Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

Dataset Distillation with Probabilistic Latent Features

T0 review · 5 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 claims that dataset distillation works better when the synthetic latent features feeding a pretrained generator are treated as random draws from a learned distribution rather than fixed points; the result is a method called…

desk verdict GLaD extension with a promising stochastic module but a central loss that doesn't match the stated likelihood; deserves review, not acceptance as is. read the letter →

arxiv 2505.06647 v2 pith:J3CQEYPO submitted 2025-05-10 cs.CV

Zhe Li , Sarah Cechnicka , Cheng Ouyang , Katharina Breininger , Peter Schüffler , Bernhard Kainz This is my paper · ORCID
classification cs.CV
keywords datasetdistillationlatentfeaturedistributionlow-rankmultivariatenormalcross-architecturegeneralizationStyleGAN-XLgradientmatchingmedicalimageclassification
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

Dataset distillation compresses a large training set into a few synthetic images that can stand in for the original data when training classifiers. This paper argues that previous latent-space distillation methods are too deterministic: they ignore the spatial structure and uncertainty inherent in natural images. SLFD instead models the latent features that produce synthetic images as samples from a low-rank multivariate normal distribution, learned by a lightweight three-layer network. The paper reports that this stochastic approach consistently improves cross-architecture accuracy on ImageNet subsets, CIFAR-10, and a histopathology benchmark, with larger gains at higher resolutions. If correct, SLFD offers a low-overhead, generator-agnostic way to make distilled datasets generalize better to unseen model architectures.

What carries the argument

The engine is a low-rank multivariate normal distribution over the generator's latent features, predicted by three linear layers that produce the mean μ(ws), the diagonal D, and the low-rank factor P, with covariance Σ = PPᵀ + D. This distribution makes the distillation stochastic and spatially correlated. The samples q_m are shifted by the classifier's deterministic output via q*_m = q_m − μ + f(I_syn), grounding sampled variability in semantic content, and a gradient-matching loss aligns synthetic and real training dynamics.

What would settle it

Re-run distillation with the stochastic loss replaced by a direct, correctly dimensioned negative log-likelihood that includes the softmax denominator; if accuracy drops or training changes substantially, Eq. (6) is not equivalent to the model being optimized.

Watch

Extended reading notes

Core claim

The central claim is that explicitly modeling spatial correlation and uncertainty in the latent space of a pretrained generator produces synthetic datasets that train classifiers more effectively across architectures. The method learns a mean, a diagonal, and a low-rank covariance factor for the latent features, samples multiple latent codes, and blends the samples with the classifier's deterministic output before computing a stochastic cross-entropy loss; gradient matching between real and synthetic batches is retained. The paper reports consistent accuracy gains over GLaD on ten ImageNet subsets at 128×128 and five at 256×256, on CIFAR-10, and on MedMNIST PathMNIST at three resolutions, with no degradation as resolution increases. It also shows the low-rank covariance performs on par with a full-rank variant, and that the method transfers to a diffusion generator (UViT), albeit with weaker results than with StyleGAN-XL.

Load-bearing premise

The load-bearing assumption is that Eq. (6) is a valid Monte Carlo approximation of the negative log-likelihood in Eq. (4); as written, Eq. (6) omits the class-normalizing softmax denominator and mixes tensors of different shapes, so the training signal may not actually match the stated probabilistic model.

Editorial extensions

If this is right

  • Distilled datasets produced by SLFD transfer to unseen architectures (ResNet18, VGG11, ViT, AlexNet) better than GLaD baselines, with gains up to about 7% under gradient matching and about 36% under distribution matching on ImageNet subsets.
  • At 256×256 resolution, SLFD does not degrade and improves accuracy (e.g., 17.7% on ImNet-A), in contrast to earlier pixel-space and latent-space methods that often lose performance at higher resolution.
  • On histopathology (PathMNIST), SLFD improves with resolution while GLaD declines, suggesting the method is suitable for high-resolution medical image distillation and data sharing.
  • The low-rank covariance gives performance on par with the full-rank version, so the stochastic benefit does not require a full covariance matrix, keeping computational overhead low.
  • The method works with both GAN and diffusion generators, although StyleGAN-XL remains the stronger backbone in the reported experiments.

