Pith. sign in

REVIEW 3 major objections 5 minor 64 references

Robust ML Auditing using Prior Knowledge

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper establishes that an auditor with a private labeled dataset can compute the probability of detecting a platform that changes its answers to look fair.

desk verdict Theorem 4.3's detection-rate formula has a sign error that makes it exceed 1; the framework is promising, but the central quantitative claim is wrong as printed. read the letter →

arxiv 2505.04796 v2 pith:UTPLHNCG submitted 2025-05-07 cs.LG

Jade Garcia Bourrée , Augustin Godinot , Martijn De Vos , Milos Vujasinovic , Sayan Biswas , Gilles Tredan , Erwan Le Merrer , Anne-Marie Kermarrec This is my paper · ORCID
classification cs.LG
keywords auditmanipulationfairwashingfairnessauditingdemographicparityauditorpriordatasetdetectionrateblack-box
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether a fairness auditor can stop a platform from fairwashing, meaning changing its answers only during the audit to appear fair, and answers yes, on a condition. The condition is that the auditor holds a private labeled dataset and treats its ground-truth labels as the center of a dataset prior: the set of models whose squared loss on that dataset is below a threshold. Under that prior, the paper derives a closed-form probability for detecting the optimally manipulating platform, depending only on the distance from the ground truth to the fair set, the threshold, and the number of audit queries. It also proves that any public prior can always be gamed, so the prior must stay secret. The paper's message is that audit robustness comes less from clever querying and more from what the auditor knows before the audit begins.

What carries the argument

The load-bearing object is the dataset prior $\mathcal{H}_a = \{h : L(h, D_a) < \tau\}$, a ball in model space centered on the auditor's labeled data. The fair set $\mathcal{F}$ is a hyperplane under demographic parity, and the manipulating platform's optimal move is the orthogonal projection $\mathrm{proj}_{\mathcal{F}}(h_p)$, which flips as few labels as possible. Detection fails exactly when that projection still lands inside the prior ball, so the detection probability is a volume ratio: the part of the ball whose projection exits the ball divided by the whole ball. Because the prior is assumed uniform, the ratio becomes the volume of a ball minus a cylinder and two spherical caps, evaluated in closed form with Wallis integrals.

What would settle it

Train a cohort of honest models on a fixed task, choose an audit dataset $\mathcal{D}_a$ and threshold $\tau$, measure how often the optimal projection of an honest model lands outside $\mathcal{H}_a$, and compare that empirical frequency with the closed-form $P_{uf}$ for several values of $n$, $\delta$, and $\tau$; a systematic gap would show that the uniform-prior assumption is not how real models distribute.

Watch

Extended reading notes

Core claim

The central claim is Theorem 4.3: for binary classification, squared-error loss, and an auditor prior that is a ball of radius $\tau$ around the labeled dataset's ground truth $h_a$, the probability that the auditor catches a platform whose optimal manipulation is the orthogonal projection of its honest model onto the demographic-parity hyperplane is $P_{uf} = 1 - \frac{1}{W_n}\left(\int_0^{\arccos(\delta/\tau)} \sin^n(\theta)\,d\theta - \frac{\delta}{\tau}\left(1-\frac{\delta^2}{\tau^2}\right)^{(n-1)/2}\right)$. Here $\delta = d(h_a, \mathcal{F})$ is the distance from the ground truth to the fair set, $n$ is the audit budget, and $W_n$ is the $n$-term Wallis integral. Two boundary cases frame the result: a fair prior ($\delta=0$) gives zero detection probability, while a tangent prior ($\delta=\tau$) gives detection probability one. The paper also proves a more general, distribution-free statement: detection probability equals the volume of the prior ball whose projection onto the fair set leaves the ball, normalized by the ball's volume.

Load-bearing premise

The formula's numbers are only objective probabilities if the honest platform model is uniformly distributed inside the prior ball and the auditor's labeled dataset gives the exact center of that ball; otherwise the detection rate is a conditional geometric ratio that depends on how models actually cluster.

Editorial extensions

If this is right

  • An auditor with a private labeled dataset can compute, before sending a single query, the probability of catching an optimally manipulating platform; no assumption about the platform's training pipeline is needed beyond the uniform prior.
  • Any prior the platform can reconstruct is worthless: with knowledge of $\mathcal{H}_a$, the platform simply answers with a model in $\mathcal{H}_a \cap \mathcal{F}$ and passes both the fairness and honesty checks.
  • A perfectly fair audit prior leaves the auditor blind ($P_{uf}=0$); to be effective the prior's ground truth must sit at positive distance from the fair set, and tightening $\tau$ down to $\delta$ gives detection probability one.
  • Detection thresholds can be calibrated from the error rate of state-of-the-art models on the task, and simulations on tabular and image data show that for easy tasks a larger audit budget drives concealable unfairness to zero, while for harder tasks some unfairness remains hidden.

