PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models

Aishan Liu; Jiaheng Liu; Jinyang Guo; Ruihao Gong; Xianglong Liu; Yang Yong; Yushi Huang; Zining Wnag

REVIEW 5 major objections 7 minor 55 references

PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models

T0 review · 5 major / 7 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 Learned sparsity allocation beats heuristics across 40-plus models

desk verdict A useful benchmark with real code and extensive experiments, but its headline allocation ranking rests on one in-group method and should be read as a hypothesis, not a conclusion. read the letter →

arxiv 2412.07268 v1 pith:EFN4QBW7 submitted 2024-12-10 cs.LG cs.MM

Zining Wnag , Jinyang Guo , Ruihao Gong , Yang Yong , Aishan Liu , Yushi Huang , Jiaheng Liu , Xianglong Liu This is my paper · ORCID

classification cs.LGcs.MM

keywords post-trainingsparsitymodelcompressionbenchmarkallocationreconstructiongranularityfine-grainedpruningsparsificationarchitecturepotential

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper introduces PTSBench, a systematic benchmark for post-training sparsity (PTS), the practice of pruning a trained neural network to a target sparsity using only a small calibration set and no retraining. The authors benchmark more than ten pluggable PTS techniques across three vision tasks and over forty off-the-shelf models, and they argue that the results settle several open practice questions: learned sparsity allocation outperforms criterion-based and heuristic allocation, block-wise reconstruction is consistently the best reconstruction granularity, sparse reconstruction input beats dense input, and attention-based models are the most sparsity-friendly. The point of the benchmark is to give researchers and practitioners a reliable ranking of PTS choices and model families so they can design better compression methods and choose sparsifiable backbones.

What carries the argument

The benchmark itself is the central instrument: PTSBench organizes evaluation into five tracks (sparsity allocation, reconstruction, neural architecture, model size robustness, and application tasks) and summarizes each with an overall metric, specifically the quadratic mean of relative task accuracies such as OM_alloc and OM_recon. The two-step PTS pipeline, sparsity allocation then activation reconstruction, is the object under study; the paper isolates each fine-grained technique (allocation strategy, error correction, reconstruction input, reconstruction granularity) by swapping one component while holding the rest fixed. FCPTS serves as the sole representative of learning-based allocation, ERK and L2Norm as criterion-based representatives, and Uniform as the heuristic baseline.

What would settle it

Run the same PTSBench protocol on a held-out set of architectures, such as EfficientNet, ConvNeXt, or a CNN-transformer hybrid, and also at sparsity rates below 50% and above 80%; if block-wise reconstruction fails to beat layer-wise reconstruction, or if a new learning-based allocation method falls below L2Norm, the paper's headline conclusions are refuted.

Watch

Extended reading notes

Core claim

PTSBench's central claim is that the two-step PTS pipeline, sparsity allocation followed by reconstruction, can be decomposed into fine-grained choices whose effects are consistent and measurable. On allocation, the overall metric ranks FCPTS, a learning-based method, first at 94.36, well ahead of the criterion-based L2Norm at 83.94 and ERK at 80.61, with the heuristic Uniform strategy last at 63.59. On reconstruction, block-wise reconstruction achieves the best overall reconstruction gain (66.76) versus layer-wise (64.95) and single-layer (62.54), and using the sparse model's own output as reconstruction input beats dense input, especially at high sparsity. Across architectures, attention-based models (ViT, MobileNetV3, DeiT) rank at the top of sparsity potential; the authors attribute this to attention concentrating information on critical neurons, and they also report that detection models sparsify better than classification models, while image generation collapses beyond 50% sparsity.

Load-bearing premise

The benchmark's headline rankings assume that FCPTS is a fair representative of learning-based sparsity allocation and that the tested models and sparsity range (50-80%) cover the regimes that matter, so if either fails, the 'learning-based beats criterion-based' conclusion may not generalize.

Editorial extensions

If this is right

If the rankings hold, practitioners should use learned allocation methods over uniform or criterion-based allocation, with an ERK-style initialization for the learned method.
Block-wise reconstruction should become the default reconstruction granularity in PTS, since it is both higher-performing and faster than layer-wise or single reconstruction.
Attention-based architectures should be the preferred backbones when a model will be sparsified for deployment, and large-scale pretraining appears to further increase sparsity friendliness.
Detection backbones tolerate sparsity better than classification models, suggesting that downstream-head parameters can absorb pruning damage.
PTS methods specifically designed for diffusion and generation models are needed, since existing general methods collapse beyond 50% sparsity.