Reading between the lines

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

  • A testable alternative explanation is that most of the gain comes from the sample-shift operation in Eq. (5) acting as a form of latent-space data augmentation, rather than from the probabilistic model itself; evaluating SLFD with the shift but without sampling, or vice versa, would separate the two.
  • The reported stability at 256×256 suggests SLFD could extend to even higher-resolution distillation, where pixel-space methods typically degrade; this extrapolation is not tested in the paper.
  • For medical data sharing, the stochastic, non-photorealistic synthetic images may carry privacy benefits beyond compactness, but the paper does not measure re-identification risk.
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

5 major / 5 minor

Summary. The paper proposes Stochastic Latent Feature Distillation (SLFD), a dataset distillation method that learns a low-rank multivariate normal distribution over the classifier outputs of synthetic images, draws samples from this distribution, and uses the resulting stochastic loss within a gradient-matching objective. The distilled latent codes are passed through a frozen StyleGAN-XL generator to produce synthetic images, and the method is evaluated on ImageNet subsets at 128x128 and 256x256, CIFAR-10, and MedMNIST by measuring cross-architecture test accuracy. The central claim is that SLFD achieves state-of-the-art cross-architecture performance, especially at high resolutions, while adding only three linear layers of computational overhead.

Significance. If the probabilistic formulation were sound, injecting structured, spatially correlated uncertainty into the distillation process would be a useful and broadly applicable contribution to dataset distillation, with potential relevance to medical imaging. The experimental scope is reasonably broad, covering multiple architectures, several ImageNet subsets, a low-resolution benchmark, and a histopathology dataset, and the paper includes ablations on the generative backbone and on the covariance rank. However, the central loss derivation in Sec. 3.3 is mathematically invalid as written, the object being modeled is classifier logits rather than latent features despite the title and abstract, and several empirical claims are not supported by the reported error bars. The contribution is therefore not currently established.

major comments (5)
  1. [Sec. 3.3.1, Eq. (6)] Equation (6) is not a valid Monte Carlo estimate of the negative log-likelihood in Eq. (4). A correct estimate would be -log(1/M \sum_m softmax_y(q*_m)), including the softmax denominator over classes; Eq. (6) uses the raw logit y_c·(q*_m)_c without class normalization, so the argument of the logarithm is not a probability. In addition, the log M term has the wrong sign for a Monte Carlo average (the correct term is -log M), and the dimensions are incompatible: q_m is in R^{ipc x Nw x C} while f(I_syn) is in R^{ipc x C}, so the shift in Eq. (5) introduces an Nw-dependent broadcasting that has no counterpart in Eq. (4). Furthermore, the outer sum over m in Eq. (6) is redundant because the inner sum uses the same index, so the 1/L factor cancels and the expression does not average over the M samples as claimed. Since this loss directly determines the gradients used in Algorithm 1, the central training mechanism of SLFD is not grounded in the stated probabilistic model.
  2. [Sec. 3.2, Eqs. (1)-(2) and Abstract/Title] The distribution is defined over classifier outputs f(I_syn), not over latent features. The abstract claims SLFD "models the joint distribution of latent features" and the method name repeats this, but Eq. (2) explicitly states f(I_syn) | G16(...) ~ N(mu(w_s), Sigma(w_s)), where mu(w_s) is in R^{B x Nw x C}. The paper should either rename the method and revise the claims or actually model the distribution of the style-code features. As written, the central conceptual claim does not match what is implemented.
  3. [Algorithm 1] The optimization of the stochastic module is unspecified. Algorithm 1 updates only the latent features w_s and f_n by backpropagation, but the mean, diagonal, and covariance factor are outputs of three fully connected layers that also have parameters. Without specifying how these FC layers are trained, the proposed module cannot learn the distribution it is supposed to model. This is a reproducibility-critical omission.
  4. [Table 4, MedMNIST] The text states that SLFD shows "a consistent improvement" over GLaD, but the table does not support this. At res=64, SLFD (36.67±0.6) is slightly below GLaD (36.88±1.1); at res=128, 36.27±1.7 versus 35.77±1.2 is within one standard deviation; only at res=256 is the gap (37.17±2.3 versus 35.64±1.7) favorable, and even there the error bars overlap substantially. The authors should report significance tests or additional seeds before claiming consistency.
  5. [Table 3, CIFAR-10] The text claims that "across both settings, our method consistently outperforms GLaD," but Table 3 lists no SLFD result for the Gr (randomly initialized generator) setting; only SLFDGt is shown. The comparison to GLaDGr is therefore missing, and the claim is not supported by the presented data.
minor comments (5)
  1. [Sec. 3.3.1, Eq. (4)] Equation (4) writes p(y|f(I_syn)_m) as if f is stochastic, but earlier the classifier is described as deterministic. The notation should distinguish the deterministic classifier output from the sampled q_m.
  2. [Sec. 4.3.1, Table 1] The prose reports gains "up to 7%" and "up to 36%" without specifying whether these are absolute or relative improvements. Please clarify the basis of these percentages.
  3. [Supplementary Material] The supplementary material contains only template placeholder text about compiling the supplementary with the main paper; it has no actual content. Either provide a real supplementary or remove the placeholder.
  4. [Sec. 4.2, Implementation details] The runtime is reported as approximately 9 hours, but it is unclear whether this covers all five ImageNet subsets at both resolutions. Please specify the exact experimental configuration used for the runtime measurement.
  5. [Figures 3 and 4] The qualitative figures would be easier to interpret if each panel was labeled with its class, since the reader cannot otherwise identify which synthetic images correspond to which class.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SLFD is an empirical distillation scheme evaluated on held-out test sets; the Eq. 6 loss inconsistency is a correctness concern, not a circular reduction.

