Pith. sign in

REVIEW 4 major objections 4 minor 47 references

EDPC: Accelerating Lossless Compression via Lightweight Probability Models and Decoupled Parallel Dataflow

T0 review · 4 major / 4 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 EDPC beats the strongest prior neural compressor by 3.2% in average compression ratio while running 2.7x faster and using up to 4x fewer parameters.

desk verdict A practical compression systems paper with a plausible 3.2% ratio gain over PAC, but the LTE parameter/memory accounting is internally inconsistent as written and needs a code check before the efficiency claims are believed. read the letter →

arxiv 2507.18969 v1 pith:SZST4UXT submitted 2025-07-25 cs.IT eess.SPmath.IT

Zeyi Lu , Xiaoxiao Ma , Yujun Huang , Minxiao Chen , Bin Chen , Baoyi An , Shu-Tao Xia This is my paper · ORCID
classification cs.ITeess.SPmath.IT MSC 68P3094A17
keywords losslesscompressionautoregressivemodelarithmeticcodingmutualinformationmulti-branchfeatureextractionlatenttransformationpipelineparallelismGPU-CPUcollaboration
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 sets out to show that lossless neural compression can improve both objectives at once: a better compression ratio and a much faster, lighter pipeline. It proposes EDPC (Efficient Dual-path Parallel Compression), which combines a two-branch byte-probability model (MBRB), a latent-space compression module (LTE), and a decoupled GPU-CPU pipeline (DPCA) that splits byte streams into segments for parallel arithmetic coding. Against the strongest prior autoregressive model, PAC, the paper reports an average 3.2% improvement in compression ratio, 2.7x faster compression, up to a 4x parameter reduction, and roughly 1.91x lower GPU memory. The motivation is that real-time transmission of large multimedia data needs neural compressors that are fast and memory-cheap, not just tight.

What carries the argument

The load-bearing object is the three-component EDPC stack. MBRB is a two-branch feature module: input is layer-normalized, passed through two separate linear paths, fused by elementwise multiplication, then through a GELU and an output projection with a residual connection; the IFR metric, a ratio of estimated mutual information between the skip connection and the fused output, justifies stopping at two branches. LTE is a compression-recovery bottleneck that down-projects features to a latent dimension, multiplies them by a learnable Feature Distribution Matrix, and up-projects back, explaining the memory and parameter reductions. DPCA is the decoupled pipeline that runs probability prediction and encoding concurrently on GPU and CPU and uses a process pool with 32 subprocesses to encode independent stream segments in parallel. These three pieces together are the paper's argument that ratio and efficiency can be improved in the same design.

What would settle it

Run EDPC on Silesia in one-segment mode versus 32-segment mode and compare total compressed sizes; if the 32-segment output is more than about 3.2% larger than the single-segment output, the claimed ratio advantage over PAC would not survive on a single uninterrupted stream.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that an autoregressive byte-level compressor does not have to choose between modeling quality and system efficiency. The MBRB replaces the single-branch feature path of earlier models with two parallel linear transformations whose outputs are multiplied elementwise, and the IFR metric, computed from mutual information estimates, is used to argue that a second branch captures most of the information-diversity benefit that additional branches offer. The LTE shrinks the high-dimensional feature map to a latent bottleneck, applies a learnable Feature Distribution Matrix in that space, and then recovers the original dimension, which the paper credits for most of the memory and parameter savings. The DPCA pipeline overlaps GPU-side probability prediction with CPU-side arithmetic coding and partitions each batch's byte stream across 32 subprocesses, which the paper credits for the speedup. The result is the paper's claim of a 3.2% average compression-ratio gain over PAC with far lower resource use.

Load-bearing premise

The load-bearing premise is that splitting a byte stream into independent segments for parallel arithmetic coding costs almost nothing in compression ratio, yet the paper reports no measurement of the context loss at segment boundaries.

