pith. sign in

arxiv: 2505.06907 · v2 · pith:E7UYYBI2new · submitted 2025-05-11 · 💻 cs.AI · cs.CV· cs.NE

A Survey on Foundation Models for Personalized Federated Intelligence

Yu Qiao , Huy Q. Le , Avi Deb Raha , Phuong-Nam Tran , Apurba Adhikary , Mengchun Zhang , Loc X. Nguyen , Eui-Nam Huh
show 2 more authors
Dusit Niyato Choong Seon Hong
This is my paper

Pith reviewed 2026-05-22 15:27 UTC · model grok-4.3

classification 💻 cs.AI cs.CVcs.NE
keywords personalized federated intelligencefoundation modelsfederated learningpersonalizationprivacyedge personalizationretrieval-augmented generation
0
0 comments X

The pith

Personalized federated intelligence adapts foundation models to individual users while preserving privacy.

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

This survey introduces personalized federated intelligence as a paradigm that merges the privacy advantages of federated learning with the broad capabilities of foundation models, with personalization as the central focus. It reviews recent progress in both areas and lays out a pipeline for achieving artificial personalized intelligence as a counterpart to general intelligence. The approach addresses how large models can be customized for end users despite their size, data sensitivity, and compute needs. A sympathetic reader would see value in systems that deliver tailored AI responses without requiring users to upload personal information to shared servers.

Core claim

The paper proposes personalized federated intelligence (PFI) as a new paradigm that integrates the privacy benefits of federated learning with the generalization capabilities of foundation models while placing personalization at its core, thereby enabling artificial personalized intelligence (API). It surveys advances in federated learning and foundation models, then details the PFI pipeline through three stages: efficient personalization at the edge, trustworthy adaptation, and adaptive refinement via retrieval-augmented generation, before outlining future directions.

What carries the argument

The PFI pipeline consisting of efficient edge personalization, trustworthy adaptation, and adaptive refinement via retrieval-augmented generation.

If this is right

  • Users receive customized responses from large models without transmitting raw personal data to central servers.
  • Edge devices perform model adaptation with reduced computational load compared to full retraining.
  • Trust mechanisms maintain model reliability and security during the personalization process.
  • Retrieval augmentation supplies dynamic, user-specific knowledge to refine outputs over time.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • PFI could extend to multi-user scenarios where indirect contributions improve shared models while each user retains private personalization.
  • Existing federated learning platforms might adopt PFI stages to lower deployment barriers for foundation model customization.
  • Testing PFI under highly heterogeneous user data distributions would reveal practical limits not covered in the survey.

Load-bearing premise

The stages of efficient edge personalization, trustworthy adaptation, and retrieval-augmented refinement can be combined without major trade-offs in privacy, performance, or scalability.

What would settle it

A working implementation of the full PFI pipeline that exhibits either privacy leakage, accuracy below that of separate federated learning or foundation model baselines, or inability to scale past a modest number of users would show the stages cannot be integrated as described.

Figures

Figures reproduced from arXiv: 2505.06907 by Apurba Adhikary, Avi Deb Raha, Choong Seon Hong, Dusit Niyato, Eui-Nam Huh, Huy Q. Le, Loc X. Nguyen, Mengchun Zhang, Phuong-Nam Tran, Yu Qiao.

Figure 1
Figure 1. Figure 1: The structure and scope of the survey. each entity’s specific domain. In other words, we envision a central server in PFI that no longer performs centralized training, but instead coordinates multiple distributed entities with heterogeneous settings and private data to collaboratively develop a shared foundation model. Following this, each entity applies domain-specific adjustments to refine the model, res… view at source ↗
Figure 2
Figure 2. Figure 2: Illustration of cross-device and cross-silo federated learning scenarios. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of horizontal federated learning, vertical federated learning, and federated transfer learning [53]. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: A taxonomy of techniques for enabling retrieval-augmented PFI. The design is organized into two complementary pillars: (1) Client-side personalization, [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: A taxonomy of techniques for enabling efficient personalized federated intelligence. It is organized into two main categories: (1) Efficient adaptation [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Illustration of the prompt tuning architecture, where trainable soft [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Architecture of a typical adapter module, consisting of a down [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Illustration of model pruning and model compression. (a) Model [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: A taxonomy of techniques for enabling trustworthy personalized federated intelligence. It is organized into four main categories: (1) mitigating bias, [PITH_FULL_IMAGE:figures/full_fig_p016_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Illustration of various approaches to mitigate bias and ensure fairness [PITH_FULL_IMAGE:figures/full_fig_p016_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Illustration of Key Challenges in PFI. security issues, all of which can hinder practical deployment in real-world scenarios. Understanding and mitigating these challenges is essential for advancing PFI systems in real￾world scenarios. In the following section, we discuss these challenges and then highlight promising future research direc￾tions, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p019_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Overview of domain heterogeneity challenges [47]. Domain hetero [PITH_FULL_IMAGE:figures/full_fig_p020_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Illustration of Future Directions in PFI. [PITH_FULL_IMAGE:figures/full_fig_p022_14.png] view at source ↗
read the original abstract

