Pith. sign in

REVIEW 3 major objections 2 minor 50 references

Trusted Routing for Blockchain-Empowered UAV Networks via Multi-Agent Deep Reinforcement Learning

T0 review · 3 major / 2 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Blockchain-based trust management combined with multi-agent double deep Q-network routing reduces total delay in time-varying UAV networks with malicious nodes.

desk verdict The submitted full text is an unrelated ocean-SST paper, so the UAV routing claims can only be assessed from the abstract, which is coherent but empirically unverifiable. read the letter →

arxiv 2508.00938 v1 pith:IELROLFO submitted 2025-07-31 eess.SY cs.AIcs.CRcs.SY

Ziye Jia , Sijie He , Qiuming Zhu , Wei Wang , Qihui Wu , Zhu Han This is my paper
classification eess.SYcs.AIcs.CRcs.SY
keywords UAVnetworksblockchaintrustmanagementmulti-agentdeepreinforcementlearningroutingdelayminimizationDec-POMDPmaliciousnodes
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 argues that routing in time-varying UAV networks with malicious nodes can be made both secure and fast by combining a blockchain-based trust management mechanism with a multi-agent deep reinforcement learning router. The routing problem is posed as an integer linear program that is intractable, then recast as a decentralized partially observable Markov decision process solvable by a multi-agent double deep Q-network. The trust mechanism dynamically scores UAVs and excludes low-trust nodes from forwarding, which the paper claims is why total delay drops by 13.39% over multi-agent PPO, 12.74% over multi-agent DQN, and 16.6% over routing without trust management. A sympathetic reader would care because UAV networks used in surveillance and disaster rescue are exposed to malicious interference, and this is a concrete proposal for keeping them reliable.

What carries the argument

The blockchain-based trust management mechanism (BTMM) is the central object; it maintains a tamper-resistant ledger of UAV trust values and uses an improved practical Byzantine fault tolerance consensus to update that ledger when UAVs join or leave. The second carrying element is the multi-agent double deep Q-network, which learns routing decisions from each UAV's partially observable state. Together they turn a global integer linear routing program into a decentralized learning problem whose objective is total delay.

What would settle it

Run the same routing algorithm in a testbed with real radios or in a higher-fidelity simulator while varying node speed and the number of malicious UAVs; if the delay gap against the no-BTMM baseline shrinks or reverses under those conditions, the claimed advantage does not generalize.

Watch

Extended reading notes

Core claim

The central claim is that a blockchain-based trust management mechanism (BTMM) integrated with a multi-agent double deep Q-network routing algorithm reduces total delay in a time-varying UAV network containing attacked nodes. The paper formulates routing as an integer linear program, proves it intractable for direct solution, and reformulates it as a decentralized partially observable Markov decision process that respects each UAV's local observability. BTMM dynamically evaluates UAV trust values and identifies low-trust UAVs, and a consensus UAV update mechanism improves the practical Byzantine fault tolerance algorithm used in the blockchain. Simulations with attacked UAVs report delay reductions of 13.39%, 12.74%, and 16.6% relative to multi-agent PPO, multi-agent DQN, and no-BTMM baselines, respectively.

Load-bearing premise

The simulation scenario—its topology, number of UAVs, attack model, and traffic pattern—is assumed to stand in for real UAV networks, so the measured delay reductions are expected to carry over outside the simulator.

Editorial extensions

If this is right

  • If the delay reductions replicate outside the simulator, UAV fleets can route around malicious nodes without a central controller.
  • The trust ledger makes routing decisions auditable, so a compromised UAV's influence can be traced and revoked.
  • The Dec-POMDP reformulation lets the same approach scale to larger swarms where full observability is impossible.
  • The consensus UAV update mechanism addresses the practical difficulty of maintaining a blockchain in a topology that changes over time.

