Search

Important advances have recently been made in the search for materials with complex multi-phase landscapes that host photoinduced metastable collective states with exotic functionalities. In almost all cases so far, the desired phases are accessed by exploiting light-matter interactions via the imaginary part of the dielectric function through above-bandgap or resonant mode excitation. Nonresonant Raman excitation of coherent modes has been experimentally observed and proposed for dynamic material control, but the resulting atomic excursion has been limited to perturbative levels. Here, we demonstrate that it is possible to overcome this challenge by employing nonresonant ultrashort pulses with low photon energies well below the bandgap. Using mid-infrared pulses, we induce ferroelectric reversal in lithium niobate and phase switching in tin selenide and characterize the large-amplitude mode displacements through femtosecond Raman scattering, second harmonic generation, and x-ray diffraction. This approach, validated by first-principle calculations, defines a novel method for synthesizing hidden phases with unique functional properties and manipulating complex energy landscapes at reduced energy consumption and ultrafast speeds., Comment: 5 figures

This paper provides a new theory of the observed co-movement between overall wage inequality and its between-firm component. We develop and solve analytically a frictionless sorting model with two-sided heterogeneity, in which firms consist of distributions of tasks, choose how many workers to employ and reward their workers both through wages and amenities. We show that, for empirically-relevant parameter ranges, overall and between-firm inequality are firmly linked: A change in any of the models' primitives increases overall wage inequality if and only if it also increases the ratio of between-firm to overall inequality. Subsequently, we calibrate the model to match the Norwegian economy and find that the increase in wage inequality from 1995 to 2014 had a different primary cause (raising span-of-control cost) than the accompanying rise in welfare inequality (increased skill variance), and that the apparent decrease in wage inequality after 2015 masked a continued increase in welfare inequality.

Vulnerability detection is a crucial yet challenging task to identify potential weaknesses in software for cyber security. Recently, deep learning (DL) has made great progress in automating the detection process. Due to the complex multi-layer structure and a large number of parameters, a DL model requires massive labeled (vulnerable or secure) source code to gain knowledge to effectively distinguish between vulnerable and secure code. In the literature, many datasets have been created to train DL models for this purpose. However, these datasets suffer from several issues that will lead to low detection accuracy of DL models. In this paper, we define three critical issues (i.e., data imbalance, low vulnerability coverage, biased vulnerability distribution) that can significantly affect the model performance and three secondary issues (i.e., errors in source code, mislabeling, noisy historical data) that also affect the performance but can be addressed through a dedicated pre-processing procedure. In addition, we conduct a study of 14 papers along with 54 datasets for vulnerability detection to confirm these defined issues. Furthermore, we discuss good practices to use existing datasets and to create new ones., Comment: 2023 IEEE European Symposium on Security and Privacy Workshops (EuroS&P;PW)

The significant increase in software production driven by automation and faster development lifecycles has resulted in a corresponding surge in software vulnerabilities. In parallel, the evolving landscape of software vulnerability detection, highlighting the shift from traditional methods to machine learning and large language models (LLMs), provides massive opportunities at the cost of resource-demanding computations. This paper thoroughly analyses LLMs' capabilities in detecting vulnerabilities within source code by testing models beyond their usual applications to study their potential in cybersecurity tasks. We evaluate the performance of six open-source models that are specifically trained for vulnerability detection against six general-purpose LLMs, three of which were further fine-tuned on a dataset that we compiled. Our dataset, alongside five state-of-the-art benchmark datasets, were used to create a pipeline to leverage a binary classification task, namely classifying code into vulnerable and non-vulnerable. The findings highlight significant variations in classification accuracy across benchmarks, revealing the critical influence of fine-tuning in enhancing the detection capabilities of small LLMs over their larger counterparts, yet only in the specific scenarios in which they were trained. Further experiments and analysis also underscore the issues with current benchmark datasets, particularly around mislabeling and their impact on model training and performance, which raises concerns about the current state of practice. We also discuss the road ahead in the field suggesting strategies for improved model training and dataset curation., Comment: Accepted to ESORICS 2024

As a critical component of modern infrastructure, data centers account for a huge amount of power consumption and greenhouse gas emission. This paper studies the electricity purchase strategy for a data center to lower its energy cost while integrating local renewable generation under uncertainty. To facilitate efficient and scalable decision-making, we propose a two-layer hierarchy where the lower layer consists of the operation of all electrical equipment in the data center and the upper layer determines the procurement and dispatch of electricity. At the lower layer, instead of device-level scheduling in real time, we propose to exploit the inherent flexibility in demand, such as thermostatically controlled loads and flexible computing tasks, and aggregate them into virtual batteries. By this means, the upper-layer decision only needs to take into account these virtual batteries, the size of which is generally small and independent of the data center scale. We further propose an online algorithm based on Lyapunov optimization to purchase electricity from the grid with a manageable energy cost, even though the prices, renewable availability, and battery specifications are uncertain and dynamic. In particular, we show that, under mild conditions, our algorithm can achieve bounded loss compared with the offline optimal cost, while strictly respecting battery operational constraints. Extensive simulation studies validate the theoretical analysis and illustrate the tradeoff between optimality and conservativeness.

