Pith. sign in

REVIEW 3 major objections 5 minor 2 cited by

PenTest++: Elevating Ethical Hacking with AI and Automation

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read An AI-augmented automation tool claims to make ethical hacking a five-phase human-overseen pipeline.

desk verdict PenTest++ is a solid proof of concept that a Python wrapper can drive standard pentest tools and the ChatGPT API through a chained lab attack, but the headline "five-phase automation" claim overstates what the prototype actually does. read the letter →

arxiv 2502.09484 v1 pith:GGVSJVSM submitted 2025-02-13 cs.CR cs.AI

Haitham S. Al-Sinani , Chris J. Mitchell This is my paper · ORCID
classification cs.CRcs.AI
keywords AI-assistedpenetrationtestinggenerativeAIChatGPTethicalhackingautomationmixed-initiativesystemscybersecuritylargelanguagemodels
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

PenTest++ is an AI-augmented command-line system that automates the routine stages of ethical hacking, including reconnaissance, scanning and enumeration, exploitation, and report generation, by scripting standard tools and sending their output to ChatGPT-4o for interpretation and next-step suggestions. The paper's central claim is that this mixed-initiative design demonstrates the feasibility of automating the five-phase penetration-testing process while keeping a human in control at each decision point. That claim is supported by two walk-through case studies in which the tool gained shell access to two deliberately vulnerable Linux VMs by chaining together nmap scans, anonymous FTP, hash cracking, directory enumeration, a PHP reverse shell, and an SSH private-key login, and then produced a structured pentest report. The point of the work is not a new vulnerability or attack technique; it is a proof of concept that GenAI plus automation can shoulder the mechanical parts of offensive security work and let testers focus on judgment, with the caution that AI hallucinations make human validation unavoidable.

What carries the argument

The load-bearing mechanism is a human-in-the-loop command pipeline. A Python driver executes penetration-testing tools such as nmap, gobuster, hashcat, John the Ripper, Hydra, and Netcat through subprocess calls, parses their output into PrettyTable tables, and sends selected logs or prompts to OpenAI's GPT-4 API. ChatGPT returns interpretations and suggested commands, which the driver displays and, on user approval, executes; the exact command being run is always shown to the user. This automate-analyse-approve-execute loop is what lets the system chain multiple attack stages while preserving oversight, and its modular design is meant to allow new scanning or exploitation tools to be added as needed.

What would settle it

A concrete experiment would run PenTest++ against a more realistic network with patched services, an intrusion detection or prevention system, rate-limited SSH, no anonymous FTP, and no world-readable NFS shares, and then count how many ChatGPT-suggested commands execute successfully without manual correction. If the suggestion-accuracy rate is low enough that the human effectively must redo the analysis, or if the chained attack fails under those conditions, the broad feasibility claim would be weakened; the paper currently reports only that 'some' recommendations required manual adjustment, with no count.

Watch

Extended reading notes

Core claim

PenTest++ is a working prototype that integrates generative AI into an otherwise scripted penetration-testing workflow. The system executes ordinary offensive-security utilities under programmatic control, parses their output into structured tables, and passes selected logs or prompt payloads to ChatGPT-4o via the GPT-4 API. ChatGPT returns interpretations, vulnerability assessments, and suggested exploitation commands, which the tool presents to the user for approval before execution. In the demonstrated case studies, this loop carried out a full attack chain against each of two target VMs, from live-host discovery and port scanning through credential cracking and reverse-shell upload to an interactive SSH session, and it automatically generated the accompanying report. The authors claim that this establishes the feasibility of automating the ethical hacking five-phase process through an AI-augmented tool, provided a human remains in the loop to validate AI outputs and approve significant actions.

Load-bearing premise

The load-bearing premise is that a small laboratory setup of two 512 MB Linux VMs on a local NAT network is representative enough of real penetration-testing environments to support the paper's efficiency, scalability, and adaptability claims; the paper itself says that generalisability to diverse operating systems, defensive mechanisms, and complex configurations remains to be validated.