Reading between the lines

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

The paper's ranking of allocation strategies depends on FCPTS being representative of all learning-based allocation; a natural test is whether a different learned method not developed by the same group preserves the ranking.
The correlation between attention mechanisms and sparsity friendliness could be probed causally by ablating the squeeze-and-excitation block in MobileNetV3 against MobileNetV2, extending the paper's observational comparison.
The claim that large-scale pretraining improves sparsity robustness suggests a testable extension: pretrain the same architecture at several data scales and measure the overall metric at fixed sparsity.
The reported collapse of generation tasks beyond 50% sparsity implies that the two-step PTS pipeline may need a fundamentally different allocation objective for diffusion models, not just tuned hyperparameters.

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.

Desk Editor's Note

read the letter

PTSBench is a genuinely useful empirical resource. The authors systematically decompose post-training sparsity into allocation and reconstruction components and benchmark them across 40+ models, three vision tasks, and multiple sparsity rates, with code released. That alone is valuable: the community can now compare pluggable techniques on a common harness. The findings that block-wise reconstruction is both better and faster, that sparse input outperforms dense input, and that model-size robustness doesn't track sparsity potential are useful, and I take them as plausible.

The soft spots are where the paper moves from observations to categorical conclusions. The headline ranking 'learning-based > criterion-based > heuristic-based' for allocation rests on exactly one learning-based method, FCPTS, which comes from the authors' group. The paper says FCPTS is the only existing method, which is true, but then the conclusion silently equates FCPTS's behavior with a property of the whole category. The spread between FCPTS's two initializations (91.69 vs 94.36 OM) is smaller than the gap to L2Norm (83.94), but a different learning-based allocator could easily land below L2Norm. That makes the category-level claim a hypothesis, not a benchmark result. The authors should either add an independent learning-based baseline or explicitly frame the conclusion as 'FCPTS outperforms'.

There are also methodological gaps. No error bars or significance tests anywhere. For a benchmark that uses 8,000 GPU-hours, I'd accept a single run, but then the rankings should be called trends, not findings. Table 4 reports identical GEN MS values (88.32) across several reconstruction techniques, which looks like the generation task simply doesn't respond to reconstruction; that deserves a comment, not a uniform number. The architecture conclusions are confounded: ViT vs DeiT differ in pretraining data, and MobileNetV3 vs MobileNetV2 differ beyond the SE block. The paper does acknowledge the training-data confound for ViT/DeiT but still attributes the gap to attention structure. The sparsity range 50-80% is defended as post-hoc, but Fig 3 shows a 90% allocation for ResNet-32, which is inconsistent with the stated collapse regime. And '20,000 epochs' of calibration is unexplained; presumably they mean iterations.

None of this kills the paper's value. It's a solid empirical contribution that deserves peer review, but it needs another learning-based allocation method or a more modest claim, plus basic error reporting. I'd cite it for the benchmark and the reconstruction results, but I would not build on the allocation ranking until it's confirmed by a second independent method.

Referee Report

5 major / 7 minor

Summary. PTSBench is a benchmark for post-training sparsity (PTS) covering 10+ fine-grained techniques, 3 vision tasks, and over 40 off-the-shelf model architectures. The paper formalizes the common two-stage PTS pipeline (sparsity allocation and reconstruction), introduces five overall metrics (OM_alloc, OM_recon, OM_arch, OM_robust, OM_task), and reports results on five tracks: sparsity allocation, reconstruction, neural architecture, model size robustness, and application tasks. The main takeaways are a category-level ranking of allocation strategies (learning-based > criterion-based > heuristic-based), reconstruction guidance (error correction helps only classification, sparse input is better, block-wise granularity is best), and model-level claims (attention-based models are more sparsity-friendly, large-scale pre-training improves sparsity potential, image generation needs specialized methods). The paper releases code and includes an explicit limitations section.

