arXiv subjects

Kun Wang

Publications and source records attributed to Kun Wang.

At least 343 records · Page 19Linked to original sources

arXiv2024

SwapNet: Efficient Swapping for DNN Inference on Edge AI Devices Beyond the Memory Budget

Executing deep neural networks (DNNs) on edge artificial intelligence (AI) devices enables various autonomous mobile computing applications. However, the memory budget of edge AI devices restricts the number and complexity of DNNs allowed in such applications. Existing solutions, such as model compression or cloud offloading, reduce the memory footprint of DNN inference at the cost of decreased model accuracy or autonomy. To avoid these drawbacks, we divide DNN into blocks and swap them in and out in order, such that large DNNs can execute within a small memory budget. Nevertheless, naive swapping on edge AI devices induces significant delays due to the redundant memory operations in the DNN development ecosystem for edge AI devices. To this end, we develop SwapNet, an efficient DNN block swapping middleware for edge AI devices. We systematically eliminate the unnecessary memory operations during block swapping while retaining compatible with the deep learning frameworks, GPU backends, and hardware architectures of edge AI devices. We further showcase the utility of SwapNet via a multi-DNN scheduling scheme. Evaluations on eleven DNN inference tasks in three applications demonstrate that SwapNet achieves almost the same latency as the case with sufficient memory even when DNNs demand 2.32x to 5.81x memory beyond the available budget. The design of SwapNet also provides novel and feasible insights for deploying large language models (LLMs) on edge AI devices in the future.

cs.LG↗

arXiv2024

Existence and nonexistence of minimizer for Thomas-Fermi-Dirac-Von Weizsäcker model on lattice graph

The focus of our paper is to investigate the possibility of a minimizer for the Thomas-Fermi-Dirac-von Weizsäcker model on the lattice graph $\mathbb{Z}^{3}$. The model is described by the following functional: \begin{equation*} E(φ)=\sum_{y\in\mathbb{Z}^{3}}\left(|\nablaφ(y)|^2+ (φ(y))^{\frac{10}{3}}-(φ(y))^{\frac{8}{3}}\right)+ \sum_{x,y\in\mathbb{Z}^{3}\atop ~\ y\neq x\hfill}\frac{φ^2(x)φ^2(y)}{|x-y|}, \end{equation*} with the additional constraint that $\sum\limits_{y\in\mathbb{Z}^{3}} φ^2(y)=m$ is sufficiently small. We also prove the nonexistence of a minimizer provided the mass $m$ is adequately large. Furthermore, we extend our analysis to a subset $Ω\subset \mathbb{Z}^{3}$ and prove the nonexistence of a minimizer for the following functional: \begin{equation*} E(Ω)=|\partialΩ|+\sum_{x,y\inΩ\atop ~y\neq x\hfill}\frac{1}{|x-y|}, \end{equation*} under the constraint that $|Ω|=V$ is sufficiently large.

math.AP↗

arXiv2024

Quantifying the unextendibility of entanglement

Entanglement is a striking feature of quantum mechanics, and it has a key property called unextendibility. In this paper, we present a framework for quantifying and investigating the unextendibility of general bipartite quantum states. First, we define the unextendible entanglement, a family of entanglement measures based on the concept of a state-dependent set of free states. The intuition behind these measures is that the more entangled a bipartite state is, the less entangled each of its individual systems is with a third party. Second, we demonstrate that the unextendible entanglement is an entanglement monotone under two-extendible quantum operations, including local operations and one-way classical communication as a special case. Normalization and faithfulness are two other desirable properties of unextendible entanglement, which we establish here. We further show that the unextendible entanglement provides efficiently computable benchmarks for the rate of exact entanglement or secret key distillation, as well as the overhead of probabilistic entanglement or secret key distillation.