Editorial extensions

If this is right

  • Routine reconnaissance, scanning, and enumeration can be delegated to a scripted pipeline while the tester selects targets and approves key actions.
  • Generative AI can reduce the cognitive load of interpreting tool output by extracting hashes, credentials, and likely next steps in a structured format.
  • Complex multi-step attack chains, such as extracting a protected archive, cracking a passphrase, exploiting a local file inclusion, and logging in with an SSH key, can be coordinated by a single tool with user confirmation at each stage.
  • Report generation can be largely automated, leaving the tester to review and refine the executive summary, findings, risk ratings, and recommendations.
  • The five-phase ethical hacking process is presented as automatable in a user-centric, mixed-initiative way, with the demonstrated phases being reconnaissance, scanning, exploitation, and documentation.

Reading between the lines

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

  • Beyond the paper: if this mixed-initiative pattern generalizes, the same automate-analyse-approve structure could be applied to less mature offensive areas such as wireless testing or mobile application assessments, where GenAI would propose tool-specific commands and the tester would validate them.
  • Beyond the paper: the privacy limitation the authors flag suggests a natural variant of PenTest++ using a local large language model instead of an online API, trading some analytical power for data containment; that variant could be tested directly against the same two VMs.
  • Beyond the paper: the two successful case studies do not by themselves isolate whether the generative model or the underlying scripts caused the success; a controlled comparison with the same automation but a static rule-based analyst would separate those contributions.
  • Beyond the paper: the authors note that quantitative metrics are future work; a measurable version of the central claim would be a suggestion-accuracy rate, such as the fraction of ChatGPT-recommended commands that execute successfully without manual correction.
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 / 5 minor

Summary. The paper presents PenTest++, a Python-based command-line tool that combines automation of common penetration-testing commands with ChatGPT-4o API calls. It describes four operation modules (reconnaissance, scanning/enumeration, exploitation, documentation) and reports two case studies in a VirtualBox environment in which the tool chains nmap, gobuster, hashcat, John the Ripper, and netcat to gain initial shell access to two Debian VMs. The authors claim that this demonstrates the feasibility of automating the ethical-hacking five-phase process while preserving human oversight, and they discuss ethical risks, limitations, and future work.

Significance. If fully supported, the paper would be a useful proof-of-concept for mixed-initiative GenAI-assisted penetration testing: the two attack chains are described in enough detail to be followed, the human-in-the-loop checkpoints are explicit, and the modular design is a reasonable basis for extension. The main value is architectural and pedagogical rather than empirical. The paper deserves credit for acknowledging in Sections 7.3 and 7.4 that generalisability and quantitative evaluation remain future work, but those acknowledgements conflict with the strength of the abstract's efficiency, scalability, and adaptability claims. No code release, benchmark, baseline, or quantitative measurements are provided; the evidence is a qualitative case-study narrative plus figures.

