Pith. sign in

REVIEW 4 major objections 5 minor 34 references

Leveraging A New GAN-based Transformer with ECDH Crypto-system for Enhancing Energy Theft Detection in Smart Grid

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read A GAN-Transformer split-learning framework with ECDH-based additive masking detects energy theft in smart grids with state-of-the-art accuracy while blocking AI-enabled reconstruction of consumers' raw energy data.

desk verdict Interesting protocol-level masking idea for split learning, but the security proof is absent and the mask-reuse issue is potentially fatal; the detection half is fine. read the letter →

arxiv 2411.18023 v1 pith:4C5PYHAK submitted 2024-11-27 cs.CR

classification cs.CR
keywords energytheftdetectionsplitlearningGAN-Transformerprivacy-preservingmachinesmartgridsecurityadditivemaskingECDHkeyexchangereconstructionattack
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to establish that split learning can be made safe against AI-enabled reconstruction attacks without resorting to heavy encryption: a client and server establish a shared key via elliptic-curve Diffie-Hellman, derive a pseudorandom mask, and add it to the intermediate tensor before it crosses the public channel. On the detection side, a generative-adversarial-network (GAN) based Transformer, with the generator split across client and server and the discriminator on the server, achieves AUC of 0.690, 0.817, and 0.970 for 10%, 20%, and 30% theft, respectively, outperforming the autoencoder, adversarial-anomaly, LSTM, and plain-transformer baselines. With the mask, an adversarial decoder's reconstruction $R^2$ drops from about 0.98 to below 0.01 across the five reported samples. If correct, this gives utilities a privacy-preserving theft detector that can run on modest edge hardware and avoids the communication cost of homomorphic encryption.

What carries the argument

The load-bearing object is the masking protocol: elliptic-curve Diffie-Hellman key exchange to produce a shared secret, a key derivation function to stretch it into a mask key, and a pseudorandom tensor generator seeded by that key to produce a mask that is added to the intermediate tensor, and to the gradient on the return path, so that only masked values travel over the public channel. The GAN-Transformer supplies the detection model: a transformer encoder split so the client computes the early layers while the server computes the decoder and a transformer-based discriminator, with an adversarial loss that forces the generator to produce realistic reconstructions. The masking is what the security claim rests on, while the model's AUC gains are what make the framework useful rather than merely private.

What would settle it

Train an adversarial decoder on a large collection of masked intermediate tensors from many sessions with the model architecture known; if the reconstruction $R^2$ rises well above 0.01 (for example above 0.5), the mask fails to defeat AI-enabled reconstruction. A second test: check whether the pseudorandom mask is deterministic per session, so an adversary who obtains one mask can subtract it from every message generated under the same seed.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that a protocol-level additive mask, generated from a shared key established by elliptic-curve Diffie-Hellman and stretched by a key derivation function, can hide the intermediate tensor of a split neural network from an AI-enabled eavesdropper while preserving detection accuracy. The client and server each generate the same pseudorandom mask, the client applies it to the intermediate tensor before transmission, and the server subtracts it after reception; the same treatment is applied to the backward-pass gradient. The paper reports that the GAN-Transformer detector matches or exceeds state-of-the-art AUC on the smart-grid dataset and that masked reconstruction $R^2$ values are near zero, and it claims a formal proof that the protocol achieves mutual authentication, key indistinguishability, unlinkability, and AI security against a network eavesdropper with full channel access. The full proof is deferred to an appendix that does not appear in this preprint.

Load-bearing premise

The whole privacy guarantee rests on the assumption that a fresh additive pseudo-random mask, derived from a shared ECDH key and a key derivation function, makes the intermediate tensor useless to an AI-enabled adversary even when the adversary can collect many masked tensors and knows the model architecture, and that assumption is not proven in the preprint; the proof is deferred to a missing appendix and the empirical check uses only five samples.

Editorial extensions

If this is right

  • If the framework is right, utilities can deploy split learning with client-side edge hardware and a server-side accelerator, detecting 30% theft at AUC 0.970.
  • An eavesdropper on the public channel who captures the masked intermediate tensor cannot reconstruct the raw consumption trace, with reported $R^2$ values below 0.01.
  • The masking scheme is faster in the reported benchmarks than AES, Simon, Speck, and homomorphic encryption, so it can be applied during training without stalling the split-learning loop.
  • The framework claims formal security for mutual authentication, key indistinguishability, unlinkability, and AI security against a full-channel eavesdropper, provided the deferred proof holds.
  • The split placement of the GAN keeps the heavy computation on the server while the client only runs the early transformer layers, which is what makes the approach viable on constrained edge devices.

