Pith. sign in

REVIEW 4 major objections 4 minor 37 references

Complex System Diagnostics Using a Knowledge Graph-Informed and Large Language Model-Enhanced Framework

T0 review · 4 major / 4 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 AI turns plant manuals into fault-diagnosis graphs with 90%+ accuracy

desk verdict A credible engineering framework for LLM-assisted DML construction, but the evaluation does not support the safety-critical readiness claim. read the letter →

arxiv 2505.21291 v1 pith:RVVXDY3J submitted 2025-05-27 cs.AI

Saman Marandi , Yu-Shu Hu , Mohammad Modarres This is my paper · ORCID
classification cs.AI
keywords LargeLanguageModelsKnowledgeGraphsDynamicMasterLogicFaultdiagnosticsPromptchainingGraph-basedRetrieval-AugmentedGenerationNuclearpowerplantsafetyFunctionalmodeling
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

This paper claims that a gated prompt-chaining workflow can build a Dynamic Master Logic (DML) knowledge graph directly from system documentation, and that an LLM agent can then answer natural-language diagnostic queries by choosing between structured graph tools and retrieval-based explanation. The authors argue this makes functional modeling practical for safety-critical systems like nuclear power plants, where manually building and maintaining DML models is costly and error-prone. They report a case study on an auxiliary feedwater system with over 90% extraction accuracy for key KG elements and consistent tool and argument extraction across five repeated runs. The promise is that combining LLM flexibility with graph-grounded reasoning reduces hallucination and yields traceable, interpretable diagnostics.

What carries the argument

The central mechanism is the KG-DML, a knowledge graph encoding a Dynamic Master Logic hierarchy, where goals, functions, subfunctions, components, and success conditions are linked by AND/OR gates. Two coordinated LLM components operate on this graph: a gated prompt-chaining workflow that builds the graph from text, and an LLM agent that decides between executing external propagation tools or retrieving subgraphs for explanation. The upward tool evaluates each success condition as $\sum_i P(\text{Success}_j \mid \text{State}_i)\,P(\text{State}_i \mid \text{Data})$ and aggregates component-level probabilities through the logical gates; the downward tool constructs minimal success path-sets by recursive AND/OR traversal. This design confines the language model to query interpretation and response presentation, keeping the actual reasoning grounded in graph logic.

What would settle it

A concrete test would be to run the construction pipeline on a system whose documentation deliberately omits one success condition that is known to be required, then check whether upward propagation still returns high success for the affected function; if it does, element-level accuracy masks a broken reasoning chain. A second check is to compare generated cut-sets and path-sets against expert-built DML baselines on the same system and see whether any missing gate changes a minimal success path.

Watch

Extended reading notes

Core claim

The central claim is that one coordinated framework can automate both the construction and the interactive use of DML functional models. System documentation is passed through a sequence of LLM calls, each followed by a validation gate, producing a hierarchical knowledge graph whose nodes are goals, functions, subfunctions, components, and success conditions connected by AND/OR gates. An LLM agent classifies each user query as diagnostic or interpretive: diagnostic queries trigger external tools that propagate success probabilities upward or generate minimal success path-sets downward through the graph, while interpretive queries retrieve and embed relevant graph segments into the prompt via Graph-RAG. Across five runs, element-level extraction accuracy stayed above 90% for every KG element type, and on a 60-query test set the agent classified task type and extracted valid tool inputs with near-perfect consistency.

Load-bearing premise

The reported accuracy figures assume that the manually defined ground truth used in Section 6.1 is correct and complete, and that element-level agreement with it is a valid proxy for diagnostic reliability in safety-critical use.

Editorial extensions

If this is right

  • Functional DML models for complex plants could be built from existing documentation rather than hand-crafted, reducing cost and enabling more widespread diagnostic modeling.
  • Because diagnostic queries route through external graph tools, the language model's reasoning is grounded in the KG rather than free-form generation, lowering the risk of fabricated fault paths.
  • The repeatability across five runs, with extraction accuracy above 90% for gates and success conditions, implies the construction pipeline is stable enough to use with human oversight.
  • Upward and downward propagation give operators concrete answers about which failures matter and which components are required for a function to succeed, both directly traceable through the graph.
  • State probabilities and success-condition attributes stored in the KG allow the framework to incorporate expert judgment or operational data, so the model can be refined without rebuilding the graph.