major comments (3)
  1. [Sec. 2.1, Sec. 2.3, Sec. 4, Sec. 6.3] The central contribution, stated as 'Demonstrating the feasibility of automating the ethical hacking five-phase process' (Sec. 2.3), is not supported by the implemented system and case studies. Sec. 2.1 defines the five phases as reconnaissance, scanning, exploitation, post-exploitation, and reporting, but Sec. 4 describes only four operation modules and the two case studies in Sec. 6.3 terminate at initial shell access. The text in Sec. 6.3 says the system 'prepares for the subsequent phase, which involves maintaining and elevating access,' and Sec. 6.4 moves directly to report generation; no privilege escalation, lateral movement, persistence, or other post-exploitation action is performed or evaluated. The abstract's alternative phase list (reconnaissance, scanning, enumeration, exploitation, documentation) highlights the ambiguity. Please either implement and evaluate the post-exploitation phase or revise the contribution and abstract to claim automation of the phases actually demonstrated.
  2. [Abstract, Sec. 7.2, Sec. 7.3, Sec. 7.4] The abstract and Sec. 7.2 claim enhanced efficiency, scalability, and adaptability, but no timing data, success-rate statistics, baseline comparison, or analysis of scalability are reported. The evaluation consists of two qualitative case studies on a single Kali machine and two 512 MB Linux VMs in a local NAT network (Sec. 5.2). Sec. 7.3 admits that 'generalisability to real-world networks with diverse operating systems (OSs), defensive mechanisms, and complex configurations remains to be validated,' and Sec. 7.4 lists quantitative evaluation as future work. These admissions are appropriate, but they directly contradict the strength of the abstract's claims. Please either provide measurements supporting the efficiency/scalability/adaptability claims or temper the claims to a qualitative proof-of-concept in a controlled lab environment.
  3. [Sec. 2.2, Sec. 7.1, Sec. 6.3] The answer to Research Question 3 ('How much human intervention is required...') is asserted rather than measured. Sec. 7.1 states that human intervention is essential and that users validate outputs, and Sec. 6.3 repeatedly says the user approved each step, but no record is given of the number, type, or frequency of interventions across the two case studies. Since 'user-centric' and 'human oversight' are central to the paper's contribution, a simple count of approvals, overrides, and AI corrections would substantiate the claim. Without such data, the degree of human intervention remains anecdotal.
minor comments (5)
  1. [Throughout] There are frequent spacing artifacts in the text, such as 'PenT est++', 'V M', 'T ool', 'F ig', and 'nmap command ( nmap -sn ...)'. Please run a proofreading pass to fix these rendering issues.
  2. [Sec. 6.4] The phrase 'The PenTest++ system also supports the conversion of this report into multiple formats, a:' contains a stray 'a:' fragment; it should read 'multiple formats:' or similar.
  3. [Sec. 5.3 vs. Sec. 5.4] Sec. 5.3 says the experiments used ChatGPT-4o, while Sec. 5.4 says the integration used the GPT-4 API. Please clarify which model/API version was actually used.
  4. [Sec. 6.3] The module is called 'Gaining Access Module' here but 'Exploitation' in Sec. 4. Please use consistent module names throughout.
  5. [Sec. 7.3] The statement that 'prior research [4] revealed that ChatGPT occasionally suggested incorrect commands' is used to support the hallucination risk, but since [4] is one of the authors' own papers, an independent citation or a direct example from this study would strengthen the point.

Circularity Check

0 steps flagged · score 2.0 of 10

No derivation-level circularity: the central PenTest++ case studies are self-contained demonstrations, and the recurring self-citations are supporting rather than load-bearing; the five-phase claim overreach is an evidence gap, not a circular reduction.

full rationale

The paper contains no fitted parameters, no prediction derived from an input by construction, and no uniqueness theorem imported from the authors' prior work. Its central demonstration is the two walk-through attack chains in Sec. 6.3, which stand on their own: each step (nmap scan, FTP retrieval, hashcat, gobuster, PHP reverse shell, NFS mount, LFI, SSH key use) is described concretely with commands and outcomes, and the supporting evidence is not logically equivalent to the paper's definitions or inputs. The self-citations that appear, such as 'As we have shown in previous research [6], ChatGPT can be leveraged to analyse scan outputs' (Sec. 6.1) and 'prior research [4] revealed that ChatGPT occasionally suggested incorrect commands' (Sec. 7.3), are used only to supply background plausibility for ChatGPT's behaviour, not to establish the paper's main claim. The related-work paragraph citing the authors' own earlier conceptual system [5] is incremental context rather than a load-bearing proof. A genuine weakness is that Sec. 2.3 claims demonstration of 'the ethical hacking five-phase process' while Sec. 2.1 includes post-exploitation among the five phases and the implemented prototype and case studies stop at initial shell access, with Sec. 6.3 only 'preparing' for the maintenance/elevation phase. This is an internal consistency and evidence-support problem for the strongest claim, not a circularity: the claim is unsupported by the shown work, but it is not equivalent to an input, a fitted value, or a self-citation chain. Accordingly, the circularity score is low; the correct critique is about scope of evidence, which the paper itself partially acknowledges in Sec. 7.3.

