Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Foundation Model Empowered Synesthesia of Machines (SoM): AI-native Intelligent Multi-Modal Sensing-Communication Integration

T0 review · 4 major / 6 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 Foundation models, pre-trained at scale and adaptable to many tasks, are the proposed backbone for 6G's sensing-communication paradigm SoM, giving two concrete design roadmaps that beat task-specific AI models in the reported case studies.

desk verdict A useful taxonomy and design roadmap for foundation-model-empowered SoM, but the load-bearing claim that LLM pretraining transfers to wireless tasks is asserted, not tested. read the letter →

arxiv 2506.07647 v1 pith:IUAJALHC submitted 2025-06-09 eess.SP

Xiang Cheng , Boxun Liu , Xuanyu Liu , Ensong Liu , Ziwei Huang This is my paper
classification eess.SP
keywords SynesthesiaofMachinesfoundationmodelslargelanguagewirelessintegratedsensingandcommunicationschannelpredictionmulti-modal6Gnetworks
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that foundation models—models pre-trained at scale and adapted to many tasks—are the right backbone for Synesthesia of Machines (SoM), the proposed 6G paradigm that couples multi-modal sensing with wireless communication. Existing task-specific AI models for SoM fail on four counts: scarce aligned datasets, limited modeling capacity, poor generalization, and no task universality. The paper's central claim is that two families of foundation models solve all four: general-purpose large language models (LLMs), adapted by fine-tuning or prompting, and domain-specific wireless foundation models, pre-trained on channel and sensing data. It proposes concrete roadmaps and reports that LLM-based path-loss and scatterer generation, multi-task transceivers, and wireless foundation models (WiCo, WiFo, WiPo) outperform task-specific baselines. If right, SoM system design becomes a choice of backbone plus an adaptation strategy rather than a bespoke network per task.

What carries the argument

The load-bearing mechanism is pre-training plus task adaptation. For LLMs, frozen GPT-2 weights act as a general-knowledge prior; only LayerNorm parameters (LLM4PG/LLM4SG) or LoRA experts (LLM4WM) are trained, converting camera, LiDAR, or CSI tokens into channel outputs. For wireless foundation models, self-supervised objectives—masked reconstruction for WiFo, cross-modal generation for WiCo, heterogeneous token reconstruction for WiPo—produce a shared backbone that is then frozen or lightly fine-tuned. The SynthSoM dataset supplies the aligned multi-modal sensing-communication pairs that both roadmaps train and evaluate on.

What would settle it

Fine-tune GPT-2 on the same SynthSoM tasks after randomly re-initializing its pretrained weights; if the randomized model matches the pretrained model's path-loss and scatterer accuracy, the claimed benefit of language pretraining is absent. Alternatively, evaluate LLM4SG on a held-out city or frequency band never seen in pretraining or fine-tuning; a sharp accuracy drop would contradict the claimed generalization.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes a systematic framework: foundation models for SoM fall into two classes—general-purpose LLMs and SoM-specific wireless foundation models—and each class addresses the four SoM challenges through a distinct mechanism. For LLMs, the mechanism is transfer: pre-trained language representations, after light fine-tuning (LayerNorm tuning for generation tasks, MoE-LoRA for multi-task transceivers), generate path loss maps, scatterers, and channel predictions from camera and LiDAR inputs better than GAN or ResNet baselines. For wireless foundation models, the mechanism is self-supervised pre-training on channel or heterogeneous multi-modal data, followed by zero-shot or few-shot adaptation; WiCo generates path loss and multipath parameters from images, WiFo predicts channels from CSI with or without RGB, and WiPo transmits modality-agnostic features for cooperative perception. Each case study reports that the foundation-model scheme beats the task-specific model on accuracy or generalization with comparable inference time and parameter count. The paper positions this as the first systematic categorization and design pipeline for FM-empowered SoM.

Load-bearing premise

The claim rests on the assumption that what a language model learned from text and images helps it understand radio signals, so that lightly fine-tuning GPT-2 produces accurate path loss, scatterers, and channel predictions from camera and LiDAR data.

Editorial extensions

If this is right

  • If FM transfer holds, scarce aligned sensing-communication datasets stop being the bottleneck: few-shot adaptation replaces large task-specific collections.
  • A single LLM or wireless foundation model can serve multiple SoM tasks, cutting deployed model count and storage overhead.
  • Zero-shot channel prediction becomes feasible in new scenarios, removing the need for per-scenario retraining.
  • SoM mechanism exploration can be automated: LLMs and WiCo generate path loss and scatterers from visual data, supporting transceiver design without full channel measurement.

Reading between the lines

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

  • The reported gains are preliminary and benchmarked on SynthSoM; a stronger test would randomize the pretrained LLM weights to see how much of the advantage comes from language pretraining versus the fine-tuning procedure itself.
  • If the transfer premise holds broadly, the same roadmap could apply to other radio tasks such as spectrum sensing and beam management, where labeled data is scarce but visual or LiDAR context is available.
  • The wireless-foundation-model roadmap may eventually make LLM-based SoM unnecessary for physical-layer tasks, since WiFo and WiCo match or exceed LLM results with fewer parameters.
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 / 6 minor

Summary. The paper proposes a systematic framework for applying foundation models (FMs) to Synesthesia of Machines (SoM), an AI-native paradigm integrating multi-modal sensing and communication. It categorizes FMs into general-purpose LLMs and domain-specific wireless foundation models, and proposes two design roadmaps. For each roadmap, it offers a step-by-step framework and case studies: LLM4PG and LLM4SG for mechanism exploration, LLM4WM for transceiver design, and WiCo, WiFo, and WiPo for mechanism exploration, transceiver design, and cooperative perception respectively. Preliminary simulations are presented to support the claim that FM-based schemes significantly outperform task-specific models.

Significance. If validated, the proposed categorization and roadmaps would provide a useful design blueprint for 6G research, organizing a fragmented body of work and highlighting where LLMs versus wireless foundation models are more appropriate. The paper makes a strength of being the first to systematically organize FM-based SoM design and includes several concrete case studies with external baselines. However, the empirical evidence is preliminary: figures show single-run comparisons, some with only qualitative image outputs, and no statistical uncertainty is reported. The central premise of Roadmap 1—that language-pretrained GPT-2 provides useful physical knowledge for wireless tasks—is asserted but not experimentally isolated. The framework itself is plausible and potentially significant, but the current evidence base is insufficient to support the strong 'significant superiority' claim in the abstract.