Reading between the lines

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

  • Editorial extension: the same ball-hyperplane geometry should transfer to any fairness metric whose constraint set is a hyperplane or low-codimension subspace of model space, such as equality of opportunity; only the direction of the normal changes.
  • Editorial extension: if the uniform-prior assumption is doubtful, the formula can be re-read as the detection probability under the least-informative prior, and one could replace it with a Monte Carlo estimate obtained by sampling models that actually pass the task.
  • Editorial extension: an auditor could actively design the labeled dataset $\mathcal{D}_a$ to push $\delta$ upward, for instance by balancing the protected groups in the audit set, thereby raising the lower bound on $P_{uf}$ at the cost of making $\delta$ itself an audited quantity.
  • Editorial extension: a direct empirical check is to train many honest models on one task, project each onto the fair set, and compare the measured detection frequency against the closed-form $P_{uf}$ across sample sizes and threshold values.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper formalizes the problem of audit manipulation (fairwashing) in black-box fairness audits. It introduces the notion of an auditor prior as a set of models the auditor considers expectable, proves that public priors cannot prevent manipulation (Theorem 3.2), and proposes a private dataset prior defined as a risk ball around the auditor's labeled data. The central theoretical result is Theorem 4.3, which gives a closed-form detection probability Puf under an assumed uniform distribution over the prior ball, together with boundary corollaries and a lower bound. The empirical section measures the concealable unfairness achievable by four manipulation methods on CelebA and ACSEmployment datasets. The main result as printed is, however, algebraically incorrect: the sign between the two terms inside the brackets in Theorem 4.3 is wrong, so the expression is not a valid probability for intermediate parameter values.

Significance. The paper addresses a timely and important problem: making fairness audits robust to strategic manipulation. If the central formula is corrected, the framework provides a computable detection probability for a private dataset prior, clean boundary cases (tau=delta gives detection probability one, delta=0 gives zero), and a useful formal separation between public and private priors. The experimental study of concealable unfairness across datasets and manipulation methods is a genuine contribution, and the authors state that their code is publicly available. However, the current Theorem 4.3 does not define a probability for intermediate delta/tau, and the erroneous sign also invalidates the claimed lower bound in Corollary 4.4. The paper's main theoretical contribution therefore needs correction before its conclusions can be accepted.

major comments (3)
  1. [Corollary 4.4] Theorem 4.3 (and the final line of its proof in Appendix A) contains a sign error. The proof correctly decomposes the volume as V# = Vball - 2Vcap - Vcylinder, which yields Puf = 1 - (1/W_n)[integral_0^{arccos(delta/tau)} sin^n(theta) dtheta + (delta/tau)(1 - delta^2/tau^2)^{(n-1)/2}]. The printed formula has a minus sign where the plus sign should be. This is not cosmetic: for n=2 and delta/tau=1/2, the printed expression evaluates to 1/3 + 3*sqrt(3)/(2*pi) ≈ 1.160, which is impossible for a probability, whereas the correct expression gives 1/3 - sqrt(3)/(2*pi) ≈ 0.058. Direct integration over the unit disk confirms the plus-sign version. Because the second term vanishes at delta=0 and delta=tau, the boundary corollaries A.4 and A.5 are unaffected, but the quantitative claim of Theorem 4.3 for intermediate delta/tau is false as printed.
  2. [Section 4.3] The theorem's probabilistic interpretation is conditional on two strong assumptions that are asserted rather than justified: (i) the uninformative prior, i.e., the platform model hp is uniformly distributed over the ball Ha, and (ii) the auditor's labeled dataset Da provides the exact ground-truth center ha of that ball, with no label noise or distribution shift. These assumptions are stated, but the takeaway in Section 4.3 that 'the auditor can always calculate a priori the probability to correctly detect a malicious platform' overstates what is actually derived: Theorem 4.3 computes a geometric volume ratio under the uniformity assumption, not an unconditional operational detection probability. I ask the authors to qualify this claim and, ideally, to include a sensitivity discussion or a simple non-uniform example showing how Puf would change.
  3. [Section 5.2] The experimental section does not test Theorem 4.3 directly: the detection score in Eq. (9) and the concealable unfairness in Eq. (8) are different objects from the Puf formula's volume ratio, and no experiment compares measured detection frequencies with the theorem's prediction. The experiments are informative about achievable manipulation in practice, but they do not validate the corrected formula. If the authors intend the experiments as an empirical check of the theory, this link should be made explicit; otherwise the paper should state more clearly that the theory and the experiments are complementary rather than confirmatory.
minor comments (5)
  1. [Table 1] In Table 1, delta is described as the 'distance between the groundtruth and the set of expectable model'; it should be the distance between the ground truth and the set of fair models F, i.e., d(ha,F).
  2. [Theorem 4.3] The statement of Theorem 4.3 should explicitly define n = |S| and state the domain 0 <= delta <= tau; currently these are only mentioned in the surrounding text.
  3. [Equation (7)] In Eq. (7), the symbol tau is reused as the tolerable demographic-parity constraint, after tau was introduced in Definition 4.1 as the risk threshold for the dataset prior. Using a different symbol (e.g., epsilon) would avoid a notational collision.
  4. [Appendix A] The appendix states that Ha is an open set, while Definition 4.1 gives Ha = {h : L(h,Da) < tau}, an open ball, yet Theorem 4.3 and the corollaries use volumes of the closed ball. The boundary has measure zero, so the distinction does not affect the results, but it should be stated explicitly to avoid confusion.
  5. [Figure 2] Figure 2 would be easier to interpret if the axes were labeled and the 'red area' were linked to the formula in Theorem 4.3; currently the geometric intuition is hard to map onto the algebra.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: Theorem 4.3 is a geometric volume computation from stated inputs, and the experimental concealable-unfairness measurements do not fit or instantiate the theorem's constants.