Significance. The benchmark addresses a practical and underserved area, and the scale of the experiments (over 8000 A800 GPU hours, 40+ models, three tasks) is a genuine asset. The paper is also commendably transparent about some limitations in Section 6, and the open-source framework is likely to be useful to the community. If the headline conclusions hold, they would provide concrete guidance for PTS method design and model selection. However, several load-bearing conclusions are currently under-supported: the category-level allocation ranking rests on a single learning-based method (FCPTS) from the authors' own group, the reconstruction conclusions rely on small differences with no error bars and with suspicious constant values in Table 4, and the architecture conclusions are drawn from confounded comparisons. These issues are fixable, but they require additional experiments or substantially softened claims before the central assertions can be accepted.

major comments (5)

[§3.1, Table 2, Table 7 (Track 1)] The takeaway 'Priority: learning-based > criterion-based > heuristic-based' is supported by exactly one learning-based algorithm, FCPTS, which is from the same research group as this paper (Ref. [10]). Table 2 shows that FCPTS(L2Norm) and FCPTS(ERK) differ by 2.67 points in OM_alloc (91.69 vs 94.36), a spread comparable to the 3.33-point gap between L2Norm and ERK (83.94 vs 80.61). Because the variance across initializations of the single learning-based method is the same order of magnitude as the difference between the two criterion-based methods, the data do not establish that 'learning-based' as a category is superior. The paper should either include at least one independent learning-based allocation method or rephrase the conclusion as 'FCPTS outperforms the evaluated criterion-based and heuristic methods'.
[§5.1.2, Table 4] The reconstruction track reports no error bars or significance tests, and several entries are suspicious: the GEN MS column is exactly 88.32 for 'w/ Correction', 'w/o Correction', 'Sparse Input', and 'Block-wise', with other GEN values also nearly constant (88.06, 88.11, 88.18). Since the headline conclusion 'block-wise reconstruction is always the best' depends on OM_recon differences of about 2 points (66.76 vs 64.95 vs 62.54), these exact-looking values and the absence of variance make it impossible to assess whether the differences are meaningful. Please report standard deviations across at least multiple seeds (or explain why they are absent) and clarify how the GEN reconstruction gains are computed at each sparsity rate.
[§4, Section 4] The benchmark restricts evaluation to sparsity rates {0.5, 0.6, 0.7, 0.8} with the justification that lower rates are 'negligible' and higher rates 'collapse'. This is a post-hoc choice that excludes the high-sparsity regime where PTS guidance is arguably most needed, and all overall metrics are averaged over this range. The paper should either validate at least a representative subset of settings at higher sparsity (e.g., 0.9 and 0.95), or explicitly reposition PTSBench as a mid-sparsity benchmark and avoid general statements about 'the sparsification ability of models' that may not transfer beyond 80%.
[§5.2.1, Table 5] The model-level conclusions confound architecture with other factors. The claim that 'attention mechanism is a more sparsity-friendly model structure' is based on comparing MobileNetV3 vs MobileNetV2, ViT vs CNNs, and DeiT vs ViT, but these models differ in multiple ways beyond attention (inverted residuals, SE blocks, activation functions, patch size, input resolution, pre-training data, distillation). Similarly, the 'training strategy' conclusion drawn from DeiT vs ViT is confounded because the two models also differ in architecture and pre-training data. Controlled comparisons are needed, such as the same architecture with and without an SE block, or a single architecture pre-trained at different scales, before these model-level conclusions can be considered established.
[Table 2, Eqs. (2)-(7)] Table 2 reports only a single GEN value per allocation method (e.g., 16.40, 78.47, 91.71) with dashes for 60/70/80, while Table 6 reports GEN relative accuracy at all four sparsity rates (78.06, 5.44, 0.18, 0). This means OM_alloc appears to average over different numbers of sparsity rates per task, making the cross-task and cross-method comparisons in Tables 2 and 7 difficult to interpret. Please specify the exact support of the mean in each of Eqs. (2)-(7) and make the per-sparsity-rate GEN results available for the allocation track.

minor comments (7)

[Table 4] 'Singe' should be 'Single'.
[§5.3] There are typos: 'sparisy' should be 'sparsity' and 'perpectives' should be 'perspectives'.
[Table 3] Time entries such as '3h34' should be written as '3h34m' for consistency with '3h32m'.
[Figure 3] The label 'layer31.2.conv1' appears to be a typo for 'layer3.1.2.conv1'.
[§4] The text says the models are calibrated for '20,000 epochs' with 1,024 images and batch size 64; this likely means iterations or steps, and the intended meaning should be clarified.
[References] The reference list appears twice in the manuscript; the duplicate should be removed.
[Eqs. (2)-(7)] The notation E^2(...) is undefined; please clarify whether E denotes the arithmetic mean and E^2 denotes the square of that mean.

