arXiv subjects

Yi Zhang

Publications and source records attributed to Yi Zhang.

At least 739 records · Page 41Linked to original sources

arXiv2023

Progress and outlook on advanced fly scans based on Mamba

Development related to PandABox-based fly scans is an important part of the active work on Mamba, the software framework for beamline experiments at the High Energy Photon Source (HEPS); presented in this paper is the progress of our development, and some outlook for advanced fly scans based on knowledge learned during the process. By treating fly scans as a collaboration between a few loosely coupled subsystems - motors / mechanics, detectors / data processing, sequencer devices like PandABox - systematic analyses of issues in fly scans are conducted. Interesting products of these analyses include a general-purpose software-based fly-scan mechanism, a general way to design undulator-monochromator fly scans, a sketch of how to practically implement online tuning of fly-scan behaviours based on processing of the data acquired, and many more. Based on the results above, an architectural discussion on >=10kHz fly scans is given.

physics.ins-det↗

arXiv2023

Ultrabroadband, high color purity multispectral color filter arrays

Multispectral imagers that capture spatial and spectral information are of growing importance in various fields, particularly in remote sensing and metrology. To enable integrated snapshot multispectral imagers and eliminate the drawbacks of traditional systems, such as bulkiness and slow scanning mechanisms, miniature, broadband multispectral filter arrays with narrow line widths, high transmission, and CMOS compatibility are essential. However, current miniature filter arrays, primarily based on diffraction nanostructures, suffer from limitations such as small working bandwidth, low transmission, poor color purity, and sensitivity to polarization and incident angles. To address these challenges, we present a high-order Fabry-Perot Multispectral Filter Array (MSFA) with selective peak suppression, leveraging subwavelength structures for filter tuning without changing the physical thickness and employing an ultrathin metal layer to exploit high-order resonances, significantly extending the working range and spectral resolution. High color purity across a broad range (400nm-1000nm) is made possible through optical absorption from polysilicon and selective suppression from a platinum layer. The fabricated color filter arrays cover wavelengths from 622nm to 960nm, with Full Width Half Maximum (FWHM) ranging from 13nm to 31nm and average transmissions exceeding 60%. Furthermore, these filters can be downscaled to sizes compatible with modern CMOS imagers, reaching dimensions as small as 1um. The introduction of a resonance combining design further extends the working range (455nm-960nm), aligning with the capabilities of silicon photodetectors. Its adaptability across wavelength ranges and potential for tunable applications hold promise for transformative imaging and display technologies across a wide spectrum.

physics.optics↗

arXiv2023

Learning threshold neurons via the "edge of stability"

Existing analyses of neural network training often operate under the unrealistic assumption of an extremely small learning rate. This lies in stark contrast to practical wisdom and empirical studies, such as the work of J. Cohen et al. (ICLR 2021), which exhibit startling new phenomena (the "edge of stability" or "unstable convergence") and potential benefits for generalization in the large learning rate regime. Despite a flurry of recent works on this topic, however, the latter effect is still poorly understood. In this paper, we take a step towards understanding genuinely non-convex training dynamics with large learning rates by performing a detailed analysis of gradient descent for simplified models of two-layer neural networks. For these models, we provably establish the edge of stability phenomenon and discover a sharp phase transition for the step size below which the neural network fails to learn "threshold-like" neurons (i.e., neurons with a non-zero first-layer bias). This elucidates one possible mechanism by which the edge of stability can in fact lead to better generalization, as threshold neurons are basic building blocks with useful inductive bias for many tasks.

cs.LG↗

arXiv2023

Privacy-Preserving Encrypted Low-Dose CT Denoising

Deep learning (DL) has made significant advancements in tomographic imaging, particularly in low-dose computed tomography (LDCT) denoising. A recent trend involves servers training powerful models with large amounts of self-collected private data and providing application programming interfaces (APIs) for users, such as Chat-GPT. To avoid model leakage, users are required to upload their data to the server model, but this way raises public concerns about the potential risk of privacy disclosure, especially for medical data. Hence, to alleviate related concerns, in this paper, we propose to directly denoise LDCT in the encrypted domain to achieve privacy-preserving cloud services without exposing private data to the server. To this end, we employ homomorphic encryption to encrypt private LDCT data, which is then transferred to the server model trained with plaintext LDCT for further denoising. However, since traditional operations, such as convolution and linear transformation, in DL methods cannot be directly used in the encrypted domain, we transform the fundamental mathematic operations in the plaintext domain into the operations in the encrypted domain. In addition, we present two interactive frameworks for linear and nonlinear models in this paper, both of which can achieve lossless operating. In this way, the proposed methods can achieve two merits, the data privacy is well protected and the server model is free from the risk of model leakage. Moreover, we provide theoretical proof to validate the lossless property of our framework. Finally, experiments were conducted to demonstrate that the transferred contents are well protected and cannot be reconstructed. The code will be released once the paper is accepted.

