Pith. sign in

REVIEW 3 major objections 6 minor 46 references

Large Language Model-driven Security Assistant for Internet of Things via Chain-of-Thought

T0 review · 3 major / 6 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 proposes ICoT, a two-stage chain-of-thought prompt that makes LLMs analyze a user's role and a vulnerability's type, location, and impact before generating personalized IoT security advice, and reports that this prompt…

desk verdict The two-stage role-and-vulnerability prompt is a sensible, clearly presented idea, but the only evidence for the central claim is an unvalidated LLM judge scoring the same model's outputs, so the headline assertion of higher accuracy and reliability is not supported. read the letter →

arxiv 2505.06307 v1 pith:VTEYGGSY submitted 2025-05-08 cs.CR cs.AI

Mingfei Zeng , Ming Xie , Xixi Zheng , Chunhai Li , Chuan Zhang , Liehuang Zhu This is my paper · ORCID
classification cs.CRcs.AI
keywords InternetofThingssecuritylargelanguagemodelchain-of-thoughtpromptingICoTvulnerabilityanalysispersonalizedadviceGPT-4oDeepSeek
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 a carefully structured prompt can make off-the-shelf large language models produce better, more personalized security advice for Internet of Things devices, with no fine-tuning. The proposed method, ICoT, is a two-stage chain-of-thought scheme: the model first analyzes the person asking (their knowledge, goals, and constraints) and the vulnerability (its type, location, and impact), and then uses that analysis to write the final answer. Across GPT-4o, GPT-4o-mini, DeepSeek-V3, and DeepSeek-R1, and for three user roles, ICoT responses score above plain LLM responses on reliability, relevance, detail, technicality, and friendliness, as judged by GPT-4o. If the claim holds, it gives a parameter-free way to upgrade LLM-based security assistants in heterogeneous IoT environments. The weight-bearing part of the evaluation is the judge, since the fixed evaluator is one of the models under test.

What carries the argument

The central object is the ICoT prompt, a two-round chain-of-thought template. Round one asks the LLM to output a JSON object describing the user's knowledge, goals, and requirements alongside the vulnerability's type, location, and impact ($S_g = f_\theta(P_{in}^{(1)})$). Round two injects that JSON as context and instructs the model to answer using it ($R = f_\theta(P_{in}^{(2)})$). The work it does is to force the model to separate who is asking from what is broken before recommending a fix, which is what the authors argue produces context-aware and personalized security advice without retraining or access to model parameters.

What would settle it

Have independent human IoT security experts blind-rate the same ICoT and LLM-only answers for factual correctness and actionability, or run a swap test in which the labels 'ICoT' and 'LLM only' are exchanged before GPT-4o scores them; if human experts find no consistent advantage, or if GPT-4o's scores track the labels instead of the content, the claim that ICoT improves understanding of IoT security issues is falsified.

Watch

Extended reading notes

Core claim

The central discovery is that decomposing a security query into a role-and-vulnerability analysis, serialized as JSON, and feeding that analysis back into the model produces responses the authors judge more reliable, relevant, detailed, technically precise, and user-friendly than the same model answering directly. The pipeline is formalized as two inferences, $S_g = f_\theta(P_{in}^{(1)})$ for the intermediate analysis and $R = f_\theta(P_{in}^{(2)})$ for the final response, where the first prompt is $[V][T][S_{in}]$ and the second is $[V][S_g][C][T]$. The reported results show positive deltas in nearly every model-role-metric cell of Table I, with the largest gains often on technicality and relevance, and the paper acknowledges that redundant output and hallucination can blunt the gains in some cases.

Load-bearing premise

The comparison's validity rests on the assumption that GPT-4o, acting as the fixed judge, scores ICoT and LLM-only answers impartially and on their factual merits; if the judge merely rewards the structured, role-aware, verbose style that ICoT deliberately produces, the reported gains are an artifact rather than a genuine improvement in understanding.