Circularity Check

0 steps flagged · score 2.0 of 10

No derivational circularity: PTSBench's conclusions are empirical measurements; the main concern is that the 'learning-based' allocation category is represented solely by the authors' own FCPTS, a representativeness/independence issue rather than a circular reduction.

full rationale

PTSBench's derivation chain is empirical rather than algebraic: the overall metrics in Eqs. (2)-(7) are explicit aggregations of measured relative accuracies or reconstruction gains, and the headline rankings (learning-based allocation, block-wise reconstruction, attention-friendly architectures) are read off Tables 2, 4, and 5 rather than derived from fitted parameters. No fitted value is renamed as a prediction, and no result is defined into existence: the benchmark's conclusions are not consequences of its metric definitions. The self-citations are to methods placed under test (FCPTS [10] as the sole learning-based allocation method, BRECQ [33] as the source of block-wise reconstruction), and the block-wise finding is verified by the measured OM_recon in Table 4, not imported from the citation. The paper's own limitation statement acknowledges the small pool of PTS algorithms, and Section 3.1's claim that FCPTS is 'the only existing method' means the category-level takeaway rests on a single self-authored algorithm; that is a legitimate generalizability and independence risk but not a circular step. Apart from these non-load-bearing self-citations, the benchmark is self-contained empirical evaluation against standard datasets and models.

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

No new theoretical entities are introduced; the central claims rest on hand-chosen evaluation ranges, a representative model sample, and the two-step PTS framing.

free parameters (3)

Sparsity rate range = {0.5, 0.6, 0.7, 0.8}
Hand-chosen after observing behavior in pilot runs; all overall metrics (Eqs. 2-7) and rankings are computed over this range, so conclusions depend on it.
Calibration budget = 1,024 images; 20,000 epochs; batch size 64
Hand-chosen reconstruction budget; the epoch count is likely a typo, and the budget directly affects all reconstruction results in track 2.
Reconstruction optimizer settings = SGD, lr=1e-4, momentum=0.9
Hand-chosen; affects reconstruction gains and thus comparisons between granularities.

assumptions (3)

domain assumption PTS methods follow the two-step paradigm of sparsity allocation followed by reconstruction
Section 2.1; the benchmark's decomposition into tracks rests on this framing and excludes methods outside it.
domain assumption Relative accuracy (sparse/dense, or reciprocal FID) is a comparable measure of sparsification ability across tasks and architectures
Equations 2-7; the overall metrics assume that relative accuracies can be averaged across heterogeneous tasks and model sizes.
ad hoc to paper The selected model families, tasks, and datasets represent the space of 'off-the-shelf' models for which PTS guidance is needed
Section 3.2 and Table 1; the generalizability of all takeaways ('attention is sparsity-friendly,' etc.) is assumed from this sample.

how reviews work

0 comments

Cite this review

Pith. "Pith review of PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models." pith.science (2026). https://pith.science/paper/EFN4QBW7

@misc{pith2026241207268,
  author       = {Pith},
  title        = {Pith review of: PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EFN4QBW7}},
  note         = {Machine review of arXiv:2412.07268}
}

read the original abstract

With the increased attention to model efficiency, post-training sparsity (PTS) has become more and more prevalent because of its effectiveness and efficiency. However, there remain questions on better practice of PTS algorithms and the sparsification ability of models, which hinders the further development of this area. Therefore, a benchmark to comprehensively investigate the issues above is urgently needed. In this paper, we propose the first comprehensive post-training sparsity benchmark called PTSBench towards algorithms and models. We benchmark 10+ PTS general-pluggable fine-grained techniques on 3 typical tasks using over 40 off-the-shelf model architectures. Through extensive experiments and analyses, we obtain valuable conclusions and provide several insights from both algorithms and model aspects. Our PTSBench can provide (1) new observations for a better understanding of the PTS algorithms, (2) in-depth and comprehensive evaluations for the sparsification ability of models, and (3) a well-structured and easy-integrate open-source framework. We hope this work will provide illuminating conclusions and advice for future studies of post-training sparsity methods and sparsification-friendly model design. The code for our PTSBench is released at \href{https://github.com/ModelTC/msbench}{https://github.com/ModelTC/msbench}.

