arXiv subjects

Jun Hu

Publications and source records attributed to Jun Hu.

At least 55 records · Page 3Linked to original sources

arXiv2026

Cross-Modal Attention Network with Dual Graph Learning in Multimodal Recommendation

Multimedia recommendation systems leverage user-item interactions and multimodal information to capture user preferences, enabling more accurate and personalized recommendations. Despite notable advancements, existing approaches still face two critical limitations: first, shallow modality fusion often relies on simple concatenation, failing to exploit rich synergic intra- and inter-modal relationships; second, asymmetric feature treatment-where users are only characterized by interaction IDs while items benefit from rich multimodal content-hinders the learning of a shared semantic space. To address these issues, we propose a Cross-modal Recursive Attention Network with dual graph Embedding (CRANE). To tackle shallow fusion, we design a core Recursive Cross-Modal Attention (RCA) mechanism that iteratively refines modality features based on cross-correlations in a joint latent space, effectively capturing high-order intra- and inter-modal dependencies. For symmetric multimodal learning, we explicitly construct users' multimodal profiles by aggregating features of their interacted items. Furthermore, CRANE integrates a symmetric dual-graph framework-comprising a heterogeneous user-item interaction graph and a homogeneous item-item semantic graph-unified by a self-supervised contrastive learning objective to fuse behavioral and semantic signals. Despite these complex modeling capabilities, CRANE maintains high computational efficiency. Theoretical and empirical analyses confirm its scalability and high practical efficiency, achieving faster convergence on small datasets and superior performance ceilings on large-scale ones. Comprehensive experiments on four public real-world datasets validate an average 5% improvement in key metrics over state-of-the-art baselines.

cs.IR↗

arXiv2026

UniBiDex: A Unified Teleoperation Framework for Robotic Bimanual Dexterous Manipulation

We present UniBiDex a unified teleoperation framework for robotic bimanual dexterous manipulation that supports both VRbased and leaderfollower input modalities UniBiDex enables realtime contactrich dualarm teleoperation by integrating heterogeneous input devices into a shared control stack with consistent kinematic treatment and safety guarantees The framework employs nullspace control to optimize bimanual configurations ensuring smooth collisionfree and singularityaware motion across tasks We validate UniBiDex on a longhorizon kitchentidying task involving five sequential manipulation subtasks demonstrating higher task success rates smoother trajectories and improved robustness compared to strong baselines By releasing all hardware and software components as opensource we aim to lower the barrier to collecting largescale highquality human demonstration datasets and accelerate progress in robot learning.

cs.RO↗

arXiv2026

Study of Class-Incremental Radio Frequency Fingerprint Recognition Without Storing Exemplars

The rapid proliferation of wireless devices makes robust identity authentication essential. Radio Frequency Fingerprinting (RFF) exploits device-specific, hard-to-forge physical-layer impairments for identification, and is promising for IoT and unmanned systems. In practice, however, new devices continuously join deployed systems while per-class training data are limited. Conventional static training and naive replay of stored exemplars are impractical due to growing class cardinality, storage cost, and privacy concerns. We propose an exemplar-free class-incremental learning framework tailored to RFF recognition. Starting from a pretrained feature extractor, we freeze the backbone during incremental stages and train only a classifier together with lightweight Adapter modules that perform small task-specific feature adjustments. For each class we fit a diagonal Gaussian Mixture Model (GMM) to the backbone features and sample pseudo-features from these fitted distributions to rehearse past classes without storing raw signals. To improve robustness under few-shot conditions we introduce a time-domain random-masking augmentation and adopt a multi-teacher distillation scheme to compress stage-wise Adapters into a single inference Adapter, trading off accuracy and runtime efficiency. We evaluate the method on large, self-collected ADS-B datasets: the backbone is pretrained on 2,175 classes and incremental experiments are run on a disjoint set of 669 classes with multiple rounds and step sizes. Against several representative baselines, our approach consistently yields higher average accuracy and lower forgetting, while using substantially less storage and avoiding raw-data retention. The proposed pipeline is reproducible and provides a practical, low-storage solution for RFF deployment in resource- and privacy-constrained environments.

