arXiv subjects

Rui Wen

Publications and source records attributed to Rui Wen.

At least 55 records · Page 3Linked to original sources

arXiv2024

CMT: Cross Modulation Transformer with Hybrid Loss for Pansharpening

Pansharpening aims to enhance remote sensing image (RSI) quality by merging high-resolution panchromatic (PAN) with multispectral (MS) images. However, prior techniques struggled to optimally fuse PAN and MS images for enhanced spatial and spectral information, due to a lack of a systematic framework capable of effectively coordinating their individual strengths. In response, we present the Cross Modulation Transformer (CMT), a pioneering method that modifies the attention mechanism. This approach utilizes a robust modulation technique from signal processing, integrating it into the attention mechanism's calculations. It dynamically tunes the weights of the carrier's value (V) matrix according to the modulator's features, thus resolving historical challenges and achieving a seamless integration of spatial and spectral attributes. Furthermore, considering that RSI exhibits large-scale features and edge details along with local textures, we crafted a hybrid loss function that combines Fourier and wavelet transforms to effectively capture these characteristics, thereby enhancing both spatial and spectral accuracy in pansharpening. Extensive experiments demonstrate our framework's superior performance over existing state-of-the-art methods. The code will be publicly available to encourage further research.

cs.CV↗

arXiv2024

Efficient Data-Free Model Stealing with Label Diversity

Machine learning as a Service (MLaaS) allows users to query the machine learning model in an API manner, which provides an opportunity for users to enjoy the benefits brought by the high-performance model trained on valuable data. This interface boosts the proliferation of machine learning based applications, while on the other hand, it introduces the attack surface for model stealing attacks. Existing model stealing attacks have relaxed their attack assumptions to the data-free setting, while keeping the effectiveness. However, these methods are complex and consist of several components, which obscure the core on which the attack really depends. In this paper, we revisit the model stealing problem from a diversity perspective and demonstrate that keeping the generated data samples more diverse across all the classes is the critical point for improving the attack performance. Based on this conjecture, we provide a simplified attack framework. We empirically signify our conjecture by evaluating the effectiveness of our attack, and experimental results show that our approach is able to achieve comparable or even better performance compared with the state-of-the-art method. Furthermore, benefiting from the absence of redundant components, our method demonstrates its advantages in attack efficiency and query budget.

cs.CR↗

arXiv2024

Estimates on the convergence of expansions at finite baryon chemical potentials

Convergence of three different expansion schemes at finite baryon chemical potentials, including the conventional Taylor expansion, the Padé approximants, and the $T'$ expansion proposed recently in lattice QCD simulations, have been investigated in a low energy effective theory within the fRG approach. It is found that the $T'$ expansion or the Padé approximants would hardly improve the convergence of expansion in comparison to the conventional Taylor expansion, within the expansion orders considered in this work. Furthermore, we find that the consistent regions of the three different expansions are in agreement with the convergence radius of the Lee-Yang edge singularities.

hep-ph↗

arXiv2024

Chiral phase transition and spin alignment of vector meson in the Polarized-Polyakov-loop Nambu-Jona-Lasinio model under rotation

By using the extrapolation method, a polarized Polykov-loop potential at finite real angular velocity is constructed from the lattice results at finite imaginary angular velocity. The chiral and deconfinement phase transitions under rotation have been simultaneously investigated in the Polarized-Polyakov-loop Nambu-Jona-Lasinio (PPNJL) model. It is observed that both critical temperatures of deconfinement and chiral phase transition increase with the angular velocity, which is in consistent with lattice results. The spin alignment of vector meson has the negative deviation of $ρ_{00} -1/3$ under rotation, and the deviation in the PPNJL model is much more significant than that in the NJL model and the quark coalescence model, which revealing the important role of rotating gluons on the quark polarization.

hep-ph↗

arXiv2024

The influence of hadronic rescatterings on the net-baryon number fluctuations

Fluctuations of conserved charges, such as the net-baryon number fluctuations, are influenced by different dynamical evolution processes. In this paper, we investigate the influence of hadronic rescatterings on different orders of cumulants of the net-baryon number distribution. At the start of hadronic rescatterings, we introduce net-baryon number distributions reconstructed based on net-baryon cumulants of different orders obtained from computation in functional renormalization group (FRG), where the distributions were constructed using the maximum entropy method. This way we introduce the critical fluctuations of Quantum Chromodynamics (QCD) into the AMPT model. Firstly, we find that hadronic rescatterings have distinct effects on cumulant ratios of different orders for the net-baryon number. Secondly, we observe that the effect of hadronic rescatterings is more significant for critical fluctuations than dynamical fluctuations, because the two-, three- and four-particle correlation functions due to critical fluctuations are weakened more significantly by hadronic rescatterings.