The rise of large language models (LLMs), such as ChatGPT, Gemini, and Grok, has reshaped the AI landscape. As prominent instances of foundational models (FMs), they exhibit remarkable capabilities in generating human-like content, pushing the boundaries towards artificial general intelligence (AGI). However, their large-scale nature, privacy sensitivity, and substantial computational demands pose significant challenges for personalized customization for end users. To bridge this gap, we present the vision of artificial personalized intelligence (API), which focuses on adapting FMs to individual users while ensuring privacy. As a central enabler of API, we propose personalized federated intelligence (PFI), a new paradigm that not only integrates the privacy benefits of federated learning (FL) with the generalization capabilities of FMs but also places personalization at its core. To this end, we first survey recent advances in FL and FMs that lay the foundation for PFI. We then explore core stages of the PFI pipeline: efficient personalization at the edge, trustworthy adaptation, and adaptive refinement via retrieval-augmented generation. Finally, we highlight future directions for enabling PFI. Overall, this survey aims to lay a foundation for the development of API as a complementary direction to AGI, with PFI as a key enabling paradigm.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper surveys advances in federated learning and foundation models to introduce Personalized Federated Intelligence (PFI) as a paradigm integrating FL privacy benefits with FM generalization capabilities while centering personalization. It outlines a PFI pipeline with three stages—efficient personalization at the edge, trustworthy adaptation, and adaptive refinement via retrieval-augmented generation—and discusses future directions toward artificial personalized intelligence (API) as a complement to AGI.

Significance. If the integration of the proposed stages holds, the survey could establish a useful conceptual roadmap for privacy-preserving, user-specific adaptation of large models, highlighting a practical direction complementary to general-purpose foundation model scaling.

major comments (1)
  1. [Core stages of the PFI pipeline] The description of the PFI pipeline (efficient edge personalization, trustworthy adaptation, and retrieval-augmented refinement) presents these stages as combinable in a coherent framework but supplies only high-level overviews without citing concrete mechanisms, communication bounds, or empirical references that address potential trade-offs in privacy leakage, edge compute overhead, or erosion of zero-shot FM capabilities.
minor comments (1)
  1. [Abstract] The abstract introduces 'artificial personalized intelligence (API)' and 'personalized federated intelligence (PFI)' in quick succession; a brief sentence clarifying their relationship would improve readability for readers new to the framing.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive review and for recognizing the potential of the PFI framework as a conceptual roadmap. We address the single major comment below and describe the revisions we will undertake.

read point-by-point responses

  1. Referee: The description of the PFI pipeline (efficient edge personalization, trustworthy adaptation, and retrieval-augmented refinement) presents these stages as combinable in a coherent framework but supplies only high-level overviews without citing concrete mechanisms, communication bounds, or empirical references that address potential trade-offs in privacy leakage, edge compute overhead, or erosion of zero-shot FM capabilities.

    Authors: We agree that the current manuscript presents the three stages at a conceptual level. As a survey introducing a new paradigm, this was intentional to establish the overall vision; however, the referee is correct that additional specificity would strengthen the paper. In the revised version we will expand each stage with citations to concrete mechanisms (e.g., parameter-efficient federated fine-tuning, differential-privacy bounds for adaptation, and retrieval-augmented generation pipelines), include references that quantify communication and compute costs, and explicitly discuss the cited trade-offs in privacy leakage, edge-device overhead, and possible degradation of zero-shot performance, drawing on the most relevant empirical studies available in the literature. revision: yes

Circularity Check

0 steps flagged

Survey proposes PFI vision without any self-referential derivations or fitted predictions

full rationale

This is a survey and vision paper that defines personalized federated intelligence (PFI) as a paradigm integrating the privacy benefits of federated learning with the generalization capabilities of foundation models while centering personalization. It surveys external advances in FL and FMs, then outlines high-level pipeline stages (efficient edge personalization, trustworthy adaptation, retrieval-augmented refinement) and future directions. No equations, parameter fits, or derivations appear that could reduce by construction to the paper's own inputs; the proposal explicitly builds on cited prior literature rather than self-citations as load-bearing uniqueness results or ansatzes. The structure is therefore self-contained against external benchmarks with no circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 1 invented entities