major comments (4)
  1. [§III.B, §III.C, Figs. 4 and 5] The claim that LLM-based schemes (LLM4PG, LLM4SG, LLM4WM) benefit from 'general knowledge transfer' of language pretraining is never isolated experimentally. The comparisons pit GPT-2 with only LayerNorm parameters (or MoE-LoRA adapters) trainable against task-specific GAN, ResNet, or Cross-Stitch baselines that differ in architecture, parameter count, and training protocol. Since the GPT-2 backbone is frozen, an observed gain could come from transformer capacity, the adaptation modules, or the training procedure rather than from language pretraining. A randomized-initialization ablation—same architecture, same tuning strategy, but with randomly initialized GPT-2 weights—is needed to attribute any improvement to the pretrained representations. Without it, Roadmap 1's core premise remains unverified.
  2. [Figs. 4, 5, 7, 8, 9, 10 and Tables II–VI] All reported comparisons are single-run results with no error bars, confidence intervals, or significance tests. Phrases such as 'achieves over 11% higher generalization accuracy' (Fig. 4) and 'significantly outperforms' (Fig. 9) are not statistically established. The paper does not report the number of random seeds, the variance across runs, or the exact fine-tuning dataset sizes and splits. To support the abstract's claim of 'significant superiority of FMs over task-specific models', the authors should provide means and standard deviations over multiple seeds, or equivalent statistical evidence, and specify training hyperparameters and dataset partitions for every case study.
  3. [§IV.B, Case Study 1 for Roadmap 2 (WiCo)] WiCo is presented as a 'wireless channel foundation model' but, as described, it is trained on a single dataset (SynthSoM) for a single generation task (path loss or multipath generation). This does not meet the paper's own definition in §II.B.1, which requires pre-training on broad wireless and multi-modal sensing data and adaptability to a wide range of tasks through few-shot or zero-shot learning. The pre-training strategy is described only as 'data augmentation and noise suppression' (Step 3), which is not a specific self-supervised objective. The authors should clarify what distinguishes WiCo from a conventional task-specific generative model, or revise the terminology.
  4. [§IV.C, WiFo case study and §III.C, LLM4WM case study] The WiFo comparison in Fig. 9 uses, as a task-specific baseline, a randomly initialized WiFo encoder with otherwise identical components; this is a good controlled ablation for the value of WiFo's pre-training. However, no analogous controlled ablation is provided for the LLM-based schemes, so the WiFo result does not compensate for the missing LLM ablation. Additionally, the WiFo fine-tuning uses 1,500 aligned CSI-RGB pairs but the training/validation split and the number of runs are not given, and the LLM4WM training details (e.g., MoE-LoRA rank, learning rate, number of epochs) are absent. These missing specifications make it impossible to reproduce or independently verify the reported gains.
minor comments (6)
  1. [§II.B.2] The phrase 'self-supervised pertaining' should be 'self-supervised pre-training'.
  2. [§III.B, Step 2 and Table II] The paper states that GPT-2 is used but does not specify which variant (e.g., GPT-2 small, medium) or how the reported total parameter counts (e.g., 275.70M for LLM4PG) arise. Clarifying the backbone variant and the exact trainable parameter set would help reproducibility.
  3. [§IV.B, Step 4 and elsewhere] The term 'adaption' is used in several places (e.g., §IV.B Step 4 and §IV.B 'adoption') where 'adaptation' is intended; please correct these typos.
  4. [Fig. 4, Fig. 9, Fig. 10] Several figures lack axis labels or legends (e.g., Figs. 4(d) and 4(e) have no labeled axes, and Figs. 9 and 10 do not define NMSE and PSNR axes fully). Adding clear captions with evaluation protocols would improve readability.
  5. [§V.B.3] The sentence 'as demonstrated in case study 1 of Section IV-C' appears to be a cross-reference error; the WiFo case study with vision enhancement is described in §IV.C, which is actually 'Case Study 2' of Roadmap 2. Please correct the reference.
  6. [§IV.B] The text repeatedly writes 'UA V images' (with an extra space) instead of 'UAV images'; please fix this typo.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's roadmaps are proposals supported by external-baseline case studies, not inputs masquerading as predictions.

full rationale

I walked the paper's claimed derivation chain and found no step where a prediction or derived characteristic is equivalent to its input by construction. The paper's contributions are taxonomic and architectural: it categorizes FMs into LLMs and wireless foundation models, maps their general capabilities to SoM challenges in Table I, and proposes roadmaps with case studies. These mappings are qualitative attributions from the FM literature rather than fitted quantities or equations, so there is no self-definitional reduction. The case studies compare the proposed schemes against external baselines (GAN, ResNet, Cross-Stitch, and same-architecture task-specific models) and against ray-tracing ground truth from the SynthSoM dataset, so the reported superiority is not a renamed fit. Self-citations to LLM4CP [66], LLM4WM [72], WiFo [104], and SynthSoM [8] are used as prior components or motivating evidence; they are externally falsifiable because the baselines and datasets are specified, and the paper's own WiCo, WiPo, and LLM4PG results are not derived solely from those citations. The main weaknesses, including the absence of a randomized-initialization ablation for GPT-2 and the lack of error bars or seeds, are empirical-evidence gaps rather than circular reductions; the paper itself partially acknowledges the multi-modal limitation of CSI-only wireless FMs in Section V.B.3. I therefore find no significant circularity and assign a score of 0.

Assumptions & free parameters 2 free parameters · 5 assumptions · 2 invented entities

The central conceptual claims rest on assumptions about transferability of pretrained representations, dataset fidelity, simulator validity, and learnability of the SoM mechanism. The simulations depend on unreported hyperparameters, and several case studies reuse the authors' prior models, so independent evidence is limited. WiCo and WiPo are new model components with no external falsifiable handle beyond the paper's own simulations.

free parameters (2)
  • Unreported training hyperparameters for case studies
    The simulations depend on learning rates, LoRA ranks, masking ratios, fine-tuning epochs, and loss weights, none of which are listed in the paper. These are chosen by the experimenters and directly affect the magnitudes of the reported gains.
  • Fine-tuning dataset sizes = 30 CSI samples (LLM4CP), 7% and 5% sample subsets (LLM4SG), 1,500 aligned CSI-RGB pairs (WiFo)
    These numbers are experimental choices used to demonstrate few-shot advantages. They are not derived from the problem and determine how impressive the generalization claims appear.
assumptions (5)
  • domain assumption LLMs pretrained primarily on text transfer their representations to non-linguistic wireless and sensing tasks.
    Invoked throughout Section III in LLM4PG, LLM4SG, and LLM4WM, and in the motivation table. No structural argument is given for why text-pretrained weights should encode radio channel physics.
  • domain assumption The SynthSoM dataset accurately represents real-world sensing-communication co-occurrence and is a reliable basis for training and evaluation.
    Used as ground truth in Sections III.B, IV.B, and IV.C. The dataset is constructed by the authors in prior work [8], and no independent validation is presented in this paper.
  • domain assumption Self-supervised pretraining at scale produces representations that generalize to SoM tasks with few or zero examples.
    Assumed in the wireless foundation model roadmap in Section IV and used to motivate WiCo, WiFo, and WiPo. The support is limited to the paper's own preliminary results.
  • domain assumption Ray-tracing simulator output is an acceptable physical ground truth for path loss and multipath parameters.
    Used in all generation case studies as the target maps. The paper does not discuss simulator calibration, measurement campaigns, or error characterization.
  • domain assumption The SoM mechanism between visual or RF sensing and channel data is learnable from paired data.
    This is the premise of all mechanism-exploration case studies in Sections III.B and IV.B. It is asserted rather than proven, and the paper does not characterize the limits of the learnable mapping.
invented entities (2)
  • WiCo (wireless channel foundation model)
    purpose: Cross-modal generation of path loss and multipath components from sensory data in SoM mechanism exploration.
    Introduced in Section IV.B as a new model family. Evidence is confined to in-paper comparisons with no external benchmark, released weights, or independent validation.
  • WiPo (wireless cooperative perception foundation model)
    purpose: Modality-agnostic feature transmission for cooperative perception under wireless channel effects.
    Introduced in Section IV.D. The paper reports in-paper simulations on ImageNet, CsiNet-Outdoor, ShapeNet, and SynthSoM, but no code or weights are released and no external replication is available.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Foundation Model Empowered Synesthesia of Machines (SoM): AI-native Intelligent Multi-Modal Sensing-Communication Integration." pith.science (2026). https://pith.science/paper/IUAJALHC

@misc{pith2026250607647,
  author       = {Pith},
  title        = {Pith review of: Foundation Model Empowered Synesthesia of Machines (SoM): AI-native Intelligent Multi-Modal Sensing-Communication Integration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IUAJALHC}},
  note         = {Machine review of arXiv:2506.07647}
}
read the original abstract

To support future intelligent multifunctional sixth-generation (6G) wireless communication networks, Synesthesia of Machines (SoM) is proposed as a novel paradigm for artificial intelligence (AI)-native intelligent multi-modal sensing-communication integration. However, existing SoM system designs rely on task-specific AI models and face challenges such as scarcity of massive high-quality datasets, constrained modeling capability, poor generalization, and limited universality. Recently, foundation models (FMs) have emerged as a new deep learning paradigm and have been preliminarily applied to SoM-related tasks, but a systematic design framework is still lacking. In this paper, we for the first time present a systematic categorization of FMs for SoM system design, dividing them into general-purpose FMs, specifically large language models (LLMs), and SoM domain-specific FMs, referred to as wireless foundation models. Furthermore, we derive key characteristics of FMs in addressing existing challenges in SoM systems and propose two corresponding roadmaps, i.e., LLM-based and wireless foundation model-based design. For each roadmap, we provide a framework containing key design steps as a guiding pipeline and several representative case studies of FM-empowered SoM system design. Specifically, we propose LLM-based path loss generation (LLM4PG) and scatterer generation (LLM4SG) schemes, and wireless channel foundation model (WiCo) for SoM mechanism exploration, LLM-based wireless multi-task SoM transceiver (LLM4WM) and wireless foundation model (WiFo) for SoM-enhanced transceiver design, and wireless cooperative perception foundation model (WiPo) for SoM-enhanced cooperative perception, demonstrating the significant superiority of FMs over task-specific models. Finally, we summarize and highlight potential directions for future research.

Figures

Figures reproduced from arXiv: 2506.07647 by the authors.

