AAAI Conference on Artificial Intelligence . Vol. 35. 2021, pp. 5664–5672

    2021

  90. [98]

    Deep learning for two-sided matching

    Sai Srivatsa Ravindranath et al. “Deep learning for two-sided matching”. In: arXiv preprint arXiv:2107.03427 (2021)

    2021 arXiv

  91. [99]

    url: https : / / www

    Reproducible and Non-Reproducible Goods - Okpedia . url: https : / / www . okpedia . com / reproducible-and-non-reproducible-goods (visited on 06/09/2025)

    2025

  92. [100]

    Exponential convergence of Langevin distribu- tions and their discrete approximations

    Gareth O Roberts and Richard L Tweedie. “Exponential convergence of Langevin distribu- tions and their discrete approximations”. In: (1996)

    1996

  93. [101]

    The characterization of implementable choice rules

    Kevin Roberts. “The characterization of implementable choice rules”. In: Aggregation and revelation of preferences 12.2 (1979), pp. 321–348

    1979

  94. [102]

    A necessary and sufficient condition for rationalizability in a quasi- linear context

    Jean-Charles Rochet. “A necessary and sufficient condition for rationalizability in a quasi- linear context”. In: Journal of mathematical Economics 16.2 (1987), pp. 191–200

    1987

  95. [103]

    Automated mechanism design: A new application area for search al- gorithms

    Tuomas Sandholm. “Automated mechanism design: A new application area for search al- gorithms”. In: International Conference on Principles and Practice of Constraint Program- ming. Springer. 2003, pp. 19–36

    2003

  96. [104]

    Automated Design of Revenue-Maximizing Combinatorial Auctions

    Tuomas Sandholm and Anton Likhodedov. “Automated Design of Revenue-Maximizing Combinatorial Auctions”. In: Operations Research 63.5 (2015), pp. 1000–1025

    2015

  97. [105]

    Strategy-proofness and Arrow’s conditions: Existence and cor- respondence theorems for voting procedures and social welfare functions

    Mark Allen Satterthwaite. “Strategy-proofness and Arrow’s conditions: Existence and cor- respondence theorems for voting procedures and social welfare functions”. In: Journal of economic theory 10.2 (1975), pp. 187–217

    1975

  98. [106]

    On cores and indivisibility

    Lloyd Shapley and Herbert Scarf. “On cores and indivisibility”. In: Journal of mathematical economics 1.1 (1974), pp. 23–37. 24

    1974

  99. [107]

    Automated Mechanism Design via Neural Networks

    Weiran Shen, Pingzhong Tang, and Song Zuo. “Automated Mechanism Design via Neural Networks”. In: Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems. 2019, pp. 215–223

    2019

  100. [108]

    Electronic reverse auctions: Promises, risks and conditions for success

    Larry R Smeltzer and Amelia S Carr. “Electronic reverse auctions: Promises, risks and conditions for success”. In: Industrial Marketing Management 32.6 (2003), pp. 481–488

    2003

  101. [109]

    Score-based generative modeling through stochastic differential equations

    Yang Song et al. “Score-based generative modeling through stochastic differential equations”. In: arXiv preprint arXiv:2011.13456 (2020)

    2020 arXiv

  102. [110]

    Truthful Aggregation of LLMs with an Application to Online Advertising

    Ermis Soumalias, Michael J Curry, and Sven Seuken. “Truthful Aggregation of LLMs with an Application to Online Advertising”. In: arXiv preprint arXiv:2405.05905 (2024)

    2024 arXiv

  103. [111]

    Analysis and Optimization of Multi- Dimensional Percentile Mechanisms

    Xin Sui, Craig Boutilier, and Tuomas Sandholm. “Analysis and Optimization of Multi- Dimensional Percentile Mechanisms.” In: IJCAI. Citeseer. 2013, pp. 367–374

    2013

  104. [112]

    Mechanism Design for LLM Fine-tuning with Multiple Reward Models

    Haoran Sun et al. “Mechanism Design for LLM Fine-tuning with Multiple Reward Models”. In: arXiv preprint arXiv:2405.16276 (2024)

    2024

  105. [113]

    Reinforcement learning: An introduction

    Richard S Sutton, Andrew G Barto, et al. Reinforcement learning: An introduction. Vol. 1

  106. [114]

    Threshold policies with tight guarantees for online selection with convex costs

    Xiaoqi Tan et al. “Threshold policies with tight guarantees for online selection with convex costs”. In: ACM Transactions on Economics and Computation 13.2 (2025), pp. 1–49

    2025

  107. [115]

    MIT press Cambridge, 1998

    1998

  108. [116]

    Market structure and equilibrium

    Heinrich Von Stackelberg. Market structure and equilibrium . Springer Science & Business Media, 2010

    2010

  109. [117]