Reading between the lines

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

  • The gated prompt-chaining design is a template for other structured-knowledge extraction tasks, such as generating fault trees or event trees from maintenance logs, wherever strict output schemas and error costs make validation gates valuable.
  • A stronger validity test than element accuracy is end-to-end diagnostic comparison: benchmark the generated model's cut-sets and path-sets against expert-built DML logic on the same plant, which the authors explicitly identify as future work.
  • If the framework scales beyond the auxiliary feedwater case, the dominant failure mode may shift from extraction errors to source-documentation gaps, since the pipeline can only structure what the text supplies.
  • The current evaluation metric of element-level agreement does not capture whether an omitted gate silently severs a fault-propagation path, so deployment would need a coverage check that flags unmodeled dependencies.
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 / 4 minor

Summary. The paper proposes a two-stage LLM-and-knowledge-graph framework for building and using Dynamic Master Logic (DML) diagnostic models from natural-language system documentation. In the construction stage, a gated prompt-chaining workflow extracts goals, functions, subfunctions, components, success conditions, and AND/OR gates into a Neo4j knowledge graph (KG-DML); in the interaction stage, an LLM agent classifies user queries and either invokes external tools for upward/downward propagation over the KG or uses Graph-RAG for explanatory queries. A case study on a simplified auxiliary feedwater system is evaluated in five construction runs and five runs of a 60-query interaction set, reporting element-level extraction accuracies above 93% and tool/argument extraction accuracies of 96-100%. The paper concludes that these results support use in safety-critical diagnostics, while acknowledging in Section 7.1 that element-level accuracy may miss critical nodes and that scalability is untested.

Significance. The framework addresses a real bottleneck: DML model construction and interrogation currently require heavy manual expertise. The paper's strengths include a transparent gated pipeline with full prompts in the appendix, a released code repository, explicit reporting of hallucinated elements, and a candid limitations section. If the accuracy figures were tied to diagnostic reasoning performance rather than element counts, the approach would be a useful step toward automated functional modeling. As written, however, the evidence is limited to one small case study with self-constructed ground truth, no baseline comparison, no confidence intervals, and no external validation, so the significance of the stated 'over 90% accuracy' is uncertain.

major comments (4)
  1. [§6.1, Table 1; §7.1] The central claim that the framework's accuracy 'supports its use in safety-critical diagnostics' is not supported by the chosen metric. The authors themselves state that the evaluation 'does not capture the semantic impact of missing critical nodes' and that omitted gates, subfunctions, or success conditions can break logical chains. Element-level extraction accuracy is therefore not a valid proxy for preservation of diagnostic reasoning. The paper should add a graph-level evaluation—for example, comparing the generated minimal success path sets or cut sets against an expert-engineered baseline—and should temper the safety-critical conclusion until such evidence is provided.
  2. [§5.2, §6.2, Table 2] The agent evaluation uses a curated set of 60 queries built by the authors, with correctness labels assigned by the same team, and reports no inter-rater reliability, no per-run variance, and no adversarial or paraphrased inputs. Section 7.1 concedes that the query set may not reflect edge cases or linguistic variation. The near-perfect classification and extraction rates could be partly an artifact of favorable query construction. I ask for an independent or at least blinded query set, a clear scoring rubric, and a report of per-run counts and failure cases.
  3. [§4.2, Eq. (1), Figs. 6-7] The diagnostic tools' outputs depend on the state likelihoods and conditional success probabilities in Equation (1), and on a 'predefined threshold' that is never specified; the evaluation, however, tests only tool selection and argument extraction, not whether the returned diagnostic conclusions are correct. Without fault-injection experiments in which true fault scenarios are known, the 'diagnostic insights' produced by upward and downward propagation remain unvalidated. At minimum, the paper should report the threshold used and run a scenario-based check of the propagated probabilities or success paths.
  4. [§5.1, §6.1] The construction evaluation consists of five runs on one small system with no baseline or ablation. Since the gated prompt-chaining workflow is the proposed innovation, its contribution to accuracy should be tested against a no-gate or single-pass extraction baseline, and per-run element counts rather than only averages should be reported so that variability in hallucinated elements (e.g., logical gates, standard deviation) can be assessed.
minor comments (4)
  1. [§4.2] The statement that the LLM agent was 'fine-tuned on a dataset' lacks details (dataset size, split, base model, fine-tuning procedure). Add these to the appendix for reproducibility.
  2. [§5.2, Eq. (1)] Clarify whether P(Success_j | State_i) is assumed conditionally independent of Data given State_i, and define the 'predefined threshold' used in upward propagation.
  3. [§5.3, Figure 8] The example interface is illustrative; adding one actual user query with the tool selected, the tool's output, and the final LLM-generated answer would make the interaction concrete.
  4. [§7.2] The sentence 'The proposed framework was validated through comprehensive evaluations' overstates what is reported in Section 6; I suggest aligning the conclusion with the limitations stated in Section 7.1.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: KG-DML construction is benchmarked against source documentation, and no output quantity reduces to a fitted input or to the framework's own outputs.

