Pith. sign in

REVIEW 3 major objections 45 references

AI Generalisation Gap In Comorbid Sleep Disorder Staging

T0 review · 3 major / 0 minor · reviewed 2026-07-13 · grok-4.5

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

Pith's one-line read Sleep-staging models trained on healthy EEG fail on stroke patients and attend to uninformative signal regions.

desk verdict We only have the abstract for the sleep-staging paper; the supplied full text is an unrelated AML GNN paper, so the central claims cannot be checked. read the letter →

arxiv 2603.23582 v2 pith:LDPPIXEH submitted 2026-03-24 cs.LG cs.AI

Saswata Bose , Suvadeep Maiti , Shivam Kumar Sharma , Mythirayee S , Tapabrata Chakraborti , Srijitesh Rajendran , Raju S. Bapi This is my paper
classification cs.LGcs.AI
keywords sleepstagingEEGdomaingeneralizationischemicstrokeGrad-CAMiSLEEPSclinicalvalidationdeeplearning
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

Automated sleep staging from single-channel EEG works well on healthy people, but this paper shows it does not transfer to ischemic stroke patients whose sleep is disrupted. The authors release a new clinically annotated stroke EEG dataset, iSLEEPS, and test a SE-ResNet plus bidirectional LSTM model. Cross-domain performance collapses. Grad-CAM maps, checked with clinical experts, show the network focusing on EEG segments that are not physiologically useful for staging in patients. Statistical comparison confirms that sleep architecture differs substantially between healthy and stroke cohorts. The practical claim is that generic healthy-trained models are not safe to deploy on this clinical population without subject-aware or disease-specific training and clinical validation.

What carries the argument

Grad-CAM attention maps on a SE-ResNet plus bidirectional LSTM single-channel EEG stager, interpreted with clinical expert feedback on the new iSLEEPS ischemic-stroke dataset, used to show that the network attends to the wrong signal regions when sleep architecture is pathologically altered.

What would settle it

Train and evaluate the same architecture on matched healthy and stroke cohorts with held-out clinical labels; if cross-domain F1 remains high and expert-reviewed Grad-CAM maps consistently highlight known stage-discriminative EEG features in patients, the claimed generalization gap and attention failure would be refuted.

Watch

Extended reading notes

Core claim

Deep learning sleep-staging models that perform well on healthy subjects generalize poorly to ischemic stroke patients; attention visualizations indicate the model relies on physiologically uninformative EEG regions in patient data, so subject-aware or disease-specific models with clinical validation are required before deployment.

Load-bearing premise

That Grad-CAM heatmaps plus expert review of those maps correctly reveal what the network uses for staging and that those regions are truly physiologically uninformative in stroke EEG.

Editorial extensions

If this is right

  • Healthy-trained automated sleep stagers should not be deployed on stroke or similarly disrupted-sleep populations without retraining or adaptation.
  • Public release of iSLEEPS enables disease-specific benchmarking that current healthy-only datasets cannot provide.
  • Explainability checks (attention maps plus clinician review) become a required gate before clinical use of EEG staging models.
  • Subject-aware or cohort-specific models are needed when sleep architecture statistics differ significantly between groups.

Reading between the lines

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

  • Similar generalization failures are likely for other neurological conditions that fragment sleep architecture, not only ischemic stroke.
  • If attention maps systematically land on uninformative regions, post-hoc XAI may also expose failure modes in other clinical EEG tasks that currently report only healthy-domain accuracy.
  • Regulatory or hospital procurement pathways for automated PSG scoring will need explicit cross-population validation criteria rather than healthy-only performance numbers.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 0 minor

Summary. The abstract claims that SE-ResNet+BiLSTM models for single-channel EEG sleep staging, while effective on healthy subjects, exhibit poor cross-domain generalization to ischemic stroke patients with disrupted sleep architecture; Grad-CAM attention maps (supported by clinical expert feedback) indicate focus on physiologically uninformative EEG regions in patient data, and statistical analyses confirm cohort differences, motivating subject-aware or disease-specific models plus the new iSLEEPS dataset. The supplied full manuscript text, however, is an unrelated work (LineMVGNN for anti-money laundering on transaction digraphs) whose sections, equations, tables, and experiments do not address EEG, sleep staging, Grad-CAM, or iSLEEPS.

