Pith. sign in

REVIEW 5 major objections 5 minor 42 references

Persuasive Prediction via Decision Calibration

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read A finite-data sender can learn a decision-calibrated predictor that matches the utility of a Bayesian sender who knows the prior.

desk verdict The paper has a real idea and a repairable-looking sign error, but as written the central guarantee does not go through. read the letter →

arxiv 2505.16141 v1 pith:LAWM3I22 submitted 2025-05-22 cs.GT

classification cs.GT
keywords Bayesianpersuasiondecisioncalibrationpersuasivepredictionprior-freeminimaxequilibriumno-regretlearningswapregretquantalresponse
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

Bayesian persuasion normally presupposes that sender and receiver share an exact prior, which is unlearnable in high-dimensional or infinite state spaces. The paper introduces persuasive prediction, where the sender learns a predictor from past data and receivers best respond to it, and adopts decision calibration as the behavioral link: a prediction is credible when it is unbiased conditioned on the receiver's best response. The central result is that with a sample count growing polynomially in the number of receiver actions and outcome dimensions, but independent of the size of the context space, the PerDecCal algorithm finds a randomized predictor that is nearly decision-calibrated and nearly maximizes sender utility among all such predictors. In the single-receiver case this matches the utility of a fully informed Bayesian sender restricted to the same predictor class.

What carries the argument

The load-bearing object is decision calibration: a randomized predictor $f \in \Delta(H)$ is $\epsilon$-decision-calibrated when, for every receiver $i$, outcome coordinate $j$, and action $a$, the expectation $|\mathbb{E}[(y_j - h(x)_j) \cdot b_i(h(x),a)]|$ is at most $\epsilon$, where $b_i$ is the receiver's strict best response. The argument converts the sender's constrained optimization problem into a zero-sum game via Lagrange multipliers: the min player chooses $f$, the max player chooses bounded dual variables $\lambda$, and the payoff is the negative sender utility plus calibration-violation penalties. PerDecCal solves this game by alternating an ERM oracle (best response for the min player) with the Hedge algorithm (no-regret for the max player), and uniform-convergence bounds transfer the empirical equilibrium to the true distribution. The bridge to Bayesian persuasion is a pair of lemmas: calibrated predictors correspond exactly to signaling schemes over posterior means, and any perfectly decision-calibrated predictor can be post-processed into a perfectly calibrated predictor with the same sender utility, which is what makes the Bayesian-benchmark comparison possible.

What would settle it

Directly substitute a perfectly calibrated constant predictor, say $h(x) = \mathbb{E}[Y]$, into Eq. (3) with $\gamma > 0$ and a Lagrange multiplier placed on a negative-sign coordinate; the calibration term is $\lambda\gamma > 0$, so the Lagrangian value exceeds the negative sender utility, contradicting the claimed equivalence with the constrained problem and the proof of Lemma 3.2.

Watch

Extended reading notes

Core claim

The paper's central claim is that decision calibration is enough to make persuasion prior-free. Formally, for a finite predictor class $H$, PerDecCal returns a predictor $\hat{f}$ with $\mathrm{DecCE}(\hat{f}) \le \gamma + \epsilon$ and $\mathbb{E}[u(a,y)b(h(x),a)] \ge \mathrm{OPT}(H,D,\gamma) - \epsilon$, with probability $1-\delta$, using $n = O(\log(|H|Ndm/\delta)/\epsilon^4)$ samples; the bound does not depend on $|X|$. The proof routes through a Lagrangian reformulation of the constrained optimization problem as a zero-sum game between a predictor and a dual player, and shows that best-response-vs-no-regret dynamics converge to an approximate equilibrium. The authors further show that in the single-receiver case the attained utility is at least $\mathrm{BayesOPT}(\mu_D, \Pi_H) - \epsilon$, so the data-driven sender matches a Bayesian sender with full knowledge of $D$ who is confined to signaling schemes induced by $H$. For infinite $H$ and quantal-responding receivers, an analogous algorithm achieves the same type of guarantee with sample complexity governed by a covering number.