full rationale

The paper's construction pipeline (prompt chaining with LLM validation gates) is evaluated by cross-checking extracted DML elements against the original system description and a manually defined ground-truth count, not against the pipeline's own output or against parameters fitted from the benchmark (Table 1). Equation (1) is the law of total probability applied to stored state/success-condition attributes; upward and downward propagation are deterministic graph traversals with no free parameters. The LLM agent's tool selection and argument extraction are scored against a manually built 60-query test set; although the fine-tuning dataset and test set are not explicitly stated to be disjoint, the paper does not claim a prediction from a fitted input, so any possible leakage is an evaluation-design concern, not a demonstrated circular step. The DML modeling paradigm is prior work by the same authors and is cited as the underlying representational framework, but it is not invoked as a uniqueness theorem and does not by itself force the reported extraction accuracies. The authors' Section 7.1 candidly concedes that element-level accuracy does not capture missing-node semantic impact; that is a validity limitation, not circularity. No equation or claimed result reduces by construction to its own input, so the analysis is self-contained against an external documentation reference.

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

No invented physical entities are introduced. The KG-DML is a structured representation, not a new physical object.

free parameters (1)
  • impact threshold
    In Section 5.2, the upward propagation tool declares a node 'impacted' if its success probability falls below a predefined threshold, but the threshold value is not specified.
assumptions (3)
  • domain assumption The DML logical gates (AND/OR) correctly represent the success logic of the studied auxiliary feedwater system.
    The entire KG-DML is built on this hierarchy, and accuracy is measured against it. Section 5.1.
  • domain assumption Component success probabilities are conditionally independent for aggregation through AND gates.
    The upward propagation algorithm computes gate-level success as a product of child probabilities, which assumes independence. This is not stated or justified in Section 5.2.
  • domain assumption The LLM-based validation gates catch most hallucinations and structural errors.
    The pipeline relies on the gates to filter bad extractions, but Section 7.1 concedes that hallucinations may pass through.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Complex System Diagnostics Using a Knowledge Graph-Informed and Large Language Model-Enhanced Framework." pith.science (2026). https://pith.science/paper/RVVXDY3J

@misc{pith2026250521291,
  author       = {Pith},
  title        = {Pith review of: Complex System Diagnostics Using a Knowledge Graph-Informed and Large Language Model-Enhanced Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RVVXDY3J}},
  note         = {Machine review of arXiv:2505.21291}
}
read the original abstract

In this paper, we present a novel diagnostic framework that integrates Knowledge Graphs (KGs) and Large Language Models (LLMs) to support system diagnostics in high-reliability systems such as nuclear power plants. Traditional diagnostic modeling struggles when systems become too complex, making functional modeling a more attractive approach. Our approach introduces a diagnostic framework grounded in the functional modeling principles of the Dynamic Master Logic (DML) model. It incorporates two coordinated LLM components, including an LLM-based workflow for automated construction of DML logic from system documentation and an LLM agent that facilitates interactive diagnostics. The generated logic is encoded into a structured KG, referred to as KG-DML, which supports hierarchical fault reasoning. Expert knowledge or operational data can also be incorporated to refine the model's precision and diagnostic depth. In the interaction phase, users submit natural language queries, which are interpreted by the LLM agent. The agent selects appropriate tools for structured reasoning, including upward and downward propagation across the KG-DML. Rather than embedding KG content into every prompt, the LLM agent distinguishes between diagnostic and interpretive tasks. For diagnostics, the agent selects and executes external tools that perform structured KG reasoning. For general queries, a Graph-based Retrieval-Augmented Generation (Graph-RAG) approach is used, retrieving relevant KG segments and embedding them into the prompt to generate natural explanations. A case study on an auxiliary feedwater system demonstrated the framework's effectiveness, with over 90% accuracy in key elements and consistent tool and argument extraction, supporting its use in safety-critical diagnostics.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 30 canonical work pages