Reading between the lines

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

  • A testable extension would be to vary the fraction of malicious nodes and measure whether the delay advantage of BTMM grows or shrinks; the paper's abstract does not report this sensitivity.
  • The trust-value signals could plausibly be reused for other decisions, such as task allocation or spectrum access, which the paper does not explore.
  • The reported 13-16% delay reductions are only as good as the simulator's attack model; a hardware-in-the-loop test would be a stronger demonstration.
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 / 2 minor

Summary. The manuscript, as submitted, consists of an abstract on trusted routing for blockchain-empowered UAV networks using multi-agent deep reinforcement learning, followed by a full text that is an unrelated paper about ocean knowledge graphs and sea-surface temperature prediction (OKG-LLM). The abstract claims that a blockchain-based trust management mechanism (BTMM) combined with a multi-agent double deep Q-network reduces total delay by 13.39%, 12.74%, and 16.6% over multi-agent PPO, multi-agent DQN, and no-BTMM baselines, respectively. However, the full text contains none of the technical content—no Dec-POMDP formulation, no reward design, no algorithm details, no blockchain or consensus mechanism, and no simulation setup—needed to evaluate or reproduce these claims. The central contribution of the paper is therefore absent from the submitted manuscript.

Significance. If the work described in the abstract were fully present, it would address a timely and relevant problem: secure routing in dynamic UAV networks with malicious nodes. The combination of blockchain-based trust management with multi-agent deep reinforcement learning is a plausible and potentially valuable direction. However, the submitted manuscript does not contain that work. The only assessable material is the abstract, which offers empirical percentages without any methodological support. There is no derivable formula, no reproducible code, no formal model, and no simulation protocol. As a result, the paper cannot contribute to the field in its current form; the claimed significance is unverifiable from the submission.

major comments (3)
  1. [Full Text vs. Abstract] The full text of the submission is entirely the OKG-LLM paper on ocean knowledge graphs for sea-surface temperature prediction, which has no connection to the abstract's UAV routing topic. None of the claimed components—BTMM, the consensus UAV update mechanism, the multi-agent double DQN, or the Dec-POMDP reformulation—appear anywhere in the supplied text. This is a load-bearing mismatch: the central claim of the abstract is unsupported by any of the manuscript's content, and the paper as submitted cannot be reviewed as a technical contribution.
  2. [Abstract] The reported delay reductions of 13.39%, 12.74%, and 16.6% are given without any simulation parameters, such as the number of UAVs, network topology, mobility model, attack ratio, traffic demand, block interval, or number of random seeds. The abstract also does not state whether 'total delay' includes the latency overhead of the blockchain consensus process. Without this information and without error bars or statistical significance tests, the empirical claim cannot be checked and may fall within run-to-run variance.
  3. [Abstract (Problem Formulation)] The abstract states that the routing problem is first formulated as an integer linear program and then reformulated as a decentralized partially observable Markov decision process, but the manuscript contains no equations, no formal definitions of state, action, observation, reward, or transition, and no algorithm pseudocode. This omission prevents any assessment of whether the multi-agent double DQN algorithm is correctly specified, whether it indeed addresses the claimed problem, and whether the baselines are compared fairly.
minor comments (2)
  1. [Abstract] There are several language issues in the abstract: 'proximal policy optimal algorithms' should read 'proximal policy optimization algorithms'; 'decreases by 13.39%, 12.74%, and 16.6% than' should be 'decreases by 13.39%, 12.74%, and 16.6% compared with'; and there are missing spaces after commas throughout the abstract.
  2. [Full Text] The full text is a self-contained paper on a different topic; even taken on its own terms, its formatting is inconsistent with the submission (e.g., variable names are garbled as symbols). More importantly, its relation to the abstract is never explained, and it does not cite or build on the claimed UAV routing work.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identifiable; the supplied full text is an unrelated SST-prediction paper, leaving the UAV routing derivation chain unavailable for inspection.

full rationale

The abstract describes an empirical simulation study with no equations, fitted parameters, or cited uniqueness theorems; the only quantitative claims are reported delay reductions over baselines in simulation. No circular reduction can be exhibited from the abstract alone. The supplied full text is a different manuscript (OKG-LLM for sea surface temperature prediction, arXiv:2508.00933v1) and contains no material on BTMM, UAV routing, multi-agent double DQN, or blockchain consensus; therefore the derivation chain for the central UAV-routing claim cannot be inspected. Absence of evidence is not circularity: I cannot quote an equation or parameter that is defined in terms of the claimed output. Per hard rule 1, speculation about hidden hyperparameter tuning or post-hoc simulation selection is not counted. The score is therefore 0, with the caveat that the central claim's support is unverifiable from this record.