Due to its high communication efficiency, over-the-air computation (AirComp) has been expected to carry out various computing tasks in the next-generation wireless networks. However, up to now, most applications of AirComp are explored in the analog domain, which limits the capability of AirComp in resisting the complex wireless environment, not to mention to integrate the AirComp technique to the existing universal communication standards, most of which are based on the digital system. In this paper, we propose a joint design of channel coding and digital modulation for digital AirComp transmission to attempt to reinforce the foundation for the application of AirComp in the digital system. Specifically, we first propose a non-binary LDPC-based channel coding scheme to enhance the error-correction capability of AirComp. Then, a digital modulation scheme is proposed to achieve the number summation from multiple transmitters via the lattice coding technique. We also provide simulation results to demonstrate the feasibility and the performance of the proposed design., Comment: This paper has been submitted to IEEE ICC 2024

As a two-dimensional (2D) dielectric material, hexagonal boron nitride (hBN) is in high demand for applications in photonics, nonlinear optics, and nanoelectronics. Unfortunately, the high-throughput preparation of macroscopic-scale, high-quality hBN flakes with controlled thickness is an ongoing challenge, limiting device fabrication and technological integration. Here, we present a metal thin-film exfoliation method to prepare hBN flakes with millimeter-scale dimension, near-unity yields, and tunable flake thickness distribution from 1-7 layers, a substantial improvement over scotch tape exfoliation. The single crystallinity and high quality of the exfoliated hBN are demonstrated with optical microscopy, atomic force microscopy, Raman spectroscopy, and second harmonic generation. We further explore a possible mechanism for the effectiveness and selectivity based on thin-film residual stress measurements, density functional theory calculations, and transmission electron microscopy imaging of the deposited metal films. We find that the magnitude of the residual tensile stress induced by thin film deposition plays a key role in determining exfoliated flake thickness in a manner which closely resembles 3D semiconductor spalling. Lastly, we demonstrate that our exfoliated, large-area hBN flakes can be readily incorporated as encapsulating layers for other 2D monolayers. Altogether, this method brings us one step closer to the high throughput, mass production of hBN-based 2D photonic, optoelectronic, and quantum devices., Comment: 21 pages, 5 figures, work completed at Stanford University

In recent years, vehicle re-identification (Re-ID) has gained increasing importance in various applications such as assisted driving systems, traffic flow management, and vehicle tracking, due to the growth of intelligent transportation systems. However, the presence of extraneous background information and occlusions can interfere with the learning of discriminative features, leading to significant variations in the same vehicle image across different scenarios. This paper proposes a method, named graph network based on dynamic similarity adjacency matrices (DSAM-GN), which incorporates a novel approach for constructing adjacency matrices to capture spatial relationships of local features and reduce background noise. Specifically, the proposed method divides the extracted vehicle features into different patches as nodes within the graph network. A spatial attention-based similarity adjacency matrix generation (SASAMG) module is employed to compute similarity matrices of nodes, and a dynamic erasure operation is applied to disconnect nodes with low similarity, resulting in similarity adjacency matrices. Finally, the nodes and similarity adjacency matrices are fed into graph networks to extract more discriminative features for vehicle Re-ID. Experimental results on public datasets VeRi-776 and VehicleID demonstrate the effectiveness of the proposed method compared with recent works., Comment: This paper has been accepted by the 20th Pacific Rim International Conference on Artificial Intelligence in 2023

Much research on Machine Learning testing relies on empirical studies that evaluate and show their potential. However, in this context empirical results are sensitive to a number of parameters that can adversely impact the results of the experiments and potentially lead to wrong conclusions (Type I errors, i.e., incorrectly rejecting the Null Hypothesis). To this end, we survey the related literature and identify 10 commonly adopted empirical evaluation hazards that may significantly impact experimental results. We then perform a sensitivity analysis on 30 influential studies that were published in top-tier SE venues, against our hazard set and demonstrate their criticality. Our findings indicate that all 10 hazards we identify have the potential to invalidate experimental findings, such as those made by the related literature, and should be handled properly. Going a step further, we propose a point set of 10 good empirical practices that has the potential to mitigate the impact of the hazards. We believe our work forms the first step towards raising awareness of the common pitfalls and good practices within the software engineering community and hopefully contribute towards setting particular expectations for empirical research in the field of deep learning testing.