eess.SP↗

arXiv2026

Edge Truncation Effect Suppression of Ultrawideband Phased Arrays for Radar Application

This letter presents a novel, effective method to suppress the edge truncation effect of ultrawideband tightly coupled dipole linear arrays. To restrain the edge truncation effect within an ultrawideband operating band, a new type of T-shaped metal strip with a resistor is further loaded on the array edges apart from extending the length of the overlapping patches. Besides, the excitation phase of the elements at the array edges is optimized. Full-wave simulation results show that the active standing wave standing ratio of the 2 x 16 tightly coupled dipole linear arrays using the proposed method is significantly optimized to less than 3.5 within a 5:1 [(1.2 to 6) GHz] bandwidth, while scanning up to +/-60° in the E-plane. The effectiveness of the proposed method is experimentally verified by a 2 x 16 linear array prototype.

physics.app-ph↗

arXiv2025

Manifold Function Encoder: Identifying Different Functions Defined on Different Manifolds

We propose the Manifold Function Encoder (MFE) for identifying different functions defined on different manifolds. Both a manifold in Euclidean space and a function defined on this manifold can be viewed as bounded linear functionals on a suitable space of continuous functions. From this perspective, we treat manifold functions as elements of the dual space. By expanding them in the dual space based on appropriate approximating sequence of bases, we obtain a corresponding method for encoding manifold functions, that is MFE. Especially, we prove that MFE achieves super-algebraic convergence based on smooth bases commonly used in spectral methods, such as Legendre polynomials and Fourier basis. We further extend MFE to handle more complex cases, including joint manifold functions of different dimensions and manifold functions with different measures. In addition, we show the approximation theory for MFE-based operator learning, in particular learning the solution mappings of PDEs defined on varying domains, together with several numerical experiments including the 2-d Poisson equation and the 3-d elasticity problem on the real-world bearing.

math.NA↗

arXiv2025

Standard symmetrizing forms and Murphy bases of the cyclotomic Hecke algebras

In this paper we compute the explicit value of the standard symmetrizing form of the cyclotomic Hecke algebra of type $G(\ell,1,n)$ on each Murphy basis element. We do this for both the non-degenerate cyclotomic Hecke algebra of type $G(\ell,1,n)$ and the degenerate cyclotomic Hecke algebra of type $G(\ell,1,n)$. As an application, we obtain some equations on the coefficients which occur in the transition matrix between the Murphy basis elements and the seminormal basis elements of the cyclotomic Hecke algebra of type $G(\ell,1,n)$.

math.RT↗

arXiv2025

NTSFormer: A Self-Teaching Graph Transformer for Multimodal Isolated Cold-Start Node Classification

Isolated cold-start node classification on multimodal graphs is challenging because such nodes have no edges and often have missing modalities (e.g., absent text or image features). Existing methods address structural isolation by degrading graph learning models to multilayer perceptrons (MLPs) for isolated cold-start inference, using a teacher model (with graph access) to guide the MLP. However, this results in limited model capacity in the student, which is further challenged when modalities are missing. In this paper, we propose Neighbor-to-Self Graph Transformer (NTSFormer), a unified Graph Transformer framework that jointly tackles the isolation and missing-modality issues via a self-teaching paradigm. Specifically, NTSFormer uses a cold-start attention mask to simultaneously make two predictions for each node: a "student" prediction based only on self information (i.e., the node's own features), and a "teacher" prediction incorporating both self and neighbor information. This enables the model to supervise itself without degrading to an MLP, thereby fully leveraging the Transformer's capacity to handle missing modalities. To handle diverse graph information and missing modalities, NTSFormer performs a one-time multimodal graph pre-computation that converts structural and feature data into token sequences, which are then processed by Mixture-of-Experts (MoE) Input Projection and Transformer layers for effective fusion. Experiments on public datasets show that NTSFormer achieves superior performance for multimodal isolated cold-start node classification.