Editorial extensions

If this is right

  • Because ICoT operates only through prompts, any existing LLM-based assistant can adopt it by changing the prompt, with no retraining and no access to model parameters.
  • The two-stage design is user-extensible: users can supply their own role templates and choose which vulnerability characteristics to analyze, so the same mechanism covers roles beyond the three tested.
  • The authors argue the method carries over to other security domains, naming industrial control systems, smart grids, and smart-home security as direct targets.
  • For non-expert users, the measured gains in friendliness and relevance translate into advice that is easier to act on, which is the practical payoff the paper emphasizes.
  • The paper's own limitations imply the gains are not uniform: verbosity, hallucinated details, and inconsistent reasoning on poorly specified vulnerabilities can reduce clarity and may offset the improvements in some cases.

Reading between the lines

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

  • A cheap falsification-style check follows directly from the setup: relabel the two answer sets before asking GPT-4o to score them; if scores follow the labels rather than the content, part of the reported gain is judge self-preference for structured, role-aware text rather than improved security understanding.
  • The role-and-vulnerability JSON profile could be carried across a multi-turn conversation, turning a one-shot assistant into a persistent advisor that remembers the user's context; the paper does not explore this.
  • Pairing ICoT with retrieval of up-to-date vulnerability databases would directly address the paper's stated weakness that descriptions can be stale or incomplete, since the prompt structure is agnostic to where the vulnerability text comes from.
  • Sending user roles, goals, and infrastructure context to a commercial LLM API raises a privacy cost the paper mentions but does not quantify; a local or privacy-preserving inference variant is a natural next step.
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

3 major / 6 minor

Summary. The paper proposes ICoT, a two-stage chain-of-thought prompting method for an LLM-based IoT security assistant. In the first stage, the LLM analyzes the user role and the vulnerability description (type, location, impact) and emits a structured JSON summary; in the second stage, it uses that summary to generate a personalized security recommendation. The method is tested on GPT-4o, GPT-4o-mini, DeepSeek-V3, and DeepSeek-R1 using a fixed GPT-4o evaluator that scores outputs on five subjective quality metrics. The abstract claims that ICoT 'significantly improves the understanding of IoT security issues' and achieves 'higher accuracy and reliability' compared with LLM-only prompting.

Significance. If the empirical claim were supported, the paper would offer a practical prompt-only way to make LLM IoT security advice more tailored and useful, with no fine-tuning. The strength of this contribution, however, is entirely dependent on the evaluation. The manuscript provides no human validation, no statistical analysis, and no comparison against the closest prior system (ChatIoT), even though that system is cited. The evaluation design has a self-judging flaw: GPT-4o scores answers that, in the GPT-4o and GPT-4o-mini conditions, the same model family generated. For these reasons, the results in Table I do not establish the claimed accuracy and reliability gains, and the central contribution is not currently supported.