Figures

Figures reproduced from arXiv: 2412.07268 by the authors.

**Figure 2.** Evaluation tracks of PTSBench. We benchmark the performance of PTS fine-grained algorithms and model sparsi [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Visualization of sparsity allocation of ResNet-32 at a sparsity rate of 90% on CIFAR-100. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Mean relative accuracy loss of different model sizes. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: The overall evaluation results of PTSBench. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 25 canonical work pages

[10]

Ruihao Gong, Yang Yong, Zining Wang, Jinyang Guo, Xiuying Wei, Yuqing Ma, and Xianglong Liu. 2024. Fast and Controllable Post-training Sparsity: Learning Optimal Sparsity Allocation with Global Constraint in Minutes. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 38. 12190–12198

work page 2024
[1]

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

2009
[2]

Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xi- aohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. 2020. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)

arXiv 2020
[3]

Gintare Karolina Dziugaite and Daniel M Roy. 2015. Neural network matrix factorization. arXiv preprint arXiv:1511.06443 (2015)

arXiv 2015
[4]

Utku Evci, Trevor Gale, Jacob Menick, Pablo Samuel Castro, and Erich Elsen
[5]

Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. 2010. The pascal visual object classes (voc) challenge. Inter- national journal of computer vision 88 (2010), 303–338

2010
[6]

Jonathan Frankle, Gintare Karolina Dziugaite, Daniel M Roy, and Michael Carbin

work page
[7]

Elias Frantar and Dan Alistarh. 2022. Optimal brain compression: A framework for accurate post-training quantization and pruning. Advances in Neural Information Processing Systems 35 (2022), 4475–4488

work page 2022

Show all 55 references

[8]

Trevor Gale, Erich Elsen, and Sara Hooker. 2019. The state of sparsity in deep neural networks. arXiv preprint arXiv:1902.09574 (2019)

2019 arXiv
[9]

Ruihao Gong, Yang Yong, Shiqiao Gu, Yushi Huang, Yunchen Zhang, Xianglong Liu, and Dacheng Tao. 2024. LLMC: Benchmarking Large Language Model Quan- tization with a Versatile Compression Toolkit. arXiv preprint arXiv:2405.06001 (2024)

2024 arXiv
[11]

Jinyang Guo, Jiaheng Liu, Zining Wang, Yuqing Ma, Ruihao Gong, Ke Xu, and Xianglong Liu. 2023. Adaptive Contrastive Knowledge Distillation for BERT Compression. In Findings of the Association for Computational Linguistics: ACL

2023
[12]

Jinyang Guo, Wanli Ouyang, and Dong Xu. 2020. Multi-Dimensional Pruning: A Unified Framework for Model Compression. In CVPR

2020
[13]

Jinyang Guo, Jianyu Wu, Zining Wang, Jiaheng Liu, Ge Yang, Yifu Ding, Ruihao Gong, Haotong Qin, and Xianglong Liu. 2024. Compressing large language models by joint sparsification and quantization. In Forty-first International Conference on Machine Learning

2024
[14]

Song Han, Jeff Pool, John Tran, and William Dally. 2015. Learning both weights and connections for efficient neural network. Advances in neural information processing systems 28 (2015)

2015
[15]

Babak Hassibi and David Stork. 1992. Second order derivatives for network pruning: Optimal brain surgeon. Advances in neural information processing systems 5 (1992)

1992
[16]

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition . 770–778

2016
[17]

Yihui He, Ji Lin, Zhijian Liu, Hanrui Wang, Li-Jia Li, and Song Han. 2018. Amc: Automl for model compression and acceleration on mobile devices. InProceedings of the European conference on computer vision (ECCV) . 784–800

2018
[18]

Yefei He, Luping Liu, Jing Liu, Weijia Wu, Hong Zhou, and Bohan Zhuang. 2023. PTQD: Accurate Post-Training Quantization for Diffusion Models. arXiv preprint arXiv:2305.10657 (2023)

2023 arXiv
[19]

Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. 2017. Gans trained by a two time-scale update rule converge to a local nash equilibrium. Advances in neural information processing systems 30 (2017)

2017
[20]

Geoffrey Hinton, Oriol Vinyals, and Jeff Dean. 2015. Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531 (2015)

2015 arXiv
[21]