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

PenTest++ introduces no fitted numerical parameters, no new formal axioms, and no postulated physical entities. The central claim rests on domain assumptions about the representativeness of the lab environment, the reliability of ChatGPT outputs under human oversight, and the completeness of the five-phase model. These are the unproved premises a replication study would need to challenge.

assumptions (3)
  • domain assumption The local NAT virtual environment with two Linux VMs is representative of real penetration testing engagements for the purpose of evaluating PenTest++'s efficiency and scalability claims.
    The entire evaluation in Sec. 5.2 and Sec. 6 runs in this lab; Sec. 7.3 explicitly states generalizability to real-world networks remains to be validated, so this assumption is load-bearing.
  • domain assumption ChatGPT-4o, with human approval at each step, provides sufficiently accurate analysis and command suggestions for the demonstrated workflows.
    The system's value depends on GenAI interpretation of scan outputs (Sec. 6.1, 6.2) and exploitation guidance (Sec. 6.3); the paper itself reports hallucinations and manual corrections (Sec. 7.3), so the reliability of ChatGPT is an unproved premise for efficiency claims.
  • domain assumption The standard penetration testing process is captured by the five phases built into PenTest++ (reconnaissance, scanning/enumeration, exploitation, post-exploitation, reporting).
    The modular workflow in Sec. 4 adopts these phases from NIST SP 800-115 (ref. 16); the paper does not justify that this decomposition is complete for modern engagements, for example cloud or wireless testing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PenTest++: Elevating Ethical Hacking with AI and Automation." pith.science (2026). https://pith.science/paper/GGVSJVSM

@misc{pith2026250209484,
  author       = {Pith},
  title        = {Pith review of: PenTest++: Elevating Ethical Hacking with AI and Automation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GGVSJVSM}},
  note         = {Machine review of arXiv:2502.09484}
}
read the original abstract

Traditional ethical hacking relies on skilled professionals and time-intensive command management, which limits its scalability and efficiency. To address these challenges, we introduce PenTest++, an AI-augmented system that integrates automation with generative AI (GenAI) to optimise ethical hacking workflows. Developed in a controlled virtual environment, PenTest++ streamlines critical penetration testing tasks, including reconnaissance, scanning, enumeration, exploitation, and documentation, while maintaining a modular and adaptable design. The system balances automation with human oversight, ensuring informed decision-making at key stages, and offers significant benefits such as enhanced efficiency, scalability, and adaptability. However, it also raises ethical considerations, including privacy concerns and the risks of AI-generated inaccuracies (hallucinations). This research underscores the potential of AI-driven systems like PenTest++ to complement human expertise in cybersecurity by automating routine tasks, enabling professionals to focus on strategic decision-making. By incorporating robust ethical safeguards and promoting ongoing refinement, PenTest++ demonstrates how AI can be responsibly harnessed to address operational and ethical challenges in the evolving cybersecurity landscape.

Figures

Figures reproduced from arXiv: 2502.09484 by the authors.

Figure 1
Figure 1. Host scanning [PITH_FULL_IMAGE:figures/full_fig_p025_1.png] view at source ↗
Figure 2
Figure 2. Port targets discovered [PITH_FULL_IMAGE:figures/full_fig_p026_2.png] view at source ↗
Figure 3
Figure 3. Password cracking [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Shell payloads [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]
Figure 5
Figure 5. Figure 5: Shell gained on target 192.168.1.7 [PITH_FULL_IMAGE:figures/full_fig_p027_5.png]
Figure 6
Figure 6. Figure 6: Shell gained on target 192.168.1.10 [PITH_FULL_IMAGE:figures/full_fig_p027_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. A Survey of LLM-Driven Penetration Testing: Taxonomy, Co-Evolution, and Open Challenges

    cs.SE 2026-07 accept novelty 5.5 of 10

    LLM pentest agents co-evolved through four bottleneck-driven phases into RLVR systems, while CTF platforms became dual evaluation/training infrastructure and three linked reliability gaps remain.

  2. PenTest2.0: Towards Autonomous Privilege Escalation Using GenAI

    cs.CR 2025-07 conditional novelty 5.0 of 10

    PenTest2.0 demonstrates that an LLM-driven agent can autonomously suggest and run privilege escalation commands on a purposely vulnerable Linux VM, reaching root in every tested configuration but achieving automatic r...