Significance. If the abstract claims hold under proper verification, the work would usefully document a clinically relevant domain-shift failure mode for automated sleep staging and release a public stroke EEG resource, strengthening the case for disease-specific validation before deployment. The mismatch between abstract and full text renders those claims currently unassessable; no machine-checked proofs, code artifacts, or parameter-free derivations for the sleep-staging results are present in the provided materials.

major comments (3)
  1. The full manuscript text supplied under paper_id 2603.23582 is the LineMVGNN AML paper (arXiv 2603.23584), not the sleep-staging work described by the title and abstract. Consequently no methods, results tables, Grad-CAM layer choices, expert-protocol details, or statistical tests for the claimed generalization gap can be inspected or reproduced.
  2. Abstract claim that Grad-CAM plus expert feedback shows focus on 'physiologically uninformative' regions cannot be evaluated: the mismatched manuscript contains no Grad-CAM setup, faithfulness controls, baseline comparisons, blinding protocol, or inter-rater metrics, leaving the central interpretability argument untestable.
  3. Abstract assertion of 'poor' cross-domain performance and 'significant' sleep-architecture differences lacks any accompanying metrics, sample sizes, baselines, or p-values in the provided materials; without the correct manuscript these load-bearing quantitative claims remain unsupported.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation chain; claims are empirical performance drops, visualizations, and cohort statistics with no reduction of outputs to inputs by construction.

full rationale

The paper (as given by its abstract and the mismatched full manuscript text supplied in the cache) advances no first-principles derivation, uniqueness theorem, or fitted-parameter-as-prediction. Its load-bearing claims are empirical: cross-domain F1 degradation between healthy and ischemic-stroke EEG, Grad-CAM attention maps judged uninformative by clinical experts, and statistical differences in sleep architecture. None of these reduce by the paper's own equations or definitions to their inputs. The supplied full text is an unrelated LineMVGNN AML paper whose own experimental results (F1 tables, ablations on views/parameter-sharing/learning-rate/embedding-size) are likewise ordinary empirical comparisons against external baselines and do not exhibit self-definitional loops, fitted inputs renamed as predictions, load-bearing self-citation uniqueness claims, or ansatz smuggling. No circular steps can therefore be quoted. Score 0 is the correct default for a self-contained empirical ML paper.

Assumptions & free parameters 0 free parameters · 3 assumptions · 1 invented entities

Abstract-only review of the sleep-staging paper. Load-bearing premises are domain assumptions about PSG/EEG staging, transfer of healthy-trained deep models, and interpretability of Grad-CAM plus expert review. No free parameters or invented physical entities are stated in the abstract. The cached full text belongs to a different paper and was not used as evidence for this claim.

assumptions (3)
  • domain assumption Automated single-channel EEG sleep staging trained on healthy subjects is expected to transfer to clinical populations unless domain shift is shown.
    Abstract contrasts healthy-subject success with clinical failure; this transfer expectation is the foil for the main claim.
  • domain assumption Grad-CAM heatmaps plus clinical expert feedback validly indicate whether the model attends to physiologically informative EEG regions.
    Central interpretability claim in the abstract depends on this without methods detail available here.
  • domain assumption Ischemic stroke cohorts have systematically different sleep architecture from healthy cohorts in ways that break standard staging models.
    Abstract states statistical and computational analyses confirm significant architecture differences; treated as an empirical premise supporting the need for disease-specific models.
invented entities (1)
  • iSLEEPS dataset
    purpose: Provide clinically annotated ischemic stroke EEG for sleep staging and generalization studies.
    New resource announced in the abstract; independent public release is promised but not verified from the provided materials.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AI Generalisation Gap In Comorbid Sleep Disorder Staging." pith.science (2026). https://pith.science/paper/LDPPIXEH

@misc{pith2026260323582,
  author       = {Pith},
  title        = {Pith review of: AI Generalisation Gap In Comorbid Sleep Disorder Staging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LDPPIXEH}},
  note         = {Machine review of arXiv:2603.23582}
}
read the original abstract