Andrew Howard, Mark Sandler, Grace Chu, Liang-Chieh Chen, Bo Chen, Mingx- ing Tan, Weijun Wang, Yukun Zhu, Ruoming Pang, Vijay Vasudevan, et al. 2019. Searching for mobilenetv3. InProceedings of the IEEE/CVF international conference on computer vision. 1314–1324

2019
[22]

Jie Hu, Li Shen, and Gang Sun. 2018. Squeeze-and-excitation networks. InProceed- ings of the IEEE conference on computer vision and pattern recognition . 7132–7141

2018
[23]

Yushi Huang, Ruihao Gong, Jing Liu, Tianlong Chen, and Xianglong Liu. 2024. Tfmq-dm: Temporal feature maintenance quantization for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 7362–7371

2024
[24]

Itay Hubara, Brian Chmiel, Moshe Island, Ron Banner, Joseph Naor, and Daniel Soudry. 2021. Accelerated sparse neural training: A provable and efficient method to find n: m transposable masks.Advances in neural information processing systems 34 (2021), 21099–21111

2021
[25]

Itay Hubara, Matthieu Courbariaux, Daniel Soudry, Ran El-Yaniv, and Yoshua Bengio. 2017. Quantized neural networks: Training neural networks with low precision weights and activations. The Journal of Machine Learning Research 18, 1 (2017), 6869–6898

2017
[26]

J. Guo, W. Ouyang, and D. Xu. 2020. Channel pruning guided by classification loss and feature importance. In AAAI

2020
[27]

Benoit Jacob, Skirmantas Kligys, Bo Chen, Menglong Zhu, Matthew Tang, Andrew Howard, Hartwig Adam, and Dmitry Kalenichenko. 2018. Quantization and2 training of neural networks for efficient integer-arithmetic-only inference. In IEEE Conf. Comput. Vis. Pattern Recog. 2704–2713

2018
[28]

Raghuraman Krishnamoorthi. 2018. Quantizing deep convolutional networks for efficient inference: A whitepaper. arXiv preprint arXiv:1806.08342 (2018)

2018 arXiv
[29]

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

2009
[30]

Aditya Kusupati, Vivek Ramanujan, Raghav Somani, Mitchell Wortsman, Prateek Jain, Sham Kakade, and Ali Farhadi. 2020. Soft threshold weight reparameter- ization for learnable sparsity. In International Conference on Machine Learning . PMLR, 5544–5555

2020
[31]

Ivan Lazarevich, Alexander Kozlov, and Nikita Malinin. 2021. Post-training deep neural network pruning via layer-wise calibration. InProceedings of the IEEE/CVF International Conference on Computer Vision . 798–805

2021
[32]

Yawei Li, Kamil Adamczewski, Wen Li, Shuhang Gu, Radu Timofte, and Luc Van Gool. 2022. Revisiting random channel pruning for neural network compres- sion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 191–201

2022
[33]

Yuhang Li, Ruihao Gong, Xu Tan, Yang Yang, Peng Hu, Qi Zhang, Fengwei Yu, Wei Wang, and Shi Gu. 2021. Brecq: Pushing the limit of post-training quantization by block reconstruction. arXiv preprint arXiv:2102.05426 (2021)

2021 arXiv
[34]

Tsung-Yi Lin, Piotr Dollár, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. 2017. Feature pyramid networks for object detection. In Proceed- ings of the IEEE conference on computer vision and pattern recognition . 2117–2125

2017
[35]

Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollár. 2017. Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision . 2980–2988. PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithm...

2017
[36]

Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C Lawrence Zitnick. 2014. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proce...

2014
[37]

Jiawei Liu, Lin Niu, Zhihang Yuan, Dawei Yang, Xinggang Wang, and Wenyu Liu. 2023. Pd-quant: Post-training quantization based on prediction difference metric. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 24427–24437

2023
[38]

Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, and Alexander C Berg. 2016. Ssd: Single shot multibox detec- tor. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part...