Figure 1
Figure 1. An illustration of the SoM framework, highlighting its three key characteristics and the interrelation of its five key research directions. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The organization of this paper. design. In this paper, inspired by existing studies on SoM￾related domains empowered by FMs, we for the first time systematically categorize FMs into two types: general-purpose foundation models, specifically LLMs, and SoM domain￾specific foundation models, namely wireless foundation mod￾els. Inspired by the superior capabilities of these two types of FMs in addressing the existing ch… view at source ↗
Figure 3
Figure 3. An illustration of the framework of roadmap 1 and the proposed schemes for the two case studies introduced. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (6 more)
Figure 5
Figure 5. Figure 5: NMSE performance comparison of the proposed LLM4WM scheme with the SM-MTL and SM-STL schemes across three SoM-related tasks. specialized model single-task learning (SM-STL) scheme and the specialized model multi-task learning (SM-MTL) scheme. Furthermore, to compare th…
Figure 6
Figure 6. Figure 6: An illustration of the framework of roadmap 2 and the proposed schemes for the three case studies introduced. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Comparisons of path loss map results. (a) Ray-tracing-based result. [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Comparisons of multipath parameter results. (a) Ray-tracing-based [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: NMSE performance comparison of WiFo-based, LLM-based, and [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: Reconstruction performance comparison of WiPo and task-specific [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Towards channel foundation models (CFMs): Motivations, methodologies and opportunities

    eess.SP 2025-07 conditional novelty 4.0 of 10

    A survey and position paper proposing channel foundation models, with experiments on two pretrained CSI models showing gains over a vanilla ViT baseline.

Reference graph

Works this paper leans on

150 extracted references · 25 canonical work pages · cited by 1 Pith paper

  1. [1]

    Towards 6G Wireless Communication Net- works: Vision, Enabling Technologies, and New Paradigm Shifts,

    X. Youet al., “Towards 6G Wireless Communication Net- works: Vision, Enabling Technologies, and New Paradigm Shifts,”Sci. China Inf. Sci., vol. 64, pp. 1–74, Nov. 2021

    2021

  2. [2]

    White paper on 6G vision and candidate technologies,

    IMT-2030 (6G) Promotion Group, “White paper on 6G vision and candidate technologies,”China, CAICT, 2021

    2021

  3. [3]

    Integrated Sensing and Communications: Toward Dual-Functional Wireless Networks for 6G and Beyond,

    F. Liuet al., “Integrated Sensing and Communications: Toward Dual-Functional Wireless Networks for 6G and Beyond,”IEEE J. Select. Areas Commun., vol. 40, no. 6, pp. 1728–1767, Jun. 2022

    2022

  4. [4]

    Intelligent Multi-Modal Sensing- Communication Integration: Synesthesia of Machines,

    X. Chenget al., “Intelligent Multi-Modal Sensing- Communication Integration: Synesthesia of Machines,” IEEE Commun. Surv. Tutorials, vol. 26, pp. 258–301, Firstquarter 2024

    2024

  5. [5]

    Seventy Years of Radar and Communications: The road from separation to integration,

    F. Liuet al., “Seventy Years of Radar and Communications: The road from separation to integration,”IEEE Signal Process Mag., vol. 40, no. 5, pp. 106–121, Jul. 2023

    2023

  6. [6]

    A Survey on Vision-Language-Action Models for Embodied AI,

    Y . Ma, Z. Song, Y . Zhuang, J. Hao, and I. King, “A Survey on Vision-Language-Action Models for Embodied AI,”arXiv preprint arXiv:2405.14093, 2024

    arXiv 2024

  7. [7]

    Multi-Modal Intelligent Channel Modeling: A New Modeling Paradigm via Synesthesia of Machines,

    L. Bai, Z. Huang, M. Sun, X. Cheng, and L. Cui, “Multi-Modal Intelligent Channel Modeling: A New Modeling Paradigm via Synesthesia of Machines,”IEEE Commun. Surv. Tutorials, 2025

    2025

  8. [8]

    SynthSoM: A Synthetic Intelligent Multi- Modal Sensing-Communication Dataset for Synesthesia of Machines (SoM),

    X. Chenget al., “SynthSoM: A Synthetic Intelligent Multi- Modal Sensing-Communication Dataset for Synesthesia of Machines (SoM),”Sci. Data, vol. 12, no. 819, May 2025

    2025

Show all 150 references
  1. [9]

    Multi-Modal Sensing Data Based Real-Time Path Loss Prediction for 6G UA V-to-Ground Communications,

    M. Sun, L. Bai, Z. Huang, and X. Cheng, “Multi-Modal Sensing Data Based Real-Time Path Loss Prediction for 6G UA V-to-Ground Communications,”IEEE Wireless Commun., vol. 13, no. 9, pp. 2462–2466, Sept. 2024. 18

    2024

  2. [10]

    Integrated Sensing and Communications Towards Proactive Beamforming in mmWave V2I via Multi-Modal Feature Fusion (MMFF),

    H. Zhang, S. Gao, X. Cheng, and L. Yang, “Integrated Sensing and Communications Towards Proactive Beamforming in mmWave V2I via Multi-Modal Feature Fusion (MMFF),” IEEE Trans. Wireless Commun., vol. 23, pp. 15 721–15 735, Nov. 2024

    2024

  3. [11]

    On the Opportunities and Risks of Foundation Models,

    R. Bommasaniet al., “On the Opportunities and Risks of Foundation Models,”arXiv preprint arXiv:2108.07258, 2021

    2021 arXiv

  4. [12]

    A Survey of Large Language Models,

    W. X. Zhaoet al., “A Survey of Large Language Models,” arXiv preprint arXiv:2303.18223, vol. 1, no. 2, 2023

    2023 arXiv

  5. [13]

    Deepseek-v3 Technical Report,

    A. Liuet al., “Deepseek-v3 Technical Report,”arXiv preprint arXiv:2412.19437, 2024

    2024 arXiv

  6. [14]

    Foundation Models for Time Series Analysis: A Tutorial and Survey,

    Y . Lianget al., “Foundation Models for Time Series Analysis: A Tutorial and Survey,” inProc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min. (KDD), Barcelona, Spain, Aug. 2024, pp. 6555–6565

    2024

  7. [15]

    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, Jul. 2023

    2023

  8. [16]

    RingMo: A Remote Sensing Foundation Model With Masked Image Modeling,

    X. Sunet al., “RingMo: A Remote Sensing Foundation Model With Masked Image Modeling,”IEEE Trans. Geosci. Remote Sens., vol. 61, pp. 1–22, Jul. 2022

    2022

  9. [17]

    When large language model agents meet 6G networks: Perception, grounding, and alignment,

    M. Xuet al., “When large language model agents meet 6G networks: Perception, grounding, and alignment,”IEEE Wireless Commun., vol. 31, no. 6, pp. 63–71, Dec. 2024

    2024

  10. [18]

    Generative AI Meets Wireless Networking: An Interactive Paradigm for Intent-Driven Communications,

    X. Qinet al., “Generative AI Meets Wireless Networking: An Interactive Paradigm for Intent-Driven Communications,” IEEE Trans. Cognit. Commun. Networking, early access 2025

    2025

  11. [19]

    Large Language Model Enhanced Multi-Agent Systems for 6G Communications,

    F. Jianget al., “Large Language Model Enhanced Multi-Agent Systems for 6G Communications,”IEEE Wireless Commun., vol. 31, no. 6, pp. 48–55, Dec. 2024

    2024

  12. [20]

    Large Language Model (LLM) for Telecom- munications: A Comprehensive Survey on Principles, Key Techniques, and Opportunities,

    H. Zhouet al., “Large Language Model (LLM) for Telecom- munications: A Comprehensive Survey on Principles, Key Techniques, and Opportunities,”IEEE Commun. Surv. Tuto- rials, early access 2024

    2024

  13. [21]

    On LLMs-Driven Synthetic Data Gener- ation, Curation, and Evaluation: A Survey,

    L. Longet al., “On LLMs-Driven Synthetic Data Gener- ation, Curation, and Evaluation: A Survey,”arXiv preprint arXiv:2406.15126, 2024

    2024 arXiv

  14. [22]

    A Survey of Multimodel Large Language Models,

    Z. Lianget al., “A Survey of Multimodel Large Language Models,” inProc. Int. Conf. Comput. Artif. Intell. Control Eng. (CAICE), Xi’an, China, Jan. 2024, pp. 405–409

    2024

  15. [23]