The paper introduces PFI as a new conceptual paradigm without new mathematical axioms, free parameters, or invented physical entities. It relies on standard assumptions from federated learning and foundation model literature.

invented entities (1)
  • Personalized Federated Intelligence (PFI) no independent evidence
    purpose: A new paradigm integrating FL privacy with FM generalization and personalization
    Introduced in the abstract as the central enabler of artificial personalized intelligence.

pith-pipeline@v0.9.0 · 5799 in / 1202 out tokens · 31790 ms · 2026-05-22T15:27:59.797990+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

299 extracted references · 299 canonical work pages · 24 internal anchors

  1. [1]

    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:2303.08774, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  2. [2]

    Gemini: A Family of Highly Capable Multimodal Models

    G. Team, R. Anil, S. Borgeaud, J.-B. Alayrac, J. Yu, R. Soricut, J. Schalkwyk, A. M. Dai, A. Hauth, K. Millican, et al. , “Gemini: a family of highly capable multimodal models,”ArXiv:2312.11805, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  3. [3]

    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:2501.12948, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  4. [4]

    Deepseek-inspired exploration of rl-based llms and synergy with wireless networks: A survey,

    Y . Qiao, P.-N. Tran, J. S. Yoon, L. X. Nguyen, and C. S. Hong, “Deepseek-inspired exploration of rl-based llms and synergy with wireless networks: A survey,” Authorea Preprints, 2025

    work page 2025

  5. [5]

    Large language models as computational linguistics tools: A comparative analysis of chatgpt and google machine transla- tions,

    M. A. AlAfnan, “Large language models as computational linguistics tools: A comparative analysis of chatgpt and google machine transla- tions,” Journal of Artificial Intelligence and Technology , 2024

    work page 2024

  6. [6]

    Chatgpt in content cre- ation: Techniques, applications, and ethical implications,

    A. Htet, S. R. Liana, T. Aung, and A. Bhaumik, “Chatgpt in content cre- ation: Techniques, applications, and ethical implications,” in Advanced Applications of Generative AI and Natural Language Processing Mod- els, pp. 43–68, IGI Global, 2024

    work page 2024

  7. [7]

    Comparative analysis of chatgpt-4 and llama: Performance evaluation on text summarization, data analysis, and question answering,

    S. R. Bogireddy and N. Dasari, “Comparative analysis of chatgpt-4 and llama: Performance evaluation on text summarization, data analysis, and question answering,” in Proc. 15th Int. Conf. Comput. Commun. Netw. Technol. (ICCCNT), 2024

    work page 2024

  8. [8]

    Examining the potential of chatgpt on biomedical information retrieval: fact-checking drug-disease associations,

    Z. Gao, L. Li, S. Ma, Q. Wang, L. Hemphill, and R. Xu, “Examining the potential of chatgpt on biomedical information retrieval: fact-checking drug-disease associations,” Annals of biomedical engineering , vol. 52, no. 8, pp. 1919–1927, 2024

    work page 1919

  9. [9]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Lo, et al. , “Segment anything,” in Proc. IEEE Int. Conf. Comput. Vis., pp. 4015–4026, 2023

    work page 2023

  10. [10]

    Robustness of sam: Segment anything under corruptions and beyond,

    Y . Qiao, C. Zhang, T. Kang, D. Kim, S. Tariq, C. Zhang, and C. S. Hong, “Robustness of sam: Segment anything under corruptions and beyond,” ArXiv:2306.07713, 2023

    work page arXiv 2023

  11. [11]

    Segment anything model aided beam prediction for the millimeter wave communication,

    A. D. Raha, A. Adhikary, M. S. Munir, Y . Qiao, and C. S. Hong, “Segment anything model aided beam prediction for the millimeter wave communication,” inProc. Asia–Pacific Netw. Operations Manage. Symp. (APNOMS), pp. 113–118, Jan. 2023

    work page 2023

  12. [12]

    Efficientvit-sam: Accelerated segment anything model without performance loss,

    Z. Zhang, H. Cai, and S. Han, “Efficientvit-sam: Accelerated segment anything model without performance loss,” in Proc. of the IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp. 7859–7863, 2024

    work page 2024

  13. [13]