Accurate sleep staging is essential for diagnosing OSA and hypopnea in stroke patients. Although PSG is reliable, it is costly, labor-intensive, and manually scored. While deep learning enables automated EEG-based sleep staging in healthy subjects, our analysis shows poor generalization to clinical populations with disrupted sleep. Using Grad-CAM interpretations, we systematically demonstrate this limitation. We introduce iSLEEPS, a newly clinically annotated ischemic stroke dataset (to be publicly released), and evaluate a SE-ResNet plus bidirectional LSTM model for single-channel EEG sleep staging. As expected, cross-domain performance between healthy and diseased subjects is poor. Attention visualizations, supported by clinical expert feedback, show the model focuses on physiologically uninformative EEG regions in patient data. Statistical and computational analyses further confirm significant sleep architecture differences between healthy and ischemic stroke cohorts, highlighting the need for subject-aware or disease-specific models with clinical validation before deployment. A summary of the paper and the code is available at https://himalayansaswatabose.github.io/iSLEEPS_Explainability.github.io/

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

45 extracted references · 8 linked inside Pith

  1. [1]

    Machine learning techniques for anti-money laundering (AML) solutions in suspicious transaction detection: A review

    Chen, Z.; Khoa, L.D.; Teoh, E.N.; Nazir, A.; Karuppiah, E.K.; Lam, K.S. Machine learning techniques for anti-money laundering (AML) solutions in suspicious transaction detection: A review. Knowl. Inf. Syst. 2018, 57, 245–285. [CrossRef]

    2018

  2. [2]

    Financial Fraud: A Review of Anomaly Detection Techniques and Recent Advances

    Hilal, W.; Gadsden, S.A.; Yawney, J. Financial Fraud: A Review of Anomaly Detection Techniques and Recent Advances. Expert Syst. Appl. 2022, 193, 116429. [CrossRef]

    2022

  3. [3]

    Financial fraud detection using graph neural networks: A systematic review

    Motie, S.; Raahemi, B. Financial fraud detection using graph neural networks: A systematic review. Expert Syst. Appl. 2024, 240, 122156. [CrossRef]

    2024

  4. [4]

    HoloNets: Spectral Convolutions do extend to Directed Graphs

    Koke, C.; Cremers, D. HoloNets: Spectral Convolutions do extend to Directed Graphs. In Proceedings of the Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, 7–11 May 2024

    2024

  5. [5]

    SigMaNet: One laplacian to rule them all

    Fiorini, S.; Coniglio, S.; Ciavotta, M.; Messina, E. SigMaNet: One laplacian to rule them all. In Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence and Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence, AAAI’23/IAAI’23/EAAI’23...

    2023

  6. [6]

    MagNet: A Neural Network for Directed Graphs

    Zhang, X.; He, Y.; Brugnone, N.; Perlmutter, M.; Hirn, M.J. MagNet: A Neural Network for Directed Graphs. In Proceedings of the NeurIPS, Online, 6–14 December 2021; Ranzato, M., Beygelzimer, A., Dauphin, Y.N., Liang, P ., Vaughan, J.W., Eds.; Curran Associates, Inc.: Red Hook, NY, USA, 2021; pp. 27003–27015

    2021

  7. [7]

    Digraph Inception Convolutional Networks

    Tong, Z.; Liang, Y.; Sun, C.; Li, X.; Rosenblum, D.S.; Lim, A. Digraph Inception Convolutional Networks. In Proceedings of the Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, Virtual, 6–12 December 2020; Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H., Eds.;...

    2020

  8. [8]

    Directed Graph Convolutional Network

    Tong, Z.; Liang, Y.; Sun, C.; Rosenblum, D.S.; Lim, A. Directed Graph Convolutional Network. arXiv 2020, arXiv:2004.13970

    arXiv 2020

Show all 45 references
  1. [9]

    Spectral-based Graph Convolutional Network for Directed Graphs

    Ma, Y.; Hao, J.; Yang, Y.; Li, H.; Jin, J.; Chen, G. Spectral-based Graph Convolutional Network for Directed Graphs. arXiv 2019, arXiv:1907.08990

    2019 arXiv

  2. [10]