Reference graph

Works this paper leans on

19 extracted references · 13 canonical work pages · cited by 2 Pith papers

  1. [1]

    In: Proceedings of 2023 IEEE/ACM 45th International Con- ference on Software Engineering (ICSE)

    Ahmed, T., Ghosh, S., Bansal, C., Zimmermann, T., Zhang, X., Rajmohan, S.: Recommending root-cause and mitigation steps for cloud incidents using large language models. In: Proceedings of 2023 IEEE/ACM 45th International Con- ference on Software Engineering (ICSE). pp. 1737–1749. IEEE (2023), https: //ieeexplore.ieee.org/abstract/document/10172904/

    work page arXiv 2023

  2. [2]

    AI-Enhanced Ethical Hacking: A Linux-Focused Experiment

    Al-Sinani, H., Mitchell, C.: AI-enhanced ethical hacking: A Linux-focused exper- iment. Technical report, Royal Holloway, University of London (2024), https: //arxiv.org/abs/2410.05105

    work page Pith review arXiv 2024

  3. [3]

    Al-Sinani, H., Mitchell, C.: Unleashing AI in ethical hacking: A prelimi- nary experimental study. Technical report, Royal Holloway, University of Lon- don (2024), https://pure.royalholloway.ac.uk/files/58692091/TechReport_ UnleashingAIinEthicalHacking.pdf Introducing PenTest++ 23

    arXiv 2024

  4. [4]

    Working- paper, arXiv (Nov 2024)

    Al-Sinani, H., Mitchell, C.: AI-augmented ethical hacking: A practical examination of manual exploitation and privilege escalation in Linux environments. Working- paper, arXiv (Nov 2024). https://doi.org/10.48550/arXiv.2411.17539, https: //arxiv.org/abs/2411.17539

    arXiv doi:10.48550/arxiv.2411.17539 2024

  5. [5]

    In: Security and Trust Management

    Al-Sinani, H., Mitchell, C., Sahli, N., Al-Siyabi, M.: Unleashing AI in ethi- cal hacking. In: Security and Trust Management. pp. 140–151. Lecture Notes in Computer Science, Springer-Verlag (Dec 2024). https://doi.org/10.1007/ 978-3-031-76371-7_10

    work page 2024

  6. [6]

    In: Proceedings of ITASEC and SERICS, Bologna, Italy (Feb 2025), https://www.chrismitchell.net/Papers/aehwaa.pdf

    Al-Sinani, H., Sahli, N., Mitchell, C., Al-Siyabi, M.: Advancing ethical hacking with AI: A Linux-based experimental study. In: Proceedings of ITASEC and SERICS, Bologna, Italy (Feb 2025), https://www.chrismitchell.net/Papers/aehwaa.pdf

    work page 2025

  7. [7]

    In: Joshi, A., Carminati, B., Verma, R.M

    Bertino, E., Kantarcioglu, M., Akcora, C.G., Samtani, S., Mittal, S., Gupta, M.: AI for security and security for AI. In: Joshi, A., Carminati, B., Verma, R.M. (eds.) CODASPY ’21: Eleventh ACM Conference on Data and Applica- tion Security and Privacy, Virtual Event, USA, April 26–28, 2021. pp. 333–

    work page 2021

  8. [8]

    In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H

    Brown, T.B., et al.: Language models are few-shot learners. In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H. (eds.) Advances in Neural Information Processing Systems 33: Annual Conference on Neu- ral Information Processing Systems 2020, NeurIPS 2020, December 6– 12, 2020, virtual (2020), https://proceedings.neurips.cc/paper/2020/hash/ 145...

    work page 2020

Show all 19 references
  1. [9]

    In: Proceedings of the SecAI 2024, the Workshop on Security and Artificial Intelligence (co-located with ESORICS 2024), Bydgoszcz, Poland

    Fujii, S., Yamagishi, R.: Feasibility study for supporting static malware analy- sis using LLM. In: Proceedings of the SecAI 2024, the Workshop on Security and Artificial Intelligence (co-located with ESORICS 2024), Bydgoszcz, Poland. p. to appear. LNCS series, Springer (2024)...