cs.LG↗

arXiv2025

Echoless Label-Based Pre-computation for Memory-Efficient Heterogeneous Graph Learning

Heterogeneous Graph Neural Networks (HGNNs) are widely used for deep learning on heterogeneous graphs. Typical end-to-end HGNNs require repetitive message passing during training, limiting efficiency for large-scale real-world graphs. Pre-computation-based HGNNs address this by performing message passing only once during preprocessing, collecting neighbor information into regular-shaped tensors, which enables efficient mini-batch training. Label-based pre-computation methods collect neighbors' label information but suffer from training label leakage, where a node's own label information propagates back to itself during multi-hop message passing - the echo effect. Existing mitigation strategies are memory-inefficient on large graphs or suffer from compatibility issues with advanced message passing methods. We propose Echoless Label-based Pre-computation (Echoless-LP), which eliminates training label leakage with Partition-Focused Echoless Propagation (PFEP). PFEP partitions target nodes and performs echoless propagation, where nodes in each partition collect label information only from neighbors in other partitions, avoiding echo while remaining memory-efficient and compatible with any message passing method. We also introduce an Asymmetric Partitioning Scheme (APS) and a PostAdjust mechanism to address information loss from partitioning and distributional shifts across partitions. Experiments on public datasets demonstrate that Echoless-LP achieves superior performance and maintains memory efficiency compared to baselines.

cs.LG↗

arXiv2025

Generalizing Graph Transformers Across Diverse Graphs and Tasks via Pre-training

Graph pre-training has been concentrated on graph-level tasks involving small graphs (e.g., molecular graphs) or learning node representations on a fixed graph. Extending graph pre-trained models to web-scale graphs with billions of nodes in industrial scenarios, while avoiding negative transfer across graphs or tasks, remains a challenge. We aim to develop a general graph pre-trained model with inductive ability that can make predictions for unseen new nodes and even new graphs. In this work, we introduce a scalable transformer-based graph pre-training framework called PGT (Pre-trained Graph Transformer). Based on the masked autoencoder architecture, we design two pre-training tasks: one for reconstructing node features and the other for reconstructing local structures. Unlike the original autoencoder architecture where the pre-trained decoder is discarded, we propose a novel strategy that utilizes the decoder for feature augmentation. Our framework, tested on the publicly available ogbn-papers100M dataset with 111 million nodes and 1.6 billion edges, achieves state-of-the-art performance, showcasing scalability and efficiency. We have deployed our framework on Tencent's online game data, confirming its capability to pre-train on real-world graphs with over 540 million nodes and 12 billion edges and to generalize effectively across diverse static and dynamic downstream tasks.

cs.LG↗

arXiv2025

Transmitter Identification via Volterra Series Based Radio Frequency Fingerprint

The growing number of wireless devices increases the need for secure network access. Radio Frequency Fingerprinting (RFF), a physical-layer authentication method, offers a promising solution as it requires no cryptography and resists spoofing. However, existing RFF approaches often lack a unified theory and effective feature extraction. Many methods use handcrafted signal features or direct neural network classification, leading to limited generalization and interpretability. In this work, we model the transmitter as a black box and analyze its impact on transmitted signals. By treating the deviation from an ideal signal as hardware-induced distortion, we represent the received signal using a Volterra series, using its kernels to capture linear and nonlinear hardware traits. To manage the high dimensionality of these kernels, we approximate them via wavelet decomposition and estimate coefficients through least-squares fitting. The resulting wavelet coefficients provide compact yet informative hardware representations, which are classified using a complex-valued neural network. Experiments on a public LoRa dataset show state-of-the-art performance, with over 98% accuracy in static channels and above 90% under multipath and Doppler effects. The proposed approach improves both interpretability and generalization across varying channel conditions.