    MotifNet: A motif-based Graph Convolutional Network for directed graphs

    Monti, F.; Otness, K.; Bronstein, M.M. MotifNet: A motif-based Graph Convolutional Network for directed graphs. arXiv 2018, arXiv:1802.01572

    2018 arXiv

  3. [11]

    Edge Directionality Improves Learning on Heterophilic Graphs

    Rossi, E.; Charpentier, B.; Giovanni, F.D.; Frasca, F.; Günnemann, S.; Bronstein, M.M. Edge Directionality Improves Learning on Heterophilic Graphs. In Proceedings of the LoG, PMLR, Virtual, 27–30 November 2023; Villar, S., Chamberlain, B., Eds.; PMLR: London, UK, 2023; Volume...

    2023

  4. [12]

    Directed Acyclic Graph Neural Networks

    Thost, V .; Chen, J. Directed Acyclic Graph Neural Networks. In Proceedings of the 9th International Conference on Learning Representations, ICLR 2021, Virtual, Austria, 3–7 May 2021

    2021

  5. [13]

    Gated Graph Sequence Neural Networks

    Li, Y.; Tarlow, D.; Brockschmidt, M.; Zemel, R.S. Gated Graph Sequence Neural Networks. In Proceedings of the 4th International Conference on Learning Representations, ICLR 2016, San Juan, PR, USA, 2–4 May 2016

    2016

  6. [14]

    Neural Message Passing for Quantum Chemistry

    Gilmer, J.; Schoenholz, S.S.; Riley, P .F.; Vinyals, O.; Dahl, G.E. Neural Message Passing for Quantum Chemistry. In Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6–11 August 2017; Precup, D., Teh, Y.W., Eds.; PMLR: Lo...

    2017

  7. [15]

    The Graph Neural Network Model

    Scarselli, F.; Gori, M.; Tsoi, A.C.; Hagenbuchner, M.; Monfardini, G. The Graph Neural Network Model. IEEE T rans. Neural Netw. 2009, 20, 61–80. [CrossRef] [PubMed]

    2009

  8. [16]

    Screen the Account for Suspicious Indicators: Recognition of a Suspicious Activity Indicator or Indicators

    Joint Financial Intelligence Unit. Screen the Account for Suspicious Indicators: Recognition of a Suspicious Activity Indicator or Indicators. 2024. Available online: https://www.jfiu.gov.hk/en/str_screen.html (accessed on 10 August 2024)

    2024

  9. [17]

    Fast incremental and personalized PageRank

    Bahmani, B.; Chowdhury, A.; Goel, A. Fast incremental and personalized PageRank. Proc. VLDB Endow. 2010, 4, 173–184. [CrossRef]

    2010

  10. [18]

    Graph Attention Networks

    Velickovic, P .; Cucurull, G.; Casanova, A.; Romero, A.; Liò, P .; Bengio, Y. Graph Attention Networks. In Proceedings of the 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, 30 April–3 May 2018

    2018

  11. [19]

    Supervised Community Detection with Line Graph Neural Networks

    Chen, Z.; Li, L.; Bruna, J. Supervised Community Detection with Line Graph Neural Networks. In Proceedings of the 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, 6–9 May 2019

    2019

  12. [20]

    Line Graph Neural Networks for Link Weight Prediction

    Liang, J.; Pu, C. Line Graph Neural Networks for Link Weight Prediction. arXiv 2023, arXiv:2309.15728

    2023 arXiv

  13. [21]

    LeL-GNN: Learnable Edge Sampling and Line Based Graph Neural Network for Link Prediction

    Morshed, M.G.; Sultana, T.; Lee, Y.K. LeL-GNN: Learnable Edge Sampling and Line Based Graph Neural Network for Link Prediction. IEEE Access 2023, 11, 56083–56097. [CrossRef]

    2023

  14. [22]

    Learning Graph Representations Through Learning and Propagating Edge Features

    Zhang, H.; Xia, J.; Zhang, G.; Xu, M. Learning Graph Representations Through Learning and Propagating Edge Features. IEEE T rans. Neural Netw. Learn. Syst. 2024, 35, 8429–8440. [CrossRef] [PubMed]

    2024

  15. [23]