Load-bearing premise

The load-bearing premise is that the Lagrangian in Eq. (3) correctly encodes the calibration constraint, but as printed the sign convention gives a perfectly calibrated predictor a positive dual penalty, so the central guarantee is not established unless that expression is a typo.

Editorial extensions

If this is right

  • With $O(\log(|H|Ndm/\delta)/\epsilon^4)$ samples the sender can produce a $\gamma+\epsilon$-decision-calibrated predictor that is $\epsilon$-optimal in sender utility, no matter how large $|X|$ is.
  • Receivers who strict best respond to the output predictor incur at most $2mL(\gamma+\epsilon)$ swap regret, so the behavioral model is internally consistent.
  • In the single-receiver case the data-driven sender's utility matches the Bayesian-persuasion benchmark restricted to signaling schemes induced by $H$, even though the prior was never estimated.
  • When receivers use $\eta$-quantal responses, the same near-optimal utility and smoothed-calibration guarantees hold for infinite hypothesis classes with bounded covering numbers.
  • The algorithm is oracle-efficient: it needs only one ERM call per iteration, so any class with a practical ERM oracle is computationally tractable.

Reading between the lines

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

  • Beyond the paper: the same Lagrangian-minimax template could turn any constraint expressible as a family of linear inequalities on a predictor, such as multi-calibration or fairness constraints, into a finite-sample optimization problem solved by no-regret dynamics.
  • Beyond the paper: the benchmark-matching result suggests decision calibration is not merely a proxy but an operational realization of the common-prior assumption, so the choice of calibration notion determines which rational receiver behavior is being assumed in prior-free persuasion.
  • Beyond the paper: replacing the exact ERM oracle with an approximate oracle, such as SGD-trained neural networks, and measuring the gap in utility and calibration would provide a direct empirical test of how much the oracle-efficiency guarantee degrades in practice.
  • Beyond the paper: varying the inverse temperature $\eta$ in the quantal-response extension interpolates between strict no-regret behavior and the Bayesian best-response limit, offering a way to experimentally separate decision-calibrated persuasion from classical prior-based persuasion.
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 / 5 minor

Summary. The paper proposes 'persuasive prediction,' a data-driven variant of Bayesian persuasion in which a sender learns from samples a decision-calibrated predictor to influence a receiver. The main algorithmic claim (Theorem 3.1) is that for a finite hypothesis class H, PerDecCal, based on a Lagrangian minimax reformulation with an ERM oracle and Hedge, achieves near-optimal sender utility among γ-decision-calibrated predictors with sample complexity independent of the feature-space size |X|. Section 4 claims this matches a Bayesian persuasion benchmark in the single-receiver case, and Section 5 extends the approach to quantal responses and infinite hypothesis classes via covering numbers. The paper also contains no-regret guarantees for receivers and a finite-sample uniform-convergence analysis.

Significance. If established, the results would offer a novel connection between decision calibration and Bayesian persuasion, with a plausible oracle-efficient algorithm and finite-sample guarantees that do not depend on |X|. The manuscript contains substantial technical components: a minimax reformulation, uniform-convergence bounds, a Lipschitz covering-number analysis for smoothed responses, and regret guarantees for receivers. However, the central proof contains a sign error in the Lagrangian that invalidates Theorem 3.1 as written, and the 'Bayesian benchmark' in Section 4 is defined to coincide with the decision-calibrated optimum, so the headline claim of matching a fully informed Bayesian sender is not supported. These are load-bearing issues, so the contribution as stated is not established.