Editorial extensions

If this is right

  • If the reported numbers hold, arithmetic coding's sequential dependency is not a hard barrier to parallel neural compression: decoupling prediction from encoding and segmenting the stream yields a 2.7x overall speedup and up to a 21.73x reduction in encoding time.
  • A two-branch multiplicative fusion captures most of the modeling benefit of more branches: moving from two to three branches adds only 0.02 on Backup and 0.03 on Silesia in the reported ratio metric while raising memory by 8.8% and cutting speed by more than 22%.
  • The LTE bottleneck reduces model parameters and GPU memory by large margins, up to 4x fewer parameters and roughly 1.91x lower memory, while losing little compression ratio, making neural compressors more viable on 12GB-class GPUs.
  • The 3.2% average ratio gain over PAC is uneven across domains: it is 7.71% on Enwik9 but zero on Image, so the method's benefit is concentrated where byte-level dependencies are text-like.

Reading between the lines

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

  • A test the paper does not report: compare EDPC's compressed size with segmentation disabled, meaning one arithmetic-coding context for the whole stream, against the 32-subprocess configuration, because segment boundaries reset context and the reported ratios may therefore overstate the model's true sequential performance.
  • The same decoupled pipeline idea could be applied to other sequential entropy coders, such as context-mixing compressors or transformer-based byte models, because DPCA is orthogonal to the choice of probability model.
  • The IFR metric, a ratio of estimated mutual information, could be reused as an architecture-search signal for choosing the number and fusion style of branches in other lossless models, not only the two-branch MBRB tested here.
  • The near-zero Image gain suggests the multi-branch benefit is tied to byte-level text-like context; an image-specific input representation, such as pixel residuals or bit-planes, might reveal whether the block helps across modalities.
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

4 major / 4 minor

Summary. The paper proposes EDPC, a lossless compression framework combining a Multi-path Byte Refinement Block (MBRB) with mutual-information-based guidance, a Latent Transformation Engine (LTE), and a Decoupled Pipeline Compression Architecture (DPCA). The central claims are that EDPC improves compression ratio by 3.2% over PAC on average, achieves 2.7x faster compression, reduces parameters by up to 4x, and lowers GPU memory by roughly 1.91x. The evaluation is conducted on large benchmark datasets against Gzip, 7z, Zstd, Dzip, TRACE, OREO, and PAC, with ablations for each component.

Significance. If the reported results are correct, EDPC would be a practically useful step in neural lossless compression, combining better byte-level modeling with a GPU-CPU pipeline. The paper has real strengths: the code is released, the experiments use standard external benchmarks with internal consistency in the speed tables, and the ablations in Tables 4 and 5 support the usefulness of the MBRB and DPCA components. However, the LTE parameter accounting is internally inconsistent, and the multi-process encoding setup leaves the compression-ratio penalty from segmentation unmeasured, so both halves of the central claim currently rest on unsupported assumptions.