Reading between the lines

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

  • Beyond the paper, a decisive stress test would be to train an adversarial decoder on thousands of masked tensors across many sessions and measure reconstruction $R^2$; the five-sample table suggests but does not bound the defence.
  • Beyond the paper, the protocol's privacy guarantee would be strengthened by refreshing the mask seed per message, since reuse of a shared mask across sessions would let an adversary subtract aligned tensors and recover structure.
  • Beyond the paper, the masking protects the intermediate tensor, but the encrypted target value and message signatures still leak session timing and volume; a traffic-analysis extension could test whether those side channels reveal theft behaviour.
  • Beyond the paper, membership-inference resistance is not covered by the stated AI-security game, so a natural extension is to check whether an adversary can tell which households contributed training data from the masked traffic.
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

4 major / 5 minor

Summary. The paper proposes a split-learning framework for energy-theft detection in smart grids, in which a GAN-based Transformer is split between a client and a server. To protect the intermediate tensor sent over the public channel, the client adds a mask derived from an ECDH-established shared key via a KDF and a pseudorandom generator. The authors claim protocol-level security against AI-enabled reconstruction attacks, improved AUC over several baselines on the Pecan Street dataset, and computational advantages over full encryption. The manuscript reports a formal security theorem, a small privacy evaluation with five samples, and complexity and energy-consumption benchmarks on a Raspberry Pi/Jetson testbed.

Significance. If rigorously established, the protocol-level masking approach would be a useful contribution: it could provide an efficient alternative to end-to-end encryption for protecting split-learning intermediates against reconstruction attacks, with a realistic embedded testbed and a novel GAN-Transformer architecture. The paper also provides useful comparisons among GAN-Transformer, LSTM, and Transformer detectors and releases code. However, the central privacy guarantee is not established in the submitted manuscript: the formal security game is underspecified, the proof is deferred to an appendix that is not present, the protocol does not state a mask-freshness condition, and the empirical privacy evaluation uses only five samples with no error bars or adversary details. The contribution is therefore conditional on substantial revisions.

major comments (4)
  1. [Section V, Definition 4, Theorem 5] Definition 4 defines AdvAI only as |Pr(b'=b)-1/2|, without specifying the adversary experiment: there is no challenger, no oracle, no description of the challenge bit b, and no event that links the adversary's reconstruction of energy-consumption data to its guess of b. As written, the statement that the advantage is negligible is not a well-defined claim. The proof of Theorem 5 is a placeholder: Game 5.1 bounds the full advantage by 'ANY(AdvMA, AdvKIND, AdvUnlink, AdvAI)' and Game 5.2 sets AdvG5.1=0 'based on Theorem 1 to Theorem 4', none of which are stated in the manuscript. The full proof is said to reside in an appendix that does not appear in the submission. The central claim of protocol-level AI security is therefore unverified.
  2. [Section IV-D, Figure 3] The protocol never specifies how the mask is refreshed across training samples. Figure 3 and the protocol steps show Φ.Init(seed=kMask) and Mask1=Φ.Gen() once per session, with no explicit per-tensor advancement or re-initialization of the pseudorandom generator. If the same Mask1 is reused for multiple samples, then an eavesdropper observing masked tensors m1_i = T_i + Mask1 and m1_j = T_j + Mask1 obtains m1_i - m1_j = T_i - T_j, which removes the mask and reveals differences of intermediate tensors. This mask-reuse condition is not analyzed anywhere in the paper. The manuscript must either specify a fresh-mask mechanism for every transmitted tensor, or analyze the protocol under the stated reuse behavior.
  3. [Section VI-E, Table III] The privacy evaluation reports R2 values for only five samples, with no error bars, no architecture or training details for the adversarial decoder, no statement of how many masked tensors the adversary observes, and no adaptive-query scenario. Five point estimates cannot support the claim that the masking 'destroys the distribution' of the inter-data against AI-enabled adversaries. The experiment should report a distribution over many runs, vary the adversary's access to masked tensors, and describe the decoder and training procedure used for the attack.
  4. [Section VI-C, Table II] The AUC comparison in Table II has no error bars, no number of independent runs, and no statistical test. The claim in Section VI-C that the proposed model 'demonstrated at least a 5% higher detection rate' is not supported by the table at the 0.3 level, where the difference to the Transformer baseline is only 0.003 (0.970 vs. 0.967). Without variance estimates or significance testing, the comparative performance claim is not assessable.
minor comments (5)
  1. [Section IV-D, Step 2] The phrase 'Defile-Hellman results' should be 'Diffie-Hellman results'.
  2. [Definitions 1-4] The term 'cleanness predicate' appears in all security definitions but is never defined, so the reader cannot verify the formal security statements.
  3. [Figure 3 and Algorithms 1-2] The message labels MCS1/MSC1 are used inconsistently in the figure and the protocol steps, and Algorithm 2 iterates 'for each (epoch)' during testing, which should be over test samples rather than epochs.
  4. [Section IV-C] The exact split point of the Transformer encoder between client and server (which layers reside on each side) is not specified, which hampers reproducibility of the split-learning framework.
  5. [Section VI-A] The code link is provided as a tinyurl; a persistent repository identifier or DOI would improve reproducibility.