major comments (3)
  1. [V.B and Fig. 3] The evaluation uses GPT-4o as a fixed judge to score answers, but for the GPT-4o and GPT-4o-mini conditions the judged answers were generated by the same model family. The judge is a closed API model with no human validation or external ground truth. The template asks for subjective scores on qualities such as Detail and Friendliness, which are likely to reward the longer, structured, role-aware ICoT outputs regardless of their factual correctness. The paper's own Section V.D admits that LLMs hallucinate and can produce redundant information, so the abstract's claim of 'higher accuracy and reliability' is not supported by this design. The numeric gains in Table I therefore cannot be read as evidence of improved security understanding.
  2. [V.C and Table I] The metric definitions are inconsistent: Section V.B defines Accuracy, Relevance, Detail, Technicality, and Friendliness, while Fig. 3 and Table I report Reliability instead of Accuracy. Table I also contains an invalid numeric entry, '3.3.66', in the Developer/DeepSeek-R1/Detail row. The text states only that scores are in [0,5], with no sample size, variance, number of repeated runs, or significance tests. Deltas as small as +0.05 (e.g., GPT-4o General User Reliability) are presented as improvements, and the abstract uses the word 'significantly' without any statistical support. The evaluation protocol is under-specified and does not allow the reader to assess whether the observed differences are real or noise.
  3. [Introduction and II.B] The paper cites ChatIoT [3] as an existing LLM-based IoT security assistant that uses retrieval-augmented generation, but it never compares ICoT against ChatIoT or any other LLM-based baseline. The only baseline is a bare LLM with no IoT-specific prompting. This cannot support the contributions section's claim that ICoT 'outperforming traditional methods' or the Discussion's claim of more accurate and in-depth analysis than traditional vulnerability detection tools. A direct comparison with the closest prior assistant, or at least a clear discussion of why such a comparison is not needed, is required to position the contribution.
minor comments (6)
  1. [Table I] The entry '3.3.66' in the Developer/DeepSeek-R1/Detail row should be corrected to a valid score in [0,5].
  2. [V.A] The dataset description does not state how many vulnerability descriptions or user queries were used, nor how they were selected from V ARIoT. Provide these counts and the selection criteria so the reader can gauge the evaluation's coverage.
  3. [V.B] The claim that judging both outputs at the same time 'minimizes the impact of any inherent randomness' is not supported; report the number of evaluation runs and the observed variance.
  4. [II.B] ChatIoT [3] is cited in the introduction but not discussed in Section II.B; add it to the related-work discussion and explicitly explain how ICoT differs from and improves upon it.
  5. [Fig. 2] Figure 2 is difficult to read: the prompt text is truncated and the structure is not self-explanatory. A cleaner diagram with the full prompt template would help.
  6. [IV, Eq. (2) and (4)] The notation '[V][T][Sin]' and '[V][Sg][C][T]' is introduced only informally; define each component explicitly (e.g., what T contains, how Sin differs from C) so the two-stage process is reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ICoT is a prompt-engineering method with no fitted parameters or derivation chain; the LLM-as-judge evaluation is a validity concern, not a reduction-by-construction.

full rationale

The paper contains no mathematical derivation chain or fitted parameters. ICoT is defined by two prompt-construction equations, P_in^(1) = [V][T][S_in] and P_in^(2) = [V][S_g][C][T] (Eqs. 2 and 4), and the outputs are just f_theta of these prompts (Eqs. 3 and 5). There is no claim that ICoT's superiority is derived from these equations; the superiority is an empirical assertion supported by Table I. The evaluation uses GPT-4o as a fixed judge (Section V.B: 'we used an independent LLM as an evaluator (fixed as GPT-4o)'), and for the GPT-4o row the judged answers were produced by the same model family. This creates a legitimate validity risk: the judge may favor ICoT's structured, verbose format, and no human or ground-truth validation is provided. The paper itself concedes in Section V.D that 'LLMs, despite their powerful reasoning capabilities, are not immune to hallucination' and that ICoT 'sometimes encourages the LLM to generate excessive redundant information.' These are serious threats to the reliability of the reported gains, but they do not make the gains equivalent to the inputs by construction. The scores are not fitted parameters, the comparison is not logically forced, and the paper even reports small or inconsistent gains in some rows (e.g., Table I, GPT-4o General User Reliability +0.08). Additionally, the metric name drifts from 'Accuracy' in Section V.B to 'Reliability' in Table I, further weakening the accuracy claim; this is a reporting inconsistency, not circularity. Citation [3], used to justify the simultaneous-presentation evaluation protocol, is not a self-citation. Therefore, under the defined circularity patterns, no step reduces to its own input; the appropriate finding is no significant circularity, with the caveat that the empirical support is weak.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No numeric free parameters are fitted. The method's load-bearing assumptions are the validity of the LLM-as-judge, the representativeness of the VARIoT sample, the reliability of the two-stage prompt, and the consistency of the five metrics. No new physical or computational entities are introduced; ICoT is a prompt design, not an invented entity.