    CensNet: Convolution with Edge-Node Switching in Graph Neural Networks

    Jiang, X.; Ji, P .; Li, S. CensNet: Convolution with Edge-Node Switching in Graph Neural Networks. In Proceedings of the IJCAI, Macao, 10–16 August 2019; Kraus, S., Ed.; AAAI Press: Washington, DC, USA, 2019; pp. 2656–2662

    2019

  16. [24]

    Line graph contrastive learning for node classification

    Li, M.; Meng, L.; Ye, Z.; Xiao, Y.; Cao, S.; Zhao, H. Line graph contrastive learning for node classification. J. King Saud Univ. Comput. Inf. Sci. 2024, 36, 102011. [CrossRef]

    2024

  17. [25]

    Semi-Supervised Classification with Graph Convolutional Networks

    Kipf, T.N.; Welling, M. Semi-Supervised Classification with Graph Convolutional Networks. In Proceedings of the 5th Interna- tional Conference on Learning Representations, ICLR 2017, Toulon, France, 24–26 April 2017

    2017

  18. [26]

    Inductive Representation Learning on Large Graphs

    Hamilton, W.L.; Ying, Z.; Leskovec, J. Inductive Representation Learning on Large Graphs. In Proceedings of the NIPS, Long Beach, CA, USA, 4–9 December 2017; Guyon, I., von Luxburg, U., Bengio, S., Wallach, H.M., Fergus, R., Vishwanathan, S.V .N., Garnett, R., Eds.; Curran Ass...

    2017

  19. [27]

    How Powerful are Graph Neural Networks? In Proceedings of the 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, 6–9 May 2019

    Xu, K.; Hu, W.; Leskovec, J.; Jegelka, S. How Powerful are Graph Neural Networks? In Proceedings of the 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, 6–9 May 2019

    2019

  20. [28]

    Beyond Homophily: Robust Graph Anomaly Detection via Neural Sparsification

    Gong, Z.; Wang, G.; Sun, Y.; Liu, Q.; Ning, Y.; Xiong, H.; Peng, J. Beyond Homophily: Robust Graph Anomaly Detection via Neural Sparsification. In Proceedings of the IJCAI, Macao, 19–25 August 2023; pp. 2104–2113

    2023

  21. [29]

    Adaptive Universal Generalized PageRank Graph Neural Network

    Chien, E.; Peng, J.; Li, P .; Milenkovic, O. Adaptive Universal Generalized PageRank Graph Neural Network. In Proceedings of the 9th International Conference on Learning Representations, ICLR 2021, Virtual, Austria, 3–7 May 2021

    2021

  22. [30]

    LaundroGraph: Self-Supervised Graph Representation Learning for Anti-Money Laundering

    Cardoso, M.; Saleiro, P .; Bizarro, P . LaundroGraph: Self-Supervised Graph Representation Learning for Anti-Money Laundering. In Proceedings of the Third ACM International Conference on AI in Finance, ICAIF ’22, New York, NY, USA, 2–4 November 2022; pp. 130–138. [CrossRef]

    2022

  23. [31]

    Cross-Stitch Networks for Multi-task Learning

    Misra, I.; Shrivastava, A.; Gupta, A.; Hebert, M. Cross-Stitch Networks for Multi-task Learning. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV , USA, 27–30 June 2016; IEEE Computer Society: Piscataway, NJ, USA, ...

    2016

  24. [32]

    Deep Residual Learning for Image Recognition

    He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV , USA, 27–30 June 2016; IEEE Computer Society: Piscataway, NJ, USA, 2016; pp. 770–7...

    2016

  25. [33]

    Ethereum Fraud Detection with Heterogeneous Graph Neural Networks

    Kanezashi, H.; Suzumura, T.; Liu, X.; Hirofuchi, T. Ethereum Fraud Detection with Heterogeneous Graph Neural Networks. arXiv 2022, arXiv:2203.12363

    2022 arXiv

  26. [34]

    Who Are the Phishers? Phishing Scam Detection on Ethereum via Network Embedding

    Wu, J.; Yuan, Q.; Lin, D.; You, W.; Chen, W.; Chen, C.; Zheng, Z. Who Are the Phishers? Phishing Scam Detection on Ethereum via Network Embedding. IEEE T rans. Syst. Man Cybern. Syst. 2022, 52, 1156–1166. [CrossRef]