    Segment and recognize anything at any granularity,

    F. Li, H. Zhang, P. Sun, X. Zou, S. Liu, C. Li, J. Yang, L. Zhang, and J. Gao, “Segment and recognize anything at any granularity,” in Proc. Eur. Conf. Comput. Vis., pp. 467–484, Springer, 2024

    work page 2024

  14. [14]

    A contemporary survey on semantic communications: Theory of mind, generative ai, and deep joint source-channel coding,

    L. X. Nguyen, A. D. Raha, P. S. Aung, D. Niyato, Z. Han, and C. S. Hong, “A contemporary survey on semantic communications: Theory of mind, generative ai, and deep joint source-channel coding,” ArXiv:2502.16468, Feb. 2025

    work page arXiv 2025

  15. [15]

    Segment anything in medical images,

    J. Ma, Y . He, F. Li, L. Han, C. You, and B. Wang, “Segment anything in medical images,” Nat. Commun., vol. 15, no. 1, p. 654, 2024

    work page 2024

  16. [16]

    SAM 2: Segment Anything in Images and Videos

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. Rädle, C. Rolland, L. Gustafson, et al., “Sam 2: Segment anything in images and videos,” ArXiv:2408.00714, 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  17. [17]

    Few-shot adaptation of training-free foundation model for 3d medical image segmentation,

    X. He, Y . Hu, Z. Zhou, M. Jarraya, and F. Liu, “Few-shot adaptation of training-free foundation model for 3d medical image segmentation,” ArXiv:2501.09138, 2025

    work page arXiv 2025

  18. [18]

    When sam2 meets video camouflaged object segmentation: A comprehensive eval- uation and adaptation,

    Y . Zhou, G. Sun, Y . Li, L. Benini, and E. Konukoglu, “When sam2 meets video camouflaged object segmentation: A comprehensive eval- uation and adaptation,” ArXiv:2409.18653, 2024

    work page arXiv 2024

  19. [19]

    Planning for agi and beyond,

    A. Sam, “Planning for agi and beyond,” February 2023

    work page 2023

  20. [20]

    A survey on federated learning,

    C. Zhang, Y . Xie, H. Bai, B. Yu, W. Li, and Y . Gao, “A survey on federated learning,” Knowl.-Based Syst., vol. 216, p. 106775, 2021

    work page 2021

  21. [21]

    Federated learning in mobile edge networks: A comprehensive survey,

    W. Y . B. Lim, N. C. Luong, D. T. Hoang, Y . Jiao, Y .-C. Liang, Q. Yang, D. Niyato, and C. Miao, “Federated learning in mobile edge networks: A comprehensive survey,”IEEE Commun. Surveys Tuts., vol. 22, no. 3, pp. 2031–2063, 2020

    work page 2031

  22. [22]

    Survey of personalization techniques for federated learning,

    V . Kulkarni, M. Kulkarni, and A. Pant, “Survey of personalization techniques for federated learning,” in Proc. 4th World Conference on Smart Trends in Systems, Security and Sustainability (WorldS4) , pp. 794–797, 2020

    work page 2020

  23. [23]

    The future of digital health with federated learning,

    N. Rieke, J. Hancox, W. Li, F. Milletari, H. R. Roth, S. Albarqouni, S. Bakas, M. N. Galtier, B. A. Landman, K. Maier-Hein, et al., “The future of digital health with federated learning,” NPJ digital medicine , vol. 3, no. 1, p. 119, 2020

    work page 2020

  24. [24]

    Federated learning for medical image analysis: A survey,

    H. Guan, P.-T. Yap, A. Bozoki, and M. Liu, “Federated learning for medical image analysis: A survey,” Pattern Recognit., p. 110424, 2024

    work page 2024

  25. [25]

    Federated learning for internet of things: A comprehensive survey,

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

    work page 2021

  26. [26]

    Threats to federated learning: A survey,

    L. Lyu, H. Yu, and Q. Yang, “Threats to federated learning: A survey,” Preprint ArXiv:2003.02133, Mar. 2020

    work page arXiv 2003

  27. [27]

    Federated learn- ing: A survey on enabling technologies, protocols, and applications,

    M. Aledhari, R. Razzak, R. M. Parizi, and F. Saeed, “Federated learn- ing: A survey on enabling technologies, protocols, and applications,” IEEE Access, vol. 8, pp. 140699–140725, 2020

    work page 2020

  28. [28]

    A survey on federated learning: The journey from centralized to distributed on-site learning and beyond,

