Pith. sign in

REVIEW 5 major objections 6 minor 100 references

Out-of-Distribution Detection with Overlap Index

T0 review · 5 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper proposes an overlap-index-based confidence score that detects out-of-distribution samples with accuracy competitive with deep detectors while requiring far less computation and memory.

desk verdict A genuinely new training-free OOD heuristic built on an elementary bound, with real empirical wins in feature spaces – but the theoretical anchor doesn't cover the score and one protocol leaks test labels. read the letter →

arxiv 2412.06168 v1 pith:WRLSUR7K submitted 2024-12-09 cs.LG stat.ML

Hao Fu , Prashanth Krishnamurthy , Siddharth Garg , Farshad Khorrami This is my paper · ORCID
classification cs.LGstat.ML
keywords out-of-distributiondetectionoverlapindexnon-parametricconfidencescoretotalvariationdistanceradialbinfunctionsmodelaccuracyboundbackdoorsmall-samplerobustness
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 tries to establish that a single lightweight statistic, derived from the overlap index between probability distributions, can serve as an out-of-distribution (OOD) detector across small UCI datasets, CIFAR-10/100, and ImageNet-scale benchmarks, with accuracy competitive with deep detectors at a fraction of the computational and memory cost. The statistic is an upper bound on the overlap index combining the distance between cluster means with the total variation distance over radial bins; with one test point against a few in-distribution samples it becomes a confidence score. The paper argues this score is non-parametric, needs no covariance inversion or neural network, runs in O(k) time per sample (about 3 ms), and inherits insensitivity to small distributional shifts and robustness to Huber epsilon-contamination. If these claims hold, reliable OOD detection no longer requires a trained deep model, which matters for small-data regimes and low-power devices.

What carries the argument

The load-bearing object is the overlap index $\eta(P,Q) = \int \min(f_P, f_Q)\,dx$, together with a new upper bound that splits it into a mean-distance term and a total-variation term over a subset $A$. The implementation replaces the intractable variation term by a maximum over radial-bin condition functions, yielding a finite-sample score with $O(k+1)$ time per query and $O(k+1)$ memory. What carries the argument empirically is that the two terms, mean separation and radial-bin occupancy difference, amplify each other: neither alone separates ID from OOD on CIFAR-10, but their sum does.

What would settle it

Take in-distribution data from a standard normal and out-of-distribution data from a normal with the same zero mean and a slightly larger variance, so that both the mean-distance term and the radial-bin occupancies nearly coincide; the paper's own observation that centering both clusters at the origin degrades performance predicts an AUROC near chance, and measuring the AUROC of Algorithm 1 on this pair would settle whether the score is anything beyond a mean-and-norm statistic.

Watch

Extended reading notes

Core claim

On its own terms, the central discovery is that the overlap index admits a computable upper bound that can be turned into a ranking score for out-of-distribution detection. Theorem III.3 states that for bounded distributions, $\eta \le 1 - \frac{\|\mu_{D_+} - \mu_{D_-}\|}{2 r_{A^c}} - \frac{r_{A^c} - r_A}{r_{A^c}} \delta_A$, and Corollary III.4 plus Algorithm 1 realize it finitely: the total-variation term is replaced by the largest absolute difference of expectations under radial-bin condition functions $g_j(x) = \mathbf{1}\{r_{j-1} \le \|x\| \le r_j\}$, and the whole expression is evaluated with one query point as $D_+$ and $m$ stored in-distribution samples as $D_-$. The paper's claim is that this finite-sample score ranks in-distribution inputs above out-of-distribution inputs across UCI datasets, CIFAR-10 and CIFAR-100 in raw and feature space, and ImageNet-scale benchmarks, with per-sample time and memory that do not scale with dimension. The same bound is also presented as a tool for estimating the overlap index itself and for bounding model accuracy under domain shift.

Load-bearing premise

The load-bearing premise is that the finite-sample bound, computed with a single query point and a handful of stored samples, ranks in-distribution inputs above out-of-distribution inputs; the paper demonstrates this empirically, while the bound's discriminating term can vanish exactly for the most extreme-norm outliers, so the ranking property is not guaranteed by the theorem.

Editorial extensions

If this is right

  • OOD detection becomes feasible on low-power devices: with $k=100$ bins the detector runs in about 3 ms per sample and stores only $O(k+1)$ values, with no trained network to maintain.
  • The detector can work in raw input space or in any pretrained feature space, and it avoids covariance inversion and Gaussian assumptions, removing two common sources of numerical instability.
  • The same computation serves three tasks: OOD detection, overlap-index estimation, and an upper bound on accuracy under domain shift, so a single deployed score can be reused for monitoring.
  • Small distributional variations and Huber $\epsilon$-contamination do not trigger the detector, which means it will ignore benign shifts while still flagging genuine OOD inputs.
  • In backdoor settings, clean samples can be treated as ID and poisoned samples as OOD, yielding high detection rates from as few as ten clean samples per class and no hyperparameter tuning.

Reading between the lines

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

  • An implicit consequence is that the method's discriminative power depends on OOD samples separating in mean or radial-bin occupancy; OOD data engineered to share both (for instance, by centering) is a blind spot the paper acknowledges in its limitation section.
  • The radial-bin condition functions are one convenient choice; replacing them with class-conditional or learned one-dimensional projections could extend the same bound to settings where norm histograms overlap.
  • Because the score costs about 3 ms per sample and needs no neural network, it could serve as a fast pre-filter that only invokes a deep detector when the score is ambiguous.
  • Theorem V.3's accuracy bound offers a cheap diagnostic: the gap between measured accuracy and the calculated upper bound in a given feature space indicates how detectable a domain shift is there, which the appendix already uses to recommend monitoring spaces.
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 / 6 minor

Summary. The paper proposes an out-of-distribution (OOD) detector based on a newly derived upper bound on the overlap index (OI). The detector scores a test point x by f(x) = ComputeBound({x}, {xi}, {gj}), combining the distance between the test point and the mean of the in-distribution (ID) samples with discrepancies of radial-bin indicator functions. The authors prove an upper bound for bounded distributions (Theorem III.3), instantiate it as a finite-sample score (Algorithm 1), and evaluate the method on UCI datasets, raw CIFAR-10/100 images, pretrained feature spaces, large-scale ImageNet-style settings, and backdoor detection. They claim the detector is non-parametric, lightweight, memory-efficient, and competitive with state-of-the-art OOD detectors.