assumptions (4)
  • domain assumption LLM-as-judge scores from GPT-4o on a 0-5 scale are a valid measure of security answer quality.
    Invoked in Section V.B and Fig. 3. No human evaluation or ground-truth labels are provided; if judge preferences correlate with answer length or structure, the comparison is biased.
  • domain assumption The VARIoT Vulnerabilities and Exploits datasets supply representative and correct vulnerability descriptions for the evaluation.
    Invoked in Section V.A. The paper gives no sample counts, no selection criteria, and no verification of label correctness.
  • domain assumption The two-stage ICoT prompt exploits the LLM's pre-existing reasoning without fine-tuning, and the JSON intermediate output is reliable enough to improve final answers.
    Invoked in Sections IV.B and IV.C. If the intermediate analysis is noisy or ignored, the method may not generalize.
  • domain assumption The five evaluation metrics are well-defined and consistently applied despite inconsistent naming in the text and figure.
    The text defines Accuracy, Relevance, Detail, Technicality, and Friendliness, while Fig. 3 and Table I use Reliability; the mismatch suggests the metric set is not fixed, and the validity of the scores is unclear.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Large Language Model-driven Security Assistant for Internet of Things via Chain-of-Thought." pith.science (2026). https://pith.science/paper/VTEYGGSY

@misc{pith2026250506307,
  author       = {Pith},
  title        = {Pith review of: Large Language Model-driven Security Assistant for Internet of Things via Chain-of-Thought},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VTEYGGSY}},
  note         = {Machine review of arXiv:2505.06307}
}
read the original abstract

The rapid development of Internet of Things (IoT) technology has transformed people's way of life and has a profound impact on both production and daily activities. However, with the rapid advancement of IoT technology, the security of IoT devices has become an unavoidable issue in both research and applications. Although some efforts have been made to detect or mitigate IoT security vulnerabilities, they often struggle to adapt to the complexity of IoT environments, especially when dealing with dynamic security scenarios. How to automatically, efficiently, and accurately understand these vulnerabilities remains a challenge. To address this, we propose an IoT security assistant driven by Large Language Model (LLM), which enhances the LLM's understanding of IoT security vulnerabilities and related threats. The aim of the ICoT method we propose is to enable the LLM to understand security issues by breaking down the various dimensions of security vulnerabilities and generating responses tailored to the user's specific needs and expertise level. By incorporating ICoT, LLM can gradually analyze and reason through complex security scenarios, resulting in more accurate, in-depth, and personalized security recommendations and solutions. Experimental results show that, compared to methods relying solely on LLM, our proposed LLM-driven IoT security assistant significantly improves the understanding of IoT security issues through the ICoT approach and provides personalized solutions based on the user's identity, demonstrating higher accuracy and reliability.

Figures

Figures reproduced from arXiv: 2505.06307 by the authors.

Figure 1
Figure 1. System architecture. B. System Architecture The general workflow of the IoT security assistant is shown in [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Full prompt example of ICoT. A. Preliminaries The LLM is an advanced and highly sophisticated model specifically designed to process textual input and generate coherent, contextually appropriate responses. When presented with an input prompt Pin, which could be in the form of a question, statement, or instruction, the LLM systematically processes this text through multiple transformer-based layers to produce an outp… view at source ↗
Figure 3
Figure 3. The evaluation template. evaluator (fixed as GPT-4o) and measured the results based on five metrics: Accuracy, Relevance, Detail, Technicality, and Friendliness, which are defined as follows: 1) Accuracy: The correctness of the answer, ensuring that it aligns with established IoT security principles and accu￾rately addresses the described vulnerabilities or threats. 2) Relevance: The extent to which the answer direc… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 38 canonical work pages

  1. [3]

    Chatiot: Large language model-based security assistant for internet of things with retrieval-augmented generation,

    Y . Dong, Y . L. Aung, S. Chattopadhyay, and J. Zhou, “Chatiot: Large language model-based security assistant for internet of things with retrieval-augmented generation,” arXiv preprint arXiv:2502.09896, 2025. 9

    arXiv 2025

  2. [1]