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

Only assumptions inferable from the abstract are listed. The missing full text prevents a complete inventory of free parameters and axioms.

free parameters (3)
  • DQN hyperparameters = not specified
    Learning rate, discount factor, exploration schedule, and network architecture are not given in the abstract and would be needed to reproduce the MADQN routing results.
  • Trust evaluation thresholds in BTMM = not specified
    The abstract says BTMM dynamically evaluates trust values and identifies low-trust UAVs, but the thresholds or update rates that determine when a UAV is considered untrusted are not stated.
  • Simulation scenario parameters = not specified
    Number of UAVs, network area, mobility model, message rates, and attack intensity are not listed, yet the reported delay reductions depend on them.
assumptions (3)
  • domain assumption The routing problem can be modeled as a Dec-POMDP with local observability.
    The abstract states the problem is reformulated into a decentralized partially observable Markov decision process; this assumes each UAV has partial observations and that the Markov property holds for the network state.
  • domain assumption Blockchain consensus and trust scoring do not add prohibitive delay or communication overhead.
    The proposed BTMM relies on a practical Byzantine fault tolerance consensus variant; for the delay reduction claim to hold, the overhead of maintaining the blockchain must be less than the routing gains.
  • ad hoc to paper Malicious nodes can be detected through behavior-based trust values.
    The abstract assumes that low-trust UAVs can be identified from dynamic trust evaluation, which presumes a correlation between honest behavior and delay-relevant routing actions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Trusted Routing for Blockchain-Empowered UAV Networks via Multi-Agent Deep Reinforcement Learning." pith.science (2026). https://pith.science/paper/IELROLFO

@misc{pith2026250800938,
  author       = {Pith},
  title        = {Pith review of: Trusted Routing for Blockchain-Empowered UAV Networks via Multi-Agent Deep Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IELROLFO}},
  note         = {Machine review of arXiv:2508.00938}
}
abstract

Due to the high flexibility and versatility, unmanned aerial vehicles (UAVs) are leveraged in various fields including surveillance and disaster rescue.However, in UAV networks, routing is vulnerable to malicious damage due to distributed topologies and high dynamics. Hence, ensuring the routing security of UAV networks is challenging. In this paper, we characterize the routing process in a time-varying UAV network with malicious nodes. Specifically, we formulate the routing problem to minimize the total delay, which is an integer linear programming and intractable to solve. Then, to tackle the network security issue, a blockchain-based trust management mechanism (BTMM) is designed to dynamically evaluate trust values and identify low-trust UAVs. To improve traditional practical Byzantine fault tolerance algorithms in the blockchain, we propose a consensus UAV update mechanism. Besides, considering the local observability, the routing problem is reformulated into a decentralized partially observable Markov decision process. Further, a multi-agent double deep Q-network based routing algorithm is designed to minimize the total delay. Finally, simulations are conducted with attacked UAVs and numerical results show that the delay of the proposed mechanism decreases by 13.39$\%$, 12.74$\%$, and 16.6$\%$ than multi-agent proximal policy optimal algorithms, multi-agent deep Q-network algorithms, and methods without BTMM, respectively.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

50 extracted references · 1 canonical work pages

  1. [1]

    Accurate medium-range global weather forecasting with 3d neural networks,

    K. Bi, L. Xie, H. Zhang, X. Chen, X. Gu, and Q. Tian, “Accurate medium-range global weather forecasting with 3d neural networks,” Nature, vol. 619, no. 7970, pp. 533–538, 2023

    2023

  2. [2]

    A survey on time-series pre-trained models,

    Q. Ma, Z. Liu, Z. Zheng, Z. Huang, S. Zhu, Z. Yu, and J. T. Kwok, “A survey on time-series pre-trained models,” IEEE Transactions on Knowledge and Data Engineering , 2024