    Attention Is All You Need,

    A. Vaswaniet al., “Attention Is All You Need,” inAdv. Neural Inf. Process. Syst. (NeurIPS), vol. 30, Long Beach, CA, USA, 2017, pp. 6000–6010

    2017

  16. [24]

    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 Trans. Knowl. Data Eng., vol. 36, no. 7, pp. 3580–3599, Jul. 2024

    2024

  17. [25]

    Scaling laws for neural language models,

    J. Kaplanet al., “Scaling laws for neural language models,” arXiv preprint arXiv:2001.08361, 2020

    2001 arXiv

  18. [26]

    Large Language Models for Mathematical Reasoning: Pro- gresses and Challenges,

    J. Ahn, R. Verma, R. Lou, D. Liu, R. Zhang, and W. Yin, “Large Language Models for Mathematical Reasoning: Pro- gresses and Challenges,”arXiv preprint arXiv:2402.00157, 2024

    2024 arXiv

  19. [27]

    What Indeed Can GPT Models Do In Chem- istry? A Comprehensive Benchmark On Eight Tasks,

    T. Guoet al., “What Indeed Can GPT Models Do In Chem- istry? A Comprehensive Benchmark On Eight Tasks,”arXiv preprint arXiv:2305.18365, 2023

    2023 arXiv

  20. [28]

    Scientific Large Language Models: A Survey on Biological & Chemical Domains,

    Q. Zhanget al., “Scientific Large Language Models: A Survey on Biological & Chemical Domains,”ACM Comput. Surv., vol. 57, no. 6, pp. 1–38, Feb. 2025

    2025

  21. [29]

    Large Language Models for Software Engineer- ing: A Systematic Literature Review,

    X. Houet al., “Large Language Models for Software Engineer- ing: A Systematic Literature Review,”ACM Trans. Software Eng. Methodol., vol. 33, no. 8, pp. 1–79, Dec. 2024

    2024

  22. [30]

    Parameter- Efficient Fine-Tuning for Large Models: A Comprehensive Survey,

    Z. Han, C. Gao, J. Liu, J. Zhang, and S. Q. Zhang, “Parameter- Efficient Fine-Tuning for Large Models: A Comprehensive Survey,”arXiv preprint arXiv:2403.14608, 2024

    2024 arXiv

  23. [31]

    A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications,

    P. Sahoo, A. K. Singh, S. Saha, V . Jain, S. Mondal, and A. Chadha, “A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications,”arXiv preprint arXiv:2402.07927, 2024

    2024 arXiv

  24. [32]

    Instruction Tuning for Large Language Models: A Survey,

    S. Zhanget al., “Instruction Tuning for Large Language Models: A Survey,”arXiv preprint arXiv:2308.10792, 2023

    2023

  25. [33]

    LLM With Tools: A Survey,

    Z. Shen, “LLM With Tools: A Survey,”arXiv preprint arXiv:2409.18807, 2024

    2024 arXiv

  26. [34]

    A Systematic Review of Synthetic Data Generation Techniques Using Generative AI,

    M. Goyal and Q. H. Mahmoud, “A Systematic Review of Synthetic Data Generation Techniques Using Generative AI,” Electronics, vol. 13, no. 17, p. 3509, Sep. 2024

    2024

  27. [35]

    Synthetic Data Generation with Large Language Models for Text Classification: Potential and Limitations,

    Z. Li, H. Zhu, Z. Lu, and M. Yin, “Synthetic Data Generation with Large Language Models for Text Classification: Potential and Limitations,”arXiv preprint arXiv:2310.07849, 2023

    2023 arXiv

  28. [36]

    Synthetic Dialogue Dataset Generation using LLM Agents,

    Y . Abdullin, D. Molla-Aliod, B. Ofoghi, J. Yearwood, and Q. Li, “Synthetic Dialogue Dataset Generation using LLM Agents,”arXiv preprint arXiv:2401.17461, 2024

    2024 arXiv

  29. [37]

    LLM-powered Data Augmentation for Enhanced Cross-lingual Performance,

    C. Whitehouse, M. Choudhury, and A. F. Aji, “LLM-powered Data Augmentation for Enhanced Cross-lingual Performance,” arXiv preprint arXiv:2305.14288, 2023

    2023 arXiv

  30. [38]

    Synthetic Data Generation with LLM for Improved Depression Predic- tion,

    A. Kang, J. Y . Chen, Z. Lee-Youngzie, and S. Fu, “Synthetic Data Generation with LLM for Improved Depression Predic- tion,”arXiv preprint arXiv:2411.17672, 2024

    2024 arXiv

  31. [39]

    Does Synthetic Data Generation of LLMs Help Clinical Text Mining?

    R. Tang, X. Han, X. Jiang, and X. Hu, “Does Synthetic Data Generation of LLMs Help Clinical Text Mining?”arXiv preprint arXiv:2303.04360, 2023

    2023 arXiv

  32. [40]

    Zero-Shot Relation Triplet Extraction via Knowledge-Driven LLM Syn- thetic Data Generation,

    L. He, H. Zhang, J. Liu, K. Sun, and Q. Zhang, “Zero-Shot Relation Triplet Extraction via Knowledge-Driven LLM Syn- thetic Data Generation,” inInt. Conf. Intell. Comput. (ICIC). Springer, 2024, pp. 329–340

    2024

  33. [41]

    LawGPT: Knowledge-Guided Data Gen- eration and Its Application to Legal LLM,

    Z. Zhouet al., “LawGPT: Knowledge-Guided Data Gen- eration and Its Application to Legal LLM,”arXiv preprint arXiv:2502.06572, 2025

    2025 arXiv

  34. [42]

    MedSyn: LLM-based Synthetic Medi- cal Text Generation Framework,

    G. Kumichevet al., “MedSyn: LLM-based Synthetic Medi- cal Text Generation Framework,” inJoint Eur. Conf. Mach. Learn. Knowl. Discovery Databases (ECML PKDD). Vil- nius,Lithuania: Springer, Sep. 2024, pp. 215–230

    2024

  35. [43]

    Constraint-Based Synthetic Data Generation for LLM Math- ematical Reasoning,

    T. Fedoseev, D. I. Dimitrov, T. Gehr, and M. Vechev, “Constraint-Based Synthetic Data Generation for LLM Math- ematical Reasoning,” inWorkshop on Math. Reasoning AI at NeurIPS’24, Vancouver, Canada, Dec. 2024

    2024

  36. [44]

    Enhancing Table Representations with LLM-powered Synthetic Data Generation,

    D. Yang, N. Monaikul, A. Ding, B. Tan, K. Mosaliganti, and G. Iyengar, “Enhancing Table Representations with LLM-powered Synthetic Data Generation,”arXiv preprint arXiv:2411.03356, 2024

    2024 arXiv

  37. [45]

    DiffusionGPT: LLM-Driven Text-to-Image Gen- eration System,

    J. Qinet al., “DiffusionGPT: LLM-Driven Text-to-Image Gen- eration System,”arXiv preprint arXiv:2401.10061, 2024

    2024

  38. [46]

    Medical Video Generation for Disease Progres- sion Simulation,

    X. Caoet al., “Medical Video Generation for Disease Progres- sion Simulation,”arXiv preprint arXiv:2411.11943, 2024

    2024 arXiv

  39. [47]

    GenG: An LLM-Based Generic Time Series Data Generation Approach for Edge Intelligence via Cross-Domain Collabo- ration,

    X. Zhou, Q. Jia, Y . Hu, R. Xie, T. Huang, and F. R. Yu, “GenG: An LLM-Based Generic Time Series Data Generation Approach for Edge Intelligence via Cross-Domain Collabo- ration,” inProc. IEEE Conf. Comput. Commun. Workshops (INFOCOM WKSHPS). Vancouver, Canada: IEEE, May 2024, pp. 1–6

    2024

  40. [48]

    HARMONIC: Harnessing LLMs for Tab- ular Data Synthesis and Privacy Protection,

    Y . Wanget al., “HARMONIC: Harnessing LLMs for Tab- ular Data Synthesis and Privacy Protection,”arXiv preprint arXiv:2408.02927, 2024

    2024 arXiv

  41. [49]

    LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts,

    H. Gani, S. F. Bhat, M. Naseer, S. Khan, and P. Wonka, “LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts,”arXiv preprint arXiv:2310.10640, 2023

    2023 arXiv

  42. [50]

    Generative AI-Enabled Vehicular Networks: Fundamentals, Framework, and Case Study,