Circularity Check

1 steps flagged · score 4.0 of 10

The formal AI-security proof is a self-referential placeholder: Definition 4 defines no AI game, and Theorem 5's proof sets the full advantage to zero via unstated Theorems 1-4, so the flagship privacy claim is asserted rather than derived.

  1. other [Section V (Security Frameworks, Definition 4; Theorem 5 proof), with the opening note that 'The full proof of security is provided in the Appendix' (no appendix present in arXiv:2411.18023v1).]
    "Definition 4 (AI Security): Let Π be a deep learning model. ... we define the advantage of A in the AI game to be: AdvAI A,Π(λ) = |P r(ˆb = b) − 1/2|. We say that Π is AI-secure if for all PPT A, AdvAI A,Π(λ) is a negligible value in a parameter λ. ... Game 5.2: In this game, the advantage of breaking any of the security properties is negligible based on Theorem 1 to Theorem 4. Thus, the advantage of A winning the full system security game is negligible: AdvG5.1 = 0."

    The formal derivation of the paper's flagship claim—full security including AI security—is not a derivation. The AI-security 'game' in Definition 4 never specifies the challenger, the oracle, the challenge bit b, or the event that connects reconstruction of energy data to b, so AdvAI is an undefined expression. The proof sketch for Theorem 5 then reduces full security to the union of the four component advantages and sets the result to zero 'based on Theorem 1 to Theorem 4', none of which are stated or proved in the manuscript. As printed, the conclusion is exactly a restatement of the asserted component securities: the proof assumes the very properties it claims to establish, and the deferred appendix that could supply independent support is absent.

full rationale

The actual protocol mechanism is not circular: the masking step m1 = TMid + Mask1 with Mask1 generated from an ECDH-derived KDF seed is a standard additive one-time-pad-style construction whose security would rest on the PRG/KDF/ECDH assumptions, not on the paper's conclusion. The empirical AUC comparison on the Pecan Street dataset is an external benchmark and is also independent. However, the paper's central and novel formal claim—protocol-level security against an AI-enabled reconstruction adversary—is supported only by a proof sketch that is circular in structure: Definition 4 defines an 'AI game' without a game, and Theorem 5's proof sets the full-system advantage to zero by referencing Theorem 1 to Theorem 4, which are not in the submission and whose full proof is deferred to a nonexistent appendix. As printed, the theorem reduces to the assertion that the four security properties hold; no reduction or game-hopping argument is actually exhibited. Additionally, the protocol text shows Phi.Init(seed=kMask) and Mask1=Phi.Gen() only once and does not specify per-sample mask advancement, so the freshness condition needed for the one-time-pad argument is not stated; this is an operational gap but not itself circularity. Score 4 reflects the self-referential proof placeholder for the flagship privacy claim, while acknowledging the underlying masking construction is not defined in terms of the target claim.