    2024

  3. [3]

    Spatial-temporal data mining for ocean science: Data, methodologies, and opportunities,

    H. Yang, W. Li, S. Wang, H. Li, J. Guan, S. Zhou, and J. Cao, “Spatial-temporal data mining for ocean science: Data, methodologies, and opportunities,” arXiv preprint arXiv:2307.10803 , 2023

    arXiv 2023

  4. [4]

    Ocean currents show global intensification of weak tropical cyclones,

    G. Wang, L. Wu, W. Mei, and S.-P. Xie, “Ocean currents show global intensification of weak tropical cyclones,” Nature, vol. 611, no. 7936, pp. 496–500, 2022

    2022

  5. [5]

    El ni˜no–southern oscillation complexity,

    A. Timmermann, S.-I. An, J.-S. Kug, F.-F. Jin, W. Cai, A. Capotondi, K. M. Cobb, M. Lengaigne, M. J. McPhaden, M. F. Stuecker et al., “El ni˜no–southern oscillation complexity,” Nature, vol. 559, no. 7715, pp. 535–545, 2018

    2018

  6. [6]

    The circulation of the eastern tropical pacific: A review,

    W. S. Kessler, “The circulation of the eastern tropical pacific: A review,” Progress in Oceanography, vol. 69, no. 2-4, pp. 181–217, 2006

    2006

  7. [7]

    Improving language understanding by generative pre-training,

    A. Radford, K. Narasimhan, T. Salimans, I. Sutskever et al., “Improving language understanding by generative pre-training,” 2018. JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. XX, XX 11

    2018

  8. [8]

    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

    arXiv 2023

Show all 50 references
  1. [9]

    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

  2. [10]

    Mixed-order relation-aware recurrent neural networks for spatio-temporal forecast- ing,

    Y . Liang, K. Ouyang, Y . Wang, Z. Pan, Y . Yin, H. Chen, J. Zhang, Y . Zheng, D. S. Rosenblum, and R. Zimmermann, “Mixed-order relation-aware recurrent neural networks for spatio-temporal forecast- ing,” IEEE Transactions on Knowledge and Data Engineering , vol. 35, no. 9, pp...

    2022

  3. [11]

    One fits all: Power general time series analysis by pretrained lm,

    T. Zhou, P. Niu, L. Sun, R. Jin et al. , “One fits all: Power general time series analysis by pretrained lm,” Advances in neural information processing systems, vol. 36, pp. 43 322–43 355, 2023

    2023

  4. [12]

    Unitime: A language-empowered unified model for cross-domain time series forecasting,

    X. Liu, J. Hu, Y . Li, S. Diao, Y . Liang, B. Hooi, and R. Zimmermann, “Unitime: A language-empowered unified model for cross-domain time series forecasting,” in Proceedings of the ACM Web Conference 2024 , 2024

    2024

  5. [13]

    Time-LLM: Time series forecasting by reprogramming large language models,

    M. Jin, S. Wang, L. Ma, Z. Chu, J. Y . Zhang, X. Shi, P.-Y . Chen, Y . Liang, Y .-F. Li, S. Pan, and Q. Wen, “Time-LLM: Time series forecasting by reprogramming large language models,” in International Conference on Learning Representations (ICLR) , 2024

    2024

  6. [14]

    Position paper: What can large language models tell us about time series analysis,

    M. Jin, Y . Zhang, W. Chen, K. Zhang, Y . Liang, B. Yang, J. Wang, S. Pan, and Q. Wen, “Position paper: What can large language models tell us about time series analysis,” in International Conference on Machine Learning (ICML 2024) , 2024

    2024

  7. [15]

    Tempo: Prompt-based generative pre-trained transformer for time series forecasting,

    D. Cao, F. Jia, S. O. Arik, T. Pfister, Y . Zheng, W. Ye, and Y . Liu, “Tempo: Prompt-based generative pre-trained transformer for time series forecasting,” in The Twelfth International Conference on Learning Representations

  8. [16]