full rationale

The paper's central claim is that SLFD improves cross-architecture test accuracy of distilled datasets. This claim is evaluated by training four classifiers that were not used during distillation (ResNet18, VGG11, ViT, AlexNet) on the synthetic images and testing on real validation sets, so the reported numbers are not forced by construction. The stochastic module's parameters (mean, diagonal, low-rank covariance factor) are optimized through gradient matching between synthetic and real classifier gradients, not fitted to the test accuracy or to any quantity that is then renamed as a prediction. The use of a StyleGAN-XL generator pretrained on ImageNet to distill ImageNet subsets is an external pretrained prior and standard practice in this line of work; it may raise a fairness or leakage concern, but it does not make the evaluation equivalent to the input. Self-citations are not load-bearing: reference [17] is only a related-work mention in the medical-applications discussion, and reference [23] is cited only for the log-sum-exp trick, not as an authority that by itself establishes the method's validity. The placeholder text in the supplementary material is a production artifact and does not function as evidence in the derivation chain. The mismatch between Eq. 4 and Eq. 6 identified in the skeptical analysis is an internal mathematical inconsistency that would affect the grounding of the training objective, but it is a correctness or reproducibility risk, not a circular step: the stochastic loss is not shown to be equivalent to its inputs or to the evaluated test accuracy by definition. No step in the paper reduces a derived result to a fitted constant, a self-referential prediction, or a self-citation chain, so the circularity score is 0.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim rests on three assumptions: the generative prior, the Gaussian form of logits, and the validity of the loss. The Gaussian form is introduced ad hoc and is not supported by data. The loss validity is questionable. The model itself has two free hyperparameters (rank and M), one of which is unreported.