full rationale

The derivation chain is self-contained rather than circular. The detection rate (Definition 4.2) is defined as P(h*_m notin Ha | hp in Ha), and Theorem 4.3 computes this probability under the explicitly stated dataset prior (Definition 4.1), L2 loss, and the uninformative uniform prior assumption. The formula is obtained by volume ratios: Puf = V#_n(tau,delta)/Vball_n(tau), with V# = Vball - 2Vcap - Vcylinder (Appendix A). The inputs delta and tau are not fitted to match any outcome; they parameterize the geometry (distance to the fair set and the risk threshold). The boundary corollaries (delta=0 implies Puf=0, delta=tau implies Puf=1) follow from the same formula rather than being used to calibrate it. The experiments measure concealable unfairness for concrete manipulation methods and set the detection threshold from model accuracies; they do not regress the formula's constants, so no fitted input is renamed as a prediction. Self-citations to prior work by the same authors appear in related-work and manipulation examples, but the central theorem rests on elementary spherical-cap volume facts cited to Li (2010), the Gamma function (NIST), and Wallis integrals, none of which presuppose the target result. One correctness caveat, distinct from circularity: the proof's own decomposition requires a plus sign between the integral and the delta/tau term in Theorem 4.3 (the printed minus sign makes the expression exceed 1 at, e.g., n=2, delta/tau=1/2). This is an algebraic error in the printed formula, not a reduction of the result to its inputs, so it does not change the circularity verdict.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central detection-rate theorem rests on a stack of modeling assumptions: binary classification with DP as the fairness metric, L2 loss, a hyperplane fair set, a prior ball centered on the auditor's labels, and a uniform distribution over that ball. The uniform-prior assumption is the most fragile because the resulting probability has no frequentist interpretation if the platform's model distribution differs. No new physical entities are introduced.

free parameters (1)
  • tau (expectability threshold) = 1 - max accuracy on the target (Section 5.4)
    Defines the radius of the prior ball Ha. In the theory it is an input; in the experiments it is calibrated from the maximum accuracy of trained models, so the resulting detection rates depend on this data-driven choice.
assumptions (5)
  • domain assumption The true platform model lies in the auditor prior: hp in Ha (Equation (4), Section 3.1).
    Ensures no false positives but presupposes the auditor's prior is well-calibrated to the real task.
  • domain assumption The fair set F has non-empty intersection with the prior Ha (Equation (4), Section 3.1).
    Otherwise the audit would be unnecessary; also required for Theorem 3.2.
  • domain assumption The auditor's labeled dataset Da provides ground-truth labels, making the prior ball centered at the ground-truth ha (Definition 4.1 and Appendix A).
    If Da labels are noisy or biased, the geometric calculations in Theorem 4.3 no longer describe the actual prior.
  • ad hoc to paper The model distribution over Ha is uniform (uninformative prior, Section 4.2 before Theorem 4.3).
    Introduced to make the detection probability computable; not derived from first principles or measured.
  • standard math Volume formulas for hyperspherical caps and Wallis integrals (Li 2010; NIST 2013).
    Used in the proof of Theorem 4.3 in Appendix A.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust ML Auditing using Prior Knowledge." pith.science (2026). https://pith.science/paper/UTPLHNCG