major comments (4)
  1. [Section 3.2, Eq. (15), Table 4, Figure 5(b)] The Feature Distribution Matrix U is defined as a learnable tensor in R^{b x F' x F'}, which makes the reported parameter and memory numbers internally inconsistent. With the configuration in Section 4.1 (batch size 4096, hidden dimension 2048, r=4 so F'=512), U alone would contain about 4096*512*512 ≈ 1.07e9 trainable parameters and occupy about 4.3 GB in fp32, already exceeding the total 1546 MB memory and 4.15e7 parameters reported for the full EDPC model. Figure 5(b)'s claim that the parameter count decreases sharply with batch size is also impossible under the standard definition of a parameter if U depends on b. The authors should either redefine U as a per-batch non-learnable transformation, report a parameter count that includes U, or provide a corrected architecture; as written, the LTE-based parameter and memory reductions are not supported.
  2. [Section 3.3 and Section 4.1, with Table 2] The multi-process encoding design divides the byte stream into independent segments encoded in parallel by subprocesses, but the paper never specifies the segment length, how the initial uniform coding of the first t bytes in Algorithm 1 is applied per segment, or how the online parameter updates in Algorithm 1 (lines 10-12) are synchronized across subprocesses. Arithmetic coding is inherently sequential and context-dependent, so splitting the stream can degrade the compression ratio. Because Table 2 reports ratios only for the parallel configuration, the central 3.2% average gain over PAC is not yet supported as a property of the full EDPC pipeline; a serial or single-segment control and a sweep over segment lengths are needed.
  3. [Section 3.1.1, Eqs. (4)-(8), Figure 2] The IFR argument is not a valid information-theoretic justification for the multi-branch design. The branch outputs X_i are deterministic functions of the same input X0, and a lower value of I(S; S+X) does not by itself imply that X contributes more diverse information; it could reflect a lossy or noisy transform. IFR is an arbitrary ratio of two estimated mutual informations, and no theorem links it to compression ratio. Since the choice of two branches is empirically supported by Table 5, this flaw does not invalidate the main compression results, but the paper should present the branch-count choice as an empirical finding and either remove or substantially revise the claimed MI grounding.
  4. [Section 4.3, Figure 5(b)] The claim that the number of model parameters decreases sharply as batch size increases is inconsistent with standard parameter accounting, in which the number of trainable parameters is independent of batch size. This appears to be an artifact of the batch-dimensioned FDM in Eq. (15). A parameter-versus-batch-size plot should be based on a definition of parameter count that does not depend on b, or the FDM must be redesigned accordingly.
minor comments (4)
  1. [Table 4] The 'low-rank factorization' ablation row is not described anywhere; a sentence explaining what this baseline is and how it differs from LTE would make the comparison interpretable.
  2. [Section 3.3 and Section 4.1] The text alternates between 'thread pool' and 'process pool' for the same mechanism; if subprocesses are intended, 'process pool' should be used consistently.
  3. [Table 2] The Image column shows EDPC and PAC both at 1.96, so the sentence in Section 4.2 that EDPC 'consistently achieves the best compression ratios across all datasets' should be qualified to acknowledge this tie or explain rounding.
  4. [Section 4.4.3, Figure 7] The reported FLOPs and parameter reductions for the FDM should state the exact shapes and batch size used; as written, the 93.02% and 43.76% figures cannot be independently verified from the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found; the central claims are empirical benchmarks, not consequences of the paper's own definitions.

full rationale

EDPC's headline numbers are measured against external baselines (PAC, TRACE, OREO, Dzip) on standard benchmarks, not derived from the paper's own equations. The IFR metric is a definitional ratio of mutual-information estimates used as a post-hoc narrative for choosing a two-branch MBRB, but the actual branch-count decision is made from measured cost/benefit trade-offs in Table 5 and does not reduce to Eq. (8). LTE and DPCA efficiency claims are empirical and are not obtained by substituting the paper's equations into themselves. I do flag two non-circular correctness concerns: (i) the batch-dependent Feature Distribution Matrix U in Eq. (15) makes the reported total parameter count of 4.15e7 at batch 4096 internally inconsistent (U alone would contain about 1.07e9 trainable entries), and (ii) the unmeasured cost of splitting byte streams into 32 independent segments (Sec. 3.3 and Sec. 4.1) is a potential threat to the compression-ratio claim. Neither is an equation-level reduction of a prediction to its input, nor does either depend on a self-citation or an imported uniqueness theorem. No load-bearing self-citation, ansatz-smuggling citation, or renaming of a known result as a derivation appears. The derivation chain is therefore self-contained in the circularity sense.

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

The central claim rests mainly on standard background in information theory and on the inherited ACM framework. The new interpretive axioms, that lower MI means more diverse features and that segment-wise parallel encoding is ratio-neutral, are asserted without proof. The main free parameters are the architecture sizes and the encoding-pool configuration.