Significance. If the central claims held, a lightweight, interpretable, non-parametric OOD score with provable robustness properties would be a useful contribution, especially in small-ID-sample regimes where deep detectors are impractical. The theoretical derivation in Theorem III.3 and Corollary III.4 appears mathematically sound within its continuous-density, bounded-support assumptions. The feature-space experiments (Tables VI) and backdoor-detection experiments (Table IX) are promising and suggest the score has empirical utility. However, the paper's theoretical anchor does not directly justify the single-point score used in the detector, the efficiency claim is inconsistent with Algorithm 1 as written, the 'Extra Information' protocol leaks OOD test-set statistics, and the raw-image CIFAR-10 results (average AUROC 68.0) do not support the broad 'competitive with state-of-the-art' claim. The contribution is interesting but needs substantial reworking of both the theoretical framing and the empirical protocol.

major comments (5)
  1. [Section III.B.3, Algorithm 1] The detector instantiates D+ as a single test point (d=1), making D+ a Dirac measure. Theorem III.3 is stated for continuous random variables with probability density functions, so the theorem does not directly apply to this setting. The computed value f(x) is therefore not established as an upper bound on the true overlap between the test point and the ID distribution; it is an empirical score whose ranking behavior is only justified by experiments. The paper should either extend the bound to point masses or empirical distributions, or explicitly reframe the score as a heuristic inspired by the bound rather than as a valid OI bound.
  2. [Section III.C.1 vs. Algorithm 1] Algorithm 1 defines rB = max_{x in B} ||x||, where B includes the test point x, and the condition functions use r_j = j rB/k. Therefore the radial bins and the values (1/m) sum_j g_j(x_i) depend on the test point through rB. The claim in Section III.C.1 that these quantities can be precomputed and stored, giving O(k+1) online time and O(k+1) memory, is not correct as written: changing rB changes all bin boundaries, so the bin averages cannot be reused across test points unless rB is fixed from the ID data alone. This affects the central efficiency argument and should be corrected or clarified.
  3. [Algorithm 1, self-normalizing rB] Because rB includes the test point, for an extreme outlier with ||x|| larger than every ID sample, the outermost radial bin containing x has rA = rB, so its coefficient (rB - rA)/(2 rB) vanishes identically. In that regime the score tends to a direction-independent constant around 1/2 - (1/2) max_j p_j, which is precisely the opposite of the desired behavior for an OOD detector. This is a load-bearing weakness in the proposed score's discrimination mechanism, and it is not addressed by Theorem III.3, which assumes a fixed normalization. The authors should either fix rB using only the ID sample set, modify the score to avoid self-normalization, or provide an empirical analysis showing that this saturation does not occur on realistic OOD inputs.
  4. [Section IV.B, Table III, 'Extra Information' protocol] The 'Extra Information' results are obtained by building a contaminated dataset from a merge of all ID and OOD samples, then subtracting the mean of 100 randomly selected samples from that merge. This gives the detector access to OOD test-set statistics, even if labels are not used directly, and makes the reported 96.2% average AUROC incomparable with the baseline methods that do not receive such information. The paper should report a version of the protocol using only ID samples for the mean subtraction, or clearly label the current protocol as transductive/using unlabeled OOD data, and should not present the inflated numbers as evidence of competitiveness with the baselines.
  5. [Section IV.B, Table III raw-image results and Abstract] The average raw CIFAR-10 AUROC of 68.0% is well below modern OOD detection methods, which typically report averages above 90% on this benchmark. The comparisons in Table III are limited to older methods (DCAE, AnoGAN, Deep SVDD, OCGAN, GradCon), and even among these the advantage is modest. The abstract's claim that the method is 'competitive with state-of-the-art OOD detectors in terms of detection accuracy' is not supported by the raw-image experiments. The authors should temper the claim to the specific settings where the method is competitive (e.g., feature-space and low-ID-sample regimes) or add modern baselines and report a complete comparison.
minor comments (6)
  1. [Section III.B.3] The notation f(x) = ComputeBound({x}, {xi}, {gj}) is introduced after Algorithm 1, but the algorithm is described as returning an upper bound 'eta'; the relationship between the returned value and the score f(x) should be made explicit.
  2. [Section V.B] The OI estimation variant introduces r' as a free parameter and fixes it to the median norm in the experiments, but the sensitivity of the estimator to r' is not discussed; a brief analysis or ablation would clarify the method's robustness.
  3. [Table XI] In the GTSRB rows, the text 'Fail: dependent data features' appears inline inside numeric columns; this should be moved to a footnote or explained in the caption, and the corresponding entries should be marked with a placeholder.
  4. [Appendix H] The caption of Figure 9 refers to 'the ratio of clean samples to the entire testing samples' while the formula (10) uses the parameter sigma; the notation should be unified.
  5. [References] References [16] and [25] are lecture-note URLs without authors or peer-reviewed venues; the authors should replace them with citable publications or at least provide full author and date information.
  6. [Table VIII caption] The caption contains the typo 'OOD Datasts'; it should read 'OOD Datasets'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the OI bound is derived, the OOD score is externally benchmarked, and the propositions are direct consequences of the defined score.