    Role of iot technologies in big data management systems: A review and smart grid case study,

    A. Al-Ali, R. Gupta, I. Zualkernan, and S. K. Das, “Role of iot technologies in big data management systems: A review and smart grid case study,” Pervasive and Mobile Computing , p. 101905, 2024

    work page 2024

  3. [2]

    Vulnerability of machine learning approaches applied in iot-based smart grid: A review,

    Z. Zhang, M. Liu, M. Sun, R. Deng, P. Cheng, D. Niyato, M.-Y . Chow, and J. Chen, “Vulnerability of machine learning approaches applied in iot-based smart grid: A review,”IEEE Internet of Things Journal, vol. 11, no. 11, pp. 18 951–18 975, 2024

    work page 2024

  4. [4]

    A compre- hensive review of cyber-attacks targeting iot systems and their security measures

    M. L. Mutleg, A. M. Mahmood, and M. M. J. Al-Nayar, “A compre- hensive review of cyber-attacks targeting iot systems and their security measures.” International Journal of Safety & Security Engineering , vol. 14, no. 4, 2024

    work page 2024

  5. [5]

    Heterogeneous iot (hetiot) security: techniques, challenges and open issues,

    S. S. Mahadik, P. M. Pawar, and R. Muthalagu, “Heterogeneous iot (hetiot) security: techniques, challenges and open issues,” Multimedia Tools and Applications, vol. 83, no. 12, pp. 35 371–35 412, 2024

    work page 2024

  6. [6]

    Firmalice-automatic detection of authentication bypass vul- nerabilities in binary firmware

    Y . Shoshitaishvili, R. Wang, C. Hauser, C. Kruegel, and G. Vi- gna, “Firmalice-automatic detection of authentication bypass vul- nerabilities in binary firmware.” in Proceedings of the 22nd An- nual Network and Distributed System Security Symposium , 2015, doi:10.14722/ndss.2015.23294

    work page arXiv 2015

  7. [7]

    Inception: System- wide security testing of real-world embedded systems software,

    N. Corteggiani, G. Camurati, and A. Francillon, “Inception: System- wide security testing of real-world embedded systems software,” in 27th USENIX security symposium , 2018, pp. 309–326

    work page 2018

  8. [8]

    Firm-afl: High-throughput greybox fuzzing of iot firmware via augmented process emulation,

    Y . Zheng, A. Davanian, H. Yin, C. Song, H. Zhu, and L. Sun, “Firm-afl: High-throughput greybox fuzzing of iot firmware via augmented process emulation,” in 28th USENIX Security Symposium, 2019, pp. 1099–1114

    work page 2019

Show all 46 references
  1. [9]

    A survey of machine and deep learning methods for internet of things (iot) security,

    M. A. Al-Garadi, A. Mohamed, A. K. Al-Ali, X. Du, I. Ali, and M. Guizani, “A survey of machine and deep learning methods for internet of things (iot) security,” IEEE Communications Surveys & Tutorials, vol. 22, no. 3, pp. 1646–1685, 2020

    2020

  2. [10]

    Machine learning in iot security: Current solutions and future challenges,

    F. Hussain, R. Hussain, S. A. Hassan, and E. Hossain, “Machine learning in iot security: Current solutions and future challenges,” IEEE Communications Surveys & Tutorials , vol. 22, no. 3, pp. 1686–1721, 2020

    2020

  3. [11]

    Corrauc: a malicious bot-iot traffic detection method in iot network using machine- learning techniques,

    M. Shafiq, Z. Tian, A. K. Bashir, X. Du, and M. Guizani, “Corrauc: a malicious bot-iot traffic detection method in iot network using machine- learning techniques,” IEEE Internet of Things Journal , vol. 8, no. 5, pp. 3242–3254, 2020

    2020

  4. [12]

    Mthael: Cross-architecture iot malware detection based on neural network ad- vanced ensemble learning,