major comments (5)
  1. [Eq. (3)] The Lagrangian term Σ_{s∈{+,-}} λ_{s,i,j,a_i} s(E_{i,j,a_i} - γ) is incorrect for the constraint |E_{i,j,a_i}| ≤ γ. For s = -1, the summand is λ_-(-E + γ), which is positive for every E < γ, including the perfectly calibrated value E = 0. Hence max_{λ ∈ R_+^{2Nmd}} L_D(f, λ) is infinite even for feasible predictors, and the minimax game (3)-(4) does not enforce the decision-calibration constraint. The correct relaxation is λ_+(E - γ) + λ_-(-E - γ).
  2. [Lemma 3.2] The proof of Lemma 3.2 relies on the same erroneous sign. In the first case it asserts -E[u] = max_λ L(f̂, λ) for a feasible f̂, which is false because max_λ L(f̂, λ) is infinite whenever E < γ. In the second case, the argmax over s of s(E - γ) selects s = - with value γ - E for any feasible E ∈ (-γ, γ), so the proof treats every feasible predictor as a violation and derives an artificial bound C(γ - E) ≤ 1 + 2ε. Consequently the DecCE bound in Lemma 3.2, and hence Theorem 3.1, is not established as written.
  3. [Algorithm 1] The update in line 6, c_t(λ_{s,i,j,a_i}) = λ_{s,i,j,a_i} s(E - γ), together with the ERM loss ℓ_λ in Definition 3.1, makes PerDecCal optimize a different objective from Eq. (2): the s = - coordinate receives positive cost whenever E < γ, so the Hedge weights push the predictor away from the feasible region rather than enforce |E| ≤ γ. A correction to the Lagrangian must be accompanied by a corresponding change to the algorithm's cost and loss functions; the algorithm as printed does not solve the stated constrained problem.
  4. [Definition 4.2 / Theorem 4.1] The 'Bayesian benchmark' Π_H is defined as the set of signaling schemes induced by the very class FDCAL(H) of perfectly decision-calibrated predictors over H. Therefore BayesOPT(μ_D, Π_H) = OPT(H, D, 0) by construction, and Theorem 4.1 is a corollary of Theorem 3.1 that renames the same optimization problem. The abstract's claim that the method 'matches the utility of a Bayesian sender who has full knowledge of the underlying prior distribution' is not supported: the comparison is only against the restricted class of schemes induced by H, not against the full Bayesian persuasion value.
  5. [Definition 2.1] Definition 2.1 defines the strict best response as arg min of the receiver's utility v_i(a'_i, h(x)), but the no-regret proof of Theorem 2.1 and the informal Eq. (1) require the receiver to maximize utility; with the printed arg min the inequality in the proof of Theorem 2.1 has the wrong direction. Example 5.1 follows the arg-min convention, so the paper is internally inconsistent about the direction of receiver preferences. The definition should be corrected to arg max (or the utility convention should be renamed to a cost).
minor comments (5)
  1. [Theorem 3.1 proof] The proof chooses C = 2/ε without stating this in Theorem 3.1 or Algorithm 1; the theorem should specify the dual bound C as an input, or the guarantee should be stated for this choice.
  2. [Lemma 3.3] The bound contains a term sqrt(ln|4H|/δ / 2n) with an unexplained factor of 4; the proof says the budget is split to δ/2, but the displayed expression should be derived explicitly.
  3. [Theorem 5.1] The swap regret bound is printed as ln m + 1 / η while Theorem C.3 gives (ln|A| + 1) / η; the notation should be reconciled.
  4. [Appendix D, proof of Theorem 3.1] The last sentence refers to 'receivers who play η-quantal response' and cites Theorem 2.1, but Theorem 2.1 is for strict best responses; the reference should be to Theorem C.3.
  5. [Lemma 4.2] The proof should explicitly justify the equality of events {f'(x) = f'_a} and {b(h(x), a) = 1}; this is intuitive because f' takes the value corresponding to the chosen action, but it is not stated.

Circularity Check

1 steps flagged · score 6.0 of 10

The 'matches the Bayesian benchmark' result is a definitional identity: Definition 4.2 builds Π_H as the image of the decision-calibrated class that PerDecCal already optimizes, so Theorem 4.1 follows by construction.

  1. self definitional [Section 4, Definition 4.2; Appendix E, proof of Theorem 4.1]
    "We define FDCAL(H) as the class of randomized predictor over H that is perfectly decision calibrated... Define Π_H be the class of all such signaling schemes π_f′. ... Note that OPT(H, D, 0) is the optimal sender utility achieved by the randomized predictors over H that are perfectly decision calibrated, i.e. FDCAL(H). By Definition 4.2 and Lemma 4.2, OPT(H, D, 0) is equal to the optimal sender utility achieved by the predictors in FCAL(H). Then by Definition 4.2 and Lemma 4.1, OPT(H, D, 0) is equal to the optimal sender utility achieved by the signaling schemes in Π_H, i.e."