    2024

  2. [10]

    IEEE Ac- cess 11, 80218–80245 (2023)

    Gupta, M., Akiri, C., Aryal, K., Parker, E., Praharaj, L.: From ChatGPT to ThreatGPT: Impact of generative AI in cybersecurity and privacy. IEEE Ac- cess 11, 80218–80245 (2023). https://doi.org/10.1109/ACCESS.2023.3300381, https://doi.org/10.1109/ACCESS.2023.3300381

    2023

  3. [11]

    WIREs Data Mining and Knowledge Discovery 9(4), e1306 (2019)

    Handa, A., Sharma, A., Shukla, S.K.: Machine learning in cybersecurity: A review. WIREs Data Mining and Knowledge Discovery 9(4), e1306 (2019). https://doi. org/10.1002/WIDM.1306, https://doi.org/10.1002/widm.1306

    2019 doi

  4. [12]

    In: IEEE European Symposium on Security and Privacy, EuroS&P 2023 — Workshops, Delft, Netherlands, July 3-7, 2023

    Harrison, J., Toreini, E., Mehrnezhad, M.: A practical deep learning-based acoustic side channel attack on keyboards. In: IEEE European Symposium on Security and Privacy, EuroS&P 2023 — Workshops, Delft, Netherlands, July 3-7, 2023. pp. 270–

    2023

  5. [13]

    Jiang, F., Xu, Z., Niu, L., Xiang, Z., Ramasubramanian, B., Li, B., Poovendran, R.: ArtPrompt: ASCII art-based jailbreak attacks against aligned LLMs. Tech. rep. (2024). https://doi.org/10.48550/ARXIV.2402.11753, https://doi.org/ 10.48550/arXiv.2402.11753

    2024 doi arXiv

  6. [14]

    In: Proceedings of the SecAI 2024, the Workshop on Security and Ar- tificial Intelligence (co-located with ESORICS 2024), Bydgoszcz, Poland

    Park, S., Lee, H., Cha, S.K.: Systematic bug reproduction with large language model. In: Proceedings of the SecAI 2024, the Workshop on Security and Ar- tificial Intelligence (co-located with ESORICS 2024), Bydgoszcz, Poland. p. to appear. LNCS series, Springer (2024), https:/...

    2024

  7. [15]

    In: Proceedings of the IEEE Symposium on Security and Privacy (SP)

    Pearce, H., Tan, B., Ahmad, B., Karri, R., Dolan-Gavitt, B.: Examining zero- shot vulnerability repair with large language models. In: Proceedings of the IEEE Symposium on Security and Privacy (SP). pp. 2339–2356. IEEE (2023), https: //ieeexplore.ieee.org/abstract/document/10179324

    2023

  8. [16]

    Special Publi- cation 800-115, National Institute of Standards and Technology (2008), https: //csrc.nist.gov/publications/detail/sp/800-115/final

    Swanson, M., Bartol, N., Sabato, J., Hash, J., Graffo, L.: Technical guide to information security testing and assessment (NIST SP 800-115). Special Publi- cation 800-115, National Institute of Standards and Technology (2008), https: //csrc.nist.gov/publications/detail/sp/800-...

    2008

  9. [17]

    In: Guyon, I., von Luxburg, U., Bengio, S., Wallach, H.M., Fergus, R., Vishwanathan, S.V.N., Garnett, R

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L., Polosukhin, I.: Attention is all you need. In: Guyon, I., von Luxburg, U., Bengio, S., Wallach, H.M., Fergus, R., Vishwanathan, S.V.N., Garnett, R. (eds.) Advances in Neural Information Pr...

    2017

  10. [280]

    https://doi.org/10.1109/EUROSPW59978.2023.00034, https: //doi.org/10.1109/EuroSPW59978.2023.00034

    IEEE (2023). https://doi.org/10.1109/EUROSPW59978.2023.00034, https: //doi.org/10.1109/EuroSPW59978.2023.00034

    2023

  11. [334]

    https://doi.org/10.1145/3422337.3450357, https://doi

    ACM (2021). https://doi.org/10.1145/3422337.3450357, https://doi. org/10.1145/3422337.3450357

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.