nucl-th↗

arXiv2023

Classification of 1+1D gapless symmetry protected phases via topological holography

Symmetry topological field theory (SymTFT) gives a holographic correspondence between systems with a global symmetry and a higher-dimensional topological field theory. In this framework, classification of gapped phases of matter in spacetime dimension 1+1D correspond to classifications of mechanisms to confine the SymTFT by condensing anyons. In this work, we extend these results to characterize gapless symmetry-protected topological states: symmetry-enriched gapless phases or critical points that exhibit edge modes protected by symmetry and topology. We establish a one-to-one correspondence between 1+1D bosonic gSPTs, and partially-confined boundaries of 2+1D SymTFTs. From general physical considerations, we determine the set of data and consistency conditions required to define a 1+1D gSPT, and show that this data precisely matches that of symmetry-preserving partial confinement (or partially gapped boundaries) of 2+1D quantum double models. We illustrate this correspondence through a dimensional reduction (thin-slab) construction, which enables a physically-intuitive derivation of how properties of the gSPT such as edge modes, emergent anomalies, and stability to perturbations arise from the SymTFT perspective.ditions required to define a 1+1D gSPT and show that they fully determine the physics of the gSPT including edge modes and emergent anomaly.

cond-mat.str-el↗

arXiv2023

Generated Distributions Are All You Need for Membership Inference Attacks Against Generative Models

Generative models have demonstrated revolutionary success in various visual creation tasks, but in the meantime, they have been exposed to the threat of leaking private information of their training data. Several membership inference attacks (MIAs) have been proposed to exhibit the privacy vulnerability of generative models by classifying a query image as a training dataset member or nonmember. However, these attacks suffer from major limitations, such as requiring shadow models and white-box access, and either ignoring or only focusing on the unique property of diffusion models, which block their generalization to multiple generative models. In contrast, we propose the first generalized membership inference attack against a variety of generative models such as generative adversarial networks, [variational] autoencoders, implicit functions, and the emerging diffusion models. We leverage only generated distributions from target generators and auxiliary non-member datasets, therefore regarding target generators as black boxes and agnostic to their architectures or application scenarios. Experiments validate that all the generative models are vulnerable to our attack. For instance, our work achieves attack AUC $>0.99$ against DDPM, DDIM, and FastDPM trained on CIFAR-10 and CelebA. And the attack against VQGAN, LDM (for the text-conditional generation), and LIIF achieves AUC $>0.90.$ As a result, we appeal to our community to be aware of such privacy leakage risks when designing and publishing generative models.

cs.CR↗

arXiv2023

The rotation effect on the thermodynamics of the QCD matter

In this study, we investigate the impact of rotation on the thermodynamic characteristics of QCD matter using the three-flavor NJL model. We examine the temperature, quark chemical potential, and angular velocity dependencies of key thermodynamic quantities, such as the trace anomaly, specific heat, speed of sound, angular momentum, and moment of inertia. As the main finding of our analysis, we observe that the speed of sound exhibits a nonmonotonic behavior as the angular velocity changes.

hep-ph↗

arXiv2023

Quantifying Privacy Risks of Prompts in Visual Prompt Learning

Large-scale pre-trained models are increasingly adapted to downstream tasks through a new paradigm called prompt learning. In contrast to fine-tuning, prompt learning does not update the pre-trained model's parameters. Instead, it only learns an input perturbation, namely prompt, to be added to the downstream task data for predictions. Given the fast development of prompt learning, a well-generalized prompt inevitably becomes a valuable asset as significant effort and proprietary data are used to create it. This naturally raises the question of whether a prompt may leak the proprietary information of its training data. In this paper, we perform the first comprehensive privacy assessment of prompts learned by visual prompt learning through the lens of property inference and membership inference attacks. Our empirical evaluation shows that the prompts are vulnerable to both attacks. We also demonstrate that the adversary can mount a successful property inference attack with limited cost. Moreover, we show that membership inference attacks against prompts can be successful with relaxed adversarial assumptions. We further make some initial investigations on the defenses and observe that our method can mitigate the membership inference attacks with a decent utility-defense trade-off but fails to defend against property inference attacks. We hope our results can shed light on the privacy risks of the popular prompt learning paradigm. To facilitate the research in this direction, we will share our code and models with the community.

cs.CR↗

arXiv2023

Last One Standing: A Comparative Analysis of Security and Privacy of Soft Prompt Tuning, LoRA, and In-Context Learning