full rationale

I walked the derivation chain in Sections III-V. Theorem III.3 is proved in Appendix A from the definitions of OI/TVD and boundedness; Corollary III.4 follows by applying the triangle inequality to condition-function expectations, so neither presupposes the OOD ranking it is used for. The confidence score f(x) = ComputeBound({x},{xi},{gj}) is then evaluated on external benchmarks (UCI, CIFAR-10/100, ImageNet-scale) with fixed k=100 radial-bin functions; the paper tunes k by ablation but does not fit any parameter to test labels, and the reported AUROCs are comparisons against independent baselines, so the detection claim is not forced by construction. Propositions V.1 and V.2 are direct algebraic bounds on the defined score, not empirical predictions, and Theorem V.3 is a conditional accuracy inequality whose proof uses the already-derived overlap bound; none of these convert an input into an output by definition. The self-citations (e.g., [1], [8], [9], [61]-[64], [75]) appear in related-work and backdoor-detection contexts and are not load-bearing for the central OI bound or the OOD benchmarks. The limitation noted in Section VI (score degrades when ID/OOD clusters are shifted to the origin) is an honest external-impossibility citation [100], not a circular step. A soundness concern remains - for d=1 the computed quantity is only a vacuous upper bound on OI and its ranking power is empirical - but vacuousness is not circularity, and the hard-rule requirement to exhibit a specific reduction to inputs is not met.

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

The central detector depends on the user-chosen k and the shell condition functions, plus the bounded-domain approximation. Theorem V.3 introduces an unstated accuracy-decomposition assumption that is not justified. No new physical or mathematical entities are postulated.

free parameters (4)
  • k = 100
    Number of radial condition functions; chosen via ablation (Fig. 3b, Fig. 5) with recommended range 50 to 200.
  • condition functions g_j = 1{r_{j-1} <= ||x|| <= r_j}
    Indicator functions on concentric shells are a modeling choice; the paper notes other condition functions are worth exploring.
  • norm choice = L2
    The bound is stated for any norm; L2 is selected for all experiments.
  • r' for OI estimation = median Euclidean norm of combined set
    Used only in the OI-estimation variant eta', not in the main detector.
assumptions (3)
  • domain assumption Distributions are supported on a bounded domain B (Theorem III.3).
    The theorem requires boundedness; the paper approximates unbounded distributions with bounded ones based on finite samples.
  • standard math Continuous random variables with probability density functions exist.
    The definitions of OI and TVD in Section III-B integrate densities f_P and f_Q.
  • ad hoc to paper Model accuracy on the overlap mass is p and on the disjoint mass is q (Theorem V.3).
    The proof assumes Acc = p*eta + q*(1-eta), but the stated conditions only give accuracy p on D and q on D* \ D, not on the overlap region.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Out-of-Distribution Detection with Overlap Index." pith.science (2026). https://pith.science/paper/WRLSUR7K

@misc{pith2026241206168,
  author       = {Pith},
  title        = {Pith review of: Out-of-Distribution Detection with Overlap Index},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WRLSUR7K}},
  note         = {Machine review of arXiv:2412.06168}
}
abstract

Out-of-distribution (OOD) detection is crucial for the deployment of machine learning models in the open world. While existing OOD detectors are effective in identifying OOD samples that deviate significantly from in-distribution (ID) data, they often come with trade-offs. For instance, deep OOD detectors usually suffer from high computational costs, require tuning hyperparameters, and have limited interpretability, whereas traditional OOD detectors may have a low accuracy on large high-dimensional datasets. To address these limitations, we propose a novel effective OOD detection approach that employs an overlap index (OI)-based confidence score function to evaluate the likelihood of a given input belonging to the same distribution as the available ID samples. The proposed OI-based confidence score function is non-parametric, lightweight, and easy to interpret, hence providing strong flexibility and generality. Extensive empirical evaluations indicate that our OI-based OOD detector is competitive with state-of-the-art OOD detectors in terms of detection accuracy on a wide range of datasets while requiring less computation and memory costs. Lastly, we show that the proposed OI-based confidence score function inherits nice properties from OI (e.g., insensitivity to small distributional variations and robustness against Huber $\epsilon$-contamination) and is a versatile tool for estimating OI and model accuracy in specific contexts.

Figures

Figures reproduced from arXiv: 2412.06168 by the authors.

Figure 1
Figure 1. Histograms of confidence scores using η, η1, and η2 with plane as the ID class and the other nine classes as the OOD class in CIFAR-10. 3) The OI-Based Confidence Score Function: The con￾fidence score function using Alg. 1 is defined as f(x) = ComputeBound({x}, {xi} m i=1, {gj} k j=1) (i.e., {x +} d = {x} with d = 1), which measures the maximum similarity between x and {xi} m i=1. If f(x) ≥ T0, then x is considered … view at source ↗
Figure 2
Figure 2. AUROC on UCI datasets. Horizontal dashed lines: the mean and [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. Histograms of confidence scores using η and η2 with plane as the ID class and the other nine classes as the OOD class in CIFAR-10. ID labels. To build the contaminated dataset, we merged all ID and OOD features, then randomly selected 100 features. During testing, each feature was subtracted from the mean vector of this contaminated dataset. The compared methods are allowed to use ImageNet-1K training data to train … view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: Performance of our approach with different numbers ( [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: (a): illustration of Proposition V.1. (b): illustration of Proposition V.2. (c,d): the absolute value of estimation errors between the ground-truth OI and η ′ for uniform distributions (c) and truncated Gaussian distributions (d). The insensitivity and robustness are i…
Figure 8
Figure 8. Figure 8: Top: original samples. Bottom: domain-shifted samples. From left [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 7
Figure 7. Figure 7: Pictures under “Triggers” are poisoned samples. Pictures under “Clean” [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 9
Figure 9. Figure 9: , a large difference reflects a large domain shift. When the domain shift vanishes, the model accuracy and the calculated upper bound accuracy are close [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

100 extracted references · 67 canonical work pages

  1. [1]

