Search for: All records

33rd USENIX Security Symposium (USENIX Security 24) 

Abstract 2D van der Waals (vdW) magnets with layer‐dependent magnetic states and/or diverse magnetic interactions and anisotropies have attracted extensive research interest. Despite the advances, a notable challenge persists in effectively manipulating the tunneling anisotropic magnetoresistance (TAMR) of 2D vdW magnet‐based magnetic tunnel junctions (MTJs). Here, this study reports the novel and anomalous tunneling magnetoresistance (TMR) oscillations and pioneering demonstration of bias and gate voltage controllable TAMR in 2D vdW MTJs, utilizing few‐layer CrPS₄. This material, inherently an antiferromagnet, transitions to a canted magnetic order upon application of external magnetic fields. Through TMR measurements, this work unveils the novel layer‐dependent oscillations in the tunneling resistance for few‐layer CrPS₄devices under both out‐of‐plane and in‐plane magnetic fields, with a pronounced controllability via gate voltage. Intriguingly, this study demonstrates that both the polarity and magnitude of TAMR in CrPS₄can be effectively tuned through either a bias or gate voltage. The mechanism behind this electrically tunable TAMR is further elucidated through first‐principles calculations. The implications of the findings are far‐reaching, providing new insights into 2D magnetism and opening avenues for the development of innovative spintronic devices based on 2D vdW magnets. 

Rapid and cost-effective detection of antibiotics in wastewater and through wastewater treatment processes is an important first step in developing effective strategies for their removal. Surface-enhanced Raman scattering (SERS) has the potential for label-free, real-time sensing of antibiotic contamination in the environment. This study reports the testing of two gold nanostructures as SERS substrates for the label-free detection of quinoline, a small-molecular-weight antibiotic that is commonly found in wastewater. The results showed that the self-assembled SERS substrate was able to quantify quinoline spiked in wastewater with a lower limit of detection (LoD) of 5.01 ppb. The SERStrate (commercially available SERS substrate with gold nanopillars) had a similar sensitivity for quinoline quantification in pure water (LoD of 1.15 ppb) but did not perform well for quinoline quantification in wastewater (LoD of 97.5 ppm) due to interferences from non-target molecules in the wastewater. Models constructed based on machine learning algorithms could improve the separation and identification of quinoline Raman spectra from those of interference molecules to some degree, but the selectivity of SERS intensification was more critical to achieve the identification and quantification of the target analyte. The results of this study are a proof-of-concept for SERS applications in label-free sensing of environmental contaminants. Further research is warranted to transform the concept into a practical technology for environmental monitoring. 

Neural network pruning is an essential technique for reducing the size and complexity of deep neural networks, enabling large-scale models on devices with limited resources. However, existing pruning approaches heavily rely on training data for guiding the pruning strategies, making them ineffective for federated learning over distributed and confidential datasets. Additionally, the memory- and computation-intensive pruning process becomes infeasible for recourse-constrained devices in federated learning. To address these challenges, we propose FedTiny, a distributed pruning framework for federated learning that generates specialized tiny models for memory-and computing-constrained devices. We introduce two key modules in FedTiny to adaptively search coarse- and finer-pruned specialized models to fit deployment scenarios with sparse and cheap local computation. First, an adaptive batch normalization selection module is designed to mitigate biases in pruning caused by the heterogeneity of local data. Second, a lightweight progressive pruning module aims to finer prune the models under strict memory and computational budgets, allowing the pruning policy for each layer to be gradually determined rather than evaluating the overall model structure. The experimental results demonstrate the effectiveness of FedTiny, which outperforms state-of-the-art approaches, particularly when compressing deep models to extremely sparse tiny models. FedTiny achieves an accuracy improvement of 2.61% while significantly reducing the computational cost by 95.91% and the memory footprint by 94.01% compared to state-of-the-art methods. 

This study focuses on a series of long-lived atmospheric bores that persisted from midnight to the late afternoon on 10 June 2023 over the coastal and offshore regions of the Yellow Sea. These bores were sustained by a favorable trapping mechanism associated with the stability of the marine boundary layer and the presence of a low-level wind component perpendicular to the propagation direction of the bores. These bores aided in maintaining and enhancing convective systems offshore with outflows that reached inland and initiated new convective cells. These convective cells eventually organized to produce severe weather under favorable thermodynamic conditions. These findings underscore the unique atmospheric conditions over the Yellow/East China Sea that promote long-lived atmospheric bores during the day, which is a potential mechanism to cause severe convective systems over the coastal region of eastern China. The complex multiscale processes influencing the development of mesoscale convective systems during the mei-yu monsoon season over this region present unique challenges for their accurate representation especially in coarse-grid weather and climate models. Given the growing density of coastal populations, further research is needed into the frequency of bores generated by convection and their subsequent impact on coastal convection. Significance StatementThis study links atmospheric bores over the Yellow/East China Sea to the generation of severe convection over the eastern coast of China. These offshore bores persisted well into the daylight hours with far longer lifetimes than their inland counterparts. This longevity was supported by the presence of a stable marine boundary layer and prevailing southerly winds associated with the sea breeze. These offshore bores influenced severe coastal daytime convection both directly and indirectly. This study reveals the complex characteristics of convective systems along the eastern coast of China. These findings underscore the need to incorporate offshore and coastal atmospheric measurements into operational forecasting and research activities.