Large Language Models (LLMs) are powerful tools for natural language processing, enabling novel applications and user experiences. However, to achieve optimal performance, LLMs often require adaptation with private data, which poses privacy and security challenges. Several techniques have been proposed to adapt LLMs with private data, such as Low-Rank Adaptation (LoRA), Soft Prompt Tuning (SPT), and In-Context Learning (ICL), but their comparative privacy and security properties have not been systematically investigated. In this work, we fill this gap by evaluating the robustness of LoRA, SPT, and ICL against three types of well-established attacks: membership inference, which exposes data leakage (privacy); backdoor, which injects malicious behavior (security); and model stealing, which can violate intellectual property (privacy and security). Our results show that there is no silver bullet for privacy and security in LLM adaptation and each technique has different strengths and weaknesses.

cs.CR↗

arXiv2023

Spiking-Diffusion: Vector Quantized Discrete Diffusion Model with Spiking Neural Networks

Spiking neural networks (SNNs) have tremendous potential for energy-efficient neuromorphic chips due to their binary and event-driven architecture. SNNs have been primarily used in classification tasks, but limited exploration on image generation tasks. To fill the gap, we propose a Spiking-Diffusion model, which is based on the vector quantized discrete diffusion model. First, we develop a vector quantized variational autoencoder with SNNs (VQ-SVAE) to learn a discrete latent space for images. In VQ-SVAE, image features are encoded using both the spike firing rate and postsynaptic potential, and an adaptive spike generator is designed to restore embedding features in the form of spike trains. Next, we perform absorbing state diffusion in the discrete latent space and construct a spiking diffusion image decoder (SDID) with SNNs to denoise the image. Our work is the first to build the diffusion model entirely from SNN layers. Experimental results on MNIST, FMNIST, KMNIST, Letters, and Cifar10 demonstrate that Spiking-Diffusion outperforms the existing SNN-based generation model. We achieve FIDs of 37.50, 91.98, 59.23, 67.41, and 120.5 on the above datasets respectively, with reductions of 58.60\%, 18.75\%, 64.51\%, 29.75\%, and 44.88\% in FIDs compared with the state-of-art work. Our code will be available at \url{https://github.com/Arktis2022/Spiking-Diffusion}.

cs.NE↗

arXiv2023

Functional renormalization group study of neutral and charged pion under magnetic fields in the quark-meson model

We calculated the masses of neutral and charged pion and pion decay constants under an extra magnetic field at zero temperature. The quantum fluctuations are integrated through the functional renormalization group. We consider the quark and meson propagators in the Landau level representation and weak-field expansion, respectively. The neutral pion mass monotonically decreases with the magnetic field, while the charged pion mass monotonically increases with the magnetic field. The pion decay constant and the quark mass show the magnetic catalysis behavior at vanishing temperature. The neutral pion mass and pion decay constant are quantitatively in agreement with the lattice QCD results in the region of $eB < 1.2 {\rm GeV}^2$, and no non-monotonic mass behavior for charged pion has been observed in this framework.

hep-ph↗

arXiv2023

Floquet codes and phases in twist-defect networks

We introduce a class of models, dubbed paired twist-defect networks, that generalize the structure of Kitaev's honeycomb model for which there is a direct equivalence between: i) Floquet codes (FCs), ii) adiabatic loops of gapped Hamiltonians, and iii) unitary loops or Floquet-enriched topological orders (FETs) many-body localized phases. This formalism allows one to apply well-characterized topological index theorems for FETs to understand the dynamics of FCs, and to rapidly assess the code properties of many FC models. As an application, we show that the Honeycomb Floquet code of Haah and Hastings is governed by an irrational value of the chiral Floquet index, which implies a topological obstruction to forming a simple, logical boundary with the same periodicity as the bulk measurement schedule. In addition, we construct generalizations of the Honeycomb Floquet code exhibiting arbitrary anyon-automorphism dynamics for general types of Abelian topological order.

quant-ph↗

arXiv2023

Reconstruction of baryon number distributions

The maximum entropy method (MEM) and the Gaussian process (GP) regression, which are both well-suited for the treatment of inverse problems, are used to reconstruct net-baryon number distributions based on a finite number of cumulants of the distribution. Baryon number distributions across the chiral phase transition are reconstructed. It is found that with the increase of the order of cumulants, distribution in the long tails, i.e., far away from the central number, would become more and more important. We also reconstruct the distribution function based on the experimentally measured cumulants at the collision energy $\sqrt{s_{NN}}=7.7$ GeV. Given the sizable error of the fourth-order cumulant measured in experiments, the calculation of MEM shows that with the increasing fourth-order cumulant, there is another peak in the distribution function developed in the region of large baryon number. This unnaturalness observed in the reconstructed distribution function might in turn be used to constrain the cumulants measured in experiments.

hep-ph↗

arXiv2023

Bulk-boundary correspondence for intrinsically-gapless SPTs from group cohomology