    2022

  27. [35]

    Anomaly Detection in Networks with Application to Financial Transaction Networks

    Elliott, A.; Cucuringu, M.; Luaces, M.M.; Reidy, P .; Reinert, G. Anomaly Detection in Networks with Application to Financial Transaction Networks. arXiv 2019, arXiv:1901.00402

    2019 arXiv

  28. [36]

    Principal Neighbourhood Aggregation for Graph Nets

    Corso, G.; Cavalleri, L.; Beaini, D.; Liò, P .; Velickovic, P . Principal Neighbourhood Aggregation for Graph Nets. In Proceedings of the Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, Virtua...

    2020

  29. [37]

    Rossmann-toolbox: A deep learning-based protocol for the prediction and design of cofactor specificity in Rossmann fold proteins

    Kami ´ nski, K.; Ludwiczak, J.; Jasi ´ nski, M.; Bukala, A.; Madaj, R.; Szczepaniak, K.; Dunin-Horkawicz, S. Rossmann-toolbox: A deep learning-based protocol for the prediction and design of cofactor specificity in Rossmann fold proteins. Brief. Bioinform. 2021, 23, bbab371. [...

    2021

  30. [38]

    Representation Learning on Graphs with Jumping Knowledge Networks

    Xu, K.; Li, C.; Tian, Y.; Sonobe, T.; Kawarabayashi, K.; Jegelka, S. Representation Learning on Graphs with Jumping Knowledge Networks. In Proceedings of the 35th International Conference on Machine Learning, ICML 2018, PMLR, Stockholm, Sweden, 10–15 July 2018; Dy, J.G., Kraus...

    2018

  31. [39]

    Strategies for Pre-training Graph Neural Net- works

    Hu, W.; Liu, B.; Gomes, J.; Zitnik, M.; Liang, P .; Pande, V .S.; Leskovec, J. Strategies for Pre-training Graph Neural Net- works. In Proceedings of the 8th International Conference on Learning Representations, ICLR 2020, Addis Ababa, Ethiopia, 26–30 April 2020

    2020

  32. [40]

    PyTorch Geometric Signed Directed: A Software Package on Graph Neural Networks for Signed and Directed Graphs

    He, Y.; Zhang, X.; Huang, J.; Rozemberczki, B.; Cucuringu, M.; Reinert, G. PyTorch Geometric Signed Directed: A Software Package on Graph Neural Networks for Signed and Directed Graphs. In Proceedings of the Second Learning on Graphs Conference (LoG 2023), PMLR 231, Virtual, N...

    2023

  33. [41]

    PyTorch: An Imperative Style, High-Performance Deep Learning Library

    Paszke, A.; Gross, S.; Massa, F.; Lerer, A.; Bradbury, J.; Chanan, G.; Killeen, T.; Lin, Z.; Gimelshein, N.; Antiga, L.; et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library. In Proceedings of the Advances in Neural Information Processing Systems 32: Ann...

    2019

  34. [42]

    Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks

    Wang, M.; Zheng, D.; Ye, Z.; Gan, Q.; Li, M.; Song, X.; Zhou, J.; Ma, C.; Yu, L.; Gai, Y.; et al. Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks. arXiv 2019, arXiv:1909.01315

    2019 arXiv

  35. [43]

    Fast Graph Representation Learning with PyTorch Geometric

    Fey, M.; Lenssen, J.E. Fast Graph Representation Learning with PyTorch Geometric. arXiv 2019, arXiv:1903.02428

    2019 arXiv

  36. [44]

    Adam: A Method for Stochastic Optimization

    Kingma, D.P .; Ba, J. Adam: A Method for Stochastic Optimization. In Proceedings of the 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, 7–9 May 2015

    2015

  37. [45]

    SGDR: Stochastic Gradient Descent with Warm Restarts

    Loshchilov, I.; Hutter, F. SGDR: Stochastic Gradient Descent with Warm Restarts. In Proceedings of the 5th International Conference on Learning Representations, ICLR 2017, Toulon, France, 24–26 April 2017. Disclaimer/Publisher’s Note: The statements, opinions and data containe...

    2017

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.