    The 'Bayesian' benchmark is not an independent optimum: Definition 4.2 constructs Π_H as the image of FDCAL(H), the exact class of perfectly decision-calibrated predictors over H that defines OPT(H,D,0). The proof of Theorem 4.1 then equates OPT(H,D,0) with BayesOPT(µ_D,Π_H) purely by invoking Definition 4.2 and Lemmas 4.2/4.1. Thus the theorem is a chain of definitional identities: the utility PerDecCal optimizes over FDCAL(H) is renamed 'Bayesian sender utility over Π_H'. No comparison to unrestricted Bayesian persuasion is made. The abstract's claim that the method 'matches the utility of a Bayesian sender who has full knowledge of the underlying prior distribution' is true only for this self-defined restricted benchmark, so the headline matching result holds by construction.

full rationale

The core algorithmic result (Theorem 3.1) is a genuine internal optimization over decision-calibrated predictors with uniform-convergence and no-regret analysis; it is not circular and does not depend on the Bayesian comparison. Lemma 4.1 and Lemma 4.2 are substantive bridge results, not self-citations. The only circular step is the Bayesian benchmark matching: Definition 4.2 defines Π_H as the signaling-scheme image of the very class FDCAL(H) that PerDecCal optimizes, so Theorem 4.1's equality OPT(H,D,0)=BayesOPT(µ_D,Π_H) is true by construction. This makes the headline 'matches the Bayesian benchmark' partially circular, although the restricted benchmark is explicitly disclosed in Section 4. Separately, the printed Lagrangian in Eq. (3) has a sign error (the s=- term makes the dual unbounded even for perfectly calibrated predictors), but that is a correctness defect, not a circularity, and does not affect this score.

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

The central algorithm depends on standard duality and no-regret tools, plus behavioral assumptions about receivers and an ERM oracle. The most fragile input is the Lagrangian dual as printed in Eq. (3), which does not encode the absolute-value calibration constraint because of a sign error; this invalidates the proof as written. No parameters are fitted to data.

assumptions (5)
  • domain assumption Receiver utilities are linear and L-Lipschitz in the outcome y (Assumption 2.1).
    Used throughout for Lipschitzness in no-regret proofs and the quantal response extension; restricts the class of receivers.
  • domain assumption There exists a randomized predictor in Δ(H) with DecCE ≤ γ (Assumption 2.2).
    Ensures the constrained problem is feasible; the authors argue it is mild because a constant predictor E[Y] is calibrated.
  • domain assumption Receivers strictly best respond to predictions (Definition 2.1) or follow quantal response (Definition 5.1).
    The no-regret guarantee and sender utility are defined w.r.t. these response models; if receivers are skeptical or strategic about the predictor, the model changes.
  • domain assumption Access to an ERM oracle over H (Definition 3.1).
    The algorithms are oracle-efficient; without such an oracle the optimization over H is not tractable in general.
  • standard math The Lagrangian in Eq. (3) is the correct dual of the constrained problem Eq. (2).
    This is a folklore convex duality claim, but as written the sign convention is wrong: the s = -1 term has +γ instead of −γ, so the printed Lagrangian does not enforce the absolute-value calibration constraint. The proof of Lemma 3.2 relies on this incorrect form.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Persuasive Prediction via Decision Calibration." pith.science (2026). https://pith.science/paper/LAWM3I22

@misc{pith2026250516141,
  author       = {Pith},
  title        = {Pith review of: Persuasive Prediction via Decision Calibration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LAWM3I22}},
  note         = {Machine review of arXiv:2505.16141}
}
abstract