    Functional replicas of proprietary three-axis attitude sensors via lstm neural networks,

    H. Fu, P. Krishnamurthy, and F. Khorrami, “Functional replicas of proprietary three-axis attitude sensors via lstm neural networks,” in 2020 IEEE Conference on Control Technology and Applications. IEEE, 2020, pp. 70–75

    2020

  2. [2]

    An optimal control strategy for execution of large stock orders using long short-term memory networks,

    A. Papanicolaou, H. Fu, P. Krishnamurthy, B. Healy, and F. Khorrami, “An optimal control strategy for execution of large stock orders using long short-term memory networks,” Journal of Computational Finance , vol. 26, no. 4, 2023

    2023

  3. [3]

    A deep neural network algorithm for linear-quadratic portfolio optimization with mgarch and small transaction costs,

    A. Papanicolaou, H. Fu, P. Krishnamurthy, and F. Khorrami, “A deep neural network algorithm for linear-quadratic portfolio optimization with mgarch and small transaction costs,” IEEE Access , vol. 11, pp. 16 774–16 792, 2023

    2023

  4. [4]

    Privacy- preserving collaborative learning through feature extraction,

    A. Sarmadi, H. Fu, P. Krishnamurthy, S. Garg, and F. Khorrami, “Privacy- preserving collaborative learning through feature extraction,” IEEE Transactions on Dependable and Secure Computing , vol. 21, no. 1, pp. 486–498, 2023

    2023

  5. [5]

    Entropic out-of-distribution detection: Seamless detection of unknown examples,

    D. Macêdo, T. I. Ren, C. Zanchettin, A. L. Oliveira, and T. Ludermir, “Entropic out-of-distribution detection: Seamless detection of unknown examples,” IEEE Transactions on Neural Networks and Learning Systems, vol. 33, no. 6, pp. 2350–2364, 2021

    2021

  6. [6]

    Out-of-distribution detection by cross- class vicinity distribution of in-distribution data,

    Z. Zhao, L. Cao, and K.-Y . Lin, “Out-of-distribution detection by cross- class vicinity distribution of in-distribution data,” IEEE Transactions on Neural Networks and Learning Systems , 2023

    2023

  7. [7]

    Resmooth: Detecting and utilizing ood samples when training with data augmentation,

    C. Wang, J. Jiang, X. Zhou, and X. Liu, “Resmooth: Detecting and utilizing ood samples when training with data augmentation,” IEEE Transactions on Neural Networks and Learning Systems , 2022

    2022

  8. [8]

    Clipscope: Enhancing zero-shot ood detection with bayesian scoring,

    H. Fu, N. Patel, P. Krishnamurthy, and F. Khorrami, “Clipscope: Enhancing zero-shot ood detection with bayesian scoring,” arXiv preprint arXiv:2405.14737, 2024

    arXiv 2024

Show all 100 references
  1. [9]

    Combining switching mechanism with re-initialization and anomaly detection for resiliency of cyber–physical systems,

    H. Fu, P. Krishnamurthy, and F. Khorrami, “Combining switching mechanism with re-initialization and anomaly detection for resiliency of cyber–physical systems,” Automatica, vol. 172, p. 111994, 2025

    2025

  2. [10]

    Estimating the support of a high-dimensional distribution,

    B. Schölkopf, J. C. Platt, J. Shawe-Taylor, A. J. Smola, and R. C. Williamson, “Estimating the support of a high-dimensional distribution,” Neural Computation, vol. 13, no. 7, pp. 1443–1471, 2001

    2001

  3. [11]

    Deep one-class classification,

    L. Ruff, R. Vandermeulen, N. Goernitz, L. Deecke, S. A. Siddiqui, A. Binder, E. Müller, and M. Kloft, “Deep one-class classification,” in Proceedings of International Conference on Machine Learning , 2018, pp. 4393–4402

    2018

  4. [12]

    Provable guarantees for understanding out- of-distribution detection,

    P. Morteza and Y . Li, “Provable guarantees for understanding out- of-distribution detection,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 8, 2022

    2022

  5. [13]

    Measuring distribution similarities between samples: A distribution-free overlapping index,

    M. Pastore and A. Calcagnì, “Measuring distribution similarities between samples: A distribution-free overlapping index,” Frontiers in Psychology, vol. 10, p. 1089, 2019

    2019

  6. [14]

    Learning noise transition matrix from only noisy labels via total variation regularization,

    Y . Zhang, G. Niu, and M. Sugiyama, “Learning noise transition matrix from only noisy labels via total variation regularization,” in Proceedings of International Conference on Machine Learning , 2021, pp. 12 501– 12 512

    2021

  7. [15]

    How to fake multiply by a gaussian matrix,

    M. Kapralov, V . Potluru, and D. Woodruff, “How to fake multiply by a gaussian matrix,” in Proceedings of International Conference on Machine Learning, 2016, pp. 2101–2110

    2016

  8. [16]

    Optimal transport and wasserstein distance,

    “Optimal transport and wasserstein distance,” https://www.stat.cmu.edu/ %7elarry/=sml/Opt.pdf, 2019, accessed: 2023-04-09

    2019

  9. [17]

    Limitations of the wasserstein mde for univariate data,

    Y . G. Yatracos, “Limitations of the wasserstein mde for univariate data,” Statistics and Computing , vol. 32, no. 6, p. 95, 2022

    2022

  10. [18]

    Robust estimation of a location parameter,

    P. J. Huber, “Robust estimation of a location parameter,” Breakthroughs in Statistics: Methodology and Distribution , pp. 492–518, 1992

    1992

  11. [19]

    A baseline for detecting misclassified and out-of-distribution examples in neural networks,

    D. Hendrycks and K. Gimpel, “A baseline for detecting misclassified and out-of-distribution examples in neural networks,” in Proceedings of International Conference on Learning Representations , 2017