@misc{pith2026250504796,
  author       = {Pith},
  title        = {Pith review of: Robust ML Auditing using Prior Knowledge},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UTPLHNCG}},
  note         = {Machine review of arXiv:2505.04796}
}
read the original abstract

Among the many technical challenges to enforcing AI regulations, one crucial yet underexplored problem is the risk of audit manipulation. This manipulation occurs when a platform deliberately alters its answers to a regulator to pass an audit without modifying its answers to other users. In this paper, we introduce a novel approach to manipulation-proof auditing by taking into account the auditor's prior knowledge of the task solved by the platform. We first demonstrate that regulators must not rely on public priors (e.g. a public dataset), as platforms could easily fool the auditor in such cases. We then formally establish the conditions under which an auditor can prevent audit manipulations using prior knowledge about the ground truth. Finally, our experiments with two standard datasets illustrate the maximum level of unfairness a platform can hide before being detected as malicious. Our formalization and generalization of manipulation-proof auditing with a prior opens up new research directions for more robust fairness audits.

Figures

Figures reproduced from arXiv: 2505.04796 by the authors.

Figure 1
Figure 1. The auditing process as conducted by an auditor, which proceeds in three steps. The platform exposes a model hp to the users. To appear fair to the auditor while not deteriorating the utility for its users, the platform manipulates its answers on the audit set S. µ(h) = P(X,A)∼D (h(X) = 1|A = 1) − P(X,A)∼D (h(X) = 1|A = 0) (1) For a platform, DP is the easiest metric to manipulate (Yan & Zhang, 2022; Ajarra et al., … view at source ↗
Figure 2
Figure 2. Representation of the auditor prior Ha, the honest plat￾form model hp and a corresponding malicious model hm on the fair F plane. The red area represents the area where platforms optimal manipulations are detected as dishonest: they fall outside of the blue region of F optimal manipulation is the projection of hp on F: h ∗ m = projF (hp) = arg min h∈F d(h, hp). (6) The distance d in Equation (6) is the value of risk… view at source ↗
Figure 3
Figure 3. The concealable unfairness by the platform for different detection scores and manipulation strategies. We highlight this for two features of the CelebA dataset (left) and for two different ML models trained on the ACSEmployment dataset (right). The horizontal red line indicates the DP of the most unfair model without manipulation. 5. Empirical Evaluation We now empirically quantify the extent to which the plat￾form … view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The concealable unfairness for different audit budgets (i.e., data samples from the labeled dataset). We highlight this for two features of the CelebA dataset (left) and for two different ML models trained on the ACSEmployment dataset (right). Optimal Label Transport (…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 42 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    Fairwashing: the risk of rationalization

    A \" vodji, U., Arai, H., Fortineau, O., Gambs, S., Hara, S., and Tapp, A. Fairwashing: the risk of rationalization. In Chaudhuri, K. and Salakhutdinov, R. (eds.), Proceedings of the 36th International Conference on Machine Learning, ICML 2019, 9-15 June 2019, Long Beach, California, USA , volume 97 of Proceedings of Machine Learning Research, pp.\ 161--1...

    work page 2019

  3. [3]

    Characterizing the risk of fairwashing

    A \" vodji, U., Arai, H., Gambs, S., and Hara, S. Characterizing the risk of fairwashing. In Ranzato, M., Beygelzimer, A., Dauphin, Y. N., Liang, P., and Vaughan, J. W. (eds.), Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, virtual, pp.\ 14822--14834, 2021

    work page 2021

  4. [4]

    Active Fourier Auditor for Estimating Distributional Properties of ML Models , 2024

    Ajarra, A., Ghosh, B., and Basu, D. Active Fourier Auditor for Estimating Distributional Properties of ML Models , 2024

    work page 2024

  5. [5]

    Albarghouthi, A., D'Antoni, L., Drews, S., and Nori, A. V. FairSquare : Probabilistic verification of program fairness. Proc. ACM Program. Lang., 1 0 (OOPSLA): 0 80:1--80:30, 2017. doi:10.1145/3133904

    doi:10.1145/3133904 2017

  6. [6]

    J., Pasliev, P., Dombrowski, A., M \" u ller, K., and Kessel, P

    Anders, C. J., Pasliev, P., Dombrowski, A., M \" u ller, K., and Kessel, P. Fairwashing explanations with off-manifold detergent. In Proceedings of the 37th International Conference on Machine Learning, ICML 2020, 13-18 July 2020, Virtual Event , volume 119 of Proceedings of Machine Learning Research, pp.\ 314--323. PMLR , 2020

    work page 2020

  7. [7]