Bayesian persuasion, a central model in information design, studies how a sender, who privately observes a state drawn from a prior distribution, strategically sends a signal to influence a receiver's action. A key assumption is that both sender and receiver share the precise knowledge of the prior. Although this prior can be estimated from past data, such assumptions break down in high-dimensional or infinite state spaces, where learning an accurate prior may require a prohibitive amount of data. In this paper, we study a learning-based variant of persuasion, which we term persuasive prediction. This setting mirrors Bayesian persuasion with large state spaces, but crucially does not assume a common prior: the sender observes covariates $X$, learns to predict a payoff-relevant outcome $Y$ from past data, and releases a prediction to influence a population of receivers. To model rational receiver behavior without a common prior, we adopt a learnable proxy: decision calibration, which requires the prediction to be unbiased conditioned on the receiver's best response to the prediction. This condition guarantees that myopically responding to the prediction yields no swap regret. Assuming the receivers best respond to decision-calibrated predictors, we design a computationally and statistically efficient algorithm that learns a decision-calibrated predictor within a randomized predictor class that optimizes the sender's utility. In the commonly studied single-receiver case, our method matches the utility of a Bayesian sender who has full knowledge of the underlying prior distribution. Finally, we extend our algorithmic result to a setting where receivers respond stochastically to predictions and the sender may randomize over an infinite predictor class.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 25 canonical work pages

  1. [1]

    Robust virtual implementation: Toward a reinterpretation of the wilson doctrine

    Georgy Artemov, Takashi Kunimoto, and Roberto Serrano. Robust virtual implementation: Toward a reinterpretation of the wilson doctrine. Journal of Economic Theory, 148 0 (2): 0 424--447, 2013

  2. [2]

    Reducing mechanism design to algorithm design via machine learning

    Maria-Florina Balcan, Avrim Blum, Jason D Hartline, and Yishay Mansour. Reducing mechanism design to algorithm design via machine learning. Journal of Computer and System Sciences, 74 0 (8): 0 1245--1270, 2008

  3. [3]

    Commitment without regrets: Online learning in stackelberg security games

    Maria-Florina Balcan, Avrim Blum, Nika Haghtalab, and Ariel D Procaccia. Commitment without regrets: Online learning in stackelberg security games. In Proceedings of the sixteenth ACM conference on economics and computation, pages 61--78, 2015

  4. [4]

    Information design: A unified perspective

    Dirk Bergemann, Benjamin Brooks, and Stephen Morris. Information design: A unified perspective. Econometrica, 91 0 (1): 0 1--47, 2023. doi:10.3982/ECTA18308

  5. [5]

    Convex optimization

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

  6. [6]

    Mechanisms for a no-regret agent: Beyond the common prior

    Modibo K Camara, Jason D Hartline, and Aleck Johnsen. Mechanisms for a no-regret agent: Beyond the common prior. In 2020 ieee 61st annual symposium on foundations of computer science (focs), pages 259--270. IEEE, 2020

  7. [7]

    The sample complexity of revenue maximization

    Richard Cole and Tim Roughgarden. The sample complexity of revenue maximization. In Proceedings of the forty-sixth annual ACM symposium on Theory of computing, pages 243--252, 2014

  8. [8]

    Efficient prior-free mechanisms for no-regret agents

    Natalie Collina, Aaron Roth, and Han Shao. Efficient prior-free mechanisms for no-regret agents. In Proceedings of the 25th ACM Conference on Economics and Computation, pages 511--541, 2024

Show all 42 references
  1. [9]

    Algorithmic price discrimination

    Rachel Cummings, Nikhil R Devanur, Zhiyi Huang, and Xiangning Wang. Algorithmic price discrimination. In Proceedings of the Fourteenth Annual ACM-SIAM Symposium on Discrete Algorithms, pages 2432--2451. SIAM, 2020

  2. [10]

    Learning in auctions: Regret is hard, envy is easy

    Constantinos Daskalakis and Vasilis Syrgkanis. Learning in auctions: Regret is hard, envy is easy. In 2016 ieee 57th annual symposium on foundations of computer science (focs), pages 219--228. IEEE, 2016

  3. [11]