cs.CR↗

arXiv2023

Application of Deep Learning Methods Combined with Physical Background in Wide Field of View Imaging Atmospheric Cherenkov Telescopes

The HADAR experiment, which will be constructed in Tibet, China, combines the wide-angle advantages of traditional EAS array detectors with the high sensitivity advantages of focused Cherenkov detectors. Its physics objective is to observe transient sources such as gamma-ray bursts and counterparts of gravitational waves. The aim of this study is to utilize the latest AI technology to enhance the sensitivity of the HADAR experiment. We have built training datasets and models with distinctive creativity by incorporating relevant physical theories for various applications. They are able to determine the kind, energy, and direction of incident particles after careful design. We have obtained a background identification accuracy of 98.6%, a relative energy reconstruction error of 10.0%, and an angular resolution of 0.22-degrees in a test dataset at 10 TeV. These findings demonstrate the enormous potential for enhancing the precision and dependability of detector data analysis in astrophysical research. Thanks to deep learning techniques, the HADAR experiment's observational sensitivity to the Crab Nebula has surpassed that of MAGIC and H.E.S.S. at energies below 0.5 TeV and remains competitive with conventional narrow-field Cherenkov telescopes at higher energies. Additionally, our experiment offers a fresh approach to dealing with strongly connected scattered data.

astro-ph.IM↗

arXiv2023

Complex semiclassical theory for non-Hermitian quantum systems

Non-Hermitian quantum systems exhibit fascinating characteristics such as non-Hermitian topological phenomena and skin effect, yet their studies are limited by the intrinsic difficulties associated with their eigenvalue problems, especially in larger systems and higher dimensions. In Hermitian systems, the semiclassical theory has played an active role in analyzing spectrum, eigenstate, phase, transport properties, etc. Here, we establish a complex semiclassical theory applicable to non-Hermitian quantum systems by an analytical continuation of the physical variables such as momentum, position, time, and energy in the equations of motion and quantization condition to the complex domain. Further, we propose a closed-orbit scheme and physical meaning under such complex variables. We demonstrate that such a framework straightforwardly yields complex energy spectra and quantum states, topological phases and transitions, and even the skin effect in non-Hermitian quantum systems, presenting an unprecedented perspective toward nontrivial non-Hermitian physics, even with larger systems and higher dimensions.

cond-mat.mes-hall↗

arXiv2023

A moving least square immersed boundary method for SPH with thin-walled structures

This paper presents a novel method for smoothed particle hydrodynamics (SPH) with thin-walled structures. Inspired by the direct forcing immersed boundary method, this method employs a moving least square method to guarantee the smoothness of velocity near the structure surface. It simplifies thin-walled structure simulations by eliminating the need for multiple layers of boundary particles, and improves computational accuracy and stability in three-dimensional scenarios. Supportive three-dimensional numerical results are provided, including the impulsively started plate and the flow past a cylinder. Results of the impulsively started test demonstrate that the proposed method obtains smooth velocity and pressure in the, as well as a good match to the references results of the vortex wake development. In addition, results of the flow past cylinder test show that the proposed method avoids mutual interference on both side of the boundary, remains stable for three-dimensional simulations while accurately calculating the forces acting on structure.

physics.flu-dyn↗

arXiv2023

Fermionic Higher-form Symmetries

In this paper, we explore a new type of global symmetries$-$the fermionic higher-form symmetries. They are generated by topological operators with fermionic parameter, which act on fermionic extended objects. We present a set of field theory examples with fermionic higher-form symmetries, which are constructed from fermionic tensor fields. They include the free fermionic tensor theories, a new type of fermionic topological quantum field theories, as well as the exotic 6d (4,0) theory. We also discuss the gauging and breaking of such global symmetries and the relation to the no global symmetry swampland conjecture.

hep-th↗

arXiv2023

CoreDiff: Contextual Error-Modulated Generalized Diffusion Model for Low-Dose CT Denoising and Generalization