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

No fundamentally new entities are introduced. The protocol uses standard cryptographic primitives (ECDH, KDF, PRG) and a standard split-learning setup. The main unstated assumptions are the security of the masking against multi-sample adversaries and the unspecified hyperparameters used in the model.

free parameters (3)
  • anomaly threshold = not specified
    Algorithm 2 uses 'threshold' to decide if an error indicates anomaly; the value is not disclosed and must be tuned on data.
  • loss weights λ_rec and λ_adv = not specified
    Algorithm 1 combines reconstruction and adversarial losses with weights λrec and λadv; values are not reported.
  • model hyperparameters (layers, heads, embedding size, learning rate) = not specified
    The transformer architecture and training hyperparameters are not described, making the model a black box.
assumptions (3)
  • standard math Elliptic curve discrete logarithm problem is hard and ECDH is secure
    The protocol relies on ECDH from [21]; this is a standard cryptographic assumption.
  • domain assumption The pseudorandom generator seeded by k_mask produces uniform and independent masks
    The protocol assumes Φ.Gen() output is indistinguishable from random; standard for PRGs but not explicitly discussed.
  • ad hoc to paper An AI-enabled adversary cannot exploit multiple masked tensors to separate mask from signal
    The security against A3 is not formally proven; the empirical test uses five samples, so this is a key unverified assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Leveraging A New GAN-based Transformer with ECDH Crypto-system for Enhancing Energy Theft Detection in Smart Grid." pith.science (2026). https://pith.science/paper/4C5PYHAK

@misc{pith2026241118023,
  author       = {Pith},
  title        = {Pith review of: Leveraging A New GAN-based Transformer with ECDH Crypto-system for Enhancing Energy Theft Detection in Smart Grid},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4C5PYHAK}},
  note         = {Machine review of arXiv:2411.18023}
}
read the original abstract

Detecting energy theft is vital for effectively managing power grids, as it ensures precise billing and prevents financial losses. Split-learning emerges as a promising decentralized machine learning technique for identifying energy theft while preserving user data confidentiality. Nevertheless, traditional split learning approaches are vulnerable to privacy leakage attacks, which significantly threaten data confidentiality. To address this challenge, we propose a novel GAN-Transformer-based split learning framework in this paper. This framework leverages the strengths of the transformer architecture, which is known for its capability to process long-range dependencies in energy consumption data. Thus, it enhances the accuracy of energy theft detection without compromising user privacy. A distinctive feature of our approach is the deployment of a novel mask-based method, marking a first in its field to effectively combat privacy leakage in split learning scenarios targeted at AI-enabled adversaries. This method protects sensitive information during the model's training phase. Our experimental evaluations indicate that the proposed framework not only achieves accuracy levels comparable to conventional methods but also significantly enhances privacy protection. The results underscore the potential of the GAN-Transformer split learning framework as an effective and secure tool in the domain of energy theft detection.

Figures

Figures reproduced from arXiv: 2411.18023 by the authors.

Figure 1
Figure 1. System Model D. Key Derivation Function The key derivation function (KDF) is a cryptography func￾tion designed to derive one or more keys from a given parameter. The main objective of KDF is to stretch keys to achieve a suitable length or convert keys into a required format. KDF usually take four different inputs: a random seed, a length, a salt s and context c. The security of KDF is captured from [22]. The advanta… view at source ↗
Figure 2
Figure 2. Threat Model A2: This type of adversary is an external attacker who may try to eavesdrop on the communication channel either physically or through cyber-attacks (Dolev-Yao Model). This enables the attacker to capture all or some messages transmit￾ted through the public channel to try inferring individuals’ private data. This adversary is passive, compromising the privacy of the customers’ data. This type of attacker… view at source ↗
Figure 3
Figure 3. Proposed Protocol for Securing Split-Learning. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Experiment Platform Note that we implemented another server using Raspberry Pi; it is for the energy consumption experiment and provides a clear and fair comparison with the client-side device. The results of the energy consumption can be found in Section VI-F B. Exper…
Figure 5
Figure 5. Figure 5: Exploratory Data Analysis for Selected Features [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Complexity Benchmark with the related encryption work [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Correlation Heatmap of Electricity Usage Features in the Pecan Street Smart Grid Dataset [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Energy Consumption [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Privacy Experiment results back by using the coefficient of determination R2 metric. The results of our experiments are shown in Table III, where we show how accurate it is to reconstruct the original numbers by measuring the coefficient of determination R2 metric. As …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 21 canonical work pages

  1. [1]