    Oracle-efficient online learning and auction design

    Miroslav Dud \' k, Nika Haghtalab, Haipeng Luo, Robert E Schapire, Vasilis Syrgkanis, and Jennifer Wortman Vaughan. Oracle-efficient online learning and auction design. Journal of the ACM (JACM), 67 0 (5): 0 1--57, 2020

  4. [12]

    Preparing for the worst but hoping for the best: Robust (bayesian) persuasion

    Piotr Dworczak and Alessandro Pavan. Preparing for the worst but hoping for the best: Robust (bayesian) persuasion. Econometrica, 90 0 (5): 0 2017--2051, 2022

  5. [13]

    From fairness to infinity: Outcome-indistinguishable (omni) prediction in evolving graphs

    Cynthia Dwork, Chris Hays, Nicole Immorlica, Juan C Perdomo, and Pranay Tankala. From fairness to infinity: Outcome-indistinguishable (omni) prediction in evolving graphs. arXiv preprint arXiv:2411.17582, 2024

  6. [14]

    Persuasive calibration

    Yiding Feng and Wei Tang. Persuasive calibration. arXiv preprint arXiv:2504.03211, 2025

  7. [15]

    Full swap regret and discretized calibration

    Maxwell Fishelson, Robert Kleinberg, Princewill Okoroafor, Renato Paes Leme, Jon Schneider, and Yifeng Teng. Full swap regret and discretized calibration. arXiv preprint arXiv:2502.09332, 2025

  8. [16]

    Regret in the on-line decision problem

    Dean P Foster and Rakesh Vohra. Regret in the on-line decision problem. Games and Economic Behavior, 29 0 (1-2): 0 7--35, 1999

  9. [17]

    Game theory, on-line prediction and boosting

    Yoav Freund and Robert E Schapire. Game theory, on-line prediction and boosting. In Proceedings of the ninth annual conference on Computational learning theory, pages 325--332, 1996

  10. [18]

    A decision-theoretic generalization of on-line learning and an application to boosting

    Yoav Freund and Robert E Schapire. A decision-theoretic generalization of on-line learning and an application to boosting. Journal of computer and system sciences, 55 0 (1): 0 119--139, 1997

  11. [19]

    Learning utilities and equilibria in non-truthful auctions

    Hu Fu and Tao Lin. Learning utilities and equilibria in non-truthful auctions. Advances in Neural Information Processing Systems, 33: 0 14231--14242, 2020

  12. [20]

    Oracle efficient online multicalibration and omniprediction

    Sumegha Garg, Christopher Jung, Omer Reingold, and Aaron Roth. Oracle efficient online multicalibration and omniprediction. In Proceedings of the 2024 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pages 2725--2792. SIAM, 2024

  13. [21]

    Omnipredictors

    Parikshit Gopalan, Adam Tauman Kalai, Omer Reingold, Vatsal Sharan, and Udi Wieder. Omnipredictors. arXiv preprint arXiv:2109.05389, 2021

  14. [22]

    Loss minimization through the lens of outcome indistinguishability

    Parikshit Gopalan, Lunjia Hu, Michael P Kim, Omer Reingold, and Udi Wieder. Loss minimization through the lens of outcome indistinguishability. arXiv preprint arXiv:2210.08649, 2022

  15. [23]

    Omnipredictors for regression and the approximate rank of convex functions

    Parikshit Gopalan, Princewill Okoroafor, Prasad Raghavendra, Abhishek Sherry, and Mihir Singhal. Omnipredictors for regression and the approximate rank of convex functions. In The Thirty Seventh Annual Conference on Learning Theory, pages 2027--2070. PMLR, 2024

  16. [24]

    Calibration error for decision making

    Lunjia Hu and Yifan Wu. Calibration error for decision making. arXiv preprint arXiv:2404.13503, 2024

  17. [25]

    Incentivizing exploration with selective data disclosure

    Nicole Immorlica, Jieming Mao, Aleksandrs Slivkins, and Zhiwei Steven Wu. Incentivizing exploration with selective data disclosure. arXiv preprint arXiv:1811.06026, 2018

  18. [26]