Low-dose computed tomography (CT) images suffer from noise and artifacts due to photon starvation and electronic noise. Recently, some works have attempted to use diffusion models to address the over-smoothness and training instability encountered by previous deep-learning-based denoising models. However, diffusion models suffer from long inference times due to the large number of sampling steps involved. Very recently, cold diffusion model generalizes classical diffusion models and has greater flexibility. Inspired by the cold diffusion, this paper presents a novel COntextual eRror-modulated gEneralized Diffusion model for low-dose CT (LDCT) denoising, termed CoreDiff. First, CoreDiff utilizes LDCT images to displace the random Gaussian noise and employs a novel mean-preserving degradation operator to mimic the physical process of CT degradation, significantly reducing sampling steps thanks to the informative LDCT images as the starting point of the sampling process. Second, to alleviate the error accumulation problem caused by the imperfect restoration operator in the sampling process, we propose a novel ContextuaL Error-modulAted Restoration Network (CLEAR-Net), which can leverage contextual information to constrain the sampling process from structural distortion and modulate time step embedding features for better alignment with the input at the next time step. Third, to rapidly generalize to a new, unseen dose level with as few resources as possible, we devise a one-shot learning framework to make CoreDiff generalize faster and better using only a single LDCT image (un)paired with NDCT. Extensive experimental results on two datasets demonstrate that our CoreDiff outperforms competing methods in denoising and generalization performance, with a clinically acceptable inference time. Source code is made available at https://github.com/qgao21/CoreDiff.

eess.IV↗

arXiv2023

Textbooks Are All You Need

We introduce phi-1, a new large language model for code, with significantly smaller size than competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for 4 days on 8 A100s, using a selection of ``textbook quality" data from the web (6B tokens) and synthetically generated textbooks and exercises with GPT-3.5 (1B tokens). Despite this small scale, phi-1 attains pass@1 accuracy 50.6% on HumanEval and 55.5% on MBPP. It also displays surprising emergent properties compared to phi-1-base, our model before our finetuning stage on a dataset of coding exercises, and phi-1-small, a smaller model with 350M parameters trained with the same pipeline as phi-1 that still achieves 45% on HumanEval.

cs.CL↗

arXiv2023

Resilient Model-Free Asymmetric Bipartite Consensus for Nonlinear Multi-Agent Systems against DoS Attacks

In this letter, we study an unified resilient asymmetric bipartite consensus (URABC) problem for nonlinear multi-agent systems with both cooperative and antagonistic interactions under denial-of-service (DoS) attacks. We first prove that the URABC problem is solved by stabilizing the neighborhood asymmetric bipartite consensus error. Then, we develop a distributed compact form dynamic linearization method to linearize the neighborhood asymmetric bipartite consensus error. By using an extended discrete state observer to enhance the robustness against unknown dynamics and an attack compensation mechanism to eliminate the adverse effects of DoS attacks, we finally propose a distributed resilient model-free adaptive control algorithm to solve the URABC problem. A numerical example validates the proposed results.

eess.SY↗

arXiv2023

User Simulation with Large Language Models for Evaluating Task-Oriented Dialogue

One of the major impediments to the development of new task-oriented dialogue (TOD) systems is the need for human evaluation at multiple stages and iterations of the development process. In an effort to move toward automated evaluation of TOD, we propose a novel user simulator built using recently developed large pretrained language models (LLMs). In order to increase the linguistic diversity of our system relative to the related previous work, we do not fine-tune the LLMs used by our system on existing TOD datasets; rather we use in-context learning to prompt the LLMs to generate robust and linguistically diverse output with the goal of simulating the behavior of human interlocutors. Unlike previous work, which sought to maximize goal success rate (GSR) as the primary metric of simulator performance, our goal is a system which achieves a GSR similar to that observed in human interactions with TOD systems. Using this approach, our current simulator is effectively able to interact with several TOD systems, especially on single-intent conversational goals, while generating lexically and syntactically diverse output relative to previous simulators that rely upon fine-tuned models. Finally, we collect a Human2Bot dataset of humans interacting with the same TOD systems with which we experimented in order to better quantify these achievements.