free parameters (5)
  • LTE compression ratio r = 4
    Set to 4 in Section 4.1 without derivation or sensitivity analysis; directly controls the latent dimension F' = F/r.
  • MBRB hidden dimensions = 2048 (local), 4096 (global)
    Chosen in Section 4.1; not justified or swept.
  • Number of branches in MBRB = 2
    Chosen after comparing 2 vs 3 branches in Table 5; the IFR metric is descriptive and does not derive the choice.
  • Context length t = 16
    Fixed at 16 in Section 4.1; standard in baselines but still a modeling choice.
  • Encoding subprocesses and chunk size = 32 subprocesses, chunk size 32
    System parameters in Section 4.1; their effect on compression ratio is not analyzed.
assumptions (5)
  • standard math Shannon entropy and mutual information definitions, and the Kraskov estimator for MI.
    Section 2.2 and Eq. (7); standard information-theoretic background.
  • domain assumption Autoregressive byte-level models trained online from random initialization can serve as compressors without pretraining.
    Algorithm 1 and Section 4.1; inherited from the NNCP/TRACE/OREO/PAC line of work.
  • ad hoc to paper Lower mutual information I(S; S+X) implies that the added feature X contributes more diverse information.
    Section 3.1.1; this interpretive step is asserted without proof and is questionable because X is a deterministic function of S.
  • ad hoc to paper Element-wise multiplication of parallel branch outputs preserves complementary information.
    Section 3.1.2, Eq. (11); no theoretical or empirical support beyond the overall ablation.
  • domain assumption Dividing the byte stream into independently encoded segments for parallel arithmetic coding does not materially degrade the compression ratio.
    Section 3.3 and Section 4.1; this assumption is load-bearing for the speed result and is not tested or quantified.
invented entities (1)
  • Feature Distribution Matrix (FDM)
    purpose: A learnable 3D tensor U in R^{b x F' x F'} used in batch matrix multiplication to model correlations in the latent space (Eq. 15).
    The tensor is introduced as a new architectural component with no external falsifiable prediction. Its stated shape conflicts with the reported total parameter count of 4.15e7 in Table 4.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EDPC: Accelerating Lossless Compression via Lightweight Probability Models and Decoupled Parallel Dataflow." pith.science (2026). https://pith.science/paper/SZST4UXT

@misc{pith2026250718969,
  author       = {Pith},
  title        = {Pith review of: EDPC: Accelerating Lossless Compression via Lightweight Probability Models and Decoupled Parallel Dataflow},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SZST4UXT}},
  note         = {Machine review of arXiv:2507.18969}
}
read the original abstract

The explosive growth of multi-source multimedia data has significantly increased the demands for transmission and storage, placing substantial pressure on bandwidth and storage infrastructures. While Autoregressive Compression Models (ACMs) have markedly improved compression efficiency through probabilistic prediction, current approaches remain constrained by two critical limitations: suboptimal compression ratios due to insufficient fine-grained feature extraction during probability modeling, and real-time processing bottlenecks caused by high resource consumption and low compression speeds. To address these challenges, we propose Efficient Dual-path Parallel Compression (EDPC), a hierarchically optimized compression framework that synergistically enhances modeling capability and execution efficiency via coordinated dual-path operations. At the modeling level, we introduce the Information Flow Refinement (IFR) metric grounded in mutual information theory, and design a Multi-path Byte Refinement Block (MBRB) to strengthen cross-byte dependency modeling via heterogeneous feature propagation. At the system level, we develop a Latent Transformation Engine (LTE) for compact high-dimensional feature representation and a Decoupled Pipeline Compression Architecture (DPCA) to eliminate encoding-decoding latency through pipelined parallelization. Experimental results demonstrate that EDPC achieves comprehensive improvements over state-of-the-art methods, including a 2.7x faster compression speed, and a 3.2% higher compression ratio. These advancements establish EDPC as an efficient solution for real-time processing of large-scale multimedia data in bandwidth-constrained scenarios. Our code is available at https://github.com/Magie0/EDPC.

Figures

Figures reproduced from arXiv: 2507.18969 by the authors.

