Despite the efficiency and scalability of machine learning systems, recent studies have demonstrated that many classification methods, especially Deep Neural Networks (DNNs), are vulnerable to adversarial examples; that is, examples that are carefully crafted to fool a well-trained classification model while being indistinguishable from natural data to human. This makes it potentially unsafe to apply DNNs or related methods in security-critical areas. Since this issue was first identified by Biggio et al. and Szegedy et al., much work has been done in this field, including the development of attack methods to generate adversarial examples and the construction of defense techniques to guard against such examples. This article aims to introduce this topic and its latest developments to the statistical community, primarily focusing on the generation and guarding of adversarial examples. Computing codes (in Python and R) used in the numerical experiments are publicly available for readers to explore the surveyed methods. It is the hope of the authors that this article will encourage more statisticians to work on this important and exciting field of generating and defending against adversarial examples.

Dataset Distillation (DD), a newly emerging field, aims at generating much smaller but efficient synthetic training datasets from large ones. Existing DD methods based on gradient matching achieve leading performance; however, they are extremely computationally intensive as they require continuously optimizing a dataset among thousands of randomly initialized models. In this paper, we assume that training the synthetic data with diverse models leads to better generalization performance. Thus we propose two model augmentation techniques, i.e. using early-stage models and parameter perturbation to learn an informative synthetic set with significantly reduced training cost. Extensive experiments demonstrate that our method achieves up to 20× speedup and comparable performance on par with state-of-the-art methods.

Large-scale Transformer models bring significant improvements for various downstream vision language tasks with a unified architecture. The performance improvements come with increasing model size, resulting in slow inference speed and increased cost for severing. While some certain predictions benefit from the full computation of the large-scale model, not all of inputs need the same amount of computation to conduct, potentially leading to computation resource waste. To handle this challenge, early exiting is proposed to adaptively allocate computational power in term of input complexity to improve inference efficiency. The existing early exiting strategies usually adopt output confidence based on intermediate layers as a proxy of input complexity to incur the decision of skipping following layers. However, such strategies cannot be applied to encoder in the widely-used unified architecture with both encoder and decoder due to difficulty of output confidence estimation in the encoder layers. It is suboptimal in term of saving computation power to ignore the early exiting in encoder component. To address this issue, we propose a novel early exiting strategy for unified vision language models, which allows to dynamically skip the layers in encoder and decoder simultaneously in term of input layer-wise similarities with multiple times of early exiting, namely MuE. By decomposing the image and text modalities in the encoder, MuE is flexible and can skip different layers in term of modalities, advancing the inference efficiency while minimizing performance drop. Experiments on the SNLI-VE and MS COCO datasets show that the proposed approach MuE can reduce expected inference time by up to 50% and 40% while maintaining 99% and 96% performance respectively.

BACKGROUND: Dysregulation of l-arginine metabolism has been proposed to occur in patients with severe asthma. The effects of l-arginine supplementation on l-arginine metabolite profiles in these patients are unknown. We hypothesized that individuals with severe asthma with low fractional exhaled nitric oxide (FeNO) would have fewer exacerbations with the addition of l-arginine to their standard asthma medications compared with placebo and would demonstrate the greatest changes in metabolite profiles. METHODS: Participants were enrolled in a single-center, crossover, double-blind l-arginine intervention trial at UCD. Subjects received placebo or l-arginine, dosed orally at 0.05 mg/kg (ideal body weight) twice daily. The primary end point was moderate asthma exacerbations. Longitudinal plasma metabolite levels were measured using mass spectrometry. A linear mixed-effect model with subject-specific intercepts was used for testing treatment effects. RESULTS: A cohort of 50 subjects was included in the final analysis. l-Arginine did not significantly decrease asthma exacerbations in the overall cohort. Higher citrulline levels and a lower arginine availability index (AAI) were associated with higher FeNO ( P = 0.005 and P = 2.51 × 10 ^–9 , respectively). Higher AAI was associated with lower exacerbation events. The eicosanoid prostaglandin H ₂ (PGH ₂ ) and N _α -acetyl-l-arginine were found to be good predictors for differentiating clinical responders and nonresponders. CONCLUSIONS: There was no statistically significant decrease in asthma exacerbations in the overall cohort with l-arginine intervention. PGH ₂ , N _α -acetyl-l-arginine, and the AAI could serve as predictive biomarkers in future clinical trials that intervene in the arginine metabolome. TRIAL REGISTRATION: ClinicalTrials.gov NCT01841281. FUNDING: This study was supported by NIH grants R01HL105573, DK097154, UL1 TR001861, and K08HL114882. Metabolomics analysis was supported in part by a grant from the University of California Tobacco-Related Disease Research Program program (TRDRP).