    R. Zhanget al., “Generative AI-Enabled Vehicular Networks: Fundamentals, Framework, and Case Study,”IEEE Network, vol. 38, no. 4, pp. 259–267, Jul. 2024

    2024

  43. [51]

    Generative AI agents with large language model for satellite networks via a mixture of experts trans- mission,

    R. Zhanget al., “Generative AI agents with large language model for satellite networks via a mixture of experts trans- mission,”IEEE J. Sel. Areas Commun., vol. 42, no. 12, pp. 3581–3596, Dec. 2024

    2024

  44. [52]

    WirelessLLM: Empowering Large Lan- guage Models Towards Wireless Intelligence,

    J. Shaoet al., “WirelessLLM: Empowering Large Lan- guage Models Towards Wireless Intelligence,”arXiv preprint arXiv:2405.17053, 2024. 19

    2024 arXiv

  45. [53]

    Adapting Foundation Models for Information Synthesis of Wireless Communication Specifications,

    M. Kotaru, “Adapting Foundation Models for Information Synthesis of Wireless Communication Specifications,”arXiv preprint arXiv:2308.04033, 2023

    2023 arXiv

  46. [54]

    Designing Network Algorithms via Large Lan- guage Models,

    Z. Heet al., “Designing Network Algorithms via Large Lan- guage Models,” inProc. ACM Workshop Hot Topics Networks (HotNets). New York, NY , USA: Association for Computing Machinery, 2024, pp. 205–212

    2024

  47. [55]

    POSTER: Simplifying the Networking of Wireless Embedded Systems using a Large Language Model,

    P. Medaranga, D. Shah, S. V . Kandala, and A. Varshney, “POSTER: Simplifying the Networking of Wireless Embedded Systems using a Large Language Model,” inProc. ACM SIGCOMM Posters Demos, Sydney, NSW, Australia, Aug. 2024, pp. 78–80

    2024

  48. [56]

    Domain Tailored Large Language Models for Log Mask Prediction in Cellular Network Diagnostics,

    S. Taheri, A. Ihalage, P. Mishra, S. Coaker, F. Muhammad, and H. Al-Raweshidy, “Domain Tailored Large Language Models for Log Mask Prediction in Cellular Network Diagnostics,” IEEE Trans. Netw. Serv. Manage., early access 2025

    2025

  49. [57]

    TelecomGPT: A Framework to Build Telecom-Specfic Large Language Models,

    H. Zouet al., “TelecomGPT: A Framework to Build Telecom-Specfic Large Language Models,”arXiv preprint arXiv:2407.09424, 2024

    2024 arXiv

  50. [58]

    Empowering Large Language Models in Wireless Communication: A Novel Dataset and Fine-Tuning Framework,

    Y . Linet al., “Empowering Large Language Models in Wireless Communication: A Novel Dataset and Fine-Tuning Framework,”arXiv preprint arXiv:2501.09631, 2025

    2025 arXiv

  51. [59]

    Mobile-LLaMA: Instruction Fine-Tuning Open-Source LLM for Network Anal- ysis in 5G Networks,

    K. B. Kan, H. Mun, G. Cao, and Y . Lee, “Mobile-LLaMA: Instruction Fine-Tuning Open-Source LLM for Network Anal- ysis in 5G Networks,”IEEE Network, vol. 38, pp. 76–83, Sep. 2024

    2024

  52. [60]

    ORANSight-2.0: Foundational LLMs for O-RAN,

    P. Gajjar and V . K. Shah, “ORANSight-2.0: Foundational LLMs for O-RAN,”arXiv preprint arXiv:2503.05200, 2025

    2025 arXiv

  53. [61]

    LLM Agents as 6G Orchestrator: A Paradigm for Task-Oriented Physical-Layer Automation,

    Z. Xiaoet al., “LLM Agents as 6G Orchestrator: A Paradigm for Task-Oriented Physical-Layer Automation,”arXiv preprint arXiv:2410.03688, 2024

    2024 arXiv

  54. [62]

    Large Language Model (LLM)-Enabled In- Context Learning for Wireless Network Optimization: A Case Study of Power Control,

    H. Zhouet al., “Large Language Model (LLM)-Enabled In- Context Learning for Wireless Network Optimization: A Case Study of Power Control,”arXiv preprint arXiv:2408.00214, 2024

    2024 arXiv

  55. [63]

    Leveraging Large Language Models for Wireless Symbol Detection via In-Context Learn- ing,

    M. Abbas, K. Kar, and T. Chen, “Leveraging Large Language Models for Wireless Symbol Detection via In-Context Learn- ing,”arXiv preprint arXiv:2409.00124, 2024

    2024 arXiv

  56. [64]

    Self- Refined Generative Foundation Models for Wireless Traffic Prediction,

    C. Hu, H. Zhou, D. Wu, X. Chen, J. Yan, and X. Liu, “Self- Refined Generative Foundation Models for Wireless Traffic Prediction,”arXiv preprint arXiv:2408.10390, 2024

    2024

  57. [65]

    Adaptive Resource Alloca- tion Optimization Using Large Language Models in Dynamic Wireless Environments,

    H. Noh, B. Shim, and H. J. Yang, “Adaptive Resource Alloca- tion Optimization Using Large Language Models in Dynamic Wireless Environments,”arXiv preprint arXiv:2502.02287, 2025

    2025 arXiv

  58. [66]

    LLM4CP: Adapting Large Language Models for Channel Prediction,

    B. Liu, X. Liu, S. Gao, X. Cheng, and L. Yang, “LLM4CP: Adapting Large Language Models for Channel Prediction,”J. Commun. Inf. Networks, vol. 9, no. 2, pp. 113–125, Jun. 2024

    2024

  59. [67]

    CSI-LLM: A Novel Downlink Channel Prediction Method Aligned with LLM Pre- Training,

    S. Fan, Z. Liu, X. Gu, and H. Li, “CSI-LLM: A Novel Downlink Channel Prediction Method Aligned with LLM Pre- Training,”arXiv preprint arXiv:2409.00005, 2024

    2024 arXiv

  60. [68]

    Beam Prediction Based on Large Language models,

    Y . Sheng, K. Huang, L. Liang, P. Liu, S. Jin, and G. Y . Li, “Beam Prediction Based on Large Language models,”IEEE Wireless Commun. Lett., early access 2025

    2025

  61. [69]

    Exploring the Potential of Large Language Models for Massive MIMO CSI Feedback,

    Y . Cui, J. Guo, C.-K. Wen, S. Jin, and E. Tong, “Exploring the Potential of Large Language Models for Massive MIMO CSI Feedback,”arXiv preprint arXiv:2501.10630, 2025

    2025 arXiv

  62. [70]

    Large AI Model for Delay-Doppler Domain Channel Prediction in 6G OTFS-Based Vehicular Networks,

    J. Xueet al., “Large AI Model for Delay-Doppler Domain Channel Prediction in 6G OTFS-Based Vehicular Networks,” arXiv preprint arXiv:2503.01116, 2025

    2025 arXiv

  63. [71]

    Netllm: Adapting Large Language Models for Networking,

    D. Wuet al., “Netllm: Adapting Large Language Models for Networking,” inProc. ACM SIGCOMM Conf. (SIGCOMM), Sydney, NSW, Australia, Aug. 2024, pp. 661–678

    2024

  64. [72]

    LLM4WM: Adapting LLM for Wireless Multi-Tasking,

    X. Liu, S. Gao, B. Liu, X. Cheng, and L. Yang, “LLM4WM: Adapting LLM for Wireless Multi-Tasking,”arXiv preprint arXiv:2501.12983, 2025

    2025 arXiv

  65. [73]

    Large Language Model En- abled Multi-Task Physical Layer Network,

    T. Zheng and L. Dai, “Large Language Model En- abled Multi-Task Physical Layer Network,”arXiv preprint arXiv:2412.20772, 2024

    2024 arXiv

  66. [74]

    Semantic Communications: Principles and Challenges,

    Z. Qin, X. Tao, J. Lu, W. Tong, and G. Y . Li, “Semantic Communications: Principles and Challenges,”arXiv preprint arXiv:2201.01389, 2021

    2021 arXiv

  67. [75]

    Language-Oriented Communication with Semantic Coding and Knowledge Distillation for Text-to-Image Generation,