    D. Vasan, M. Alazab, S. Venkatraman, J. Akram, and Z. Qin, “Mthael: Cross-architecture iot malware detection based on neural network ad- vanced ensemble learning,” IEEE Transactions on Computers , vol. 69, no. 11, pp. 1654–1667, 2020

    2020

  5. [13]

    Deep learning based cross architecture internet of things malware detection and classification,

    R. Chaganti, V . Ravi, and T. D. Pham, “Deep learning based cross architecture internet of things malware detection and classification,” Computers & Security , vol. 120, p. 102779, 2022

    2022

  6. [14]

    Federated learning for internet of things: A comprehensive survey,

    D. C. Nguyen, M. Ding, P. N. Pathirana, A. Seneviratne, J. Li, and H. V . Poor, “Federated learning for internet of things: A comprehensive survey,” IEEE Communications Surveys & Tutorials , vol. 23, no. 3, pp. 1622–1658, 2021

    2021

  7. [15]

    Internet of things (iot)-based smart healthcare system for efficient diagnostics of health parameters of patients in emergency care,

    A. Balasundaram, S. Routray, A. Prabu, P. Krishnan, P. P. Malla, and M. Maiti, “Internet of things (iot)-based smart healthcare system for efficient diagnostics of health parameters of patients in emergency care,” IEEE Internet of Things Journal , vol. 10, no. 21, pp. 18 563–1...

    2023

  8. [16]

    Applications of the internet of things (iot) in smart logistics: A comprehensive survey,

    Y . Song, F. R. Yu, L. Zhou, X. Yang, and Z. He, “Applications of the internet of things (iot) in smart logistics: A comprehensive survey,”IEEE Internet of Things Journal , vol. 8, no. 6, pp. 4250–4274, 2020

    2020

  9. [17]

    Industrial internet of things (iiot): opportunities, challenges, and requirements in manufacturing businesses in emerging economies,

    O. Peter, A. Pradhan, and C. Mbohwa, “Industrial internet of things (iiot): opportunities, challenges, and requirements in manufacturing businesses in emerging economies,” Procedia Computer Science , vol. 217, pp. 856–865, 2023

    2023

  10. [18]

    Federated learning for internet of things: Recent advances, taxonomy, and open challenges,

    L. U. Khan, W. Saad, Z. Han, E. Hossain, and C. S. Hong, “Federated learning for internet of things: Recent advances, taxonomy, and open challenges,” IEEE Communications Surveys & Tutorials , vol. 23, no. 3, pp. 1759–1799, 2021

    2021

  11. [19]

    Landscape of iot security,

    E. Schiller, A. Aidoo, J. Fuhrer, J. Stahl, M. Zi ¨orjen, and B. Stiller, “Landscape of iot security,” Computer Science Review , vol. 44, p. 100467, 2022

    2022

  12. [20]

    Ddos in the iot: Mirai and other botnets,

    C. Kolias, G. Kambourakis, A. Stavrou, and J. V oas, “Ddos in the iot: Mirai and other botnets,” Computer, vol. 50, no. 7, pp. 80–84, 2017

    2017

  13. [21]

    The rise of ransomware and emerging security challenges in the internet of things,

    I. Yaqoob, E. Ahmed, M. H. ur Rehman, A. I. A. Ahmed, M. A. Al-garadi, M. Imran, and M. Guizani, “The rise of ransomware and emerging security challenges in the internet of things,” Computer Networks, vol. 129, pp. 444–458, 2017

    2017

  14. [22]

    Machine learning-enabled iot security: Open issues and challenges under advanced persistent threats,

    Z. Chen, J. Liu, Y . Shen, M. Simsek, B. Kantarci, H. T. Mouftah, and P. Djukic, “Machine learning-enabled iot security: Open issues and challenges under advanced persistent threats,” ACM Computing Surveys, vol. 55, no. 5, pp. 1–37, 2022

    2022

  15. [23]