quant-ph↗

arXiv2023

Smuon contribution to muon g-2 in Grand Unified supersymmetric theories

In GUT-scale constrained (GUTc) supersymmetric (SUSY) models, the mass of smuon $\tildeμ_1$ is typically heavier than that of stau $\tildeτ_1$, and stau co-annihilation is a typical annihilation mechanism of dark matter. However, light smuon is more favored by the muon $g-2$ anomaly, thus smuon-neutralino loop contribution to muon $g-2$ is usually smaller than that of sneutrino-chargino. Inspired by the latest muon $g-2$ results, we take the GUTc- Next-to-Minimal Supersymmetric Model (NMSSM) as an example, where the gaugino (Higgs) masses are not unified to the usual parameter $M_{1/2}$ ($M_0$), exploring its possibility of light smuon and its contribution to muon $g-2$. After complicated calculations and discussions, we conclude that in GUTc-NMSSM the smuon can be lighter than stau. In this light-smuon scenario, the contribution of smuon-neutralino loop to the muon $g-2$ can be larger than that of the sneutrino-chargino loop. The annihilation mechanisms of dark matter are dominated by multiple slepton or chargino co-annihilation. In our calculations, we consider also other latest related constraints like Higgs data, SUSY searches, dark matter relic density and direct detections, etc.

hep-ph↗

arXiv2023

Light dark matter confronted with the 95 GeV diphoton excess

The correlation between Higgs-like scalars and light dark matter is an interesting topic, especially now that a $125 GeV$ Higgs was discovered and dark matter (DM) searches got negative results. The $95 GeV$ excess reported by the CMS collaboration with $132 fb^{-1}$ data recently, and the DM search results by XENONnT and LZ collaborations motivate us to revise that. In this work, we study that in the GUT-scale constrained (GUTc) Next-to-Minimal Supersymmetric Model (NMSSM), where most parameters are input at the GUT scale, but with scalar and gaugino masses not unified there. In the calculation we also consider other recent experimental constraints, such as Higgs data, Supersymmetry (SUSY) searches, DM relic density, etc. After detailed analysis and discussion, we find that: (i) The light DM can be bino- or singlino-dominated, but can be mixed with minor components of Higgsino. (ii) Both cases can get right relic density and sizable Higgs invisible decay, by adjusting the dimensionless parameters $λ, κ$, or suitably mixing with Higgsino. (iii) Both cases can have four funnel annihilation mechanisms, i.e., annihilating through $Z, a_1, h_2, h_1$. (iv) Samples with right relic density usually get weak signal of Higgs invisible decay at future lepton collider, but the $95 GeV$ scalar can have sizable $b\bar{b}$ signal.

hep-ph↗

arXiv2023

Fidelity Estimation of Entangled Measurements with Local States

We propose an efficient protocol to estimate the fidelity of an $n$-qubit entangled measurement device, requiring only qubit state preparations and classical data post-processing. It works by measuring the eigenstates of Pauli operators, which are strategically selected according to their importance weights and collectively contributed by all measurement operators. We rigorously analyze the protocol's performance and demonstrate that its sample complexity is uniquely determined by the number of Pauli operators possessing non-zero expectation values with respect to the target measurement. Moreover, from a resource-theoretic perspective, we introduce the stabilizer Rényi entropy of quantum measurements as a precise metric to quantify the inherent difficulty of estimating measurement fidelity.

quant-ph↗

arXiv2023

A New Smoothing Technique for Bang-Bang Optimal Control Problems

Bang-bang control is ubiquitous for Optimal Control Problems (OCPs) where the constrained control variable appears linearly in the dynamics and cost function. Based on the Pontryagin's Minimum Principle, the indirect method is widely used to numerically solve OCPs because it enables to derive the theoretical structure of the optimal control. However, discontinuities in the bang-bang control structure may result in numerical difficulties for gradient-based indirect method. In this case, smoothing or regularization procedures are usually applied to eliminating the discontinuities of bang-bang controls. Traditional smoothing or regularization procedures generally modify the cost function by adding a term depending on a small parameter, or introducing a small error into the state equation. Those procedures may complexify the numerical algorithms or degenerate the convergence performance. To overcome these issues, we propose a bounded smooth function, called normalized L2-norm function, to approximate the sign function in terms of the switching function. The resulting optimal control is smooth and can be readily embedded into the indirect method. Then, the simplicity and improved performance of the proposed method over some existing methods are numerically demonstrated by a minimal-time oscillator problem and a minimal-fuel low-thrust trajectory optimization problem that involves many revolutions.

math.OC↗

arXiv2023

Attend Who is Weak: Enhancing Graph Condensation via Cross-Free Adversarial Training