    H. Nam, J. Park, J. Choi, M. Bennis, and S.-L. Kim, “Language-Oriented Communication with Semantic Coding and Knowledge Distillation for Text-to-Image Generation,” inProc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP). Seoul, Korea: IEEE, Apr. 2024, pp. 13 506–13 510

    2024

  68. [76]

    Semantic Importance- Aware Communications Using Pre-Trained Language Models,

    S. Guo, Y . Wang, S. Li, and N. Saeed, “Semantic Importance- Aware Communications Using Pre-Trained Language Models,” IEEE Commun. Lett., vol. 27, no. 9, pp. 2328–2332, Sep. 2023

    2023

  69. [77]

    Large Language Model Enabled Semantic Communication Systems,

    Z. Wanget al., “Large Language Model Enabled Semantic Communication Systems,”arXiv preprint arXiv:2407.14112, 2024

    2024 arXiv

  70. [78]

    On Large Language Model Based Joint Source Channel Coding for Semantic Communication,

    S. R. Pokhrel and A. Walid, “On Large Language Model Based Joint Source Channel Coding for Semantic Communication,” inInt. Conf. Found. Large Lang. Models (FLLM). Dubai, United Arab Emirates: IEEE, 2024, pp. 322–329

    2024

  71. [79]

    Semantic communications using foundation models: Design approaches and open issues,

    P. Jiang, C.-K. Wen, X. Yi, X. Li, S. Jin, and J. Zhang, “Semantic communications using foundation models: Design approaches and open issues,”IEEE Wireless Commun., vol. 31, no. 3, pp. 76–84, Jun. 2024

    2024

  72. [80]

    Large AI Model Empowered Multimodal Semantic Communications,

    F. Jianget al., “Large AI Model Empowered Multimodal Semantic Communications,”IEEE Commun. Mag., Jan. 2025

    2025

  73. [81]

    Generative Al-aided Joint Training-free Secure Semantic Communications via Multi-modal Prompts,

    H. Duet al., “Generative Al-aided Joint Training-free Secure Semantic Communications via Multi-modal Prompts,” inProc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP). Seoul, Korea: IEEE, Apr. 2024, pp. 12 896–12 900

    2024

  74. [82]

    Semantic Communication Based on Large Language Model for Underwater Image Transmission,

    W. Chenet al., “Semantic Communication Based on Large Language Model for Underwater Image Transmission,”arXiv preprint arXiv:2408.12616, 2024

    2024 arXiv

  75. [83]

    Latency-Aware Generative Semantic Communi- cations With Pre-Trained Diffusion Models,

    L. Qiao, M. B. Mashhadi, Z. Gao, C. H. Foh, P. Xiao, and M. Bennis, “Latency-Aware Generative Semantic Communi- cations With Pre-Trained Diffusion Models,”IEEE Wireless Commun. Lett., vol. 13, no. 10, pp. 2652–2656, Oct. 2024

    2024

  76. [84]

    LaMoSC: Large Language Model-Driven Semantic Communication Sys- tem for Visual Transmission,

    Y . Zhao, Y . Yue, S. Hou, B. Cheng, and Y . Huang, “LaMoSC: Large Language Model-Driven Semantic Communication Sys- tem for Visual Transmission,”IEEE Trans. Cognit. Commun. Networking, vol. 10, no. 6, pp. 2005–2018, Dec. 2024

    2005

  77. [85]

    Per- sonalizing Semantic Communication: A Foundation Model Approach,

    Z. Chen, H. H. Yang, K. F. E. Chong, and T. Q. Quek, “Per- sonalizing Semantic Communication: A Foundation Model Approach,” inIEEE Workshop Signal Process. Adv. Wireless Commun. (SPAWC). Lucca, Italy: IEEE, Sep. 2024, pp. 846– 850

    2024

  78. [86]

    M 3SC: A Generic Dataset for Mixed Multi- Modal (MMM) Sensing and Communication Integration,

    X. Chenget al., “M 3SC: A Generic Dataset for Mixed Multi- Modal (MMM) Sensing and Communication Integration,” China Commun., vol. 20, no. 11, pp. 13–29, Nov. 2023

    2023

  79. [87]

    UniAudio: An Audio Foundation Model Toward Universal Audio Generation,

    D. Yanget al., “UniAudio: An Audio Foundation Model Toward Universal Audio Generation,”arXiv preprint arXiv:2310.00704, 2023

    2023 arXiv

  80. [88]

    MaskGIT: Masked Generative Image Transformer,

    H. Chang, H. Zhang, L. Jiang, C. Liu, and W. T. Freeman, “MaskGIT: Masked Generative Image Transformer,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), New Orleans, LA, USA, Jun. 2022, pp. 11 315–11 325

    2022

  81. [89]

    A Vision-Language Foundation Model for The Generation of Realistic Chest X-ray Images,

    C. Bluethgenet al., “A Vision-Language Foundation Model for The Generation of Realistic Chest X-ray Images,”Nat. Biomed. Eng., vol. 9, pp. 494–506, Aug. 2024

    2024

  82. [90]

    Mitigating Data Scarcity in Time Series Analysis: A Foun- dation Model with Series-Symbol Data Generation,

    W. Wang, K. Wu, Y . B. Li, D. Wang, X. Zhang, and J. Liu, “Mitigating Data Scarcity in Time Series Analysis: A Foun- dation Model with Series-Symbol Data Generation,”arXiv preprint arXiv:2502.15466, 2025

    2025 arXiv

  83. [91]

    VILA-U: a Unified Foundation Model Inte- grating Visual Understanding and Generation,

    Y . Wuet al., “VILA-U: a Unified Foundation Model Inte- grating Visual Understanding and Generation,”arXiv preprint arXiv:2409.04429, 2024

    2024 arXiv

  84. [92]

    MR- Gen: Diffusion-based Controllable Data Engine for MRI Seg- mentation towards Unannotated Modalities,

    H. Wu, Z. Zhao, Y . Zhang, W. Xie, and Y . Wang, “MR- Gen: Diffusion-based Controllable Data Engine for MRI Seg- mentation towards Unannotated Modalities,”arXiv preprint arXiv:2412.04106, 2024

    2024 arXiv

  85. [93]

    Revolutionizing wireless networks with self- 20 supervised learning: A pathway to intelligent communica- tions,

    Z. Yanget al., “Revolutionizing wireless networks with self- 20 supervised learning: A pathway to intelligent communica- tions,”IEEE Wireless Commun., early access 2025

    2025

  86. [94]

    Low- dimensional Representation Learning for Wireless CSI-based Localisation,

    A. Salihu, S. Schwarz, A. Pikrakis, and M. Rupp, “Low- dimensional Representation Learning for Wireless CSI-based Localisation,” inInt. Conf. Wireless Mobile Comput. Netw. Commun. (WiMob). IEEE, Oct. 2020, pp. 1–6

    2020

  87. [95]

    Triplet-Based Wireless Channel Charting: Architecture and Experiments,

    P. Ferrand, A. Decurninge, L. G. Ordonez, and M. Guillaud, “Triplet-Based Wireless Channel Charting: Architecture and Experiments,”IEEE J. Sel. Areas Commun., vol. 39, no. 8, pp. 2361–2373, Aug. 2021

    2021

  88. [96]

    Contrastive Self-Supervised Learning for Wireless Power Control,

    N. Naderializadeh, “Contrastive Self-Supervised Learning for Wireless Power Control,” inProc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP). IEEE, 2021, pp. 4965– 4969

    2021

  89. [97]

    Self- Supervised Deep Learning for mmWave Beam Steering Ex- ploiting Sub-6 GHz Channels,

    I. Chafaa, R. Negrel, E. V . Belmega, and M. Debbah, “Self- Supervised Deep Learning for mmWave Beam Steering Ex- ploiting Sub-6 GHz Channels,”IEEE Trans. Wireless Com- mun., vol. 21, no. 10, pp. 8803–8816, Oct. 2022

    2022

  90. [98]

    Self-Supervised RF Signal Representation Learning for NextG Signal Classification With Deep Learn- ing,

    K. Davaslioglu, S. Boztas ¸, M. C. Ertem, Y . E. Sagduyu, and E. Ayanoglu, “Self-Supervised RF Signal Representation Learning for NextG Signal Classification With Deep Learn- ing,”IEEE Wireless Commun. Lett., vol. 12, no. 1, pp. 65–69, Jan. 2022

    2022

  91. [99]

    A Transformer based Self-supervised Learning Framework for Robust Time-frequency Localization in Concurrent Cognitive Scenario,