cs.CR↗

arXiv2025

Large Language Diffusion Models

The capabilities of large language models (LLMs) are widely regarded as relying on autoregressive models (ARMs). We challenge this notion by introducing LLaDA, a diffusion model trained from scratch under the pre-training and supervised fine-tuning (SFT) paradigm. LLaDA employs a forward data masking process and a reverse generation process, parameterized by a Transformer to predict masked tokens. It provides a principled generative approach for probabilistic inference by optimizing a likelihood lower bound. Across extensive benchmarks on general tasks, math, code, and so on, LLaDA demonstrates strong scalability and performs comparably to our self-constructed ARM baselines. Remarkably, LLaDA 8B is competitive with strong LLMs like LLaMA3 8B in in-context learning and, after SFT, exhibits impressive instruction-following abilities in case studies such as multi-turn dialogue. Moreover, LLaDA addresses the reversal curse, surpassing GPT-4o in a reversal poem completion task. Our findings show the promise of diffusion models for language modeling at scale and challenge the common assumption that core LLM capabilities discussed above inherently depend on ARMs. Project page and codes: https://ml-gsai.github.io/LLaDA-demo/.

cs.CL↗

arXiv2025

First experimental proof of PET imaging based on multi-anode MCP-PMTs with Cherenkov radiator-integrated window

Improving the coincidence time resolution (CTR) of time-of-flight positron emission tomography (TOF-PET) systems to achieve a higher signal-to-noise ratio (SNR) gain or even direct positron emission imaging (dPEI) is of paramount importance for many advanced new clinical applications of PET imaging. This places higher demands on the timing performance of all aspects of PET systems. One effective approach is to use microchannel plate photomultiplier tubes (MCP-PMTs) for prompt Cherenkov photon detection. In this study, we developed a dual-module Cherenkov PET imaging experimental platform, utilising our proprietary 8 * 8-anode Cherenkov radiator-integrated window MCP-PMTs in combination with custom-designed multi-channel electronics, and designed a specific calibration and correction method for the platform. Using this platform, a CTR of 103 ps FWHM was achieved. We overcame the limitations of single-anode detectors in previous experiments, significantly enhanced imaging efficiency and achieved module-level Cherenkov PET imaging for the first time. Imaging experiments involving radioactive sources and phantoms of various shapes and types were conducted, which preliminarily validated the feasibility and advancement of this imaging method. In addition, the effects of normalisation correction and the interaction probability between the gamma rays and the MCP on the images and experimental results were analysed and verified.

physics.ins-det↗

arXiv2025

LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models

While Masked Diffusion Models (MDMs), such as LLaDA, present a promising paradigm for language modeling, there has been relatively little effort in aligning these models with human preferences via reinforcement learning. The challenge primarily arises from the high variance in Evidence Lower Bound (ELBO)-based likelihood estimates required for preference optimization. To address this issue, we propose Variance-Reduced Preference Optimization (VRPO), a framework that formally analyzes the variance of ELBO estimators and derives bounds on both the bias and variance of preference optimization gradients. Building on this theoretical foundation, we introduce unbiased variance reduction strategies, including optimal Monte Carlo budget allocation and antithetic sampling, that significantly improve the performance of MDM alignment. We demonstrate the effectiveness of VRPO by applying it to LLaDA, and the resulting model, LLaDA 1.5, outperforms its SFT-only predecessor consistently and significantly across mathematical (GSM8K +4.7), code (HumanEval +3.0, MBPP +1.8), and alignment benchmarks (IFEval +4.0, Arena-Hard +4.3). Furthermore, LLaDA 1.5 demonstrates a highly competitive mathematical performance compared to strong language MDMs and ARMs. Project page: https://ml-gsai.github.io/LLaDA-1.5-Demo/.