  1. [1]

    Function modeling using the system state flow diagram,

    U. Yildirim, F. Campean, and H. Williams, “Function modeling using the system state flow diagram,” Artif. Intell. Eng. Des. Anal. Manuf., vol. 31, no. 4, pp. 413–435, Nov. 2017, doi: 10.1017/S0890060417000294

    work page doi:10.1017/s0890060417000294 2017

  2. [2]

    A Comparison of Three Functional Modeling Methods,

    M. Modarres, R. Irehvije, and M. Lind, “A Comparison of Three Functional Modeling Methods,” Jun. 1995

    work page 1995

  3. [3]

    Time-dependent system knowledge representation based on dynamic master logic diagrams,

    Y.-S. Hu and M. Modarres, “Time-dependent system knowledge representation based on dynamic master logic diagrams,” Control Eng. Pract., vol. 4, no. 1, pp. 89–98, Jan. 1996, doi: 10.1016/0967- 0661(95)00211-5

    work page doi:10.1016/0967- 1996

  4. [4]

    Attention Is All You Need,

    A. Vaswani et al., “Attention Is All You Need,” Aug. 02, 2023, arXiv: arXiv:1706.03762. doi: 10.48550/arXiv.1706.03762

    arXiv doi:10.48550/arxiv.1706.03762 2023

  5. [5]

    Knowledge Graphs

    A. Hogan et al., “Knowledge Graphs,” 2020, doi: 10.48550/ARXIV .2003.02320

    work page Pith review arXiv doi:10.48550/arxiv.2003.02320 2020

  6. [6]

    C. A. Ericson, Hazard analysis techniques for system safety, 2. ed. Hoboken, NJ: Wiley, 2016

    work page 2016

  7. [7]

    Fault tree handbook with aerospace applications,

    M. Stamatelatos and W. E. Vesley, “Fault tree handbook with aerospace applications,” 2002. [Online]. Available: https://api.semanticscholar.org/CorpusID:61105226

    work page 2002

  8. [8]

    The application of event tree analysis in a work accident at maintenance operations,

    R. Mareş and M. P. Stelea, “The application of event tree analysis in a work accident at maintenance operations,” MATEC Web Conf., vol. 121, p. 11013, 2017, doi: 10.1051/matecconf/201712111013

    work page arXiv 2017

Show all 37 references
  1. [9]

    Event-tree analysis using binary decision diagrams,

    J. D. Andrews and S. J. Dunnett, “Event-tree analysis using binary decision diagrams,” IEEE Trans. Reliab., vol. 49, no. 2, pp. 230–238, Jun. 2000, doi: 10.1109/24.877343

    doi

  2. [10]

    A new risk assessment method based on belief rule base and fault tree analysis,

    H.-L. Zhu, S.-S. Liu, Y .-Y . Qu, X.-X. Han, W. He, and Y . Cao, “A new risk assessment method based on belief rule base and fault tree analysis,” Proc. Inst. Mech. Eng. Part O J. Risk Reliab., vol. 236, no. 3, pp. 420–438, Jun. 2022, doi: 10.1177/1748006X211011457

    2022 doi

  3. [11]

    Complex system reliability modelling with Dynamic Object Oriented Bayesian Networks (DOOBN),

    P. Weber and L. Jouffe, “Complex system reliability modelling with Dynamic Object Oriented Bayesian Networks (DOOBN),” Reliab. Eng. Syst. Saf., vol. 91, no. 2, pp. 149–162, Feb. 2006, doi: 10.1016/j.ress.2005.03.006

    2006 doi

  4. [12]

    Application of goal tree-success tree model as the knowledge-base of operator advisory systems,

    I. S. Kim and M. Modarres, “Application of goal tree-success tree model as the knowledge-base of operator advisory systems,” Nucl. Eng. Des., vol. 104, no. 1, pp. 67–81, Oct. 1987, doi: 10.1016/0029- 5493(87)90304-9

    1987 doi

  5. [13]

    Challenges in Functional Modelling for Safety and Risk Analysis,

    J. Wu, X. Zhang, M. Song, and M. Lind, “Challenges in Functional Modelling for Safety and Risk Analysis,” in Proceeding of the 33rd European Safety and Reliability Conference, Research Publishing Services, 2023, pp. 1892–1899. doi: 10.3850/978-981-18-8071-1_P132-cd

    2023 doi

  6. [14]