Figure 1
Figure 1. We propose EDPC, an efficient dual-path parallel [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Comparison of mutual information and Informa [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Overview of EDPC (best viewed in color). The left part shows the Probability Prediction Model, which efficiently [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Comparison of different encoding strategies: (a) [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: GPU memory usage (a) and total model parameters [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Compression ratios of EDPC across different batch [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 8
Figure 8. Figure 8: Comparison of encoding time between EDPC and [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 7
Figure 7. Figure 7: Actual performance comparison of with LTE and [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 40 canonical work pages

  1. [1]

    Kuffo, and Peter A

    Azim Afroozeh, Leonardo X. Kuffo, and Peter A. Boncz. 2023. ALP: Adaptive Lossless floating-Point Compression. Proc. ACM Manag. Data 1, 4 (2023), 230:1– 230:26

    work page 2023

  2. [2]

    Fabrice Bellard. 2019. NNCP: Lossless Data Compression with Neural Networks. https://bellard.org/nncp/

    work page 2019

  3. [3]

    Martin Burtscher and Paruj Ratanaworabhan. 2009. FPC: A High-Speed Com- pressor for Double-Precision Floating-Point Data. IEEE Trans. Computers 58, 1 (2009), 18–31

    work page 2009

  4. [4]

    Miao Cao, Lishun Wang, Huan Wang, Guoqing Wang, and Xin Yuan. 2024. To- wards Real-time Video Compressive Sensing on Mobile Devices. In MM, Jianfei Cai, Mohan S. Kankanhalli, Balakrishnan Prabhakaran, Susanne Boll, Ramanathan Subramanian, Liang Zheng, Vivek K. Singh, Pablo César, Lexing Xie, and Dong Xu (Eds.). ACM, 11080–11088

    work page 2024

  5. [5]

    Minxiao Chen, Haitao Yuan, Nan Jiang, Zhihan Zheng, Sai Wu, Ao Zhou, and Shangguang Wang. 2025. RLOMM: An Efficient and Robust Online Map Matching Framework with Reinforcement Learning. CoRR abs/2502.06825 (2025)

    work page Pith review arXiv 2025

  6. [6]

    Xinyu Chen, Jiannan Tian, Ian Beaver, Cynthia Freeman, Yan Yan, Jianguo Wang, and Dingwen Tao. 2024. FCBench: Cross-Domain Benchmarking of Lossless Compression for Floating-point Data. Proc. VLDB Endow. 17, 6 (2024), 1418–1431

    work page 2024

  7. [7]

    Y. Collet. 2016. Zstd GitHub Repository from Facebook. https://github.com/ facebook/zstd

    work page 2016

  8. [8]

    Grégoire Delétang, Anian Ruoss, Paul-Ambroise Duquenne, Elliot Catt, Tim Genewein, Christopher Mattern, Jordi Grau-Moya, Li Kevin Wenliang, Matthew Aitchison, Laurent Orseau, Marcus Hutter, and Joel Veness. 2024. Language Modeling Is Compression. In ICLR. OpenReview.net

    work page 2024

Show all 47 references
  1. [9]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Im- ageNet: A large-scale hierarchical image database. In CVPR. IEEE Computer Society, 248–255

    2009

  2. [10]

    Sebastian Deorowicz. 1985. Silesia Dataset. http://sun.aei.polsl.pl/sdeor/index. php?page=silesia

    1985

  3. [11]

    Peter Deutsch. 1996. GZIP File Format Specification Version 4.3 . RFC 1952. RFC Editor. 1–12 pages. https://doi.org/10.17487/RFC1952

    1996 doi

  4. [12]

    Jarek Duda. 2009. Asymmetric numeral systems. CoRR abs/0902.0271 (2009). http://arxiv.org/abs/0902.0271

    2009 arXiv

  5. [13]

    Mohit Goyal, Kedar Tatwawadi, Shubham Chandak, and Idoia Ochoa. 2021. DZip: improved general-purpose loss less compression based on novel neural network modeling. In DCC, Ali Bilgin, Michael W. Marcellin, Joan Serra-Sagristà, and James A. Storer (Eds.). IEEE, 153–162

    2021

  6. [14]

    Cynthia Huang, Yuqing Xie, Zhiying Jiang, Jimmy Lin, and Ming Li. 2023. Approx- imating Human-Like Few-shot Learning with GPT-based Compression. CoRR abs/2308.06942 (2023)

    2023 arXiv

  7. [15]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. 2015. Adam: A Method for Stochastic Opti- mization. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings , Yoshua Bengio and Yann LeCun (Eds.)