Testing deep learning-based systems is crucial but challenging due to the required time and labor for labeling collected raw data. To alleviate the labeling effort, multiple test selection methods have been proposed where only a subset of test data needs to be labeled while satisfying testing requirements. However, we observe that such methods with reported promising results are only evaluated under simple scenarios, e.g., testing on original test data. This brings a question to us: are they always reliable? In this paper, we explore when and to what extent test selection methods fail for testing. Specifically, first, we identify potential pitfalls of 11 selection methods from top-tier venues based on their construction. Second, we conduct a study on five datasets with two model architectures per dataset to empirically confirm the existence of these pitfalls. Furthermore, we demonstrate how pitfalls can break the reliability of these methods. Concretely, methods for fault detection suffer from test data that are: 1) correctly classified but uncertain, or 2) misclassified but confident. Remarkably, the test relative coverage achieved by such methods drops by up to 86.85%. On the other hand, methods for performance estimation are sensitive to the choice of intermediate-layer output. The effectiveness of such methods can be even worse than random selection when using an inappropriate layer., Comment: 12 pages

Representing source code in a generic input format is crucial to automate software engineering tasks, e.g., applying machine learning algorithms to extract information. Visualizing code representations can further enable human experts to gain an intuitive insight into the code. Unfortunately, as of today, there is no universal tool that can simultaneously visualise different types of code representations. In this paper, we introduce a tool, CodeLens, which provides a visual interaction environment that supports various representation methods and helps developers understand and explore them. CodeLens is designed to support multiple programming languages, such as Java, Python, and JavaScript, and four types of code representations, including sequence of tokens, abstract syntax tree (AST), data flow graph (DFG), and control flow graph (CFG). By using CodeLens, developers can quickly visualize the specific code representation and also obtain the represented inputs for models of code. The Web-based interface of CodeLens is available at http://www.codelens.org. The demonstration video can be found at http://www.codelens.org/demo.

The costly human effort required to prepare the training data of machine learning (ML) models hinders their practical development and usage in software engineering (ML4Code), especially for those with limited budgets. Therefore, efficiently training models of code with less human effort has become an emergent problem. Active learning is such a technique to address this issue that allows developers to train a model with reduced data while producing models with desired performance, which has been well studied in computer vision and natural language processing domains. Unfortunately, there is no such work that explores the effectiveness of active learning for code models. In this paper, we bridge this gap by building the first benchmark to study this critical problem - active code learning. Specifically, we collect 11 acquisition functions~(which are used for data selection in active learning) from existing works and adapt them for code-related tasks. Then, we conduct an empirical study to check whether these acquisition functions maintain performance for code data. The results demonstrate that feature selection highly affects active learning and using output vectors to select data is the best choice. For the code summarization task, active code learning is ineffective which produces models with over a 29.64\% gap compared to the expected performance. Furthermore, we explore future directions of active code learning with an exploratory study. We propose to replace distance calculation methods with evaluation metrics and find a correlation between these evaluation-based distance methods and the performance of code models., Comment: 12 pages, ongoing work

The development of many optical quantum technologies depends on the availability of solid-state single quantum emitters with near-perfect optical coherence. However, a standing issue that limits systematic improvement is the significant sample heterogeneity and lack of mechanistic understanding of microscopic energy flow at the single emitter level and ultrafast timescales. Here we develop solution-phase single-particle pump-probe spectroscopy with photon correlation detection that captures sample-averaged dynamics in single molecules and/or defect states with unprecedented clarity at femtosecond resolution. We apply this technique to single quantum emitters in two-dimensional hexagonal boron nitride, which suffers from significant heterogeneity and low quantum efficiency. From millisecond to nanosecond timescales, the translation diffusion, metastable-state-related bunching shoulders, rotational dynamics, and antibunching features are disentangled by their distinct photon-correlation timescales, which collectively quantify the normalized two-photon emission quantum yield. Leveraging its femtosecond resolution, spectral selectivity and ultralow noise (two orders of magnitude improvement over solid-state methods), we visualize electron-phonon coupling in the time domain at the single defect level, and discover the acceleration of polaronic formation driven by multi-electron excitation. Corroborated with results from a theoretical polaron model, we show how this translates to sample-averaged photon fidelity characterization of cascaded emission efficiency and optical decoherence time. Our work provides a framework for ultrafast spectroscopy in single emitters, molecules, or defects prone to photoluminescence intermittency and heterogeneity, opening new avenues of extreme-scale characterization and synthetic improvements for quantum information applications., Comment: 35 pages, 4 figures

The next era of program understanding is being propelled by the use of machine learning to solve software problems. Recent studies have shown surprising results of source code learning, which applies deep neural networks (DNNs) to various critical software tasks, e.g., bug detection and clone detection. This success can be greatly attributed to the utilization of massive high-quality training data, and in practice, data augmentation, which is a technique used to produce additional training data, has been widely adopted in various domains, such as computer vision. However, in source code learning, data augmentation has not been extensively studied, and existing practice is limited to simple syntax-preserved methods, such as code refactoring. Essentially, source code is often represented in two ways, namely, sequentially as text data and structurally as graph data, when it is used as training data in source code learning. Inspired by these analogy relations, we take an early step to investigate whether data augmentation methods that are originally used for text and graphs are effective in improving the training quality of source code learning. To that end, we first collect and categorize data augmentation methods in the literature. Second, we conduct a comprehensive empirical study on four critical tasks and 11 DNN architectures to explore the effectiveness of 12 data augmentation methods (including code refactoring and 11 other methods for text and graph data). Our results identify the data augmentation methods that can produce more accurate and robust models for source code learning, including those based on mixup (e.g., SenMixup for texts and Manifold-Mixup for graphs), and those that slightly break the syntax of source code (e.g., random swap and random deletion for texts).