    2017

  12. [20]

    A simple unified framework for detecting out-of-distribution samples and adversarial attacks,

    K. Lee, K. Lee, H. Lee, and J. Shin, “A simple unified framework for detecting out-of-distribution samples and adversarial attacks,” in Proceedings of Advances in Neural Information Processing Systems , vol. 31, 2018, pp. 1–9

    2018

  13. [21]

    Energy-based out-of-distribution detection,

    W. Liu, X. Wang, J. Owens, and Y . Li, “Energy-based out-of-distribution detection,” in Proceedings of Advances in Neural Information Processing Systems, vol. 33, 2020, pp. 21 464–21 475

    2020

  14. [22]

    Kullback-leibler divergence-based out-of-distribution detection with flow-based generative models,

    Y . Zhang, J. Pan, W. Liu, Z. Chen, K. Li, J. Wang, Z. Liu, and H. Wei, “Kullback-leibler divergence-based out-of-distribution detection with flow-based generative models,” IEEE Transactions on Knowledge and Data Engineering, 2023

    2023

  15. [23]

    On the empirical estimation of integral probability metrics,

    B. K. Sriperumbudur, K. Fukumizu, A. Gretton, B. Schölkopf, and G. R. Lanckriet, “On the empirical estimation of integral probability metrics,” Electronic Journal of Statistics , vol. 6, pp. 1550–1599, 2012

    2012

  16. [24]

    Solving linear programs in the current matrix multiplication time,

    M. B. Cohen, Y . T. Lee, and Z. Song, “Solving linear programs in the current matrix multiplication time,” Journal of the ACM , vol. 68, no. 1, pp. 1–39, 2021

    2021

  17. [25]

    Lecture notes 27 36-705,

    “Lecture notes 27 36-705,” https://www.stat.cmu.edu/~larry/=stat705/ Lecture27.pdf, 2020, accessed: 2023-01-09

    2020

  18. [26]

    Nonparametric estimation of the coefficient of overlapping—theory and empirical application,

    F. Schmid and A. Schmidt, “Nonparametric estimation of the coefficient of overlapping—theory and empirical application,” Computational Statistics & Data Analysis , vol. 50, no. 6, pp. 1583–1596, 2006

    2006

  19. [27]

    A review of kernel density estimation with applications to econometrics,

    A. Z. Zambom and D. Ronaldo, “A review of kernel density estimation with applications to econometrics,” International Econometric Review , vol. 5, no. 1, pp. 20–42, 2013

    2013

  20. [28]

    Positive and unlabeled examples help learning,

    F. D. Comité, F. Denis, R. Gilleron, and F. Letouzey, “Positive and unlabeled examples help learning,” in Proceedings of International Conference on Algorithmic Learning Theory , 1999, pp. 219–230

    1999

  21. [29]

    One-class classification: Concept learning in the absence of counter-examples,

    D. M. J. Tax, “One-class classification: Concept learning in the absence of counter-examples,” Ph.D. dissertation, Technische Universiteit Delft, 2002

    2002

  22. [30]

    One-class classification: taxonomy of study and review of techniques,

    S. S. Khan and M. G. Madden, “One-class classification: taxonomy of study and review of techniques,” The Knowledge Engineering Review , vol. 29, no. 3, pp. 345–374, 2014

    2014

  23. [31]

    Ocgan: One-class novelty detection using gans with constrained latent representations,

    P. Perera, R. Nallapati, and B. Xiang, “Ocgan: One-class novelty detection using gans with constrained latent representations,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 2898–2906

    2019

  24. [32]

    Backpropa- gated gradient representations for anomaly detection,

    G. Kwon, M. Prabhushankar, D. Temel, and G. AlRegib, “Backpropa- gated gradient representations for anomaly detection,” in Proceedings of European Conference on Computer Vision , 2020, pp. 206–226

    2020

  25. [33]

    Deep semi-supervised anomaly detection,

    L. Ruff, R. A. Vandermeulen, N. Görnitz, A. Binder, E. Müller, K.-R. Müller, and M. Kloft, “Deep semi-supervised anomaly detection,” in Proceedings of International Conference on Learning Representations , 2020

    2020

  26. [34]

    Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings,

    P. Bergmann, M. Fauser, D. Sattlegger, and C. Steger, “Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 4183–4192

    2020

  27. [35]

    Cutpaste: Self-supervised learning for anomaly detection and localization,

    C.-L. Li, K. Sohn, J. Yoon, and T. Pfister, “Cutpaste: Self-supervised learning for anomaly detection and localization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 9664–9674

    2021

  28. [36]

    Autoencoders-a comparative analysis in the realm of anomaly detection,

    S. Schneider, D. Antensteiner, D. Soukup, and M. Scheutz, “Autoencoders-a comparative analysis in the realm of anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 1986–1992

    2022

  29. [37]

    Adtr: Anomaly detection transformer with feature reconstruction,

    Z. You, K. Yang, W. Luo, L. Cui, Y . Zheng, and X. Le, “Adtr: Anomaly detection transformer with feature reconstruction,” in Proceedings of Neural Information Processing , 2023, pp. 298–310

    2023

  30. [38]

    Deep anomaly detection using geometric transformations,

    I. Golan and R. El-Yaniv, “Deep anomaly detection using geometric transformations,” in Proceedings of Advances in Neural Information Processing Systems, vol. 31, 2018, pp. 1–9

    2018

  31. [39]

    Latent outlier exposure for anomaly detection with contaminated data,

    C. Qiu, A. Li, M. Kloft, M. Rudolph, and S. Mandt, “Latent outlier exposure for anomaly detection with contaminated data,” in Proceedings of International Conference on Machine Learning , 2022, pp. 18 153– 18 167

    2022

  32. [40]

    Panda: Adapting pretrained features for anomaly detection and segmentation,

    T. Reiss, N. Cohen, L. Bergman, and Y . Hoshen, “Panda: Adapting pretrained features for anomaly detection and segmentation,” in Pro- JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 10 ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 20...