    Fairness seen as global sensitivity analysis

    B \'e nesse, C., Gamboa, F., Loubes, J.-M., and Boissin, T. Fairness seen as global sensitivity analysis. Machine Learning, 113 0 (5): 0 3205--3232, 2024. ISSN 1573-0565. doi:10.1007/s10994-022-06202-y

    doi:10.1007/s10994-022-06202-y 2024

  8. [9]

    Provable Fairness for Neural Network Models using Formal Verification , 2022

    Borca-Tasciuc , G., Guo, X., Bak, S., and Skiena, S. Provable Fairness for Neural Network Models using Formal Verification , 2022

    work page 2022

Show all 64 references
  1. [10]

    and Bie, T

    Buyl, M. and Bie, T. D. Optimal transport of classifiers to fairness. In Koyejo, S., Mohamed, S., Agarwal, A., Belgrave, D., Cho, K., and Oh, A. (eds.), Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, NeurI...

    2022

  2. [11]

    Building Classifiers with Independency Constraints

    Calders, T., Kamiran, F., and Pechenizkiy, M. Building Classifiers with Independency Constraints . In 2009 IEEE International Conference on Data Mining Workshops , pp.\ 13--18, 2009. doi:10.1109/ICDMW.2009.83

    2009 doi

  3. [12]

    and Haas, C

    Caton, S. and Haas, C. Fairness in Machine Learning : A Survey . ACM Comput. Surv., 56 0 (7): 0 166:1--166:38, 2024. ISSN 0360-0300. doi:10.1145/3616865

    2024 doi

  4. [13]

    Cen, S. H. and Alur, R. From Transparency to Accountability and Back : A Discussion of Access and Evidence in AI Auditing , 2024

    2024

  5. [14]

    Cherian, J. J. and Cand \`e s, E. J. Statistical Inference for Fairness Auditing . Journal of Machine Learning Research, 25 0 (149): 0 1--49, 2024. ISSN 1533-7928

    2024

  6. [15]

    Auditing fairness by betting

    Chugg, B., Cortes - Gomez, S., Wilder, B., and Ramdas, A. Auditing fairness by betting. In Oh, A., Naumann, T., Globerson, A., Saenko, K., Hardt, M., and Levine, S. (eds.), Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing...

    2023

  7. [16]

    Algorithmic accountability act of 2022

    Congress, U. Algorithmic accountability act of 2022. https://www.congress.gov/bill/117th-congress/house-bill/6580, 2022

    2022

  8. [17]

    D., and Buolamwini, J

    Costanza-Chock, S., Raji, I. D., and Buolamwini, J. Who audits the auditors? recommendations from a field scan of the algorithmic auditing ecosystem. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, pp.\ 1571--1583, 2022

    2022

  9. [18]

    Enforcing the digital markets act: institutional choices, compliance, and antitrust

    Cr \'e mer, J., Dinielli, D., Heidhues, P., Kimmelman, G., Monti, G., Podszun, R., Schnitzer, M., Scott Morton, F., and De Streel, A. Enforcing the digital markets act: institutional choices, compliance, and antitrust. Journal of Antitrust Enforcement, 11 0 (3): 0 315--349, 2023

    2023

  10. [19]

    Fairness auditing with multi-agent collaboration

    de Vos, M., Dhasade, A., Garcia Bourr \'e e, J., Kermarrec, A.-M., Le Merrer, E., Rottembourg, B., and Tredan, G. Fairness auditing with multi-agent collaboration. In ECAI 2024, pp.\ 1116--1123. IOS Press, 2024

    2024

  11. [20]

    a \" a n \

    Deng, W. H., Guo, B. B., DeVrio, A., Shen, H., Eslami, M., and Holstein, K. Understanding practices, challenges, and opportunities for user-engaged algorithm auditing in industry practice. In Schmidt, A., V \" a \" a n \" a nen, K., Goyal, T., Kristensson, P. O., Peters, A., M...

    2023

  12. [21]

    Evaluating fairness using permutation tests

    DiCiccio, C., Vasudevan, S., Basu, K., Kenthapadi, K., and Agarwal, D. Evaluating fairness using permutation tests. In Gupta, R., Liu, Y., Tang, J., and Prakash, B. A. (eds.), KDD '20: The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Virtual Event, CA, US...

    2020

  13. [22]

    Retiring adult: New datasets for fair machine learning

    Ding, F., Hardt, M., Miller, J., and Schmidt, L. Retiring adult: New datasets for fair machine learning. In Ranzato, M., Beygelzimer, A., Dauphin, Y. N., Liang, P., and Vaughan, J. W. (eds.), Advances in Neural Information Processing Systems 34: Annual Conference on Neural Inf...