free parameters (2)
  • rank R of covariance factor = 10 (from Table 6)
    Chosen for the low-rank approximation; no sensitivity analysis beyond comparing to full-rank.
  • number of Monte Carlo samples M = not reported
    Used in Eq. 6; the paper never states its value, yet it directly affects the loss and training cost.
assumptions (3)
  • domain assumption StyleGAN-XL provides a valid generative prior for distilled images
    The entire method relies on a pretrained StyleGAN-XL generator to map latent features to images; this is inherited from GLaD and not re-validated.
  • ad hoc to paper Classifier logits for synthetic images are multivariate normal conditioned on ws
    Eq. 2 posits f(Isyn)|G16 ~ N(mu(ws), Sigma(ws)) without empirical or theoretical justification; this is the core modeling assumption.
  • ad hoc to paper Eq. 6 is a valid Monte Carlo estimate of negative log-likelihood
    The paper asserts this via Eq. 4 but the explicit formula lacks softmax normalization and has dimension mismatches; the derivation is incomplete.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dataset Distillation with Probabilistic Latent Features." pith.science (2026). https://pith.science/paper/J3CQEYPO

@misc{pith2026250506647,
  author       = {Pith},
  title        = {Pith review of: Dataset Distillation with Probabilistic Latent Features},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J3CQEYPO}},
  note         = {Machine review of arXiv:2505.06647}
}
read the original abstract

As deep learning models grow in complexity and the volume of training data increases, reducing storage and computational costs becomes increasingly important. Dataset distillation addresses this challenge by synthesizing a compact set of synthetic data that can effectively replace the original dataset in downstream classification tasks. While existing methods typically rely on mapping data from pixel space to the latent space of a generative model, we propose a novel stochastic approach that models the joint distribution of latent features. This allows our method to better capture spatial structures and produce diverse synthetic samples, which benefits model training. Specifically, we introduce a low-rank multivariate normal distribution parameterized by a lightweight network. This design maintains low computational complexity and is compatible with various matching networks used in dataset distillation. After distillation, synthetic images are generated by feeding the learned latent features into a pretrained generator. These synthetic images are then used to train classification models, and performance is evaluated on real test set. We validate our method on several benchmarks, including ImageNet subsets, CIFAR-10, and the MedMNIST histopathological dataset. Our approach achieves state-of-the-art cross architecture performance across a range of backbone architectures, demonstrating its generality and effectiveness.

Figures

Figures reproduced from arXiv: 2505.06647 by the authors.

Figure 1
Figure 1. Overview of the proposed SLFD framework. (a) The distillation backbone is based on GLaD [ [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Illustration Eq. 5 about how we sample probabilities. refined probabilities are computed by shifting each sample by the classifier’s output: q ∗ m = qm − µ(ws) + f(Isyn) (5) This step adjusts each sample to reflect both the stochas￾tic variability from the latent space and the semantic grounding from the classifier as in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Example images distilled from ImageNet subsets. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Synthetic images from MedMNIST [32] for different output resolutions. Rapeseed Lorikeet [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Results of ablation study. Distilled images with the dif [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. The Evolution of Dataset Distillation: Toward Scalable and Generalizable Solutions

    cs.CV 2025-02 conditional novelty 3.0 of 10

    A 2023-2025 survey of dataset distillation that organizes matching, generative, decoupling, and selective methods and tabulates ImageNet-scale accuracy comparisons.

Reference graph

Works this paper leans on

42 extracted references · 31 canonical work pages · cited by 1 Pith paper

  1. [1]

    Im- age2stylegan: How to embed images into the stylegan latent space? In Proceedings of the IEEE/CVF international con- ference on computer vision (ICCV), pages 4432–4441, 2019

    Rameen Abdal, Yipeng Qin, and Peter Wonka. Im- age2stylegan: How to embed images into the stylegan latent space? In Proceedings of the IEEE/CVF international con- ference on computer vision (ICCV), pages 4432–4441, 2019. 3

    work page 2019

  2. [2]

    All are worth words: A vit back- bone for diffusion models

    Fan Bao, Shen Nie, Kaiwen Xue, Yue Cao, Chongxuan Li, Hang Su, and Jun Zhu. All are worth words: A vit back- bone for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 22669–22679, 2023. 7, 8

    work page 2023

  3. [3]

    Dataset distillation by matching training trajectories

    George Cazenavette, Tongzhou Wang, Antonio Torralba, Alexei A Efros, and Jun-Yan Zhu. Dataset distillation by matching training trajectories. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4750–4759, 2022. 1, 2, 6

    work page 2022

  4. [4]