    2021

  33. [41]

    Multiresolution knowledge distillation for anomaly detection,

    M. Salehi, N. Sadjadi, S. Baselizadeh, M. H. Rohban, and H. R. Rabiee, “Multiresolution knowledge distillation for anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 14 902–14 912

    2021

  34. [42]

    The unreasonable effectiveness of deep features as a perceptual metric,

    R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual metric,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 586–595

    2018

  35. [43]

    Ecod: Unsupervised outlier detection using empirical cumulative distribution functions,

    Z. Li, Y . Zhao, X. Hu, N. Botta, C. Ionescu, and G. Chen, “Ecod: Unsupervised outlier detection using empirical cumulative distribution functions,” IEEE Transactions on Knowledge and Data Engineering , 2022

    2022

  36. [44]

    Deep isolation forest for anomaly detection,

    H. Xu, G. Pang, Y . Wang, and Y . Wang, “Deep isolation forest for anomaly detection,” IEEE Transactions on Knowledge and Data Engineering, 2023

    2023

  37. [45]

    Vim: Out-of-distribution with virtual-logit matching,

    H. Wang, Z. Li, L. Feng, and W. Zhang, “Vim: Out-of-distribution with virtual-logit matching,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 4921–4930

    2022

  38. [46]

    On the importance of gradients for detecting distributional shifts in the wild,

    R. Huang, A. Geng, and Y . Li, “On the importance of gradients for detecting distributional shifts in the wild,” in Proceedings of Advances in Neural Information Processing Systems , 2021, pp. 677–689

    2021

  39. [47]

    Towards unknown-aware learning with virtual outlier synthesis,

    X. Du, Z. Wang, M. Cai, and S. Li, “Towards unknown-aware learning with virtual outlier synthesis,” in Proceedings of International Conference on Learning Representations , 2022

    2022

  40. [48]

    Towards total recall in industrial anomaly detection,

    K. Roth, L. Pemula, J. Zepeda, B. Schölkopf, T. Brox, and P. Gehler, “Towards total recall in industrial anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 14 318–14 328

    2022

  41. [49]

    Yolood: Utilizing object detection concepts for multi- label out-of-distribution detection,

    A. Zolfi, G. Amit, A. Baras, S. Koda, I. Morikawa, Y . Elovici, and A. Shabtai, “Yolood: Utilizing object detection concepts for multi- label out-of-distribution detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 1–8

    2024

  42. [50]

    Rethinking out-of- distribution (ood) detection: Masked image modeling is all you need,

    J. Li, P. Chen, Z. He, S. Yu, S. Liu, and J. Jia, “Rethinking out-of- distribution (ood) detection: Masked image modeling is all you need,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 11 578–11 589

    2023

  43. [51]

    Reverse kl-divergence training of prior networks: Improved uncertainty and adversarial robustness,

    A. Malinin and M. Gales, “Reverse kl-divergence training of prior networks: Improved uncertainty and adversarial robustness,” in Proceed- ings of Advances in Neural Information Processing Systems , vol. 32, 2019, pp. 1–9

    2019

  44. [52]

    Action elimination and stopping conditions for the multi-armed bandit and rein- forcement learning problems

    E. Even-Dar, S. Mannor, Y . Mansour, and S. Mahadevan, “Action elimination and stopping conditions for the multi-armed bandit and rein- forcement learning problems.” Journal of Machine Learning Research , vol. 7, no. 6, 2006

    2006

  45. [53]

    Optimal mass transport: Signal processing and machine-learning applications,

    S. Kolouri, S. R. Park, M. Thorpe, D. Slepcev, and G. K. Rohde, “Optimal mass transport: Signal processing and machine-learning applications,” IEEE Signal Processing Magazine , vol. 34, no. 4, pp. 43–59, 2017

    2017

  46. [54]

    A new robust partial p-wasserstein-based metric for comparing distributions,

    S. Raghvendra, P. Shirzadian, and K. Zhang, “A new robust partial p-wasserstein-based metric for comparing distributions,” arXiv preprint arXiv:2405.03664, 2024

    2024 arXiv

  47. [55]

    Computing all optimal partial transports,

    A. Phatak, S. Raghvendra, C. Tripathy, and K. Zhang, “Computing all optimal partial transports,” in International Conference on Learning Representations, 2023

    2023

  48. [56]

    A combinatorial algorithm for approximating the optimal transport in the parallel and mpc settings,

    N. Lahn, S. Raghvendra, and K. Zhang, “A combinatorial algorithm for approximating the optimal transport in the parallel and mpc settings,” Advances in Neural Information Processing Systems, vol. 36, pp. 21 675– 21 686, 2023

    2023

  49. [57]

    Fast computation of wasserstein barycenters,

    M. Cuturi and A. Doucet, “Fast computation of wasserstein barycenters,” in Proceedings of International Conference on Machine Learning , 2014, pp. 685–693

    2014

  50. [58]

    The overlapping coefficient as a measure of agreement between probability distributions and point estimation of the overlap of two normal densities,

    H. F. Inman and E. L. Bradley Jr, “The overlapping coefficient as a measure of agreement between probability distributions and point estimation of the overlap of two normal densities,” Communications in Statistics-theory and Methods , vol. 18, no. 10, pp. 3851–3874, 1989

    1989

  51. [59]

    Badnets: Evaluating backdooring attacks on deep neural networks,

    T. Gu, K. Liu, B. Dolan-Gavitt, and S. Garg, “Badnets: Evaluating backdooring attacks on deep neural networks,” IEEE Access, vol. 7, pp. 47 230–47 244, 2019

    2019

  52. [60]