2016
[39]

Zhuang Liu, Jianguo Li, Zhiqiang Shen, Gao Huang, Shoumeng Yan, and Chang- shui Zhang. 2017. Learning efficient convolutional networks through network slimming. In Proceedings of the IEEE international conference on computer vision . 2736–2744

2017
[40]

Chengtao Lv, Hong Chen, Jinyang Guo, Yifu Ding, and Xianglong Liu. 2024. PTQ4SAM: Post-Training Quantization for Segment Anything. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 15941– 15951

2024
[41]

Ningning Ma, Xiangyu Zhang, Hai-Tao Zheng, and Jian Sun. 2018. Shufflenet v2: Practical guidelines for efficient cnn architecture design. In Proceedings of the European conference on computer vision (ECCV) . 116–131

2018
[42]

Hesham Mostafa and Xin Wang. 2019. Parameter efficient training of deep convo- lutional neural networks by dynamic sparse reparameterization. In International Conference on Machine Learning . PMLR, 4646–4655

2019
[43]

Markus Nagel, Mart van Baalen, Tijmen Blankevoort, and Max Welling. 2019. Data-free quantization through weight equalization and bias correction. In Pro- ceedings of the IEEE/CVF International Conference on Computer Vision . 1325–1334

2019
[44]

Markus Nagel, Marios Fournarakis, Rana Ali Amjad, Yelysei Bondarenko, Mart Van Baalen, and Tijmen Blankevoort. 2021. A white paper on neural network quantization. arXiv preprint arXiv:2106.08295 (2021)

2021 arXiv
[45]

Ilija Radosavovic, Raj Prateek Kosaraju, Ross Girshick, Kaiming He, and Piotr Dollár. 2020. Designing network design spaces. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 10428–10436

2020
[46]

Alex Renda, Jonathan Frankle, and Michael Carbin. 2020. Comparing rewinding and fine-tuning in neural network pruning. arXiv preprint arXiv:2003.02389 (2020)

2020 arXiv
[47]

Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 10684–10695

2022
[48]

Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, and Liang- Chieh Chen. 2018. Mobilenetv2: Inverted residuals and linear bottlenecks. In Proceedings of the IEEE conference on computer vision and pattern recognition . 4510–4520

2018
[49]

Karen Simonyan and Andrew Zisserman. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014)

2014 arXiv
[50]

Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, and Hervé Jégou. 2021. Training data-efficient image transformers & distillation through attention. In International conference on machine learning . PMLR, 10347–10357

2021
[51]

Chaoqi Wang, Guodong Zhang, and Roger Grosse. 2020. Picking winning tickets before training by preserving gradient flow. arXiv preprint arXiv:2002.07376 (2020)

2020 arXiv
[52]

Xiuying Wei, Ruihao Gong, Yuhang Li, Xianglong Liu, and Fengwei Yu. 2022. Qdrop: Randomly dropping quantization for extremely low-bit post-training quantization. arXiv preprint arXiv:2203.05740 (2022)

2022 arXiv
[53]

Fisher Yu, Ari Seff, Yinda Zhang, Shuran Song, Thomas Funkhouser, and Jianxiong Xiao. 2015. Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop. arXiv preprint arXiv:1506.03365 (2015)

2015 arXiv
[2019]

arXiv preprint arXiv:1903.01611 (2019)

Stabilizing the lottery ticket hypothesis. arXiv preprint arXiv:1903.01611 (2019)

2019 arXiv
[2020]

In International Conference on Machine Learning

Rigging the lottery: Making all tickets winners. In International Conference on Machine Learning. PMLR, 2943–2952

Pith tools

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

[1] [10]

Ruihao Gong, Yang Yong, Zining Wang, Jinyang Guo, Xiuying Wei, Yuqing Ma, and Xianglong Liu. 2024. Fast and Controllable Post-training Sparsity: Learning Optimal Sparsity Allocation with Global Constraint in Minutes. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 38. 12190–12198

work page 2024

[2] [1]

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

2009

[3] [2]

Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xi- aohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. 2020. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)

arXiv 2020

[4] [3]

Gintare Karolina Dziugaite and Daniel M Roy. 2015. Neural network matrix factorization. arXiv preprint arXiv:1511.06443 (2015)

arXiv 2015

[5] [4]

Utku Evci, Trevor Gale, Jacob Menick, Pablo Samuel Castro, and Erich Elsen

[6] [5]

Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. 2010. The pascal visual object classes (voc) challenge. Inter- national journal of computer vision 88 (2010), 303–338

2010

[7] [6]

Jonathan Frankle, Gintare Karolina Dziugaite, Daniel M Roy, and Michael Carbin

work page

[8] [7]

Elias Frantar and Dan Alistarh. 2022. Optimal brain compression: A framework for accurate post-training quantization and pruning. Advances in Neural Information Processing Systems 35 (2022), 4475–4488

work page 2022