    S. Abdulrahman, H. Tout, H. Ould-Slimane, A. Mourad, C. Talhi, and M. Guizani, “A survey on federated learning: The journey from centralized to distributed on-site learning and beyond,” IEEE Internet Things J., vol. 8, no. 7, pp. 5476–5497, 2021

    work page 2021

  29. [29]

    Federated learning on non-iid data: A survey,

    H. Zhu, J. Xu, S. Liu, and Y . Jin, “Federated learning on non-iid data: A survey,” Neurocomputing, vol. 465, pp. 371–390, 2021

    work page 2021

  30. [30]

    Model aggregation techniques in federated learning: A comprehensive survey,

    P. Qi, D. Chiaro, A. Guzzo, M. Ianni, G. Fortino, and F. Piccialli, “Model aggregation techniques in federated learning: A comprehensive survey,” Future Generation Computer Systems , January 2023

    work page 2023

  31. [31]

    When foundation model meets federated learning: Motivations, challenges, and future directions,

    W. Zhuang, C. Chen, and L. Lyu, “When foundation model meets federated learning: Motivations, challenges, and future directions,” ArXiv:2306.15546, 2023. 27

    work page arXiv 2023

  32. [32]

    Advances and open challenges in federated foundation models,

    C. Ren, H. Yu, H. Peng, X. Tang, B. Zhao, L. Yi, A. Z. Tan, Y . Gao, A. Li, X. Li, et al. , “Advances and open challenges in federated foundation models,” IEEE Commun. Surveys Tuts. , 2025

    work page 2025

  33. [33]

    A survey on efficient federated learning methods for foundation model training,

    H. Woisetschläger, A. Erben, S. Wang, R. Mayer, and H.-A. Jacobsen, “A survey on efficient federated learning methods for foundation model training,” in Proc. of the Thirty-Third International Joint Conference on Artificial Intelligence , IJCAI, Jeju, Korea, 2024

    work page 2024

  34. [34]

    A comprehensive survey on pretrained foundation models: A history from bert to chatgpt,

    C. Zhou, Q. Li, C. Li, J. Yu, Y . Liu, G. Wang, K. Zhang, C. Ji, Q. Yan, L. He, et al. , “A comprehensive survey on pretrained foundation models: A history from bert to chatgpt,” International Journal of Machine Learning and Cybernetics , pp. 1–65, 2024

    work page 2024

  35. [35]

    Pushing large language models to the 6g edge: Vision, challenges, and oppor- tunities,

    Z. Lin, G. Qu, Q. Chen, X. Chen, Z. Chen, and K. Huang, “Pushing large language models to the 6g edge: Vision, challenges, and oppor- tunities,” ArXiv:2309.16739, 2023

    work page arXiv 2023

  36. [36]

    Zhao, Optimizing Memory Usage when Training Deep Neural Networks

    X. Zhao, Optimizing Memory Usage when Training Deep Neural Networks. PhD thesis, Université de Bordeaux, 2024

    work page 2024

  37. [37]

    Unveiling security, privacy, and ethical concerns of chatgpt,

    X. Wu, R. Duan, and J. Ni, “Unveiling security, privacy, and ethical concerns of chatgpt,” J. Intell. Inf. Syst. , vol. 2, no. 2, pp. 102–115, Mar. 2024

    work page 2024

  38. [38]

    A pathway towards responsible ai generated content.,

    L. Lyu, C. Chen, and J. Fu, “A pathway towards responsible ai generated content.,” in IJCAI, pp. 7033–7038, 2023

    work page 2023

  39. [39]

    Chameleon: Dual memory replay for online continual learning on edge devices,

    S. Aggarwal, K. Binici, and T. Mitra, “Chameleon: Dual memory replay for online continual learning on edge devices,” IEEE Trans. Comput.- Aided Design Integr. Circuits Syst. , 2023

    work page 2023

  40. [40]

    Communication-efficient learning of deep networks from decentral- ized data,

    B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communication-efficient learning of deep networks from decentral- ized data,” in Artificial Intelligence and Statistics (AISTATS) , Laud- erdale, FL, April 2017

    work page 2017

  41. [41]

    Federated hybrid training and self-adversarial distillation: Towards robust edge networks,

    Y . Qiao, A. Adhikary, K. Kim, E.-N. Huh, Z. Han, and C. S. Hong, “Federated hybrid training and self-adversarial distillation: Towards robust edge networks,” ArXiv:2412.19354, 2024

    work page arXiv 2024

  42. [42]

    An efficient federated learning framework for training semantic communication systems,