    Trojaning attack on neural networks,

    Y . Liu, S. Ma, Y . Aafer, W.-C. Lee, J. Zhai, W. Wang, and X. Zhang, “Trojaning attack on neural networks,” in Proceedings of the Network and Distributed Systems Security Symposium , 2018

    2018

  53. [61]

    A feature-based on-line detector to remove adversarial-backdoors by iterative demarcation,

    H. Fu, A. K. Veldanda, P. Krishnamurthy, S. Garg, and F. Khorrami, “A feature-based on-line detector to remove adversarial-backdoors by iterative demarcation,” IEEE Access, vol. 10, pp. 5545 – 5558, 2022

    2022

  54. [62]

    Differential analysis of triggers and benign features for black-box dnn backdoor detection,

    H. Fu, P. Krishnamurthy, S. Garg, and F. Khorrami, “Differential analysis of triggers and benign features for black-box dnn backdoor detection,” IEEE Transactions on Information Forensics and Security , vol. 18, pp. 4668–4680, 2023

    2023

  55. [63]

    Mitigating backdoor attacks on deep neural networks,

    H. Fu, A. Sarmadi, P. Krishnamurthy, S. Garg, and F. Khorrami, “Mitigating backdoor attacks on deep neural networks,” in Embedded Machine Learning for Cyber-Physical, IoT, and Edge Computing: Use Cases and Emerging Challenges . Springer, 2023, pp. 395–431

    2023

  56. [64]

    Enhancing the safety of machine learning systems and cyber- physical systems against cyber attacks,

    H. Fu, “Enhancing the safety of machine learning systems and cyber- physical systems against cyber attacks,” Ph.D. dissertation, New York University Tandon School of Engineering, 2024

    2024

  57. [65]

    Poisoning attacks against support vector machines,

    B. Biggio, B. Nelson, and P. Laskov, “Poisoning attacks against support vector machines,” in Proceedings of the International Conference on Machine Learning, 2012, pp. 1467–1474

    2012

  58. [66]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” in Proceedings of Advances in Neural Information Processing Systems , vol. 27, 2014, pp. 1–9

    2014

  59. [67]

    Wanet - imperceptible warping-based backdoor attack,

    T. A. Nguyen and A. T. Tran, “Wanet - imperceptible warping-based backdoor attack,” in Proceedings of the International Conference on Learning Representations, 2021

    2021

  60. [68]

    Invisible backdoor attack with sample-specific triggers,

    Y . Li, Y . Li, B. Wu, L. Li, R. He, and S. Lyu, “Invisible backdoor attack with sample-specific triggers,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 16 463–16 472

    2021

  61. [69]

    Rethinking the backdoor attacks’ triggers: A frequency perspective,

    Y . Zeng, W. Park, Z. M. Mao, and R. Jia, “Rethinking the backdoor attacks’ triggers: A frequency perspective,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 16 473–16 481

    2021

  62. [70]

    Reflection backdoor: A natural backdoor attack on deep neural networks,

    Y . Liu, X. Ma, J. Bailey, and F. Lu, “Reflection backdoor: A natural backdoor attack on deep neural networks,” in Proceedings of European Conference on Computer Vision , 2020, pp. 182–199

    2020

  63. [71]

    Neural cleanse: Identifying and mitigating backdoor attacks in neural networks,

    B. Wang, Y . Yao, S. Shan, H. Li, B. Viswanath, H. Zheng, and B. Y . Zhao, “Neural cleanse: Identifying and mitigating backdoor attacks in neural networks,” in Proceedings of IEEE Symposium on Security and Privacy, 2019, pp. 707–723

    2019

  64. [72]

    Fine-pruning: Defending against backdooring attacks on deep neural networks,

    K. Liu, B. Dolan-Gavitt, and S. Garg, “Fine-pruning: Defending against backdooring attacks on deep neural networks,” in Proceedings of International Symposium on Research in Attacks, Intrusions, and Defenses, 2018, pp. 273–294

    2018

  65. [73]

    Data-free backdoor removal based on channel lipschitzness,

    R. Zheng, R. Tang, J. Li, and L. Liu, “Data-free backdoor removal based on channel lipschitzness,” in Proceedings of the European Conference on Computer Vision , 2022, pp. 175–191

    2022

  66. [74]

    Strip: A defence against trojan attacks on deep neural networks,

    Y . Gao, C. Xu, D. Wang, S. Chen, D. C. Ranasinghe, and S. Nepal, “Strip: A defence against trojan attacks on deep neural networks,” in Proceedings of the Annual Computer Security Applications Conference , 2019, pp. 113–125

    2019

  67. [75]

    Nnoculation: Catching badnets in the wild,

    A. K. Veldanda, K. Liu, B. Tan, P. Krishnamurthy, F. Khorrami, R. Karri, B. Dolan-Gavitt, and S. Garg, “Nnoculation: Catching badnets in the wild,” in Proceedings of the ACM Workshop on Artificial Intelligence and Security, 2021, pp. 49–60

    2021

  68. [76]

    Learning multiple layers of features from tiny images,

    A. Krizhevsky, G. Hinton et al., “Learning multiple layers of features from tiny images,” 2009

    2009

  69. [77]

    UCI machine learning repository,

    D. Dua and C. Graff, “UCI machine learning repository,” "http://archive. ics.uci.edu/ml", 2017, accessed: 2022-04-09

    2017

  70. [78]

    Describing textures in the wild,

    M. Cimpoi, S. Maji, I. Kokkinos, S. Mohamed, and A. Vedaldi, “Describing textures in the wild,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2014, pp. 3606–3613

    2014

  71. [79]

    Reading digits in natural images with unsupervised feature learning,

    Y . Netzer, T. Wang, A. Coates, A. Bissacco, B. Wu, and A. Y . Ng, “Reading digits in natural images with unsupervised feature learning,” in Proceedings of NIPS Workshop on Deep Learning and Unsupervised Feature Learning, 2011