    2021

  14. [23]

    Attribution-based Explanations that Provide Recourse Cannot be Robust

    Fokkema, H., de Heide, R., and van Erven, T. Attribution-based Explanations that Provide Recourse Cannot be Robust . Journal of Machine Learning Research, 24 0 (360): 0 1--37, 2023. ISSN 1533-7928

    2023

  15. [24]

    Faking fairness via stealthily biased sampling

    Fukuchi, K., Hara, S., and Maehara, T. Faking fairness via stealthily biased sampling. In The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, The Thirty-Second Innovative Applications of Artificial Intelligence Conference, IAAI 2020, The Tenth AAAI Symposi...

    2020

  16. [25]

    On the relevance of APIs facing fairwashed audits, 2023

    Garcia Bourr \'e e, J., Le Merrer, E., Tredan, G., and Rottembourg, B. On the relevance of APIs facing fairwashed audits, 2023

    2023

  17. [26]

    Ghosh, B., Basu, D., and Meel, K. S. Justicia: A stochastic SAT approach to formally verify fairness. In Thirty-Fifth AAAI Conference on Artificial Intelligence, AAAI 2021, Thirty-Third Conference on Innovative Applications of Artificial Intelligence, IAAI 2021, The Eleventh S...

    2021

  18. [27]

    Ghosh, B., Basu, D., and Meel, K. S. Algorithmic fairness verification with graphical models. In Thirty-Sixth AAAI Conference on Artificial Intelligence, AAAI 2022, Thirty-Fourth Conference on Innovative Applications of Artificial Intelligence, IAAI 2022, The Twelveth Symposiu...

    2022

  19. [28]

    Under manipulations, are some AI models harder to audit? In 2024 IEEE Conference on Secure and Trustworthy Machine Learning ( SaTML ) , pp.\ 644--664, 2024

    Godinot, A., Le Merrer, E., Tr \'e dan, G., Penzo, C., and Ta \"i ani, F. Under manipulations, are some AI models harder to audit? In 2024 IEEE Conference on Secure and Trustworthy Machine Learning ( SaTML ) , pp.\ 644--664, 2024. doi:10.1109/SaTML59370.2024.00038

    2024

  20. [29]

    Equality of opportunity in supervised learning

    Hardt, M., Price, E., and Srebro, N. Equality of opportunity in supervised learning. In Lee, D. D., Sugiyama, M., von Luxburg, U., Guyon, I., and Garnett, R. (eds.), Advances in Neural Information Processing Systems 29: Annual Conference on Neural Information Processing System...

    2016

  21. [30]

    Wasserstein fair classification

    Jiang, R., Pacchiano, A., Stepleton, T., Jiang, H., and Chiappa, S. Wasserstein fair classification. In Globerson, A. and Silva, R. (eds.), Proceedings of the Thirty-Fifth Conference on Uncertainty in Artificial Intelligence, UAI 2019, Tel Aviv, Israel, July 22-25, 2019 , volu...

    2019

  22. [31]

    Decision theory for discrimination-aware classification

    Kamiran, F., Karim, A., and Zhang, X. Decision theory for discrimination-aware classification. In 2012 IEEE 12th international conference on data mining, pp.\ 924--929. IEEE, 2012

    2012

  23. [32]

    P., Ghorbani, A., and Zou, J

    Kim, M. P., Ghorbani, A., and Zou, J. Multiaccuracy: Black-box post-processing for fairness in classification. In Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society, pp.\ 247--254, 2019

    2019

  24. [33]

    Fooling SHAP with stealthily biased sampling

    Laberge, G., A \" vodji, U., Hara, S., Marchand, M., and Khomh, F. Fooling SHAP with stealthily biased sampling. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, May 1-5, 2023 . OpenReview.net, 2023

    2023

  25. [34]

    S., Pandit, A., Kalicki, C

    Lam, M. S., Pandit, A., Kalicki, C. H., Gupta, R., Sahoo, P., and Metaxa, D. Sociotechnical Audits : Broadening the Algorithm Auditing Lens to Investigate Targeted Advertising . Proc. ACM Hum.-Comput. Interact., 7 0 (CSCW2): 0 360:1--360:37, 2023. doi:10.1145/3610209

    2023 doi

  26. [35]

    and Tr \'e dan, G

    Le Merrer, E. and Tr \'e dan, G. Remote explainability faces the bouncer problem. Nature Machine Intelligence, 2 0 (9): 0 529--539, 2020. ISSN 2522-5839. doi:10.1038/s42256-020-0216-z

    2020 doi

  27. [36]