    Generalizing dataset distillation via deep generative prior

    George Cazenavette, Tongzhou Wang, Antonio Torralba, Alexei A Efros, and Jun-Yan Zhu. Generalizing dataset distillation via deep generative prior. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 3739–3748, 2023. 1, 2, 3, 5, 6, 7

    work page 2023

  5. [5]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 5

    2009

  6. [6]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, et al. An image is worth 16x16 words: Trans- formers for image recognition at scale. arxiv 2020. arXiv preprint arXiv:2010.11929, 2010. 7

    arXiv 2020

  7. [7]

    Minimizing the accumulated trajectory error to improve dataset distillation

    Jiawei Du, Yidi Jiang, Vincent YF Tan, Joey Tianyi Zhou, and Haizhou Li. Minimizing the accumulated trajectory error to improve dataset distillation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 3749–3758, 2023. 2

    work page 2023

  8. [8]

    Reliable mutual distillation for medical image seg- mentation under imperfect annotations

    Chaowei Fang, Qian Wang, Lechao Cheng, Zhifan Gao, Chengwei Pan, Zhen Cao, Zhaohui Zheng, and Dingwen Zhang. Reliable mutual distillation for medical image seg- mentation under imperfect annotations. IEEE Transactions on Medical Imaging, 2023. 2

    work page 2023

Show all 42 references
  1. [9]

    On calibration of modern neural networks

    Chuan Guo, Geoff Pleiss, Yu Sun, and Kilian Q Weinberger. On calibration of modern neural networks. In International conference on machine learning, pages 1321–1330. PMLR,

  2. [10]

    Vision trans- formers for small histological datasets learned through knowledge distillation

    Neel Kanwal, Trygve Eftestøl, Farbod Khoraminia, Tahlita CM Zuiverloon, and Kjersti Engan. Vision trans- formers for small histological datasets learned through knowledge distillation. In Pacific-Asia Conference on Knowledge Discovery and Data Mining , pages 167–179. Springer, 2023. 2

    2023

  3. [11]

    What uncertainties do we need in bayesian deep learning for computer vision? Advances in neural information processing systems, 30, 2017

    Alex Kendall and Yarin Gal. What uncertainties do we need in bayesian deep learning for computer vision? Advances in neural information processing systems, 30, 2017. 1

    2017

  4. [12]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky, Geoffrey Hinton, et al. Learning multiple layers of features from tiny images. 2009. 5

    2009

  5. [13]

    A comprehensive survey to dataset distillation

    Shiye Lei and Dacheng Tao. A comprehensive survey to dataset distillation. arXiv preprint arXiv:2301.05603, 2023. 2

    2023 arXiv

  6. [14]

    Soft-label anonymous gastric x-ray image distillation

    Guang Li, Ren Togo, Takahiro Ogawa, and Miki Haseyama. Soft-label anonymous gastric x-ray image distillation. In 2020 IEEE International Conference on Image Processing (ICIP), pages 305–309. IEEE, 2020. 2

    2020

  7. [15]

    Compressed gastric image generation based on soft-label dataset distillation for medical data sharing.Computer Meth- ods and Programs in Biomedicine, 227:107189, 2022

    Guang Li, Ren Togo, Takahiro Ogawa, and Miki Haseyama. Compressed gastric image generation based on soft-label dataset distillation for medical data sharing.Computer Meth- ods and Programs in Biomedicine, 227:107189, 2022. 2

    2022

  8. [16]

    Dataset distillation for medical dataset sharing

    Guang Li, Ren Togo, Takahiro Ogawa, and Miki Haseyama. Dataset distillation for medical dataset sharing. arXiv preprint arXiv:2209.14603, 2022. 2

    2022 arXiv

  9. [17]

    Image distillation for safe data sharing in histopathology

    Zhe Li and Bernhard Kainz. Image distillation for safe data sharing in histopathology. arXiv preprint arXiv:2406.13536,

    arXiv

  10. [18]