    2011

  72. [80]

    Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop,

    F. Yu, A. Seff, Y . Zhang, S. Song, T. Funkhouser, and J. Xiao, “Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop,” arXiv preprint arXiv:1506.03365 , 2015

    2015 arXiv

  73. [81]

    Turkergaze: Crowdsourcing saliency with webcam based eye tracking,

    P. Xu, K. A. Ehinger, Y . Zhang, A. Finkelstein, S. R. Kulkarni, and J. Xiao, “Turkergaze: Crowdsourcing saliency with webcam based eye tracking,” arXiv preprint arXiv:1504.06755 , 2015

    2015 arXiv

  74. [82]

    Mnist handwritten digit database,

    Y . LeCun, C. Cortes, and C. Burges, “Mnist handwritten digit database,” ATT Labs [Online]. Available: http://yann.lecun.com/exdb/mnist , vol. 2, 2010

    2010

  75. [83]

    The german traffic sign recognition benchmark: A multi-class classification competition,

    J. Stallkamp, M. Schlipsing, J. Salmen, and C. Igel, “The german traffic sign recognition benchmark: A multi-class classification competition,” in Proceedings of International Joint Conference on Neural Networks , 2011, pp. 1453–1460

    2011

  76. [84]

    Face recognition in unconstrained videos with matched background similarity,

    L. Wolf, T. Hassner, and I. Maoz, “Face recognition in unconstrained videos with matched background similarity,” in Proceedings of the IEEE JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 11 Conference on Computer Vision and Pattern Recognition , 2011, pp. 529–534

    2021

  77. [85]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L. Li, Kai Li, and Li Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2009, pp. 248–255

    2009

  78. [86]

    Winner-take-all autoencoders,

    A. Makhzani and B. J. Frey, “Winner-take-all autoencoders,” in Proceedings of Advances in Neural Information Processing Systems , vol. 28, 2015, p. 2791–2799

    2015

  79. [87]

    Unsupervised anomaly detection with generative adversarial networks to guide marker discovery,

    T. Schlegl, P. Seeböck, S. M. Waldstein, U. Schmidt-Erfurth, and G. Langs, “Unsupervised anomaly detection with generative adversarial networks to guide marker discovery,” in Proceedings of International Conference on Information Processing in Medical Imaging , 2017, pp. 146–157

    2017

  80. [88]

    Gradient-based learning applied to document recognition,

    Y . LeCun, L. Bottou, Y . Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE , vol. 86, no. 11, pp. 2278–2324, 1998

    1998

  81. [89]

    Lof: identifying density-based local outliers,

    M. M. Breunig, H.-P. Kriegel, R. T. Ng, and J. Sander, “Lof: identifying density-based local outliers,” in Proceedings of the ACM SIGMOD International Conference on Management of Data , 2000, pp. 93–104

    2000

  82. [90]

    Isolation forest,

    F. T. Liu, K. M. Ting, and Z.-H. Zhou, “Isolation forest,” in Proceedings of IEEE International Conference on Data Mining , 2008, pp. 413–422

    2008

  83. [91]

    A fast algorithm for the minimum covariance determinant estimator,

    P. J. Rousseeuw and K. V . Driessen, “A fast algorithm for the minimum covariance determinant estimator,” Technometrics, vol. 41, no. 3, pp. 212–223, 1999

    1999

  84. [92]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in Proceedings of International Conference on Machine Learning , 2021, pp. 8748–8763

    2021

  85. [93]

    The inaturalist species classifi- cation and detection dataset,

    G. Van Horn, O. Mac Aodha, Y . Song, Y . Cui, C. Sun, A. Shepard, H. Adam, P. Perona, and S. Belongie, “The inaturalist species classifi- cation and detection dataset,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2018, pp. 8769–8778

    2018

  86. [94]

    Sun database: Large-scale scene recognition from abbey to zoo,

    J. Xiao, J. Hays, K. A. Ehinger, A. Oliva, and A. Torralba, “Sun database: Large-scale scene recognition from abbey to zoo,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2010, pp. 3485–3492

    2010

  87. [95]

    Places: A 10 million image database for scene recognition,

    B. Zhou, A. Lapedriza, A. Khosla, A. Oliva, and A. Torralba, “Places: A 10 million image database for scene recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 40, no. 6, pp. 1452– 1464, 2017

    2017

  88. [96]

    Enhancing the reliability of out-of- distribution image detection in neural networks,

    S. Liang, Y . Li, and R. Srikant, “Enhancing the reliability of out-of- distribution image detection in neural networks,” in Proceedings of International Conference on Learning Representations , 2018

    2018

  89. [97]

    Non-parametric outlier synthesis,

    L. Tao, X. Du, J. Zhu, and Y . Li, “Non-parametric outlier synthesis,” in Proceedings of The International Conference on Learning Represen- tations, 2022

    2022

  90. [98]

    Targeted backdoor attacks on deep learning systems using data poisoning,

    X. Chen, C. Liu, B. Li, K. Lu, and D. Song, “Targeted backdoor attacks on deep learning systems using data poisoning,” arXiv preprint arXiv:1712.05526, 2017

    2017 arXiv

  91. [99]

    Abs: Scanning neural networks for back-doors by artificial brain stimulation,

    Y . Liu, W.-C. Lee, G. Tao, S. Ma, Y . Aafer, and X. Zhang, “Abs: Scanning neural networks for back-doors by artificial brain stimulation,” in Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security , 2019, pp. 1265–1282

    2019

  92. [100]

    Understanding failures in out-of-distribution detection with deep generative models,

    L. Zhang, M. Goldstein, and R. Ranganath, “Understanding failures in out-of-distribution detection with deep generative models,” in Proceedings of International Conference on Machine Learning , 2021, pp. 12 427–12 436. APPENDIX A. Proof of Theorem III.3 Definition A.1 (Total V...

    2021

Pith tools

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