    Higrn: a hierarchical graph recurrent network for global sea surface temperature prediction,

    H. Yang, W. Li, S. Hou, J. Guan, and S. Zhou, “Higrn: a hierarchical graph recurrent network for global sea surface temperature prediction,” ACM Transactions on Intelligent Systems and Technology, vol. 14, no. 4, pp. 1–19, 2023

    2023

  9. [17]

    Cross- region graph convolutional network with periodicity shift adaptation for wide-area sst prediction,

    H. Peng, W. Li, C. Jin, Y . Zhang, J. Guan, H. Yang, and S. Zhou, “Cross- region graph convolutional network with periodicity shift adaptation for wide-area sst prediction,” ACM Transactions on Intelligent Systems and Technology, 2025

    2025

  10. [18]

    Mustc: A multi-stage spatio–temporal clustering method for uncovering the regionality of global sst,

    H. Peng, W. Li, C. Jin, H. Yang, and J. Guan, “Mustc: A multi-stage spatio–temporal clustering method for uncovering the regionality of global sst,” Atmosphere, vol. 14, no. 9, p. 1358, 2023

    2023

  11. [19]

    Physical knowledge-enhanced deep neural network for sea surface temperature prediction,

    Y . Meng, F. Gao, E. Rigall, R. Dong, J. Dong, and Q. Du, “Physical knowledge-enhanced deep neural network for sea surface temperature prediction,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–13, 2023

    2023

  12. [20]

    Multi-scale adaptive graph neural network for multivariate time series forecasting,

    L. Chen, D. Chen, Z. Shang, B. Wu, C. Zheng, B. Wen, and W. Zhang, “Multi-scale adaptive graph neural network for multivariate time series forecasting,” IEEE Transactions on Knowledge and Data Engineering , vol. 35, no. 10, pp. 10 748–10 761, 2023

    2023

  13. [21]

    Causal- former: An interpretable transformer for temporal causal discovery,

    L. Kong, W. Li, H. Yang, Y . Zhang, J. Guan, and S. Zhou, “Causal- former: An interpretable transformer for temporal causal discovery,” IEEE Transactions on Knowledge and Data Engineering , vol. 37, no. 1, pp. 102–115, 2025

    2025

  14. [22]

    On evaluating the predictability of sea surface temperature using entropy,

    C. Jin, H. Peng, H. Yang, W. Li, and J. Guan, “On evaluating the predictability of sea surface temperature using entropy,”Remote Sensing, vol. 15, no. 8, p. 1956, 2023

    1956

  15. [23]

    The ecmwf ensemble prediction system: Methodology and validation,

    F. Molteni, R. Buizza, T. N. Palmer, and T. Petroliagis, “The ecmwf ensemble prediction system: Methodology and validation,” Quarterly journal of the royal meteorological society , vol. 122, no. 529, pp. 73– 119, 1996

    1996

  16. [24]

    Spatiotemporal attention network for chl-a prediction with sparse multifactor observations,

    X. Jiang, Y . Liu, S. Wang, W. Li, and J. Guan, “Spatiotemporal attention network for chl-a prediction with sparse multifactor observations,” IEEE Geoscience and Remote Sensing Letters , vol. 22, pp. 1–5, 2025

    2025

  17. [25]

    Multivariate time series forecasting with dynamic graph neural odes,

    M. Jin, Y . Zheng, Y .-F. Li, S. Chen, B. Yang, and S. Pan, “Multivariate time series forecasting with dynamic graph neural odes,” IEEE Trans- actions on Knowledge and Data Engineering , vol. 35, no. 9, pp. 9168– 9180, 2022

    2022

  18. [26]

    Informer: Beyond efficient transformer for long sequence time-series forecasting,

    H. Zhou, S. Zhang, J. Peng, S. Zhang, J. Li, H. Xiong, and W. Zhang, “Informer: Beyond efficient transformer for long sequence time-series forecasting,” in Proceedings of the AAAI conference on artificial intel- ligence, vol. 35, no. 12, 2021, pp. 11 106–11 115

    2021

  19. [27]