Intrinsically gapless symmetry protected topological phases (igSPT) are gapless systems with SPT edge states with properties that could not arise in a gapped system with the same symmetry and dimensionality. igSPT states arise from gapless systems in which an anomaly in the low-energy (IR) symmetry group emerges from an extended anomaly-free microscopic (UV) symmetry We construct a general framework for constructing lattice models for igSPT phases with emergent anomalies classified by group cohomology, and establish a direct connection between the emergent anomaly, group-extension, and topological edge states by gauging the extending symmetry. In many examples, the edge-state protection has a physically transparent mechanism: the extending UV symmetry operations pump lower dimensional SPTs onto the igSPT edge, tuning the edge to a (multi)critical point between different SPTs protected by the IR symmetry. In two- and three- dimensional systems, an additional possibility is that the emergent anomaly can be satisfied by an anomalous symmetry-enriched topological order, which we call a quotient-symmetry enriched topological order (QSET) that is sharply distinguished from the non-anomalous UV SETs by an edge phase transition. We construct exactly solvable lattice models with QSET order.

cond-mat.str-el↗

arXiv2022

Dynamic Backdoor Attacks Against Machine Learning Models

Machine learning (ML) has made tremendous progress during the past decade and is being adopted in various critical real-world applications. However, recent research has shown that ML models are vulnerable to multiple security and privacy attacks. In particular, backdoor attacks against ML models have recently raised a lot of awareness. A successful backdoor attack can cause severe consequences, such as allowing an adversary to bypass critical authentication systems. Current backdooring techniques rely on adding static triggers (with fixed patterns and locations) on ML model inputs which are prone to detection by the current backdoor detection mechanisms. In this paper, we propose the first class of dynamic backdooring techniques against deep neural networks (DNN), namely Random Backdoor, Backdoor Generating Network (BaN), and conditional Backdoor Generating Network (c-BaN). Triggers generated by our techniques can have random patterns and locations, which reduce the efficacy of the current backdoor detection mechanisms. In particular, BaN and c-BaN based on a novel generative network are the first two schemes that algorithmically generate triggers. Moreover, c-BaN is the first conditional backdooring technique that given a target label, it can generate a target-specific trigger. Both BaN and c-BaN are essentially a general framework which renders the adversary the flexibility for further customizing backdoor attacks. We extensively evaluate our techniques on three benchmark datasets: MNIST, CelebA, and CIFAR-10. Our techniques achieve almost perfect attack performance on backdoored data with a negligible utility loss. We further show that our techniques can bypass current state-of-the-art defense mechanisms against backdoor attacks, including ABS, Februus, MNTD, Neural Cleanse, and STRIP.

cs.CR↗

arXiv2022

Hyper-order baryon number fluctuations at finite temperature and density

Fluctuations of conserved charges are sensitive to the QCD phase transition and a possible critical endpoint in the phase diagram at finite density. In this work, we compute the baryon number fluctuations up to tenth order at finite temperature and density. This is done in a QCD-assisted effective theory that accurately captures the quantum- and in-medium effects of QCD at low energies. A direct computation at finite density allows us to assess the applicability of expansions around vanishing density. By using different freeze-out scenarios in heavy-ion collisions, we translate these results into baryon number fluctuations as a function of collision energy. We show that a non-monotonic energy dependence of baryon number fluctuations can arise in the non-critical crossover region of the phase diagram. Our results compare well with recent experimental measurements of the kurtosis and the sixth-order cumulant of the net-proton distribution from the STAR collaboration. They indicate that the experimentally observed non-monotonic energy dependence of fourth-order net-proton fluctuations is highly non-trivial. It could be an experimental signature of an increasingly sharp chiral crossover and may indicate a QCD critical point. The physics implications and necessary upgrades of our analysis are discussed in detail.

hep-ph↗

arXiv2022

Finding Influential Instances for Distantly Supervised Relation Extraction

Distant supervision (DS) is a strong way to expand the datasets for enhancing relation extraction (RE) models but often suffers from high label noise. Current works based on attention, reinforcement learning, or GAN are black-box models so they neither provide meaningful interpretation of sample selection in DS nor stability on different domains. On the contrary, this work proposes a novel model-agnostic instance sampling method for DS by influence function (IF), namely REIF. Our method identifies favorable/unfavorable instances in the bag based on IF, then does dynamic instance sampling. We design a fast influence sampling algorithm that reduces the computational complexity from $\mathcal{O}(mn)$ to $\mathcal{O}(1)$, with analyzing its robustness on the selected sampling function. Experiments show that by simply sampling the favorable instances during training, REIF is able to win over a series of baselines that have complicated architectures. We also demonstrate that REIF can support interpretable instance selection.

cs.LG↗