cs.LG↗

arXiv2025

LLaDA-MoE: A Sparse MoE Diffusion Language Model

We introduce LLaDA-MoE, a large language diffusion model with the Mixture-of-Experts (MoE) architecture, trained from scratch on approximately 20T tokens. LLaDA-MoE achieves competitive performance with significantly reduced computational overhead by maintaining a 7B-parameter capacity while activating only 1.4B parameters during inference. Our empirical evaluation reveals that LLaDA-MoE achieves state-of-the-art performance among diffusion language models with larger parameters, surpassing previous diffusion language models LLaDA, LLaDA 1.5, and Dream across multiple benchmarks. The instruct-tuned model LLaDA-MoE-7B-A1B-Instruct demonstrates capabilities comparable to Qwen2.5-3B-Instruct in knowledge understanding, code generation, mathematical reasoning, agent and alignment tasks, despite using fewer active parameters. Our results show that integrating a sparse MoE architecture into the training objective of masked diffusion language models still brings out MoE's strengths under efficient inference with few active parameters, and opens ample room for further exploration of diffusion language models. LLaDA-MoE models are available at Huggingface.

cs.CL↗

arXiv2025

Finite Element Complexes with Traces Structures: A unified framework for cohomology and bounded interpolation

This paper considers the cohomology and bounded interpolation of nonstandard finite element complexes, e.g. Stokes, Hessian, Elasticity, divdiv. Compared to the standard finite element exterior calculus, the main challenge is the existence of extra smoothness. This paper provides a unified framework for finite element complexes with extra smoothness. The trace structure is introduced to derive the bubble complexes in different dimensions (vertices, edges, faces). It is shown that if the bubble complexes in different dimensions are all exact, then the finite element has the correct cohomology. Moreover, the $L^2$ bounded interpolation can be constructed.

math.NA↗

arXiv2025

The CEPC Clock Issue and Finetuning of the Circumference

The CEPC clock issue is related with the RF frequency coordination between the various accelerator systems and may affect the operation modes of both the accelerator and the detector. The timing structure of CEPC has been restudied with the collaboration of the accelerator team and the detector team. After discussions between two sides, the CEPC bunch structure is set such that the spacings between adjacent bunches in any CEPC operation mode are integer numbers of 23.08 ns. The master CEPC clock will be provided by the accelerator to the detector systems with a frequency of 43.33 MHz, synchronous to the beam. The CEPC detector system relies on the clock to sample physics signal at the right time. It was found that if the circumference of CEPC is slightly changed to 99955.418 m, not only is the orbit length closer to 100 km, but also the detector would benefit more for the first 10-year operation.

physics.acc-ph↗

arXiv2025

JUNO 20-inch PMT and electronics system characterization using large pulses of PMT dark counts at the Pan-Asia testing platform

The main goal of the JUNO experiment is to determine the neutrino mass ordering with a 20kt liquid-scintillator detector. The 20-inch PMT and its 1F3 (one for three) electronics are crucial to realize the excellent energy resolution of at least 3% at 1MeV. The knowledge on the PMT and 1F3 electronics response is critical for detector performance understanding. A study of the JUNO 20-inch PMT and 1F3 electronics system characterization is presented using large pulses of PMT dark count at the Pan-Asia testing platform in China. Thanks to its broad amplitude range and high rate, the large pulse signals are also used to investigate the PMT after pulse response.

physics.ins-det↗

arXiv2025

The coupling of mixed and primal finite element methods for the coupled body-plate problem

This paper considers the coupled problem of a three-dimensional elastic body and a two-dimensional plate, which are rigidly connected at their interface. The plate consists of a plane elasticity model along the longitudinal direction and a plate bending model with Kirchhoff assumptions along the transverse direction. The Hellinger-Reissner formulation is adopted for the body by introducing the stress as an auxiliary variable, while the primal formulation is employed for the plate. The well-posedness of the weak formulation is established. This approach enables direct stress approximations and allows for non-matching meshes at the interface since the continuity condition of displacements acts as a natural boundary condition for the body. Under certain assumptions, discrete stability and error estimates are derived for both conforming and nonconforming finite element methods. Two specific pairs of conforming and nonconforming finite elements are shown to satisfy the required assumptions, respectively. Furthermore, the problem is reduced to an interface problem based on the domain decomposition, which can be solved effectively by a conjugate gradient iteration. Numerical experiments are conducted to validate the theoretical results.

math.NA↗