Node injection attack on Graph Neural Networks (GNNs) is an emerging and practical attack scenario that the attacker injects malicious nodes rather than modifying original nodes or edges to affect the performance of GNNs. However, existing node injection attacks ignore extremely limited scenarios, namely the injected nodes might be excessive such that they may be perceptible to the target GNN. In this paper, we focus on an extremely limited scenario of single node injection evasion attack, i.e., the attacker is only allowed to inject one single node during the test phase to hurt GNN's performance. The discreteness of network structure and the coupling effect between network structure and node features bring great challenges to this extremely limited scenario. We first propose an optimization-based method to explore the performance upper bound of single node injection evasion attack. Experimental results show that 100%, 98.60%, and 94.98% nodes on three public datasets are successfully attacked even when only injecting one node with one edge, confirming the feasibility of single node injection evasion attack. However, such an optimization-based method needs to be re-optimized for each attack, which is computationally unbearable. To solve the dilemma, we further propose a Generalizable Node Injection Attack model, namely G-NIA, to improve the attack efficiency while ensuring the attack performance. Experiments are conducted across three well-known GNNs. Our proposed G-NIA significantly outperforms state-of-the-art baselines and is 500 times faster than the optimization-based method when inferring. CCS CONCEPTS• Information systems → Data mining.
In "Distributed quantum sensing with mode-entangled spin-squeezed atomic states" Nature (2022) [1], Malia et. al. claim to improve the precision of a network of clocks by using entanglement. In particular, by entangling a clock network with up to four nodes, a precision 11.6 dB better than the quantum projection noise limit (i.e. precision without any entanglement) is reported. These claims are incorrect, Malia et. al. do not achieve an improved precision with entanglement. Here we show their demonstration is more than two orders of magnitude worse than the quantum projection noise limit.The central message in "Distributed quantum sensing with mode-entangled spin-squeezed atomic states" Nature (2022) [1] is that by entangling atoms in an atomic clock network, a precision is demonstrated that is impossible to attain using the same number of atoms and time without entanglement. Should we accept this message? Putting aside the impressive technical achievements in the paper, we can objectively assess whether the experimental data are in agreement with the claim.The final two figures of the paper present the supporting data. In Fig. 3, ∆( θ) is plotted as a function of the number of clocks M, each with N = 45, 000 atoms, where the black line (1/ √ MN ) is supposed to denote a limit that cannot be surpassed without entanglementthe quantum projection noise limit (QPN). One point should be made clear, ∆( θ) > 1/ √ MN is very definitively not the limit without entanglement. To see this requires an explanation of what ∆( θ) is. Despite the authors calling ∆( θ) the 'measured sensitivity', it is not a sensitivity at all. θ is simply the difference in values between two measurements (for M = 1).
We demonstrate that releasing atoms into free space from an optical lattice does not deteriorate cavitygenerated spin squeezing for metrological purposes. In this work, an ensemble of 500 000 spin-squeezed atoms in a high-finesse optical cavity with near-uniform atom-cavity coupling is prepared, released into free space, recaptured in the cavity, and probed. Up to ∼10 dB of metrologically relevant squeezing is retrieved for 700 μs free-fall times, and decaying levels of squeezing are realized for up to 3 ms free-fall times. The degradation of squeezing results from loss of atom-cavity coupling homogeneity between the initial squeezed state generation and final collective state readout. A theoretical model is developed to quantify this degradation and this model is experimentally validated.
No abstract
Object detection in three-dimensional (3D) space attracts much interest from academia and industry since it is an essential task in AI-driven applications such as robotics, autonomous driving, and augmented reality. As the basic format of 3D data, the point cloud can provide detailed geometric information about the objects in the original 3D space. However, due to 3D data's sparsity and unorderedness, specially designed networks and modules are needed to process this type of data. Attention mechanism has achieved impressive performance in diverse computer vision tasks; however, it is unclear how attention modules would affect the performance of 3D point cloud object detection and what sort of attention modules could fit with the inherent properties of 3D data. This work investigates the role of the attention mechanism in 3D point cloud object detection and provides insights into the potential of different attention modules. To achieve that, we comprehensively investigate classical 2D attentions, novel 3D attentions, including the latest point cloud transformers on SUN RGB-D and ScanNetV2 datasets. Based on the detailed experiments and analysis, we conclude the effects of different attention modules. This paper is expected to serve as a reference source for benefiting attention-embedded 3D point cloud object detection. The code and trained models are available at: https://github.com/ShiQiu0419/ attentions_in_3D_detection.
Thin layer transition metal dichalcogenides (TMDs) have shown great potential in the field of electronics and optoelectronics devices due to their unique electronic and optical properties. Multilayer WSe 2 is an indirect bandgap semiconductor and generally optically inactive. To improve the optical properties of multilayer TMDs, heating as a simple and effective method was widely chosen. Herein, we analyze high-temperature photoluminescence (PL) enhancement results on excitons in three-layer and four-layer WS 2 and WSe 2 with a theoretical analysis of their spectral behavior. Both direct and indirect exciton emissions in WSe 2 are relatively enhanced, which is different from the hightemperature exciton emission behavior of WS 2 in the same layer. The PL enhancement in WSe 2 could be attributed to the transfer of thermally activated electrons from the Λ valley to K valley involving K → K and K → Γ transition at high temperature, where the conduction band extreme is located at the Λ valley. We quantitatively describe this enhancement phenomenon and demonstrate that the observed spectral behavior reflects the competition between intervalley carrier transfer and high-temperature PL quenching. An excellent agreement between calculated and measured intensities is obtained. This research provides a deeper understanding of the thermally induced intervalley carrier transfer model in multilayer TMDs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.