Standard Neural Networks can learn mathematical operations, but they do not extrapolate. Extrapolation means that the model can apply to larger numbers, well beyond those observed during training. Recent architectures tackle arithmetic operations and can extrapolate; however, the equally important problem of quantitative reasoning remains unaddressed. In this work, we propose a novel architectural element, the Neural Status Register (NSR), for quantitative reasoning over numbers. Our NSR relaxes the discrete bit logic of physical status registers to continuous numbers and allows end-to-end learning with gradient descent. Experiments show that the NSR achieves solutions that extrapolate to numbers many orders of magnitude larger than those in the training set. We successfully train the NSR on number comparisons, piecewise discontinuous functions, counting in sequences, recurrently finding minimums, finding shortest paths in graphs, and comparing digits in images.
A neural network model that learns differences in diagnosis strategies among radiologists has an improved area under the curve for aneurysm status classification in magnetic resonance angiography image series
National Institute of Radiological Sciences, QST(量子科学技术研究开发机构放射线医学综合研究所);MNES corporation(MNES公司);QST Hospital, QST(量子科学技术研究开发机构医院)
TAS-GNN: A Status-Aware Signed Graph Neural Network for Anomaly Detection in Bitcoin Trust Systems
Yeshiva University(耶鲁大学);Yale University(耶鲁大学);University of California, San Diego(加州大学圣地亚哥分校);North Carolina State University(北卡罗来纳州立大学);University of Pennsylvania(宾夕法尼亚大学)
Q-Capsule: A Localized Capsule-Based Quantum Neural Architecture for Barren Plateau Mitigation
Florida Atlantic University(佛罗里达大西洋大学);Khulna University of Engineering & Technology(库尔纳工程与技术大学);Southern Illinois University Carbondale(南伊利诺伊大学卡本代尔分校)