cs.CL↗

arXiv2023

Conversation Style Transfer using Few-Shot Learning

Conventional text style transfer approaches focus on sentence-level style transfer without considering contextual information, and the style is described with attributes (e.g., formality). When applying style transfer in conversations such as task-oriented dialogues, existing approaches suffer from these limitations as context can play an important role and the style attributes are often difficult to define in conversations. In this paper, we introduce conversation style transfer as a few-shot learning problem, where the model learns to perform style transfer by observing only a few example dialogues in the target style. We propose a novel in-context learning approach to solve the task with style-free dialogues as a pivot. Human evaluation shows that by incorporating multi-turn context, the model is able to match the target style while having better appropriateness and semantic correctness compared to utterance/sentence-level style transfer. Additionally, we show that conversation style transfer can also benefit downstream tasks. For example, in multi-domain intent classification tasks, the F1 scores improve after transferring the style of training data to match the style of the test data.

cs.CL↗

arXiv2023

A Survey on Privacy in Graph Neural Networks: Attacks, Preservation, and Applications

Graph Neural Networks (GNNs) have gained significant attention owing to their ability to handle graph-structured data and the improvement in practical applications. However, many of these models prioritize high utility performance, such as accuracy, with a lack of privacy consideration, which is a major concern in modern society where privacy attacks are rampant. To address this issue, researchers have started to develop privacy-preserving GNNs. Despite this progress, there is a lack of a comprehensive overview of the attacks and the techniques for preserving privacy in the graph domain. In this survey, we aim to address this gap by summarizing the attacks on graph data according to the targeted information, categorizing the privacy preservation techniques in GNNs, and reviewing the datasets and applications that could be used for analyzing/solving privacy issues in GNNs. We also outline potential directions for future research in order to build better privacy-preserving GNNs.

cs.LG↗

arXiv2023

Real spectra in one-dimensional single-band non-Hermitian Hamiltonians

In general, the energy spectrum of a non-Hermitian system turns out to be complex, which is not so satisfactory since the time evolution of eigenstates with complex eigenvalues is either exponentially growing or decaying. Here we provide a sufficient and necessary condition of the real spectrum under open boundary conditions for one-dimensional non-Hermitian tight-binding Hamiltonians. The necessity is directly related to the fact that the generalized Brillouin zone in one dimension is a closed loop with the origin in its interior. For some simple models, we demonstrate the sufficiency by analytically determining when the preimage of the characteristic polynomial contains a loop and showing that this loop is just the generalized Brillouin zone itself. For more complex models, the numerical results are shown. Our results indicate that real spectra are more common than one may have expected in non-Hermitian systems and are helpful for designing non-Hermitian models with real spectra.

cond-mat.mtrl-sci↗

arXiv2023

Wave breaking for the generalized Fornberg-Whitham equation

This paper aims to show that the Cauchy problem of the Burgers equation with a weakly dispersive perturbation involving the Bessel potential (generalization of the Fornberg-Whitham equation) can exhibit wave breaking for initial data with large slope. We also comment on the dispersive properties of the equation.

math.AP↗

arXiv2023

Efficient and quantum-adaptive machine learning with fermion neural networks

Classical artificial neural networks have witnessed widespread successes in machine-learning applications. Here, we propose fermion neural networks (FNNs) whose physical properties, such as local density of states or conditional conductance, serve as outputs, once the inputs are incorporated as an initial layer. Comparable to back-propagation, we establish an efficient optimization, which entitles FNNs to competitive performance on challenging machine-learning benchmarks. FNNs also directly apply to quantum systems, including hard ones with interactions, and offer in-situ analysis without preprocessing or presumption. Following machine learning, FNNs precisely determine topological phases and emergent charge orders. Their quantum nature also brings various advantages: quantum correlation entitles more general network connectivity and insight into the vanishing gradient problem, quantum entanglement opens up novel avenues for interpretable machine learning, etc.

quant-ph↗

arXiv2023

Supervised Stochastic Neighbor Embedding Using Contrastive Learning

Stochastic neighbor embedding (SNE) methods $t$-SNE, UMAP are two most popular dimensionality reduction methods for data visualization. Contrastive learning, especially self-supervised contrastive learning (SSCL), has showed great success in embedding features from unlabeled data. The conceptual connection between SNE and SSCL has been exploited. In this work, within the scope of preserving neighboring information of a dataset, we extend the self-supervised contrastive approach to the fully-supervised setting, allowing us to effectively leverage label information. Clusters of samples belonging to the same class are pulled together in low-dimensional embedding space, while simultaneously pushing apart clusters of samples from different classes.

cs.LG↗