In this paper, we study the \textit{graph condensation} problem by compressing the large, complex graph into a concise, synthetic representation that preserves the most essential and discriminative information of structure and features. We seminally propose the concept of Shock Absorber (a type of perturbation) that enhances the robustness and stability of the original graphs against changes in an adversarial training fashion. Concretely, (I) we forcibly match the gradients between pre-selected graph neural networks (GNNs) trained on a synthetic, simplified graph and the original training graph at regularly spaced intervals. (II) Before each update synthetic graph point, a Shock Absorber serves as a gradient attacker to maximize the distance between the synthetic dataset and the original graph by selectively perturbing the parts that are underrepresented or insufficiently informative. We iteratively repeat the above two processes (I and II) in an adversarial training fashion to maintain the highly-informative context without losing correlation with the original dataset. More importantly, our shock absorber and the synthesized graph parallelly share the backward process in a free training manner. Compared to the original adversarial training, it introduces almost no additional time overhead. We validate our framework across 8 datasets (3 graph and 5 node classification datasets) and achieve prominent results: for example, on Cora, Citeseer and Ogbn-Arxiv, we can gain nearly 1.13% to 5.03% improvements compare with SOTA models. Moreover, our algorithm adds only about 0.2% to 2.2% additional time overhead over Flicker, Citeseer and Ogbn-Arxiv. Compared to the general adversarial training, our approach improves time efficiency by nearly 4-fold.

cs.LG↗

arXiv2023

Abnormal traffic detection system in SDN based on deep learning hybrid models

Software defined network (SDN) provides technical support for network construction in smart cities, However, the openness of SDN is also prone to more network attacks. Traditional abnormal traffic detection methods have complex algorithms and find it difficult to detect abnormalities in the network promptly, which cannot meet the demand for abnormal detection in the SDN environment. Therefore, we propose an abnormal traffic detection system based on deep learning hybrid model. The system adopts a hierarchical detection technique, which first achieves rough detection of abnormal traffic based on port information. Then it uses wavelet transform and deep learning techniques for fine detection of all traffic data flowing through suspicious switches. The experimental results show that the proposed detection method based on port information can quickly complete the approximate localization of the source of abnormal traffic. the accuracy, precision, and recall of the fine detection are significantly improved compared with the traditional method of abnormal traffic detection in SDN.

cs.NI↗

arXiv2023

Real-time optimal control for attitude-constrained solar sailcrafts via neural networks

This work is devoted to generating optimal guidance commands in real time for attitude-constrained solar sailcrafts in coplanar circular-to-circular interplanetary transfers. Firstly, a nonlinear optimal control problem is established, and necessary conditions for optimality are derived by the Pontryagin's Minimum Principle. Under some assumptions, the attitude constraints are rewritten as control constraints, which are replaced by a saturation function so that a parameterized system is formulated to generate an optimal trajectory via solving an initial value problem. This approach allows for the efficient generation of a dataset containing optimal samples, which are essential for training Neural Networks (NNs) to achieve real-time implementation. However, the optimal guidance command may suddenly change from one extreme to another, resulting in discontinuous jumps that generally impair the NN's approximation performance. To address this issue, we use two co-states that the optimal guidance command depends on, to detect discontinuous jumps. A procedure for preprocessing these jumps is then established, thereby ensuring that the preprocessed guidance command remains smooth. Meanwhile, the sign of one co-state is found to be sufficient to revert the preprocessed guidance command back into the original optimal guidance command. Furthermore, three NNs are built and trained offline, and they cooperate together to precisely generate the optimal guidance command in real time. Finally, numerical simulations are presented to demonstrate the developments of the paper.

math.OC↗

arXiv2023

To be or not to be? an exploration of continuously controllable prompt engineering

As the use of large language models becomes more widespread, techniques like parameter-efficient fine-tuning and other methods for controlled generation are gaining traction for customizing models and managing their outputs. However, the challenge of precisely controlling how prompts influence these models is an area ripe for further investigation. In response, we introduce ControlPE (Continuously Controllable Prompt Engineering). ControlPE enables finer adjustments to prompt effects, complementing existing prompt engineering, and effectively controls continuous targets. This approach harnesses the power of LoRA (Low-Rank Adaptation) to create an effect akin to prompt weighting, enabling fine-tuned adjustments to the impact of prompts. Our methodology involves generating specialized datasets for prompt distillation, incorporating these prompts into the LoRA model, and carefully adjusting LoRA merging weight to regulate the influence of prompts. This provides a dynamic and adaptable tool for prompt control. Through our experiments, we have validated the practicality and efficacy of ControlPE. It proves to be a promising solution for control a variety of prompts, ranging from generating short responses prompts, refusal prompts to chain-of-thought prompts.