    L. X. Nguyen, H. Q. Le, Y . L. Tun, P. S. Aung, Y . K. Tun, Z. Han, and C. S. Hong, “An efficient federated learning framework for training semantic communication systems,” IEEE Trans. Veh. Technol., vol. 73, no. 10, pp. 15872–15877, 2024

    work page 2024

  43. [43]

    SemSpaceFL: A collaborative hierarchical federated learning framework for semantic communication in 6G LEO satellites,

    L. X. Nguyen, S. S. Hassan, Y . M. Park, Y . K. Tun, Z. Han, and C. S. Hong, “SemSpaceFL: A collaborative hierarchical federated learning framework for semantic communication in 6G LEO satellites,”Preprint Arxiv:2505.00966, May 2025

    work page arXiv 2025

  44. [44]

    Logit calibration and feature contrast for robust federated learning on non-iid data,

    Y . Qiao, C. Zhang, A. Adhikary, and C. S. Hong, “Logit calibration and feature contrast for robust federated learning on non-iid data,” IEEE Trans. Netw. Sci. Eng. , vol. 12, no. 2, pp. 636–652, March 2025

    work page 2025

  45. [45]

    Towards communication-efficient adversarial federated learning for robust edge intelligence,

    Y . Qiao, H. Q. Le, A. Adhikary, and C. S. Hong, “Towards communication-efficient adversarial federated learning for robust edge intelligence,” ArXiv:2501.15257, 2025

    work page arXiv 2025

  46. [46]

    Towards personalized federated learning,

    A. Z. Tan, H. Yu, L. Cui, and Q. Yang, “Towards personalized federated learning,” IEEE Trans. Neural Netw. Learn. Syst. , vol. 34, no. 12, pp. 9587–9603, 2022

    work page 2022

  47. [47]

    Fedccl: Federated dual-clustered feature contrast under domain heterogeneity,

    Y . Qiao, H. Q. Le, M. Zhang, A. Adhikary, C. Zhang, and C. S. Hong, “Fedccl: Federated dual-clustered feature contrast under domain heterogeneity,” Inf. Fusion, vol. 113, p. 102645, Jan. 2025

    work page 2025

  48. [48]

    Adaptive personalized federated learning,

    Y . Deng, M. M. Kamani, and M. Mahdavi, “Adaptive personalized federated learning,” ArXiv:2003.13461, 2020

    work page arXiv 2003

  49. [49]

    Federated learning for healthcare: Systematic review and architecture proposal,

    R. S. Antunes, C. André da Costa, A. Küderle, I. A. Yari, and B. Eskofier, “Federated learning for healthcare: Systematic review and architecture proposal,” ACM Trans. Intell. Syst. Technol., vol. 13, no. 4, pp. 1–23, May 2022

    work page 2022

  50. [50]

    Privacy-preserving technology to help millions of people: Federated prediction model for stroke prevention,

    C. Ju, R. Zhao, J. Sun, X. Wei, B. Zhao, Y . Liu, H. Li, T. Chen, X. Zhang, D. Gao, et al. , “Privacy-preserving technology to help millions of people: Federated prediction model for stroke prevention,” ArXiv:2006.10517, 2020

    work page arXiv 2006

  51. [51]

    Federated learning for open banking,

    G. Long, Y . Tan, J. Jiang, and C. Zhang, “Federated learning for open banking,” in Federated Learning: Privacy and Incentive , pp. 240–254, Springer, Nov. 2020

    work page 2020

  52. [52]

    Fate: An industrial grade platform for collaborative learning with data protection,

    Y . Liu, T. Fan, T. Chen, Q. Xu, and Q. Yang, “Fate: An industrial grade platform for collaborative learning with data protection,” Journal of Machine Learning Research , vol. 22, no. 226, pp. 1–6, 2021

    work page 2021

  53. [53]

    Vertical federated learning: Concepts, advances, and challenges,

    Y . Liu, Y . Kang, T. Zou, Y . Pu, Y . He, X. Ye, Y . Ouyang, Y .-Q. Zhang, and Q. Yang, “Vertical federated learning: Concepts, advances, and challenges,” IEEE Trans. Knowl. Data Eng. , 2024

    work page 2024

  54. [54]

    Intellect-1 technical report,

    S. Jaghouar, J. M. Ong, M. Basra, F. Obeid, J. Straube, M. Keiblinger, E. Bakouch, L. Atkins, M. Panahi, C. Goddard, et al. , “Intellect-1 technical report,” ArXiv:2412.01152, 2024

    work page arXiv 2024

  55. [55]