    Few-shot dataset dis- tillation via translative pre-training

    Songhua Liu and Xinchao Wang. Few-shot dataset dis- tillation via translative pre-training. In Proceedings of the IEEE/CVF international conference on computer vision (ICCV), 2023. 2

    2023

  11. [19]

    Dataset distillation via factorization

    Songhua Liu, Kai Wang, Xingyi Yang, Jingwen Ye, and Xin- chao Wang. Dataset distillation via factorization. Advances in Neural Information Processing Systems (NeurIPS) , 35: 1100–1113, 2022. 2

    2022

  12. [20]

    Dream: Efficient dataset distillation by repre- sentative matching

    Yanqing Liu, Jianyang Gu, Kai Wang, Zheng Zhu, Wei Jiang, and Yang You. Dream: Efficient dataset distillation by repre- sentative matching. In Proceedings of the IEEE/CVF inter- national conference on computer vision (ICCV), 2023. 2

    2023

  13. [21]

    Predictive uncertainty es- timation via prior networks

    Andrey Malinin and Mark Gales. Predictive uncertainty es- timation via prior networks. Advances in neural information processing systems, 31, 2018. 2

    2018

  14. [22]

    Reverse kl-divergence training of prior networks: Improved uncertainty and adver- sarial robustness

    Andrey Malinin and Mark Gales. Reverse kl-divergence training of prior networks: Improved uncertainty and adver- sarial robustness. Advances in Neural Information Process- ing Systems, 32, 2019. 2

    2019

  15. [23]

    Stochastic seg- mentation networks: Modelling spatially correlated aleatoric uncertainty

    Miguel Monteiro, Lo ¨ıc Le Folgoc, Daniel Coelho de Castro, Nick Pawlowski, Bernardo Marques, Konstantinos Kamnit- sas, Mark van der Wilk, and Ben Glocker. Stochastic seg- mentation networks: Modelling spatially correlated aleatoric uncertainty. Advances in neural information ...

    2020

  16. [24]

    Dataset meta-learning from kernel ridge-regression

    Timothy Nguyen, Zhourong Chen, and Jaehoon Lee. Dataset meta-learning from kernel ridge-regression. arXiv preprint arXiv:2011.00050, 2020. 2

    2011 arXiv

  17. [25]

    Liu, Yuri A

    Ahmad Sajedi, Samir Khaki, Ehsan Amjadian, Lucy Z. Liu, Yuri A. Lawryshyn, and Konstantinos N. Plataniotis. Datadam: Efficient dataset distillation with attention match- ing. In Proceedings of the IEEE/CVF international confer- ence on computer vision (ICCV), 2023. 2

    2023

  18. [26]

    Stylegan- xl: Scaling stylegan to large diverse datasets

    Axel Sauer, Katja Schwarz, and Andreas Geiger. Stylegan- xl: Scaling stylegan to large diverse datasets. In ACM SIG- GRAPH 2022 conference proceedings, pages 1–10, 2022. 1, 2

    2022

  19. [27]

    Eviden- tial deep learning to quantify classification uncertainty

    Murat Sensoy, Lance Kaplan, and Melih Kandemir. Eviden- tial deep learning to quantify classification uncertainty. Ad- vances in neural information processing systems , 31, 2018. 2

    2018

  20. [28]

    Generative teaching networks: Accelerating neural architecture search by learning to gener- ate synthetic training data

    Felipe Petroski Such, Aditya Rawal, Joel Lehman, Kenneth Stanley, and Jeffrey Clune. Generative teaching networks: Accelerating neural architecture search by learning to gener- ate synthetic training data. In International Conference on Machine Learning, pages 9206–9216. PMLR, 2020. 2

    2020

  21. [29]

    Cafe: Learning to condense dataset by align- ing features

    Kai Wang, Bo Zhao, Xiangyu Peng, Zheng Zhu, Shuo Yang, Shuo Wang, Guan Huang, Hakan Bilen, Xinchao Wang, and Yang You. Cafe: Learning to condense dataset by align- ing features. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12196– ...

