In recent years, unmanned aerial vehicles (UAVs) have proven their effectiveness in surveillance due to their superior mobility. By utilizing multiple UAVs with collaborated learning, surveillance of a huge area while...
详细信息
In this paper, we introduce a novel class of fast, beam search-based adversarial attack (BEAST) for Language Models (LMs). BEAST employs interpretable parameters, enabling attackers to balance between attack speed, su...
详细信息
In this paper, we introduce a novel class of fast, beam search-based adversarial attack (BEAST) for Language Models (LMs). BEAST employs interpretable parameters, enabling attackers to balance between attack speed, success rate, and the readability of adversarial prompts. The computational efficiency of BEAST facilitates us to investigate its applications on LMs for jailbreaking, eliciting hallucinations, and privacy attacks. Our gradient-free targeted attack can jailbreak aligned LMs with high attack success rates within one minute. For instance, BEAST can jailbreak Vicuna-7B-v1.5 under one minute with a success rate of 89% when compared to a gradient-based baseline that takes over an hour to achieve 70% success rate using a single Nvidia RTX A6000 48GB GPU. BEAST can also generate adversarial suffixes for successful jailbreaks that can transfer to unseen prompts and unseen models such as GPT-4-Turbo. Additionally, we discover a unique outcome wherein our untargeted attack induces hallucinations in LM chatbots. Through human evaluations, we find that our untargeted attack causes Vicuna-7B-v1.5 to produce ∼15% more incorrect outputs when compared to LM outputs in the absence of our attack. We also learn that 22% of the time, BEAST causes Vicuna to generate outputs that are not relevant to the original prompt. Further, we use BEAST to generate adversarial prompts in a few seconds that can boost the performance of existing membership inference attacks for LMs. We believe that our fast attack, BEAST, has the potential to accelerate research in LM security and privacy. Copyright 2024 by the author(s)
Perovskite solar cells represent a revolutionary class of photovoltaic devices that have gained substantial attention for their exceptional performance and potential to provide an affordable and efficient solution for...
详细信息
Perovskite solar cells represent a revolutionary class of photovoltaic devices that have gained substantial attention for their exceptional performance and potential to provide an affordable and efficient solution for harnessing solar energy. These cells utilize perovskite-structured materials, typically hybrid organicinorganic lead halide compounds, as the light-absorbing layer.
This study introduces a data-driven approach for state and output feedback control addressing the constrained output regulation problem in unknown linear discrete-time systems. Our method ensures effective tracking pe...
详细信息
This study introduces a data-driven approach for state and output feedback control addressing the constrained output regulation problem in unknown linear discrete-time systems. Our method ensures effective tracking performance while satisfying the state and input constraints, even when system matrices are not available. We first establish a sufficient condition necessary for the existence of a solution pair to the regulator equation and propose a data-based approach to obtain the feedforward and feedback control gains for state feedback control using linear programming. Furthermore, we design a refined Luenberger observer to accurately estimate the system state, while keeping the estimation error within a predefined set. By combining output regulation theory, we develop an output feedback control strategy. The stability of the closed-loop system is rigorously proved to be asymptotically stable by further leveraging the concept of λ-contractive sets.
Supply chain management and Hyperledger are two interconnected domains. They leverage blockchain technology to enhance efficiency, transparency, and security in supply chain operations. Together, they provide a decent...
详细信息
Breast cancer, marked by uncontrolled cell growth in breast tissue, is the most common cancer among women and a second-leading cause of cancer-related deaths. Among its types, ductal and lobular carcinomas are the mos...
详细信息
Breast cancer, marked by uncontrolled cell growth in breast tissue, is the most common cancer among women and a second-leading cause of cancer-related deaths. Among its types, ductal and lobular carcinomas are the most prevalent, with invasive ductal carcinoma accounting for about 70–80% of cases and invasive lobular carcinoma for about 10–15%. Accurate identification is crucial for effective treatment but can be time-consuming and prone to interobserver variability. AI can rapidly analyze pathological images, providing precise, cost-effective identification, thus reducing the pathologists’ workload. This study utilizes a deep learning framework for advanced, automatic breast cancer detection and subtype identification. The framework comprises three key components: detecting cancerous patches, identifying cancer subtypes (ductal and lobular carcinoma), and predicting patient-level outcomes from whole slide images (WSI). The validation process includes visualization using Score-CAM to highlight cancer-affected areas prominently. Datasets include 111 WSIs (85 malignant from the Warwick HER2 dataset and 26 benign from pathologists). For subtype detection, there are 57 ductal and 8 lobular carcinoma cases. A total of 28,428 annotated patches were reviewed by two expert pathologists. Four pre-trained models—DenseNet-201, MobileNetV2, an ensemble of these two, and a Vision Transformer-based model—were fine-tuned and tested on the patches. Patient-level results were predicted using a majority voting technique based on the percentage of each patch type in the WSI. The Vision Transformer-based model outperformed other models in patch classification, achieving an accuracy of 96.74% for cancerous patch detection and 89.78% for cancer subtype classification. For WSI-based cancer classification, the majority voting method attained an F1-score of 99.06 and 96.13% for WSI-based cancer subtype classification. The proposed deep learning-based framework for advanced breast cancer det
Pretrained language models leverage selfsupervised learning to use large amounts of unlabeled text for learning contextual representations of sequences. However, in the domain of medical conversations, the availabilit...
详细信息
Integrated sensing and communication (ISAC) is a promising solution to mitigate the increasing congestion of the wireless spectrum. In this paper, we investigate the short packet communication regime within an ISAC sy...
详细信息
We propose a first-order sampling method called the Metropolis-adjusted Preconditioned Langevin Algorithm for approximate sampling from a target distribution whose support is a proper convex subset of Rd. Our proposed...
详细信息
Combining optical and electronic systems could enable information processing that is a million times faster than existing gigahertz technology. Imagine leveraging nature’s fastest processes to power the electronics i...
详细信息
Combining optical and electronic systems could enable information processing that is a million times faster than existing gigahertz technology. Imagine leveraging nature’s fastest processes to power the electronics in semiconductor chips, quantum sensors and quantum computers. Such transformative speed would not only greatly improve the performance of technology, but unveil new vistas for fundamental science as well.
暂无评论