    Photon: Federated llm pre-training,

    L. Sani, A. Iacob, Z. Cao, R. Lee, B. Marino, Y . Gao, D. Cai, Z. Li, W. Zhao, X. Qiu, et al. , “Photon: Federated llm pre-training,” ArXiv:2411.02908, 2024

    work page arXiv 2024

  56. [56]

    Mp-fedcl: Multiprototype federated contrastive learn- ing for edge intelligence,

    Y . Qiao, M. S. Munir, A. Adhikary, H. Q. Le, A. D. Raha, C. Zhang, and C. S. Hong, “Mp-fedcl: Multiprototype federated contrastive learn- ing for edge intelligence,” IEEE Internet Things J. , vol. 11, no. 5, pp. 8604–8623, September 2023

    work page 2023

  57. [57]

    An encouraging design for data owners to join multiple co-existing federated learning,

    L. X. Nguyen, L. Zou, H. Q. Le, and C. S. Hong, “An encouraging design for data owners to join multiple co-existing federated learning,” in Proc. Asia–Pacific Netw. Operations Manage. Symp. (APNOMS) , Sep. 2022

    work page 2022

  58. [58]

    Edge-based federated deep reinforcement learning for iot traffic management,

    A. Jarwan and M. Ibnkahla, “Edge-based federated deep reinforcement learning for iot traffic management,” IEEE Internet Things J. , vol. 10, no. 5, pp. 3799–3813, 2022

    work page 2022

  59. [59]

    Knowledge distillation in federated learning: Where and how to distill?,

    Y . Qiao, C. Zhang, H. Q. Le, A. D. Raha, A. Adhikary, and C. S. Hong, “Knowledge distillation in federated learning: Where and how to distill?,” in Proc. Asia–Pacific Netw. Operations Manage. Symp. (APNOMS), Sejong, South Korea, September 2023

    work page 2023

  60. [60]

    Knowledge distillation assisted robust federated learning: Towards edge intelligence,

    Y . Qiao, A. Adhikary, K. T. Kim, C. Zhang, and C. S. Hong, “Knowledge distillation assisted robust federated learning: Towards edge intelligence,” in Proc. IEEE Int. Conf. Commun. (ICC) , Denver, CO, June 2024

    work page 2024

  61. [61]

    Fedmekt: Distillation-based embedding knowledge transfer for multimodal fed- erated learning,

    H. Q. Le, M. N. Nguyen, C. M. Thwal, Y . Qiao, et al. , “Fedmekt: Distillation-based embedding knowledge transfer for multimodal fed- erated learning,” Neural Networks, vol. 183, p. 107017, Mar. 2025

    work page 2025

  62. [62]

    Fedproto: Federated prototype learning across heterogeneous clients,

    Y . Tan, G. Long, L. Liu, T. Zhou, Q. Lu, J. Jiang, and C. Zhang, “Fedproto: Federated prototype learning across heterogeneous clients,” in Proc. AAAI Conf. Artif. Intell. , New York, USA, Feb. 2022

    work page 2022

  63. [63]

    Towards robust federated learning via logits calibration on non-iid data,

    Y . Qiao, A. Adhikary, C. Zhang, and C. S. Hong, “Towards robust federated learning via logits calibration on non-iid data,” in Proc. IEEE/IFIP Netw. Operations Manage. Symp (NOMS) , Seoul, South Korea, May 2024

    work page 2024

  64. [64]

    Multi-level personalized federated learning on heterogeneous and long-tailed data,

    R. Zhang, Y . Chen, C. Wu, and F. Wang, “Multi-level personalized federated learning on heterogeneous and long-tailed data,” IEEE Trans. Mobile Comput., June 2024

    work page 2024

  65. [65]

    Federated machine learning: Concept and applications,

    Q. Yang, Y . Liu, T. Chen, and Y . Tong, “Federated machine learning: Concept and applications,” ACM Trans. Intell. Syst. Technol. , vol. 10, no. 2, pp. 1–19, 2019

    work page 2019

  66. [66]

    Hifgl: A hierarchical framework for cross-silo cross-device federated graph learning,

    Z. Guo, D. Yao, Q. Yang, and H. Liu, “Hifgl: A hierarchical framework for cross-silo cross-device federated graph learning,” in Proc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min. , pp. 968–979, 2024

    work page 2024

  67. [67]

    On privacy and personaliza- tion in cross-silo federated learning,

    K. Liu, S. Hu, S. Z. Wu, and V . Smith, “On privacy and personaliza- tion in cross-silo federated learning,” Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 35, pp. 5925–5940, 2022

    work page 2022

  68. [68]