    Fedformer: Frequency enhanced decomposed transformer for long-term series fore- casting,

    T. Zhou, Z. Ma, Q. Wen, X. Wang, L. Sun, and R. Jin, “Fedformer: Frequency enhanced decomposed transformer for long-term series fore- casting,” in International conference on machine learning . PMLR, 2022, pp. 27 268–27 286

    2022

  20. [28]

    Timemixer++: A general time series pattern machine for universal predictive analysis,

    S. Wang, J. Li, X. Shi, Z. Ye, B. Mo, W. Lin, S. Ju, Z. Chu, and M. Jin, “Timemixer++: A general time series pattern machine for universal predictive analysis,” ICLR, 2025

    2025

  21. [29]

    Are transformers effective for time series forecasting?

    A. Zeng, M. Chen, L. Zhang, and Q. Xu, “Are transformers effective for time series forecasting?” in Proceedings of the AAAI conference on artificial intelligence, vol. 37, no. 9, 2023, pp. 11 121–11 128

    2023

  22. [30]

    Frequency-domain mlps are more effective learners in time series forecasting,

    K. Yi, Q. Zhang, W. Fan, S. Wang, P. Wang, H. He, N. An, D. Lian, L. Cao, and Z. Niu, “Frequency-domain mlps are more effective learners in time series forecasting,” Advances in Neural Information Processing Systems, vol. 36, pp. 76 656–76 679, 2023

    2023

  23. [31]

    Llm4hrs: Llm-based spatiotemporal imputation model for highly sparse remote sensing data,

    S. Wang, W. Li, H. Yang, J. Guan, X. Liu, Y . Zhang, R. Qin, and S. Zhou, “Llm4hrs: Llm-based spatiotemporal imputation model for highly sparse remote sensing data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 63, pp. 1–17, 2025

    2025

  24. [32]

    Promptcast: A new prompt-based learning paradigm for time series forecasting,

    H. Xue and F. D. Salim, “Promptcast: A new prompt-based learning paradigm for time series forecasting,” arXiv preprint arXiv:2210.08964, 2022

    2022 arXiv

  25. [33]

    Urbangpt: Spatio-temporal large language models,

    Z. Li, L. Xia, J. Tang, Y . Xu, L. Shi, L. Xia, D. Yin, and C. Huang, “Urbangpt: Spatio-temporal large language models,” in Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2024, pp. 5351–5362

    2024

  26. [34]

    A foundation model of transcription across human cell types,

    X. Fu, S. Mo, A. Buendia, A. P. Laurent, A. Shao, M. d. M. Alvarez- Torres, T. Yu, J. Tan, J. Su, R. Sagatelian et al. , “A foundation model of transcription across human cell types,” Nature, pp. 1–9, 2025

    2025

  27. [35]

    Protgpt2 is a deep unsupervised language model for protein design,

    N. Ferruz, S. Schmidt, and B. H ¨ocker, “Protgpt2 is a deep unsupervised language model for protein design,” Nature communications , vol. 13, no. 1, p. 4348, 2022

    2022

  28. [36]

    K2: A foundation language model for geoscience knowledge understanding and utilization,

    C. Deng, T. Zhang, Z. He, Q. Chen, Y . Shi, Y . Xu, L. Fu, W. Zhang, X. Wang, C. Zhou et al. , “K2: A foundation language model for geoscience knowledge understanding and utilization,” in Proceedings of the 17th ACM International Conference on Web Search and Data Mining, 2024,...

    2024

  29. [37]

    Foundation model for material science,

    S. Takeda, A. Kishimoto, L. Hamada, D. Nakano, and J. R. Smith, “Foundation model for material science,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 13, 2023, pp. 15 376– 15 383

    2023

  30. [38]

    Leveraging network structure for efficient dynamic negative sampling in network embedding,

    C. Wang, Z. Zhu, P. Meng, and Y . Qiu, “Leveraging network structure for efficient dynamic negative sampling in network embedding,”Information Sciences, vol. 606, pp. 853–863, 2022

    2022

  31. [39]

    Fathomgpt: A natural language interface for interactively exploring ocean science data,

    N. Khanal, C. M. Yu, J.-C. Chiu, A. Chaudhary, Z. Zhang, K. Katija, and A. G. Forbes, “Fathomgpt: A natural language interface for interactively exploring ocean science data,” in Proceedings of the 37th Annual ACM Symposium on User Interface Software and Technology , ser. UIST...