    Privacy-aware split learning based energy theft detection for smart grids,

    A. Alromih, J. A. Clark, and P. Gope, “Privacy-aware split learning based energy theft detection for smart grids,” in International Conference on Information and Communications Security . Springer, 2022, pp. 281– 300

  2. [2]

    Woodhead and B

    L. Woodhead and B. S. D. Unit. (2023) Electricity theft at record levels in england and wales - bbc analysis. [Online]. Available: https://www.bbc.com/news/uk-england-66847243

  3. [3]

    Guidelines for smart grid cyber security: V ol. 2, privacy and the smart grid,

    N. S. Grid, “Guidelines for smart grid cyber security: V ol. 2, privacy and the smart grid,” Guideline, Aug, vol. 6, 2010

  4. [4]

    Electricity theft: a comparative analysis,

    T. B. Smith, “Electricity theft: a comparative analysis,” Energy policy, vol. 32, no. 18, pp. 2067–2076, 2004

  5. [5]

    p2detect: Electricity theft detection with privacy preservation for both data and model in smart grid,

    L. Wu, H. Shi, S. Fu, Y . Luo, and M. Xu, “p2detect: Electricity theft detection with privacy preservation for both data and model in smart grid,” IEEE Transactions on Smart Grid , vol. 14, no. 3, pp. 2301–2312, 2022

  6. [6]

    Security and privacy challenges in the smart grid,

    P. McDaniel and S. McLaughlin, “Security and privacy challenges in the smart grid,” IEEE Security & Privacy , vol. 7, no. 3, pp. 75–77, 2009

  7. [7]

    Privacy-preserving and efficient decentralized federated learning-based energy theft detector,

    M. I. Ibrahem, M. Mahmoud, M. M. Fouda, B. M. ElHalawany, and W. Alasmary, “Privacy-preserving and efficient decentralized federated learning-based energy theft detector,” in GLOBECOM 2022-2022 IEEE Global Communications Conference . IEEE, 2022, pp. 287–292

  8. [8]

    Privacy-preserving electricity theft detection based on blockchain,

    Z. Zhao, Y . Liu, Z. Zeng, Z. Chen, and H. Zhou, “Privacy-preserving electricity theft detection based on blockchain,” IEEE Transactions on Smart Grid, 2023

Show all 34 references
  1. [9]

    Feddp: A privacy-protecting theft detection scheme in smart grids using federated learning,

    M. M. Ashraf, M. Waqas, G. Abbas, T. Baker, Z. H. Abbas, and H. Alasmary, “Feddp: A privacy-protecting theft detection scheme in smart grids using federated learning,” Energies, vol. 15, no. 17, p. 6241, 2022

  2. [10]

    Feddetect: A novel privacy-preserving federated learning framework for energy theft detection in smart grid,

    M. Wen, R. Xie, K. Lu, L. Wang, and K. Zhang, “Feddetect: A novel privacy-preserving federated learning framework for energy theft detection in smart grid,” IEEE Internet of Things Journal , vol. 9, no. 8, pp. 6069–6080, 2021

  3. [11]

    Designing a blockchain- enabled privacy-preserving energy theft detection system for smart grid neighborhood area network,

    A. Muzumdar, C. Modi, and C. Vyjayanthi, “Designing a blockchain- enabled privacy-preserving energy theft detection system for smart grid neighborhood area network,” Electric Power Systems Research, vol. 207, p. 107884, 2022

  4. [12]

    Privacy and security in federated learning: A survey,

    R. Gosselin, L. Vieu, F. Loukil, and A. Benoit, “Privacy and security in federated learning: A survey,” Applied Sciences , vol. 12, no. 19, p. 9901, 2022

  5. [13]

    A review of privacy-preserving fed- erated learning for the internet-of-things,

    C. Briggs, Z. Fan, and P. Andras, “A review of privacy-preserving fed- erated learning for the internet-of-things,” Federated Learning Systems: Towards Next-Generation AI, pp. 21–50, 2021

  6. [14]

    A survey of privacy attacks in machine learning,

    M. Rigaki and S. Garcia, “A survey of privacy attacks in machine learning,” ACM Computing Surveys , 2020

  7. [15]

    {Updates-Leak}: Data set inference and reconstruction attacks in online learning,

    A. Salem, A. Bhattacharya, M. Backes, M. Fritz, and Y . Zhang, “{Updates-Leak}: Data set inference and reconstruction attacks in online learning,” in 29th USENIX security symposium (USENIX Security 20) , 2020, pp. 1291–1308