    Gradient-based learning applied to document recognition

    Lecun, Y., Bottou, L., Bengio, Y., and Haffner, P. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86 0 (11): 0 2278--2324, 1998. doi:10.1109/5.726791

    1998 doi

  28. [37]

    Concise formulas for the area and volume of a hyperspherical cap

    Li, S. Concise formulas for the area and volume of a hyperspherical cap. Asian Journal of Mathematics & Statistics, 4 0 (1): 0 66--70, 2010

    2010

  29. [38]

    Deep learning face attributes in the wild

    Liu, Z., Luo, P., Wang, X., and Tang, X. Deep learning face attributes in the wild. In 2015 IEEE International Conference on Computer Vision, ICCV 2015, Santiago, Chile, December 7-13, 2015 , pp.\ 3730--3738. IEEE Computer Society, 2015. doi:10.1109/ICCV.2015.425

    2015 doi

  30. [39]

    Too relaxed to be fair

    Lohaus, M., Perrot, M., and von Luxburg, U. Too relaxed to be fair. In Proceedings of the 37th International Conference on Machine Learning, ICML 2020, 13-18 July 2020, Virtual Event , volume 119 of Proceedings of Machine Learning Research, pp.\ 6360--6369. PMLR , 2020

    2020

  31. [40]

    Online fairness auditing through iterative refinement

    Maneriker, P., Burley, C., and Parthasarathy, S. Online fairness auditing through iterative refinement. In Singh, A. K., Sun, Y., Akoglu, L., Gunopulos, D., Yan, X., Kumar, R., Ozcan, F., and Ye, J. (eds.), Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery a...

    2023

  32. [41]

    A survey on bias and fairness in machine learning

    Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., and Galstyan, A. A survey on bias and fairness in machine learning. ACM computing surveys (CSUR), 54 0 (6): 0 1--35, 2021

    2021

  33. [42]

    Reasons to doubt the impact of ai risk evaluations

    Mukobi, G. Reasons to doubt the impact of ai risk evaluations. ArXiv preprint, abs/2408.02565, 2024

    2024 arXiv

  34. [43]

    Can auditing eliminate bias from algorithms?, 2021

    Ng, A. Can auditing eliminate bias from algorithms?, 2021. Accessed: 2025-01-07

    2021

  35. [44]

    Nist digital library of mathematical functions, 2013

    NIST . Nist digital library of mathematical functions, 2013. Release 1.0.6 of 2013-05-06

    2013

  36. [45]

    Responsible and regulatory conform machine learning for medicine: a survey of challenges and solutions

    Petersen, E., Potdevin, Y., Mohammadi, E., Zidowitz, S., Breyer, S., Nowotka, D., Henn, S., Pechmann, L., Leucker, M., Rostalski, P., et al. Responsible and regulatory conform machine learning for medicine: a survey of challenges and solutions. IEEE Access, 10: 0 58375--58418, 2022

    2022

  37. [46]

    M., Hanna, A., and Paullada, A

    Raji, D., Denton, E., Bender, E. M., Hanna, A., and Paullada, A. Ai and the everything in the whole wide world benchmark. In Vanschoren, J. and Yeung, S. (eds.), Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks, volume 1, 2021

    2021

  38. [47]

    Raji, I. D. The Anatomy of AI Audits : Form , Process , and Consequences . In Bullock, J. B., Chen, Y.-C., Himmelreich, J., Hudson, V. M., Korinek, A., Young, M. M., and Zhang, B. (eds.), The Oxford Handbook of AI Governance , pp.\ 0. Oxford University Press, 2024. doi:10.1093...

    2024

  39. [48]

    D., Xu, P., Honigsberg, C., and Ho, D

    Raji, I. D., Xu, P., Honigsberg, C., and Ho, D. Outsider Oversight : Designing a Third Party Audit Ecosystem for AI Governance . In Proceedings of the 2022 AAAI / ACM Conference on AI , Ethics , and Society , AIES '22, pp.\ 557--571, New York, NY, USA, 2022. Association for Co...

    2022

  40. [49]

    Ribeiro, M. H. Is Facebook `` Standard '' Algorithm Polarizing ? https://doomscrollingbabel.manoel.xyz/p/is-facebook-standard-algorithm-polarizing, 2024

    2024

  41. [50]

    Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead

    Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature machine intelligence, 1 0 (5): 0 206--215, 2019

    2019

  42. [51]

    S., Yaghini, M., Dullerud, N., Wyllie, S

    Shamsabadi, A. S., Yaghini, M., Dullerud, N., Wyllie, S. C., A \" vodji, U., Alaagib, A., Gambs, S., and Papernot, N. Washing the unwashable : On the (im)possibility of fairwashing detection. In Koyejo, S., Mohamed, S., Agarwal, A., Belgrave, D., Cho, K., and Oh, A. (eds.), Ad...