    Man-in-the-middle attack mitigation in internet of medical things,

    O. Salem, K. Alsubhi, A. Shaafi, M. Gheryani, A. Mehaoua, and R. Boutaba, “Man-in-the-middle attack mitigation in internet of medical things,” IEEE Transactions on Industrial Informatics , vol. 18, no. 3, pp. 2053–2062, 2021

    2021

  16. [24]

    Static analysis for discovering iot vulnerabilities,

    P. Ferrara, A. K. Mandal, A. Cortesi, and F. Spoto, “Static analysis for discovering iot vulnerabilities,” International Journal on Software Tools for Technology Transfer, vol. 23, no. 1, pp. 71–88, 2021

    2021

  17. [25]

    When iot meet llms: Applications and challenges,

    ˙I. K ¨ok, O. Demirci, and S. ¨Ozdemir, “When iot meet llms: Applications and challenges,” in 2024 IEEE International Conference on Big Data . IEEE, 2024, pp. 7075–7084

    2024

  18. [26]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologi...

    2019

  19. [27]

    Llama: Open and efficient foundation language models,

    H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.-A. Lachaux, T. Lacroix, B. Rozi `ere, N. Goyal, E. Hambro, F. Azhar et al. , “Llama: Open and efficient foundation language models,” arXiv preprint arXiv:2302.13971, 2023

    2023 arXiv

  20. [28]

    Deepseek-v3 technical report,

    A. Liu, B. Feng, B. Xue, B. Wang, B. Wu, C. Lu, C. Zhao, C. Deng, C. Zhang, C. Ruan et al., “Deepseek-v3 technical report,” arXiv preprint arXiv:2412.19437, 2024

    2024 arXiv

  21. [29]

    Generative ai for internet of things security: Challenges and opportunities,

    Y . L. Aung, I. Christian, Y . Dong, X. Ye, S. Chattopadhyay, and J. Zhou, “Generative ai for internet of things security: Challenges and opportunities,” arXiv preprint arXiv:2502.08886 , 2025

    2025 arXiv

  22. [30]

    Llmif: Augmented large language model for fuzzing iot devices,

    J. Wang, L. Yu, and X. Luo, “Llmif: Augmented large language model for fuzzing iot devices,” in 2024 IEEE Symposium on Security and Privacy. IEEE, 2024, pp. 881–896

    2024

  23. [31]

    Large language model guided protocol fuzzing,

    R. Meng, M. Mirchev, M. B ¨ohme, and A. Roychoudhury, “Large language model guided protocol fuzzing,” in Proceedings of the 31st Annual Network and Distributed System Security Symposium , vol. 2024, 2024

    2024

  24. [32]

    From one thousand pages of specification to unveiling hidden bugs: Large language model assisted fuzzing of matter iot devices,

    X. Ma, L. Luo, and Q. Zeng, “From one thousand pages of specification to unveiling hidden bugs: Large language model assisted fuzzing of matter iot devices,” in 33rd USENIX Security Symposium , 2024, pp. 4783–4800

    2024

  25. [33]

    Revolutionizing cyber threat detection with large language models: A privacy-preserving bert-based lightweight model for iot/iiot devices,

    M. A. Ferrag, M. Ndhlovu, N. Tihanyi, L. C. Cordeiro, M. Debbah, T. Lestable, and N. S. Thandi, “Revolutionizing cyber threat detection with large language models: A privacy-preserving bert-based lightweight model for iot/iiot devices,” IEEE Access , vol. 12, pp. 23 733–23 750, 2024

    2024

  26. [34]

    Ids-agent: An llm agent for explainable intrusion detection in iot networks,

    Y . Li, Z. Xiang, N. D. Bastian, D. Song, and B. Li, “Ids-agent: An llm agent for explainable intrusion detection in iot networks,” in NeurIPS 2024 Workshop on Open-World Agents , 2024

    2024

  27. [35]

    Evaluating large language models for enhanced intrusion detection in internet of things networks,

    E. Nwafor, U. Baskota, M. S. Parwez, J. Blackstone, and H. Olufowobi, “Evaluating large language models for enhanced intrusion detection in internet of things networks,” in GLOBECOM 2024-2024 IEEE Global Communications Conference. IEEE, 2024, pp. 3358–3363

