In java, some object attributes are mutable, while others are immutable (with the "final" modifier attached to them). Objects that have at least one mutable attribute may be referred to as "mutable"...
详细信息
When done manually by engineers at Amazon and other companies, refactoring legacy code in order to eliminate uses of deprecated APIs is an error-prone and time-consuming process. In this paper, we investigate to which...
详细信息
tabulapdf is an R package that utilizes the Tabula java library to import tables from PDF files directly into R. This tool can reduce time and effort in data extraction processes in fields like investigative journalis...
详细信息
Large language Models (LLMs) are gaining popularity among software engineers. A crucial aspect of developing effective code generation LLMs is to evaluate these models using a robust benchmark. Evaluation benchmarks w...
详细信息
This paper presents the results of finetuning large language models (LLMs) for the task of detecting vulnerabilities in java source code. We leverage Wizard- Coder, a recent improvement of the state-of-the-art LLM Sta...
详细信息
A song’s backbone is its chord progressions, a series of chords that improve the harmony and add to the overall composition. For individuals ranging from beginners to creative artists, comprehending and implementing ...
详细信息
Modern day studies show a high degree of correlation between high yielding crop varieties and plants with upright leaf angles. It is observed that plants with upright leaf angles intercept more light than those withou...
详细信息
Code language Models (CLMs), particularly those leveraging deep learning, have achieved significant success in code intelligence domain. However, the issue of security, particularly backdoor attacks, is often overlook...
详细信息
Code language Models (CLMs), particularly those leveraging deep learning, have achieved significant success in code intelligence domain. However, the issue of security, particularly backdoor attacks, is often overlooked in this process. The previous research has focused on designing backdoor attacks for CLMs, but effective defenses have not been adequately addressed. In particular, existing defense methods from natural language processing, when directly applied to CLMs, are not effective enough and lack generality, working well in some models and scenarios but failing in others, thus fall short in consistently mitigating backdoor attacks. To bridge this gap, we first confirm the phenomenon of "early learning" as a general occurrence during the training of CLMs. This phenomenon refers to that a model initially focuses on the main features of training data but may become more sensitive to backdoor triggers over time, leading to overfitting and susceptibility to backdoor attacks. We then analyze that overfitting to backdoor triggers results from the use of the cross-entropy loss function, where the unboundedness of cross-entropy leads the model to increasingly concentrate on the features of the poisoned data. Based on this insight, we propose a general and effective loss function DeCE (Deceptive Cross-Entropy) by blending deceptive distributions and applying label smoothing to limit the gradient to bounded, which prevents the model from overfitting to backdoor triggers and then enhances the security of CLMs against backdoor attacks. To evaluate the effectiveness of our defense method, we select four code-related tasks as our experiments scenes and conduct experimental analyses on both natural language and two programminglanguages (java and Python). Our experiments across multiple models with different sizes (from 125M to 7B) and poisoning ratios demonstrate the applicability and effectiveness of DeCE in enhancing the security of CLMs. The findings emphasize the potent
Context: Downloading the source code of open-source java projects and building them on a local computer using Maven, Gradle, or Ant is a common activity performed by researchers and practitioners. Multiple studies so ...
详细信息
We offer a preliminary description and evaluation of an Android application that can be used to characterize user exposure to electromagnetic fields emitted by an 802.11ax mobile device. The system used consists of a ...
详细信息
ISBN:
(纸本)9781665403085
We offer a preliminary description and evaluation of an Android application that can be used to characterize user exposure to electromagnetic fields emitted by an 802.11ax mobile device. The system used consists of a DUT (Device Under Test - Huawei P40 Pro) connected to a wireless router. A measurement system composed of a small isotropic antenna connected to a spectrum analyzer and remotely controlled via a custom designed Python application was used to measure the field strength. Higher field levels were observed during file upload as compared to file download. Based on the measurements recorded by the Android application we performed an analysis of the field changes according to the number of transmitted/received bits and RSSI. We observed direct field strength variation with upload/download speed. We have also introduced the dosimetric indicator of power density/unit of information and proved its validity in highlighting different usage profiles. Further investigations have to be carried out in order to provide large scale data on Wi-Fi user exposure in the case of realistic operating scenarios.
暂无评论