    2022

  43. [52]

    S., Wyllie, S

    Shamsabadi, A. S., Wyllie, S. C., Franzese, N., Dullerud, N., Gambs, S., Papernot, N., Wang, X., and Weller, A. Confidential-profitt: Confidential proof of fair training of trees. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, ...

    2023

  44. [53]

    H., and Nguyen, V

    Si, N., Murthy, K., Blanchet, J. H., and Nguyen, V. A. Testing group fairness via optimal transport projections. In Meila, M. and Zhang, T. (eds.), Proceedings of the 38th International Conference on Machine Learning, ICML 2021, 18-24 July 2021, Virtual Event , volume 139 of P...

    2021

  45. [54]

    Fooling LIME and SHAP : Adversarial Attacks on Post hoc Explanation Methods

    Slack, D., Hilgard, S., Jia, E., Singh, S., and Lakkaraju, H. Fooling LIME and SHAP : Adversarial Attacks on Post hoc Explanation Methods . In Proceedings of the AAAI / ACM Conference on AI , Ethics , and Society , AIES '20, pp.\ 180--186, New York, NY, USA, 2020. Association ...

    2020

  46. [55]

    Distill-and- Compare : Auditing Black-Box Models Using Transparent Model Distillation

    Tan, S., Caruana, R., Hooker, G., and Lou, Y. Distill-and- Compare : Auditing Black-Box Models Using Transparent Model Distillation . In Proceedings of the 2018 AAAI / ACM Conference on AI , Ethics , and Society , AIES '18, pp.\ 303--310, New York, NY, USA, 2018. Association f...

    2018

  47. [56]

    Taskesen, B., Blanchet, J., Kuhn, D., and Nguyen, V. A. A Statistical Test for Probabilistic Fairness . In Proceedings of the 2021 ACM Conference on Fairness , Accountability , and Transparency , FAccT '21, pp.\ 648--665, New York, NY, USA, 2021. Association for Computing Mach...

    2021

  48. [57]

    Facebook made big mistake in data it provided to researchers, undermining academic work

    Timberg, C. Facebook made big mistake in data it provided to researchers, undermining academic work. Washington Post, 2021. ISSN 0190-8286

    2021

  49. [58]

    Regulation (eu) 2022/1925 of the european parliament and of the council of 14 september 2022 on contestable and fair markets in the digital sector (digital markets act)

    Union, E. Regulation (eu) 2022/1925 of the european parliament and of the council of 14 september 2022 on contestable and fair markets in the digital sector (digital markets act). https://eur-lex.europa.eu/eli/reg/2022/1925/oj/eng, 2022

    2022

  50. [59]

    Trustless audits without revealing data or models

    Waiwitlikhit, S., Stoica, I., Sun, Y., Hashimoto, T., and Kang, D. Trustless audits without revealing data or models. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024 . OpenReview.net, 2024

    2024

  51. [60]

    Neural network credit scoring models

    West, D. Neural network credit scoring models. Computers & operations research, 27 0 (11-12): 0 1131--1152, 2000

    2000

  52. [61]

    XAudit : A Theoretical Look at Auditing with Explanations , 2022

    Yadav, C., Moshkovitz, M., and Chaudhuri, K. XAudit : A Theoretical Look at Auditing with Explanations , 2022

    2022

  53. [62]

    R., Boneh, D., and Chaudhuri, K

    Yadav, C., Chowdhury, A. R., Boneh, D., and Chaudhuri, K. Fairproof : Confidential and certifiable fairness for neural networks. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024 . OpenReview.net, 2024

    2024

  54. [63]

    and Zhang, C

    Yan, T. and Zhang, C. Active fairness auditing. In Chaudhuri, K., Jegelka, S., Song, L., Szepesv \' a ri, C., Niu, G., and Sabato, S. (eds.), International Conference on Machine Learning, ICML 2022, 17-23 July 2022, Baltimore, Maryland, USA , volume 162 of Proceedings of Machi...

    2022

  55. [64]

    and Gordon, G

    Zhao, H. and Gordon, G. J. Inherent tradeoffs in learning fair representations. J. Mach. Learn. Res., 23: 0 57:1--57:26, 2022

    2022

  56. [65]

    Leveling down in computer vision: Pareto inefficiencies in fair deep classifiers

    Zietlow, D., Lohaus, M., Balakrishnan, G., Kleindessner, M., Locatello, F., Sch \" o lkopf, B., and Russell, C. Leveling down in computer vision: Pareto inefficiencies in fair deep classifiers. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New O...

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.