    2015

  8. [16]

    Jürgen Klüver. 2011. A mathematical theory of communication: Meaning, infor- mation, and topology. Complex. 16, 3 (2011), 10–26

    2011

  9. [17]

    B. Knoll. 2016. Tensorflow-compress. https://github.com/byronknoll/tensorflow- compress

    2016

  10. [18]

    Alexander Kraskov, Harald Stögbauer, and Peter Grassberger. 2004. Estimating mutual information. Physical Review E 69, 6 (June 2004). doi:10.1103/physreve. 69.066138

    2004 doi

  11. [19]

    Ruiyuan Li, Zheng Li, Yi Wu, Chao Chen, and Yu Zheng. 2023. Elf: Erasing-based Lossless Floating-Point Compression. Proc. VLDB Endow. 16, 7 (2023), 1763–1776

    2023

  12. [20]

    Panagiotis Liakos, Katia Papakonstantinopoulou, and Yannis Kotidis. 2022. Chimp: Efficient Lossless Floating Point Compression for Time Series Databases. Proc. VLDB Endow. 15, 11 (2022), 3058–3070

    2022

  13. [21]

    Chunwei Liu, John Paparrizos, and Aaron J. Elmore. 2024. AdaEdge: A Dynamic Compression Selection Framework for Resource Constrained Devices. In ICDE. IEEE, 1506–1519

    2024

  14. [22]

    So, and Quoc V

    Hanxiao Liu, Zihang Dai, David R. So, and Quoc V. Le. 2021. Pay Attention to MLPs. In Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, virtual, Marc’Aurelio Ranzato, Alina ...

    2021

  15. [23]

    Huidong Ma, Sun Hui, Liping Yi, Ding Yanfeng, xiaoguang Liu, and Gang Wang

  16. [24]

    Huidong Ma, Hui Sun, Liping Yi, Xiaoguang Liu, and Gang Wang. 2025. Multi- source Data Lossless Compression via Parallel Expansion Mapping and xLSTM. In ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). 1–5

    2025

  17. [25]

    https://openreview.net/forum?id=BgSrNpCMqO

  18. [26]

    Yu Mao, Yufei Cui, Tei-Wei Kuo, and Chun Jason Xue. 2022. Accelerating General- purpose Lossless Compression via Simple and Scalable Parameterization. In MM, João Magalhães, Alberto Del Bimbo, Shin’ichi Satoh, Nicu Sebe, Xavier Alameda- Pineda, Qin Jin, Vincent Oria, and Laura...

    2022

  19. [27]

    Matt Mahoney. 2006. Large Text Compression Benchmark. http://mattmahoney. net/dc/text.html

    2006

  20. [28]

    Yu Mao, Jingzong Li, Yufei Cui, and Chun Jason Xue. 2023. Faster and Stronger Lossless Compression with Optimized Autoregressive Framework. In DAC. IEEE, 1–6

    2023

  21. [29]

    Yu Mao, Yufei Cui, Tei-Wei Kuo, and Chun Jason Xue. 2022. TRACE: A Fast Transformer-based General-Purpose Lossless Compressor. In WWW, Frédérique Laforest, Raphaël Troncy, Elena Simperl, Deepak Agarwal, Aristides Gionis, Ivan Herman, and Lionel Médini (Eds.). ACM, 1829–1838

    2022

  22. [30]

    Yu Mao, Jun Wang, Nan Guan, and Chun Jason Xue. 2025. WISE: A Framework for Gigapixel Whole-Slide-Image Lossless Compression. CoRR abs/2503.18074 (2025)

    2025 arXiv

  23. [31]

    Yu Mao, Holger Pirk, and Chun Jason Xue. 2025. Lossless Compression of Large Language Model-Generated Text via Next-Token Prediction.CoRR abs/2505.06297 (2025)

    2025 arXiv

  24. [32]

    I. Pavlov. 1999. 7zip: File archiver. http://www.7-zip.org

    1999

  25. [33]

    Willie Neiswanger, Ke Alexander Wang, and Stefano Ermon. 2021. Bayesian Algorithm Execution: Estimating Computable Properties of Black-box Functions Using Mutual Information. In Proceedings of the 38th International Conference on Machine Learning, ICML 2021, 18-24 July 2021, V...