  8. [16]

    Membership inference attacks against machine learning models,

    R. Shokri, M. Stronati, C. Song, and V . Shmatikov, “Membership inference attacks against machine learning models,” in 2017 IEEE symposium on security and privacy (SP) . IEEE, 2017, pp. 3–18

  9. [17]

    D. S. P. S. Inc., Dataport [2023]

  10. [18]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” Advances in neural information processing systems , vol. 27, 2014

  11. [19]

    Ganomaly: Semi- supervised anomaly detection via adversarial training,

    S. Akcay, A. Atapour-Abarghouei, and T. P. Breckon, “Ganomaly: Semi- supervised anomaly detection via adversarial training,” 2018

  12. [20]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, 2017

  13. [21]

    ” new directions in cryptography

    W. Diffie and M. Hellman, “” new directions in cryptography” ieee transactions on information theory, v. it-22, n. 6,” 1976

  14. [22]

    Cryptographic extraction and key derivation: The hkdf scheme,

    H. Krawczyk, “Cryptographic extraction and key derivation: The hkdf scheme,” in Annual Cryptology Conference . Springer, 2010, pp. 631– 648

  15. [23]

    Electricity theft detection in ami using customers’ consumption patterns,

    P. Jokar, N. Arianpoo, and V . C. Leung, “Electricity theft detection in ami using customers’ consumption patterns,” IEEE Transactions on Smart Grid, vol. 7, no. 1, pp. 216–226, 2015

  16. [24]

    Variational autoencoder based anomaly detection using reconstruction probability,

    J. An and S. Cho, “Variational autoencoder based anomaly detection using reconstruction probability,” Special lecture on IE , vol. 2, no. 1, pp. 1–18, 2015

  17. [25]

    Lstm-based encoder-decoder for multi-sensor anomaly de- tection,

    P. Malhotra, A. Ramakrishnan, G. Anand, L. Vig, P. Agarwal, and G. Shroff, “Lstm-based encoder-decoder for multi-sensor anomaly de- tection,” arXiv preprint arXiv:1607.00148 , 2016

  18. [26]

    Tranad: Deep transformer networks for anomaly detection in multivariate time series data,

    S. Tuli, G. Casale, and N. R. Jennings, “Tranad: Deep transformer networks for anomaly detection in multivariate time series data,” arXiv preprint arXiv:2201.07284, 2022

  19. [27]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly et al., “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929 , 2020

  20. [28]

    Random oracles are practical: A paradigm for designing efficient protocols,

    M. Bellare and P. Rogaway, “Random oracles are practical: A paradigm for designing efficient protocols,” in Proceedings of the 1st ACM Conference on Computer and Communications Security , 1993, pp. 62– 73

  21. [29]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural computation, vol. 9, no. 8, pp. 1735–1780, 1997

  22. [30]

    The coefficient of determi- nation r-squared is more informative than smape, mae, mape, mse and rmse in regression analysis evaluation,

    D. Chicco, M. J. Warrens, and G. Jurman, “The coefficient of determi- nation r-squared is more informative than smape, mae, mape, mse and rmse in regression analysis evaluation,” PeerJ Computer Science, vol. 7, p. e623, 2021

  23. [31]

    The block cipher rijndael,

    J. Daemen and V . Rijmen, “The block cipher rijndael,” in Interna- tional Conference on Smart Card Research and Advanced Applications . Springer, 1998, pp. 277–284

  24. [32]

    The simon and speck families of lightweight block ciphers,

    R. Beaulieu, D. Shors, J. Smith, S. Treatman-Clark, B. Weeks, and L. Wingers, “The simon and speck families of lightweight block ciphers,” cryptology eprint archive , 2013

  25. [33]

    Fully homomorphic encryption using ideal lattices,

    C. Gentry, “Fully homomorphic encryption using ideal lattices,” in Proceedings of the forty-first annual ACM symposium on Theory of computing, 2009, pp. 169–178

  26. [34]

    A systematic review of homomorphic encryption and its contributions in healthcare industry,

    K. Munjal and R. Bhatia, “A systematic review of homomorphic encryption and its contributions in healthcare industry,” Complex & Intelligent Systems, vol. 9, no. 4, pp. 3759–3786, 2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.