    R. Zhao, Y . Ruan, Y . Li, T. Li, R. Zhang, and P. Xiao, “A Transformer based Self-supervised Learning Framework for Robust Time-frequency Localization in Concurrent Cognitive Scenario,”IEEE Trans. Wireless Commun., 2025

    2025

  92. [100]

    A Self-Supervised Learning-Based Channel Estimation for IRS- Aided Communication Without Ground Truth,

    Z. Zhang, T. Ji, H. Shi, C. Li, Y . Huang, and L. Yang, “A Self-Supervised Learning-Based Channel Estimation for IRS- Aided Communication Without Ground Truth,”IEEE Trans. Wireless Commun., vol. 22, no. 8, pp. 5446–5460, Aug. 2023

    2023

  93. [101]

    Leveraging Self-Supervised Learning for MIMO-OFDM Channel Representation and Generation,

    Z. Liuet al., “Leveraging Self-Supervised Learning for MIMO-OFDM Channel Representation and Generation,”arXiv preprint arXiv:2407.07702, 2024

    2024 arXiv

  94. [102]

    Realistic Channel Models Pre-training,

    Y . Huangfuet al., “Realistic Channel Models Pre-training,” inIEEE Globecom Workshops (GC Wkshps). Hawaii, USA: IEEE, Dec. 2019, pp. 1–6

    2019

  95. [103]

    Large Wireless Model (LWM): A Foundation Model for Wireless Channels,

    S. Alikhani, G. Charan, and A. Alkhateeb, “Large Wireless Model (LWM): A Foundation Model for Wireless Channels,” arXiv preprint arXiv:2411.08872, 2024

    2024 arXiv

  96. [104]

    WiFo: Wire- less Foundation Model for Channel Prediction,

    B. Liu, S. Gao, X. Liu, X. Cheng, and L. Yang, “WiFo: Wire- less Foundation Model for Channel Prediction,”Sci. China Inf. Sci., early access 2025

    2025

  97. [105]

    BERT4MIMO: A Foundation Model using BERT Architecture for Massive MIMO Channel State Information Prediction,

    F. O. Catak, M. Kuzlu, and U. Cali, “BERT4MIMO: A Foundation Model using BERT Architecture for Massive MIMO Channel State Information Prediction,”arXiv preprint arXiv:2501.01802, 2025

    2025 arXiv

  98. [106]

    Prompt-Enabled Large AI Models for CSI Feedback,

    J. Guo, Y . Cui, C.-K. Wen, and S. Jin, “Prompt-Enabled Large AI Models for CSI Feedback,”arXiv preprint arXiv:2501.10629, 2025

    2025 arXiv

  99. [107]

    Self-Supervised and Invariant Representations for Wireless Localization,

    A. Salihu, M. Rupp, and S. Schwarz, “Self-Supervised and Invariant Representations for Wireless Localization,”IEEE Trans. Wireless Commun., vol. 23, no. 8, pp. 8281–8296, Aug. 2024

    2024

  100. [108]

    Building 6G Radio Foundation Models with Transformer Architec- tures,

    A. Aboulfotouh, A. Eshaghbeigi, and H. Abou-Zeid, “Building 6G Radio Foundation Models with Transformer Architec- tures,”arXiv preprint arXiv:2411.09996, 2024

    2024 arXiv

  101. [109]

    Finding the missing data: A bert-inspired approach against package loss in wireless sensing,

    Z. Zhao, T. Chen, F. Meng, H. Li, X. Li, and G. Zhu, “Finding the missing data: A bert-inspired approach against package loss in wireless sensing,” inProc. IEEE Conf. Comput. Commun. Workshops (INFOCOM WKSHPS). IEEE, 2024, pp. 1–6

    2024

  102. [110]

    Mining Limited Data Sufficiently: A BERT-inspired Approach for CSI Time Series Application in Wireless Communication and Sensing,

    Z. Zhao, F. Meng, H. Li, X. Li, and G. Zhu, “Mining Limited Data Sufficiently: A BERT-inspired Approach for CSI Time Series Application in Wireless Communication and Sensing,” arXiv preprint arXiv:2412.06861, 2024

    2024

  103. [111]

    A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency,

    J. Jiang, W. Yu, Y . Li, Y . Gao, and S. Xu, “A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency,”arXiv preprint arXiv:2502.11965, 2025

    2025 arXiv

  104. [112]

    6G-Oriented CSI-Based Multi-Modal Pre- Ttaining and Downstream Task Adaptation Paradigm,

    T. Jiaoet al., “6G-Oriented CSI-Based Multi-Modal Pre- Ttaining and Downstream Task Adaptation Paradigm,” inInt. Conf. Commun. Workshops (ICC Workshops). Denver, CO, USA: IEEE, Jun. 2024, pp. 1389–1394

    2024

  105. [113]

    Timer: Generative Pre-trained Transformers Are Large Time Series Models,

    Y . Liu, H. Zhang, C. Li, X. Huang, J. Wang, and M. Long, “Timer: Generative Pre-trained Transformers Are Large Time Series Models,”arXiv preprint arXiv:2402.02368, 2024

    2024 arXiv

  106. [114]

    Time-MoE: Billion-Scale Time Series Foun- dation Models with Mixture of Experts,

    X. Shiet al., “Time-MoE: Billion-Scale Time Series Foun- dation Models with Mixture of Experts,”arXiv preprint arXiv:2409.16040, 2024

    2024 arXiv

  107. [115]

    When Scaling Meets LLM Finetuning: The Effect of Data, Model and Finetuning Method,

    B. Zhang, Z. Liu, C. Cherry, and O. Firat, “When Scaling Meets LLM Finetuning: The Effect of Data, Model and Finetuning Method,”arXiv preprint arXiv:2402.17193, 2024

    2024 arXiv

  108. [116]

    A Survey on Knowledge Distillation of Large Language Models,

    X. Xuet al., “A Survey on Knowledge Distillation of Large Language Models,”arXiv preprint arXiv:2402.13116, 2024

    2024 arXiv

  109. [117]

    Addressing the Curse of Mobility in Massive MIMO With Prony-Based Angular-Delay Domain Channel Predictions,

    H. Yin, H. Wang, Y . Liu, and D. Gesbert, “Addressing the Curse of Mobility in Massive MIMO With Prony-Based Angular-Delay Domain Channel Predictions,”IEEE J. Sel. Areas Commun., vol. 38, no. 12, pp. 2903–2917, Dec. 2020

    2020

  110. [118]

    A LiDAR-aided channel model for vehicular intelligent sensing-communication integration,

    Z. Huang, L. Bai, M. Sun, and X. Cheng, “A LiDAR-aided channel model for vehicular intelligent sensing-communication integration,”IEEE Trans. Intell. Transp. Syst., vol. 25, no. 12, pp. 20 105–20 119, Dec. 2024

    2024

  111. [119]

    Scatterer recognition for multi-modal intelligent vehicular channel modeling via Synesthesia of Machines,

    Z. Huang, L. Bai, Z. Han, and X. Cheng, “Scatterer recognition for multi-modal intelligent vehicular channel modeling via Synesthesia of Machines,”IEEE Wireless Commun., early access 2025

    2025

  112. [120]

    Llm4sg: Large language models for scatterer generation via synesthesia of machines,

    Z. Han, L. Bai, Z. Huang, and X. Cheng, “Llm4sg: Large language models for scatterer generation via synesthesia of machines,”arXiv preprint arXiv:2505.17879, 2025

    2025 arXiv

  113. [121]

    Language Models are Unsupervised Mul- titask Learners,

    A. Radfordet al., “Language Models are Unsupervised Mul- titask Learners,”OpenAI blog, vol. 1, no. 8, p. 9, Feb. 2019

    2019

  114. [122]

    Parameter-Efficient Tuning on Layer Normalization for Pre-trained Language Models,

    W. Qi, Y .-P. Ruan, Y . Zuo, and T. Li, “Parameter-Efficient Tuning on Layer Normalization for Pre-trained Language Models,”arXiv preprint arXiv:2211.08682, 2022

    2022 arXiv

  115. [123]

    Deep Learning-Based Channel Estimation,

    M. Soltani, V . Pourahmadi, A. Mirzaei, and H. Sheikhzadeh, “Deep Learning-Based Channel Estimation,”IEEE Commun. Lett., vol. 23, no. 4, pp. 652–655, Apr. 2019

    2019

  116. [124]