    Evaluating system behavior through Dynamic Master Logic Diagram (DMLD) modeling,

    Y.-S. Hu and M. Modarres, “Evaluating system behavior through Dynamic Master Logic Diagram (DMLD) modeling,” Reliab. Eng. Syst. Saf., vol. 64, no. 2, pp. 241–269, May 1999, doi: 10.1016/S0951-8320(98)00066-0

    1999 doi

  7. [15]

    Function-centered modeling of engineering systems using the goal tree–success tree technique and functional primitives,

    M. Modarres and S. W. Cheon, “Function-centered modeling of engineering systems using the goal tree–success tree technique and functional primitives,” Reliab. Eng. Syst. Saf., vol. 64, no. 2, pp. 181–200, May 1999, doi: 10.1016/S0951-8320(98)00062-3

    1999 doi

  8. [16]

    Logic-Based Hierarchies for Modeling Behavior of Complex Dynamic Systems with Applications,

    Y.-S. Hu and M. Modarres, “Logic-Based Hierarchies for Modeling Behavior of Complex Dynamic Systems with Applications,” in Fuzzy Systems and Soft Computing in Nuclear Engineering, vol. 38, D. Ruan, Ed., in Studies in Fuzziness and Soft Computing, vol. 38. , Heidelberg: Physica...

    doi

  9. [17]

    Functional modeling of complex systems with applications,

    M. Modarres, “Functional modeling of complex systems with applications,” in Annual Reliability and Maintainability. Symposium. 1999 Proceedings (Cat. No.99CH36283), Washington, DC, USA: IEEE, 1999, pp. 418–425. doi: 10.1109/RAMS.1999.744153

    1999

  10. [18]

    Reliability assessment of generic geared wind turbines by GTST- MLD model and Monte Carlo simulation,

    Y . F. Li, S. Valla, and E. Zio, “Reliability assessment of generic geared wind turbines by GTST- MLD model and Monte Carlo simulation,” Renew. Energy, vol. 83, pp. 222–233, Nov. 2015, doi: 10.1016/j.renene.2015.04.035

    2015 doi

  11. [19]

    Z. Hao, F. Di Maio, and E. Zio, “A sequential decision problem formulation and deep reinforcement learning solution of the optimization of O&M of cyber-physical energy systems (CPESs) for reliable and safe power production and supply,” Reliab. Eng. Syst. Saf., vol. 235, p. 109...

    2023

  12. [20]

    Simulation-Based Goal Tree Success Tree for the Risk Analysis of Cyber-Physical Systems

    F. D. Maio, “Simulation-Based Goal Tree Success Tree for the Risk Analysis of Cyber-Physical Systems”

  13. [21]

    Extended GTST‐MLD for Aerospace System Safety Analysis,

    C. Guo, S. Gong, L. Tan, and B. Guo, “Extended GTST‐MLD for Aerospace System Safety Analysis,” Risk Anal., vol. 32, no. 6, pp. 1060–1071, Jun. 2012, doi: 10.1111/j.1539-6924.2011.01718.x

    2012

  14. [22]

    Modarres and N

    M. Modarres and N. Kececi, Software Development Life Cycle Model to Ensure Software Quality. 1998

    1998

  15. [23]

    Integrated Economic Risk Management in a Nuclear Power Plant,

    R. N. M. Hunt and M. Modarres, “Integrated Economic Risk Management in a Nuclear Power Plant,” in Uncertainty in Risk Assessment, Risk Management, and Decision Making, V . T. Covello, L. B. Lave, A. Moghissi, and V . R. R. Uppuluri, Eds., Boston, MA: Springer US, 1987, pp. 435...

    1987 doi

  16. [24]

    Applications of Master Plant Logic Diagram (MPLD) PC-Based Program in Probabilistic Risk Assessment,

    M. Modarres, J. H. Zamanali, and J. Wang, “Applications of Master Plant Logic Diagram (MPLD) PC-Based Program in Probabilistic Risk Assessment,” Feb. 1991

    1991

  17. [25]

    Integrating LLMs for Explainable Fault Diagnosis in Complex Systems,

    A. J. Dave, T. N. Nguyen, and R. B. Vilim, “Integrating LLMs for Explainable Fault Diagnosis in Complex Systems,” Feb. 08, 2024, arXiv: arXiv:2402.06695. doi: 10.48550/arXiv.2402.06695

    2024 doi arXiv

  18. [26]