    2021

  26. [34]

    Aécio S. R. Santos, Flip Korn, and Juliana Freire. 2024. Efficiently Estimating Mutual Information Between Attributes Across Tables. In 40th IEEE International Conference on Data Engineering, ICDE 2024, Utrecht, The Netherlands, May 13-16,

    2024

  27. [35]

    Karol J. Piczak. 2015. ESC: Dataset for Environmental Sound Classification. In MM, Xiaofang Zhou, Alan F. Smeaton, Qi Tian, Dick C. A. Bulterman, Heng Tao Shen, Ketan Mayer-Patel, and Shuicheng Yan (Eds.). ACM, 1015–1018

    2015

  28. [36]

    Claude E. Shannon. 1948. A mathematical theory of communication. Bell Syst. Tech. J. 27, 3 (1948), 379–423

    1948

  29. [37]

    Claude Elwood Shannon. 1948. A Mathematical Theory of Communication. The Bell System Technical Journal 27, 3 (1948), 379–423

    1948

  30. [38]

    Gomez, Lukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is All you Need. In NIPS, Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna M. Wallach, Rob Fergus, S. V. N. Vishwanathan, and Rom...

    2017

  31. [39]

    Chandra Shekhara Kaushik Valmeekam, Krishna Narayanan, Dileep Kalathil, Jean-François Chamberland, and Srinivas Shakkottai. 2023. LLMZip: Lossless Text Compression using Large Language Models. CoRR abs/2306.04050 (2023)

    2023 arXiv

  32. [40]

    Runhui Wang and Dong Deng. 2020. DeltaPQ: Lossless Product Quantization Code Compression for High Dimensional Similarity Search. Proc. VLDB Endow. 13, 13 (2020), 3603–3616

    2020

  33. [41]

    Pei Wang, Junqi Wang, Pushpi Paranamana, and Patrick Shafto. 2020. A mathe- matical theory of cooperative communication. In Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, December 6-12, 2020,...

    2020

  34. [42]

    Sa Yan, Nuowen Kan, Chenglin Li, Wenrui Dai, Junni Zou, and Hongkai Xiong

  35. [43]

    Dongzhu Xu, Anfu Zhou, Xinyu Zhang, Guixian Wang, Xi Liu, Congkai An, Yiming Shi, Liang Liu, and Huadong Ma. 2020. Understanding Operational 5G: A First Measurement Study on Its Coverage, Performance and Energy Consumption. In SIGCOMM ’20, Henning Schulzrinne and Vishal Misra ...

    2020

  36. [44]

    Chen Zhang, Qingyuan Xie, Mingyue Wang, Yu Guo, and Xiaohua Jia. 2024. Optimal Compression for Encrypted Key-Value Store in Cloud Systems. IEEE Trans. Computers 73, 3 (2024), 928–941

    2024

  37. [46]

    Yang Yang, Ming Cheng, Yingqiu Ding, and Weiming Zhang. 2023. A Visually Meaningful Image Encryption Scheme Based on Lossless Compression SPIHT Coding. IEEE Trans. Serv. Comput. 16, 4 (2023), 2387–2401

    2023

  38. [2024]

    In MM, Jianfei Cai, Mohan S

    Task-Oriented Multi-Bitstream Optimization for Image Compression and Transmission via Optimal Transport. In MM, Jianfei Cai, Mohan S. Kankan- halli, Balakrishnan Prabhakaran, Susanne Boll, Ramanathan Subramanian, Liang Zheng, Vivek K. Singh, Pablo César, Lexing Xie, and Dong X...

  39. [2025]

    In THE WEB CONFERENCE

    MSDZip: Universal Lossless Compression for Multi-source Data via Stepwise-parallel and Learning-based Prediction. In THE WEB CONFERENCE

Pith tools

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