cs.CL↗

arXiv2023

Detecting and Eliminating Quantum Noise of Quantum Measurements

In this work, we propose a two-stage procedure to systematically address quantum noise inherent in quantum measurements. The idea behind it is intuitive: we first detect and then eliminate quantum noise so that the classical noise assumption is satisfied and measurement error mitigation works. In the first stage, inspired by coherence witness in the resource theory of quantum coherence, we design an efficient method to detect quantum noise. It works by fitting the difference between two measurement statistics to the Fourier series, where the statistics are obtained using maximally coherent states with relative phase and maximally mixed states as inputs. The fitting coefficients quantitatively benchmark quantum noise. In the second stage, we design various methods to eliminate quantum noise, inspired by the Pauli twirling technique. They work by executing randomly sampled Pauli gates before the measurement device and conditionally flipping the measurement outcomes in such a way that the effective measurement device contains only classical noise. We demonstrate the feasibility of the two-stage procedure numerically on Baidu Quantum Platform. Remarkably, the results show that quantum noise in measurement devices is significantly suppressed, and the quantum computation accuracy is substantially improved. We highlight that the two-stage procedure complements existing measurement error mitigation techniques, and they together form a standard toolbox for manipulating measurement errors in near-term quantum devices.

quant-ph↗

arXiv2023

Discovery of four new EL CVn-type binaries in the Gaia eclipsing binaries

In this paper, we performed a search for EL CVn-type binaries based on the Gaia and TESS data. Through the combination of the Gaia DR3 eclipsing binary catalogue and the Gaia DR3 spectroscopic binary catalogue, we have identified 13 stars exhibiting EL CVn-like characteristics. Among these stars, nine have already been identified as EL CVn binaries, while the remaining four are newly discovered candidates. The photometric solutions and absolute parameters of the four binary systems were obtained by utilizing TESS photometry and orbital elements from Gaia DR3 spectroscopic binary catalogue. Light-curve modeling has revealed that all four binary systems with a detached configuration exhibit very low mass ratios, approximately $q\simeq$0.1. The derived physical parameters indicate that the four binary systems are highly probable candidates for being newly discovered EL CVn-type binaries. A preliminary frequency analysis was conducted on the residuals of the light curves, after the removal of binarity effects. The results suggest that the primary component star of TIC 100011519, TIC 219485855 and TIC 464641792 might be a $δ$ Sct pulsator. However, TIC 399725538 does not exhibit any intrinsic pulsations.

astro-ph.SR↗

arXiv2023

Deciphering Spatio-Temporal Graph Forecasting: A Causal Lens and Treatment

Spatio-Temporal Graph (STG) forecasting is a fundamental task in many real-world applications. Spatio-Temporal Graph Neural Networks have emerged as the most popular method for STG forecasting, but they often struggle with temporal out-of-distribution (OoD) issues and dynamic spatial causation. In this paper, we propose a novel framework called CaST to tackle these two challenges via causal treatments. Concretely, leveraging a causal lens, we first build a structural causal model to decipher the data generation process of STGs. To handle the temporal OoD issue, we employ the back-door adjustment by a novel disentanglement block to separate invariant parts and temporal environments from input data. Moreover, we utilize the front-door adjustment and adopt the Hodge-Laplacian operator for edge-level convolution to model the ripple effect of causation. Experiments results on three real-world datasets demonstrate the effectiveness and practicality of CaST, which consistently outperforms existing methods with good interpretability.

cs.LG↗

arXiv2023

Adaptive Reduced-Attitude Control for Spacecraft Boresight Alignment with Safety Constraints and Accuracy Requirements