    Calibrated forecasting and persuasion

    Atulya Jain and Vianney Perchet. Calibrated forecasting and persuasion. arXiv preprint arXiv:2406.15680, 2024

  19. [27]

    Bayesian persuasion

    Emir Kamenica and Matthew Gentzkow. Bayesian persuasion. American Economic Review, 101 0 (6): 0 2590--2615, 2011

  20. [28]

    U-calibration: Forecasting for an unknown agent

    Bobby Kleinberg, Renato Paes Leme, Jon Schneider, and Yifeng Teng. U-calibration: Forecasting for an unknown agent. In The Thirty Sixth Annual Conference on Learning Theory, pages 5143--5145. PMLR, 2023

  21. [29]

    Persuasion with unknown beliefs

    Svetlana Kosterina. Persuasion with unknown beliefs. Theoretical Economics, 17 0 (3): 0 1075--1107, 2022

  22. [30]

    Information design with unknown prior

    Tao Lin and Ce Li. Information design with unknown prior. arXiv preprint arXiv:2410.05533, 2024

  23. [31]

    Sample efficient omniprediction and downstream swap regret for non-linear losses

    Jiuyao Lu, Aaron Roth, and Mirah Shi. Sample efficient omniprediction and downstream swap regret for non-linear losses. arXiv preprint arXiv:2502.12564, 2025

  24. [32]

    Simultaneous swap regret minimization via kl-calibration

    Haipeng Luo, Spandan Senapati, and Vatsal Sharan. Simultaneous swap regret minimization via kl-calibration. arXiv preprint arXiv:2502.16387, 2025

  25. [33]

    Quantal choice analysis: A survey

    Daniel McFadden et al. Quantal choice analysis: A survey. Annals of economic and social measurement, 5 0 (4): 0 363--390, 1976

  26. [34]

    Quantal response equilibria for normal form games

    Richard D McKelvey and Thomas R Palfrey. Quantal response equilibria for normal form games. Games and economic behavior, 10 0 (1): 0 6--38, 1995

  27. [35]

    On the pseudo-dimension of nearly optimal auctions

    Jamie H Morgenstern and Tim Roughgarden. On the pseudo-dimension of nearly optimal auctions. Advances in Neural Information Processing Systems, 28, 2015

  28. [36]

    High-dimensional prediction for sequential decision making

    Georgy Noarov, Ramya Ramalingam, Aaron Roth, and Stephan Xie. High-dimensional prediction for sequential decision making. arXiv preprint arXiv:2310.17651, 2023

  29. [37]

    Near-optimal algorithms for omniprediction

    Princewill Okoroafor, Robert Kleinberg, and Michael P Kim. Near-optimal algorithms for omniprediction. arXiv preprint arXiv:2501.17205, 2025

  30. [38]

    Full implementation and belief restrictions

    Mariann Oll \'a r and Antonio Penta. Full implementation and belief restrictions. American Economic Review, 107 0 (8): 0 2243--2277, 2017

  31. [39]

    Forecasting for swap regret for all downstream agents

    Aaron Roth and Mirah Shi. Forecasting for swap regret for all downstream agents. In Proceedings of the 25th ACM Conference on Economics and Computation, pages 466--488, 2024

  32. [40]

    A sample complexity measure with applications to learning optimal auctions

    Vasilis Syrgkanis. A sample complexity measure with applications to learning optimal auctions. Advances in Neural Information Processing Systems, 30, 2017

  33. [41]

    Dimension-free decision calibration for nonlinear loss functions

    Jingwu Tang, Jiayun Wu, Zhiwei Steven Wu, and Jiahao Zhang. Dimension-free decision calibration for nonlinear loss functions. arXiv preprint arXiv:2504.15615, 2025

  34. [42]

    Calibrating predictions to decisions: A novel approach to multi-class calibration

    Shengjia Zhao, Michael Kim, Roshni Sahoo, Tengyu Ma, and Stefano Ermon. Calibrating predictions to decisions: A novel approach to multi-class calibration. Advances in Neural Information Processing Systems, 34: 0 22313--22324, 2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.