    FD-LLM: Large Language Model for Fault Diagnosis of Machines,

    H. A. A. M. Qaid, B. Zhang, D. Li, S.-K. Ng, and W. Li, “FD-LLM: Large Language Model for Fault Diagnosis of Machines,” Dec. 02, 2024, arXiv: arXiv:2412.01218. doi: 10.48550/arXiv.2412.01218

    2024 doi arXiv

  19. [27]

    Root-KGD: A Novel Framework for Root Cause Diagnosis Based on Knowledge Graph and Industrial Data,

    J. Chen, J. Qian, X. Zhang, and Z. Song, “Root-KGD: A Novel Framework for Root Cause Diagnosis Based on Knowledge Graph and Industrial Data,” Jun. 19, 2024, arXiv: arXiv:2406.13664. doi: 10.48550/arXiv.2406.13664

    2024 doi arXiv

  20. [28]

    An automatic machine fault identification method using the knowledge graph–embedded large language model,

    P. Wu, X. Mou, L. Gong, H. Tu, L. Qiu, and B. Yang, “An automatic machine fault identification method using the knowledge graph–embedded large language model,” Int. J. Adv. Manuf. Technol., Apr. 2025, doi: 10.1007/s00170-025-15555-2

    2025 doi

  21. [29]

    Research on knowledge graph-driven equipment fault diagnosis method for intelligent manufacturing,

    C. Cai, Z. Jiang, H. Wu, J. Wang, J. Liu, and L. Song, “Research on knowledge graph-driven equipment fault diagnosis method for intelligent manufacturing,” Int. J. Adv. Manuf. Technol., vol. 130, no. 9–10, pp. 4649–4662, Feb. 2024, doi: 10.1007/s00170-024-12998-x

    2024 doi

  22. [30]

    Joint Knowledge Graph and Large Language Model for Fault Diagnosis and Its Application in Aviation Assembly,

    P. Liu, L. Qian, X. Zhao, and B. Tao, “Joint Knowledge Graph and Large Language Model for Fault Diagnosis and Its Application in Aviation Assembly,” IEEE Trans. Ind. Inform., vol. 20, no. 6, pp. 8160–8169, Jun. 2024, doi: 10.1109/TII.2024.3366977

    2024

  23. [31]

    A Fault Analysis and Reasoning Method for Vehicle Information Systems Based on Knowledge Graphs,

    T. Sun, F. Zeng, and X. Liu, “A Fault Analysis and Reasoning Method for Vehicle Information Systems Based on Knowledge Graphs,” in 2024 IEEE 24th International Conference on Software Quality, Reliability, and Security Companion (QRS-C), Cambridge, United Kingdom: IEEE, Jul. 20...

    2024

  24. [32]

    Knowledge Graph-Based In-Context Learning for Advanced Fault Diagnosis in Sensor Networks,

    X. Xie, J. Wang, Y . Han, and W. Li, “Knowledge Graph-Based In-Context Learning for Advanced Fault Diagnosis in Sensor Networks,” Sensors, vol. 24, no. 24, p. 8086, Dec. 2024, doi: 10.3390/s24248086

    2024 doi

  25. [33]

    Neo4j GitHub Repository,

    Neo4j, Inc., “Neo4j GitHub Repository,” GitHub. [Online]. Available: https://github.com/neo4j/neo4j

  26. [34]

    Modarres, M

    M. Modarres, M. Kaminskiy, and V . Krivtsov, Reliability engineering and risk analysis: a practical guide, Third edition. Boca Raton: CRC Press, Taylor & Francis Group, CRC Press is an imprint of the Taylor & Francis Group, an informa business, 2017

    2017

  27. [35]

    Particle Filters: A Hands-On Tutorial,

    J. Elfring, E. Torta, and R. Van De Molengraft, “Particle Filters: A Hands-On Tutorial,” Sensors, vol. 21, no. 2, p. 438, Jan. 2021, doi: 10.3390/s21020438

    2021 doi

  28. [36]

    Doucet, N

    A. Doucet, N. Freitas, and N. Gordon, Eds., Sequential Monte Carlo Methods in Practice. New York, NY: Springer New York, 2001. doi: 10.1007/978-1-4757-3437-9. 21 Appendix Figure 9, Step 1 and Gate 1 Prompts Figure 10, Step 2 and Gate 2 Prompts 22 Figure 11, Step 3 and Gate 3 Prompts

    2001 doi

  29. [933]

    doi: 10.1109/QRS-C63300.2024.00123. 20

    2024

Pith tools

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