    2024

  28. [36]

    Distributed threat intelligence at the edge devices: A large language model-driven approach,

    S. M. Hasan, A. M. Alotaibi, S. Talukder, and A. R. Shahid, “Distributed threat intelligence at the edge devices: A large language model-driven approach,” in 2024 IEEE 48th Annual Computers, Software, and Appli- cations Conference. IEEE, 2024, pp. 1496–1497

    2024

  29. [37]

    Llm-tikg: Threat intel- ligence knowledge graph construction utilizing large language model,

    Y . Hu, F. Zou, J. Han, X. Sun, and Y . Wang, “Llm-tikg: Threat intel- ligence knowledge graph construction utilizing large language model,” Computers & Security , vol. 145, p. 103999, 2024

    2024

  30. [38]

    Chain-of-thought prompting elicits reasoning in large language models,

    J. Wei, X. Wang, D. Schuurmans, M. Bosma, F. Xia, E. Chi, Q. V . Le, D. Zhou et al. , “Chain-of-thought prompting elicits reasoning in large language models,” Advances in Neural Information Processing Systems , vol. 35, pp. 24 824–24 837, 2022

    2022

  31. [39]

    Tree of thoughts: Deliberate problem solving with large language models,

    S. Yao, D. Yu, J. Zhao, I. Shafran, T. Griffiths, Y . Cao, and K. Narasimhan, “Tree of thoughts: Deliberate problem solving with large language models,” Advances in Neural Information Processing Systems , vol. 36, pp. 11 809–11 822, 2023

    2023

  32. [40]

    Chain of preference optimization: Improving chain-of-thought reasoning in llms,

    X. Zhang, C. Du, T. Pang, Q. Liu, W. Gao, and M. Lin, “Chain of preference optimization: Improving chain-of-thought reasoning in llms,” Advances in Neural Information Processing Systems , vol. 37, pp. 333– 356, 2024

    2024

  33. [41]

    Gpt-4 technical report,

    J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F. L. Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkat et al. , “Gpt-4 technical report,” arXiv preprint arXiv:2303.08774 , 2023

    2023 arXiv

  34. [42]

    Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning,

    D. Guo, D. Yang, H. Zhang, J. Song, R. Zhang, R. Xu, Q. Zhu, S. Ma, P. Wang, X. Bi et al., “Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning,” arXiv preprint arXiv:2501.12948 , 2025

    2025 arXiv

  35. [43]

    Automatic actionable information processing and trust man- agement towards safer internet of things,

    M. Janiszewski, A. Felkner, P. Lewandowski, M. Rytel, and H. Ro- manowski, “Automatic actionable information processing and trust man- agement towards safer internet of things,” Sensors, vol. 21, no. 13, p. 4359, 2021

    2021

  36. [44]

    Reasoning models don’t always say what they think

    Y . Chen, J. Benton, A. Radhakrishnan, J. U. C. Denison, J. Schulman, A. Somani, P. Hase, M. W. F. R. V . Mikulik, S. Bowman, J. L. J. Kaplan et al. , “Reasoning models don’t always say what they think.” [Online]. Available: https://assets.anthropic.com/m/71876fabef0f0ed4/ ori...

  37. [45]

    Privacy- preserving data in iot-based cloud systems: A comprehensive survey with ai integration,

    D. Dhinakaran, S. Sankar, D. Selvaraj, and S. E. Raja, “Privacy- preserving data in iot-based cloud systems: A comprehensive survey with ai integration,” arXiv preprint arXiv:2401.00794 , 2024

    2024 arXiv

  38. [46]

    Internet of things (iot) of smart homes: privacy and security,

    T. Magara and Y . Zhou, “Internet of things (iot) of smart homes: privacy and security,” Journal of Electrical and Computer Engineering , vol. 2024, no. 1, p. 7716956, 2024

    2024

Pith tools

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