    2022

  22. [30]

    Dataset distillation

    Tongzhou Wang, Jun-Yan Zhu, Antonio Torralba, and Alexei A Efros. Dataset distillation. arXiv preprint arXiv:1811.10959, 2018. 2

    2018 arXiv

  23. [31]

    Bayesian deep learn- ing and a probabilistic perspective of generalization

    Andrew G Wilson and Pavel Izmailov. Bayesian deep learn- ing and a probabilistic perspective of generalization. Ad- vances in neural information processing systems , 33:4697– 4708, 2020. 2

    2020

  24. [32]

    Medm- nist v2-a large-scale lightweight benchmark for 2d and 3d biomedical image classification

    Jiancheng Yang, Rui Shi, Donglai Wei, Zequan Liu, Lin Zhao, Bilian Ke, Hanspeter Pfister, and Bingbing Ni. Medm- nist v2-a large-scale lightweight benchmark for 2d and 3d biomedical image classification. Scientific Data, 10(1):41,

  25. [33]

    Bootstrapping semi-supervised medical image segmentation with anatomical-aware contrastive dis- tillation

    Chenyu You, Weicheng Dai, Yifei Min, Lawrence Staib, and James S Duncan. Bootstrapping semi-supervised medical image segmentation with anatomical-aware contrastive dis- tillation. In International Conference on Information Pro- cessing in Medical Imaging, pages 641–653. Spring...

    2023

  26. [34]

    Dataset distillation: A comprehensive review

    Ruonan Yu, Songhua Liu, and Xinchao Wang. Dataset distillation: A comprehensive review. arXiv preprint arXiv:2301.07014, 2023. 2

    2023 arXiv

  27. [35]

    Accelerating dataset distillation via model augmenta- tion

    Lei Zhang, Jie Zhang, Bowen Lei, Subhabrata Mukherjee, Xiang Pan, Bo Zhao, Caiwen Ding, Yao Li, and Dongkuan Xu. Accelerating dataset distillation via model augmenta- tion. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 119...

    2023

  28. [36]

    Dataset condensation with differ- entiable siamese augmentation

    Bo Zhao and Hakan Bilen. Dataset condensation with differ- entiable siamese augmentation. In International Conference on Machine Learning (ICML), pages 12674–12685. PMLR,

  29. [37]

    Synthesizing informative train- ing samples with gan

    Bo Zhao and Hakan Bilen. Synthesizing informative train- ing samples with gan. NeurIPS 2022 Workshop on Synthetic Data for Empowering ML Research, 2022. 2

    2022

  30. [38]

    Dataset condensation with distri- bution matching

    Bo Zhao and Hakan Bilen. Dataset condensation with distri- bution matching. Proceedings of the IEEE/CVF Winter Con- ference on Applications of Computer Vision (WACV), 2023. 1, 2, 6

    2023

  31. [39]

    Dataset condensation with gradient matching

    Bo Zhao, Konda Reddy Mopuri, and Hakan Bilen. Dataset condensation with gradient matching. International Confer- ence on Learning Representations (ICLR), 2020. 2, 6

    2020

  32. [40]

    Im- proved distribution matching for dataset condensation

    Ganlong Zhao, Guanbin Li, Yipeng Qin, and Yizhou Yu. Im- proved distribution matching for dataset condensation. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR) , pages 7856–7865,

  33. [41]

    Rethinking data dis- tillation: Do not overlook calibration

    Dongyao Zhu, Bowen Lei, Jie Zhang, Yanbo Fang, Ruqi Zhang, Yiqun Xie, and Dongkuan Xu. Rethinking data dis- tillation: Do not overlook calibration. In Proceedings of the IEEE/CVF international conference on computer vision (ICCV), 2023. 2 Dataset Distillation with Probabilisti...

    2023

  34. [42]

    1; • The main paper can forward reference sub-sections within the supplementary explicitly (e.g

    Rationale Having the supplementary compiled together with the main paper means that: • The supplementary can back-reference sections of the main paper, for example, we can refer to Sec. 1; • The main paper can forward reference sub-sections within the supplementary explicitly ...

Pith tools

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