    A practical cross-device federated learning framework over 5g networks,

    W. Yang, N. Wang, Z. Guan, L. Wu, X. Du, and M. Guizani, “A practical cross-device federated learning framework over 5g networks,” IEEE Wireless Communications, vol. 29, no. 6, pp. 128–134, 2022

    work page 2022

  69. [69]

    Foundation models defining a new era in vision: a survey and outlook,

    M. Awais, M. Naseer, S. Khan, R. M. Anwer, H. Cholakkal, M. Shah, M.-H. Yang, and F. S. Khan, “Foundation models defining a new era in vision: a survey and outlook,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 47, no. 4, pp. 2245–2264, 2025

    work page 2025

  70. [70]

    The ai index 2021 annual report,

    D. Zhang, S. Mishra, E. Brynjolfsson, J. Etchemendy, D. Ganguli, B. Grosz, T. Lyons, J. Manyika, J. C. Niebles, M. Sellitto, et al., “The ai index 2021 annual report,” ArXiv:2103.06312, 2021

    work page arXiv 2021

  71. [71]

    On the Opportunities and Risks of Foundation Models

    R. Bommasani, D. A. Hudson, E. Adeli, R. Altman, S. Arora, S. von Arx, M. S. Bernstein, et al. , “On the opportunities and risks of foundation models,” ArXiv:2108.07258, Jul. 2021

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  72. [72]

    Transformers in vision: A survey,

    S. Khan, M. Naseer, M. Hayat, S. W. Zamir, F. S. Khan, and M. Shah, “Transformers in vision: A survey,” ACM Comput. Surv. , vol. 54, no. 10s, pp. 1–41, 2022

    work page 2022

  73. [73]

    An introduction to deep reinforcement learning,

    V . François-Lavet, P. Henderson, R. Islam, M. G. Bellemare, J. Pineau, et al. , “An introduction to deep reinforcement learning,” Foundations and Trends® in Machine Learning, vol. 11, no. 3-4, pp. 219–354, 2018

    work page 2018

  74. [74]

    Improv- ing language understanding by generative pre-training,

    A. Radford, K. Narasimhan, T. Salimans, I. Sutskever, et al., “Improv- ing language understanding by generative pre-training,” Preprint, 2018

    work page 2018

  75. [75]

    Visual reasoning eval- uation of grok, deepseek janus, gemini, qwen, mistral, and chatgpt,

    N. Jegham, M. Abdelatti, and A. Hendawi, “Visual reasoning eval- uation of grok, deepseek janus, gemini, qwen, mistral, and chatgpt,” ArXiv:2502.16428, 2025

    work page arXiv 2025

  76. [76]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, et al., “Learning transferable visual models from natural language supervision,” in Proc. Int. Conf. Mach. Learn., pp. 8748–8763, PMLR, 2021

    work page 2021

  77. [77]

    Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection

    S. Liu, Z. Zeng, T. Ren, F. Li, H. Zhang, J. Yang, C. Li, J. Yang, H. Su, J. Zhu, et al., “Grounding dino: Marrying dino with grounded pre-training for open-set object detection,” ArXiv:2303.05499, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  78. [78]

    Clip-vip: Adapting pre-trained image-text model to video-language representa- tion alignment,

    H. Xue, Y . Sun, B. Liu, J. Fu, R. Song, H. Li, and J. Luo, “Clip-vip: Adapting pre-trained image-text model to video-language representa- tion alignment,” ArXiv:2209.06430, 2022. 28

    work page arXiv 2022

  79. [79]

    Dino is also a semantic guider: Exploiting class-aware affinity for weakly supervised semantic segmentation,

    Y . Wu, X. Li, J. Li, K. Yang, P. Zhu, and S. Zhang, “Dino is also a semantic guider: Exploiting class-aware affinity for weakly supervised semantic segmentation,” in Proc. of the 32nd ACM International Conference on Multimedia , pp. 1389–1397, 2024

    work page 2024

  80. [80]

    Sam-event- adapter: Adapting segment anything model for event-rgb semantic segmentation,

    B. Yao, Y . Deng, Y . Liu, H. Chen, Y . Li, and Z. Yang, “Sam-event- adapter: Adapting segment anything model for event-rgb semantic segmentation,” in Proc. IEEE International Conference on Robotics and Automation (ICRA) , pp. 9093–9100, 2024

    work page 2024

Showing first 80 references.