    Counterspeculation, auctions, and competitive sealed tenders

    William Vickrey. “Counterspeculation, auctions, and competitive sealed tenders”. In: The Journal of finance 16.1 (1961), pp. 8–37

    1961

  110. [118]

    Deep contract design via discontinuous networks

    Tonghan Wang et al. “Deep contract design via discontinuous networks”. In: Advances in Neural Information Processing Systems 36 (2024)

    2024

  111. [119]

    GemNet: Menu-Based, Strategy- Proof Multi-Bidder Auctions Through Deep Learning

    Tonghan Wang, Yanchen Jiang, and David C Parkes. “GemNet: Menu-Based, Strategy- Proof Multi-Bidder Auctions Through Deep Learning”. In: Proceedings of the 25th ACM Conference on Economics and Computation . 2024, p. 1100

    2024

  112. [120]

    Dominant-strategy versus bayesian multi-item auctions: Maximum revenue determination and comparison

    Andrew Chi-Chih Yao. “Dominant-strategy versus bayesian multi-item auctions: Maximum revenue determination and comparison”. In: Proceedings of the 2017 ACM Conference on Economics and Computation . 2017, pp. 3–20

    2017

  113. [121]

    The GroupMax neural network approximation of convex functions

    Xavier Warin. “The GroupMax neural network approximation of convex functions”. In: IEEE Transactions on Neural Networks and Learning Systems (2023)

    2023

  114. [122]

    Deep sets

    Manzil Zaheer et al. “Deep sets”. In: Advances in neural information processing systems 30 (2017)

    2017

  115. [123]

    Error bounds for approximations with deep ReLU networks

    Dmitry Yarotsky. “Error bounds for approximations with deep ReLU networks”. In: Neural networks 94 (2017), pp. 103–114

    2017

  116. [125]

    Automated dynamic mechanism design

    Hanrui Zhang and Vincent Conitzer. “Automated dynamic mechanism design”. In: Advances in Neural Information Processing Systems 34 (2021), pp. 27785–27797. 25 A Preface to the Appendix Before delving into appendix, we first make some changes in notations, which makes the techni...

    2021

  117. [126]

    M m satisfies partial convexity

  118. [127]

    M m satisfies no-buy-no-pay

  119. [128]

    xd i (t) ∈ argmaxxi∈Xivi(xi; ti) − pm i (xi; t−i)

  120. [129]

    Proof of (1)

    pd i (t) = pm i (xd i (t); t−i). Proof of (1). Note that pm i (xi; t−i)−ci(xi) = supti∈Ti⟨ti, xi⟩− ˜ud i (t) is exactly the Fenchel conjugate of −˜ud i (t) w.r.t. ti. Therefore, it’s convex by nature of Fenchel conjugate [14]. This completes the partial convexity. Proof of (2)...

  121. [130]

    ui j(xi; t−i) is convex on xi

  122. [131]

    0 ≤ ∇xiui j(xi; t−i) ≤ ¯ti

  123. [132]

    l∞(ui 1(·; t−i), ui 2(·; t−i)) ≤ δ

  124. [133]

    xi j(ti; t−i) ∈ argmaxxi∈Xi⟨xi, ti⟩ −ui j(xi; t−i), where xi j(t) is the i’th element of xj(t). Among them, statement (1) is true because uj(x) − ⟨x−i, t−i⟩ is jointly convex on (xi, x−i), and taking the minimum value on partial variables do not affect the convexity of remaini...

  125. [134]

    max” is replaced with “differentialmax

    := Z t∈T |xi 1(t) − xi 2(t)|dt = Z t−i∈T−i dt−i Z ti∈Ti |xi 1(ti; t−i) − xi 2(ti; t−i)|dti ≤ Z t−i∈T−i dt−i √ 2δ · [¯xi + ¯ti] 2 = √ 2δ · [¯xi + ¯ti] · Q j̸=i ¯tj 2 = √ 2δ · [¯xi/¯ti + 1] · Q i∈[n] ¯ti 2 and l1(x1, x2) = X i∈[n] l1(xi 1, xi 2) ≤ √ 2δ · [P i∈[n] ¯xi/¯ti + n] · ...

  126. [135]

    algorithmic boundary

    discovers that the performance of LotteryAMA highly depends on different initializations, and we discover similar phenomena on MD-MenuNet and MD-GemNet. The instability of initialization appears in the experiment stated below: The optimal mechanism in SU 2 includes 3 determini...

  127. [136]

    permutation equivariance property

    shows that there is always an approach to reduce any mechanism design problem to another equivalent problem with outcome separability. The reduction can be done by penalizing infeasible outcomes onto the designer’s objective. 65 E.2 Open Problems E.2.1 Sampling in Stream. One ...

  128. [2003]

    Proceedings. IEEE. 2003, pp. 574–583

    2003

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.