This paper investigates the boresight alignment control problem under safety constraints and performance requirements, involving pointing-forbidden constraint, attitude angular velocity limitation, and pointing accuracy requirement. Meanwhile, the parameter uncertainty issue is taken into account simultaneously. To address this problem, we propose a modified composite framework integrating the Artificial Potential Field (APF) methodology and the Prescribed Performance Control (PPC) scheme. The APF scheme ensures safety, while the PPC scheme is employed to realize an accuracy-guaranteed control. A Switched Prescribed Performance Function (SPPF) is proposed to facilitate the integration, which monitors various constraints and further establishes compatibility between safety and performance concerns by leveraging a special PPC freezing mechanism. To further address the parameter uncertainty, we introduce the Immersion-and-Invariance (I\&I) adaptive control technique to derive an adaptive APF-PPC composite controller, further guaranteeing the closed-loop system's asymptotic convergence. Finally, numerical simulations are carried out to validate the effectiveness of the proposed scheme.

eess.SY↗

arXiv2023

Real2Sim2Real Transfer for Control of Cable-driven Robots via a Differentiable Physics Engine

Tensegrity robots, composed of rigid rods and flexible cables, exhibit high strength-to-weight ratios and significant deformations, which enable them to navigate unstructured terrains and survive harsh impacts. They are hard to control, however, due to high dimensionality, complex dynamics, and a coupled architecture. Physics-based simulation is a promising avenue for developing locomotion policies that can be transferred to real robots. Nevertheless, modeling tensegrity robots is a complex task due to a substantial sim2real gap. To address this issue, this paper describes a Real2Sim2Real (R2S2R) strategy for tensegrity robots. This strategy is based on a differentiable physics engine that can be trained given limited data from a real robot. These data include offline measurements of physical properties, such as mass and geometry for various robot components, and the observation of a trajectory using a random control policy. With the data from the real robot, the engine can be iteratively refined and used to discover locomotion policies that are directly transferable to the real robot. Beyond the R2S2R pipeline, key contributions of this work include computing non-zero gradients at contact points, a loss function for matching tensegrity locomotion gaits, and a trajectory segmentation technique that avoids conflicts in gradient evaluation during training. Multiple iterations of the R2S2R process are demonstrated and evaluated on a real 3-bar tensegrity robot.

cs.RO↗

arXiv2023

K-ST: A Formal Executable Semantics of the Structured Text Language for PLCs

Programmable Logic Controllers (PLCs) are responsible for automating process control in many industrial systems (e.g. in manufacturing and public infrastructure), and thus it is critical to ensure that they operate correctly and safely. The majority of PLCs are programmed in languages such as Structured Text (ST). However, a lack of formal semantics makes it difficult to ascertain the correctness of their translators and compilers, which vary from vendor-to-vendor. In this work, we develop K-ST, a formal executable semantics for ST in the K framework. Defined with respect to the IEC 61131-3 standard and PLC vendor manuals, K-ST is a high-level reference semantics that can be used to evaluate the correctness and consistency of different ST implementations. We validate K-ST by executing 509 ST programs extracted from Github and comparing the results against existing commercial compilers (i.e., CODESYS, CX-Programmer, and GX Works2). We then apply K-ST to validate the implementation of the open source OpenPLC platform, comparing the executions of several test programs to uncover five bugs and nine functional defects in the compiler.

cs.PL↗

arXiv2023

Transport Anisotropy in One-dimensional Graphene Superlattice in the High Kronig-Penney Potential Limit

One-dimensional graphene superlattice subjected to strong Kronig-Penney (KP) potential is promising for achieving electron lensing effect, while previous studies utilizing the modulated dielectric gates can only yield a moderate, spatially dispersed potential profile. Here, we realize high KP potential modulation of graphene via nanoscale ferroelectric domain gating. Graphene transistors are fabricated on PbZr$_{0.2}$Ti$_{0.8}$O$_{3}$ back-gates patterned with periodic, 100-200 nm wide stripe domains. Due to band reconstruction, the h-BN top-gating induces satellite Dirac points in samples with current along the superlattice vector $\hat{s}$, a feature absent in samples with current perpendicular to $\hat{s}$. The satellite Dirac point position scales with the superlattice period ($L$) as $\propto L^β$, with $β= -1.18 \pm 0.06$. These results can be well explained by the high KP potential scenario, with the Fermi velocity perpendicular to $\hat{s}$ quenched to about 1% of that for pristine graphene. Our study presents a promising material platform for realizing electron supercollimation and investigating flat band phenomena.

cond-mat.mes-hall↗