    2024

  32. [40]

    Oceangpt: A large language model for ocean science tasks,

    Z. Bi, N. Zhang, Y . Xue, Y . Ou, D. Ji, G. Zheng, and H. Chen, “Oceangpt: A large language model for ocean science tasks,” inProceed- ings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , 2024, pp. 3357–3372

    2024

  33. [41]

    Knowledge graph quality management: a compre- hensive survey,

    B. Xue and L. Zou, “Knowledge graph quality management: a compre- hensive survey,” IEEE Transactions on Knowledge and Data Engineer- ing, vol. 35, no. 5, pp. 4969–4988, 2022

    2022

  34. [42]

    Unifying large language models and knowledge graphs: A roadmap,

    S. Pan, L. Luo, Y . Wang, C. Chen, J. Wang, and X. Wu, “Unifying large language models and knowledge graphs: A roadmap,” IEEE Transactions on Knowledge and Data Engineering , vol. 36, no. 7, pp. 3580–3599, 2024

    2024

  35. [43]

    Knowledge- aware parsimony learning: A perspective from relational graphs,

    Q. Yao, Y . Zhang, Y . Wang, N. Yin, J. Kwok, and Q. Yang, “Knowledge- aware parsimony learning: A perspective from relational graphs,” arXiv preprint arXiv:2407.00478, 2024

    2024 arXiv

  36. [44]

    Knowledge graphs: Opportunities and challenges,

    C. Peng, F. Xia, M. Naseriparsa, and F. Osborne, “Knowledge graphs: Opportunities and challenges,” Artificial Intelligence Review , vol. 56, no. 11, pp. 13 071–13 102, 2023

    2023

  37. [45]

    Uukg: Unified urban knowledge graph dataset for urban spatiotemporal prediction,

    Y . Ning, H. Liu, H. Wang, Z. Zeng, and H. Xiong, “Uukg: Unified urban knowledge graph dataset for urban spatiotemporal prediction,” Advances in Neural Information Processing Systems , vol. 36, pp. 62 442–62 456, 2023

    2023

  38. [46]

    From local to global: A graph rag approach to query-focused summarization,

    D. Edge, H. Trinh, N. Cheng, J. Bradley, A. Chao, A. Mody, S. Truitt, D. Metropolitansky, R. O. Ness, and J. Larson, “From local to global: A graph rag approach to query-focused summarization,” arXiv preprint arXiv:2404.16130, 2024

    2024 arXiv

  39. [47]

    Knowledge graph-guided retrieval augmented generation,

    X. Zhu, Y . Xie, Y . Liu, Y . Li, and W. Hu, “Knowledge graph-guided retrieval augmented generation,”arXiv preprint arXiv:2502.06864, 2025

    2025 arXiv

  40. [48]

    Reversible instance normalization for accurate time-series forecasting against distri- bution shift,

    T. Kim, J. Kim, Y . Tae, C. Park, J.-H. Choi, and J. Choo, “Reversible instance normalization for accurate time-series forecasting against distri- bution shift,” in International Conference on Learning Representations , 2021

    2021

  41. [49]

    Making large language models perform better in knowledge graph completion,

    Y . Zhang, Z. Chen, L. Guo, Y . Xu, W. Zhang, and H. Chen, “Making large language models perform better in knowledge graph completion,” JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. XX, XX 12 in Proceedings of the 32nd ACM International Conference on Multime- dia, 2024, pp. 233–242

    2024

  42. [50]

    A decoder-only foundation model for time-series forecasting,

    A. Das, W. Kong, R. Sen, and Y . Zhou, “A decoder-only foundation model for time-series forecasting,” in Forty-first International Confer- ence on Machine Learning , 2024

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.