    Attention Aided CSI Wireless Localization,

    A. Salihu, S. Schwarz, and M. Rupp, “Attention Aided CSI Wireless Localization,” inIEEE Workshop Signal Process. Adv. Wireless Commun. (SPAWC). Oulu, Finland: IEEE, Jul. 2022, pp. 1–5

    2022

  117. [125]

    Deep Residual Learning for Image Recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Las Vegas, NV , USA, Jun. 2016, pp. 770–778

    2016

  118. [126]

    Cross- Stitch Networks for Multi-task Learning,

    I. Misra, A. Shrivastava, A. Gupta, and M. Hebert, “Cross- Stitch Networks for Multi-task Learning,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Las Vegas, NV , USA, Jun. 2016, pp. 3994–4003

    2016

  119. [127]

    Datasets for Large Language Models: A Comprehensive Survey,

    Y . Liu, J. Cao, C. Liu, K. Ding, and L. Jin, “Datasets for Large Language Models: A Comprehensive Survey,”arXiv preprint arXiv:2402.18041, 2024

    2024 arXiv

  120. [128]

    Vision Meets Robotics: The Kitti Dataset,

    A. Geiger, P. Lenz, C. Stiller, and R. Urtasun, “Vision Meets Robotics: The Kitti Dataset,”Int. J. Rob. Res., vol. 32, no. 11, pp. 1231–1237, Nov. 2013

    2013

  121. [129]

    DeepSense 6G: A Large-Scale Real- World Multi-Modal Sensing and Communication Dataset,

    A. Alkhateebet al., “DeepSense 6G: A Large-Scale Real- World Multi-Modal Sensing and Communication Dataset,” IEEE Commun. Mag., vol. 61, no. 9, pp. 122–128, Sep. 2023

    2023

  122. [130]

    ViWi: A Deep Learning Dataset Framework for Vision-Aided Wire- less Communications,

    M. Alrabeiah, A. Hredzak, Z. Liu, and A. Alkhateeb, “ViWi: A Deep Learning Dataset Framework for Vision-Aided Wire- less Communications,” inProc. IEEE Veh. Technol. Conf. (VTC2020-Spring). Antwerp, Belgium: IEEE, May 2020, pp. 1–5

    2020

  123. [131]

    Cross-modal learning for image-guided point cloud shape completion,

    E. Aiello, D. Valsesia, and E. Magli, “Cross-modal learning for image-guided point cloud shape completion,”Advances in Neural Information Processing Systems, vol. 35, pp. 37 349– 37 362, 2022

    2022

  124. [132]

    Autofed: Heterogeneity-aware federated multimodal learning for robust 21 autonomous driving,

    T. Zheng, A. Li, Z. Chen, H. Wang, and J. Luo, “Autofed: Heterogeneity-aware federated multimodal learning for robust 21 autonomous driving,” inProceedings of the 29th annual in- ternational conference on mobile computing and networking, 2023, pp. 1–15

    2023

  125. [133]

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale,

    A. Dosovitskiyet al., “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale,”arXiv preprint arXiv:2010.11929, 2020

    2010 arXiv

  126. [134]

    Scaling Proprioceptive-Visual Learning with Heterogeneous Pre- trained Transformers,

    L. Wang, X. Chen, J. Zhao, and K. He, “Scaling Proprioceptive-Visual Learning with Heterogeneous Pre- trained Transformers,” inAdv. Neural Inf. Process. Syst. (NeurIPS), vol. 37, Vancouver, Canada, Dec. 2024, pp. 124 420–124 450

    2024

  127. [135]

    Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity,

    W. Fedus, B. Zoph, and N. Shazeer, “Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity,”J. Mach. Learn. Res., vol. 23, no. 120, pp. 1–39, Jan. 2022

    2022

  128. [136]

    Mixtral of Experts,

    A. Q. Jianget al., “Mixtral of Experts,”arXiv preprint arXiv:2401.04088, 2024

    2024 arXiv

  129. [137]

    DeepSeekMoE: Towards Ultimate Expert Spe- cialization in Mixture-of-Experts Language Models,

    D. Daiet al., “DeepSeekMoE: Towards Ultimate Expert Spe- cialization in Mixture-of-Experts Language Models,”arXiv preprint arXiv:2401.06066, 2024

    2024 arXiv

  130. [138]

    Glu Variants Improve Transformer,

    N. Shazeer, “Glu Variants Improve Transformer,”arXiv preprint arXiv:2002.05202, 2020

    2002 arXiv

  131. [139]

    LLaMA: Open and Efficient Foundation Language Models,

    H. Touvronet al., “LLaMA: Open and Efficient Foundation Language Models,”arXiv preprint arXiv:2302.13971, 2023

    2023 arXiv

  132. [140]

    FlashAt- tention: Fast and Memory-Efficient Exact Attention with IO- Awareness,

    T. Dao, D. Fu, S. Ermon, A. Rudra, and C. R ´e, “FlashAt- tention: Fast and Memory-Efficient Exact Attention with IO- Awareness,” inAdv. Neural Inf. Process. Syst. (NeurIPS), vol. 35, Vancouver, Canada, Jun. 2022, pp. 16 344–16 359

    2022

  133. [141]

    GQA: Training Generalized Multi- Query Transformer Models from Multi-Head Checkpoints,

    J. Ainslie, J. Lee-Thorp, M. De Jong, Y . Zemlyanskiy, F. Lebr´on, and S. Sanghai, “GQA: Training Generalized Multi- Query Transformer Models from Multi-Head Checkpoints,” arXiv preprint arXiv:2305.13245, 2023

    2023 arXiv

  134. [142]

    Swin Transformer: Hierarchical Vision Trans- former using Shifted Windows,

    Z. Liuet al., “Swin Transformer: Hierarchical Vision Trans- former using Shifted Windows,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Nashville, TN, USA, Jun. 2021, pp. 10 012–10 022

    2021

  135. [143]

    Imagenet: A Large-Scale Hierarchical Image Database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei- Fei, “Imagenet: A Large-Scale Hierarchical Image Database,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR). Miami, FL, USA: IEEE, Jun. 2009, pp. 248–255

    2009

  136. [144]

    Deep Learning for Joint Channel Estimation and Feedback in Massive MIMO Systems,

    J. Guo, T. Chen, S. Jin, G. Y . Li, X. Wang, and X. Hou, “Deep Learning for Joint Channel Estimation and Feedback in Massive MIMO Systems,”Digital Commun. Networks, vol. 10, no. 1, pp. 83–93, Feb. 2024

    2024

  137. [145]

    Masked Autoencoders Are Scalable Vision Learners,

    K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick, “Masked Autoencoders Are Scalable Vision Learners,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), New Orleans, LA, USA, Jun. 2022, pp. 16 000–16 009

    2022

  138. [146]

    A Simple Framework for Contrastive Learning of Visual Representa- tions,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A Simple Framework for Contrastive Learning of Visual Representa- tions,” inInt. Conf. Mach. Learn. (ICML), Jul. 2020, pp. 1597– 1607

    2020

  139. [147]

    Cloud-Edge-Terminal Collaborative AIGC for Autonomous Driving,

    J. Zhang, Z. Wei, B. Liu, X. Wang, Y . Yu, and R. Zhang, “Cloud-Edge-Terminal Collaborative AIGC for Autonomous Driving,”IEEE Wireless Commun., vol. 31, no. 4, pp. 40–47, 2024

    2024

  140. [148]

    The LuViRA Dataset: Synchronized Vision, Radio, and Audio Sensors for Indoor Localization,

    I. Yamanet al., “The LuViRA Dataset: Synchronized Vision, Radio, and Audio Sensors for Indoor Localization,” inProc. Int. Conf. Robot. Automat. (ICRA). Yokohama, Japan: IEEE, May 2024, pp. 11 920–11 926

    2024

  141. [149]

    A Distributed Massive MIMO Channel Sounder for

    F. Euchner, M. Gauger, S. D ¨orner, and S. ten Brink, “A Distributed Massive MIMO Channel Sounder for ”Big CSI Data”-driven Machine Learning,” inProc. Int. ITG Workshop Smart Antennas (WSA), Eurecom, France, Nov. 2021, pp. 1–6

    2021

  142. [150]

    Parameter-Efficient Transfer Learning for NLP,

    N. Houlsbyet al., “Parameter-Efficient Transfer Learning for NLP,” inProc. Int. Conf. Mach. Learn. (ICML), Los Angeles, CA, USA, Jul. 2019, pp. 2790–2799

    2019

Pith tools

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