Superconductivity in KCa2Fe4As4F2 with Separate Double Fe2As2 Layers.

We report the synthesis, crystal structure, and physical properties of a quinary iron arsenide fluoride, KCa2Fe4As4F2. The new compound crystallizes in a body-centered tetragonal lattice (space group I4/mmm, a = 3.8684(2) Å, c = 31.007(1) Å, Z = 2) that contains double Fe2As2 conducting layers separated by insulating Ca2F2 layers. Our measurements of electrical resistivity, direct-current magnetic susceptibility, and heat capacity demonstrate bulk superconductivity at 33 K in KCa2Fe4As4F2.

Anomalous Eu valence state and superconductivity in undoped Eu3Bi2S4F4.

We have synthesized a novel europium bismuth sulfofluoride, Eu3Bi2S4F4, by solid-state reactions in sealed evacuated quartz ampules. The compound crystallizes in a tetragonal lattice (space group I4/mmm, a = 4.0771(1) Å, c = 32.4330(6) Å, and Z = 2), in which CaF2-type Eu3F4 layers and NaCl-like BiS2 bilayers stack alternately along the crystallographic c axis. There are two crystallographically distinct Eu sites, Eu(1) and Eu(2) at the Wyckoff positions 4e and 2a, respectively. Our bond valence sum calculation, based on the refined structural data, indicates that Eu(1) is essentially divalent, while Eu(2) has an average valence of ∼ +2.64(5). This anomalous Eu valence state is further confirmed and supported, respectively, by Mössbauer and magnetization measurements. The Eu(3+) components donate electrons into the conduction bands that are mainly composed of Bi 6px and 6py states. Consequently, the material itself shows metallic conduction and superconducts at 1.5 K without extrinsic chemical doping.

Achieving small-batch accuracy with large-batch scalability via Hessian-aware learning rate adjustment.

We consider synchronous data-parallel neural network training with a fixed large batch size. While the large batch size provides a high degree of parallelism, it degrades the generalization performance due to the low gradient noise scale. We propose a general learning rate adjustment framework and three critical heuristics that tackle the poor generalization issue. The key idea is to adjust the learning rate based on geometric information of loss landscape and encourage the model to converge into a flat minimum that is known to better generalize to the unknown data. Our empirical study demonstrates that the Hessian-aware learning rate schedule remarkably improves the generalization performance in large-batch training. For CIFAR-10 classification with ResNet20, our method achieves 92.31% accuracy using 16,384 batch size, which is close to 92.83% achieved using 128 batch size, at a negligible extra computational cost.

L-BGNN: Layerwise Trained Bipartite Graph Neural Networks.

Learning low-dimensional representations of bipartite graphs enables e-commerce applications, such as recommendation, classification, and link prediction. A layerwise-trained bipartite graph neural network (L-BGNN) embedding method, which is unsupervised, efficient, and scalable, is proposed in this work. To aggregate the information across and within two partitions of a bipartite graph, a customized interdomain message passing (IDMP) operation and an intradomain alignment (IDA) operation are adopted by the proposed L-BGNN method. Furthermore, we develop a layerwise training algorithm for L-BGNN to capture the multihop relationship of large bipartite networks and improve training efficiency. We conduct extensive experiments on several datasets and downstream tasks of various scales to demonstrate the effectiveness and efficiency of the L-BGNN method as compared with state-of-the-art methods. Our codes are publicly available at https://github.com/TianXieUSC/L-BGNN.

Accelerated Distributed Approximate Newton Method.

Distributed second-order optimization, as an effective strategy for training large-scale machine learning systems, has been widely investigated due to its low communication complexity. However, the existing distributed second-order optimization algorithms, including distributed approximate Newton (DANE), accelerated inexact DANE (AIDE), and statistically preconditioned accelerated gradient (SPAG), are all required to precisely solve an expensive subproblem up to the target precision. Therefore, this causes these algorithms to suffer from high computation costs and this hinders their development. In this article, we design a novel distributed second-order algorithm called the accelerated distributed approximate Newton (ADAN) method to overcome the high computation costs of the existing ones. Compared with DANE, AIDE, and SPAG, which are constructed based on the relative smooth theory, ADAN's theoretical foundation is built upon the inexact Newton theory. The different theoretical foundations lead to handle the expensive subproblem efficiently, and steps required to solve the subproblem are independent of the target precision. At the same time, ADAN resorts to the acceleration and can effectively exploit the objective function's curvature information, making ADAN to achieve a low communication complexity. Thus, ADAN can achieve both the communication and computation efficiencies, while DANE, AIDE, and SPAG can achieve only the communication efficiency. Our empirical study also validates the advantages of ADAN over extant distributed second-order algorithms.

Superconductivity at 35 K by self doping in RbGd2Fe4As4O2.

We report synthesis, crystal structure and physical properties of a novel quinary compound RbGd2Fe4As4O2. The new iron oxyarsenide is isostructural to the fluo-arsenide KCa2Fe4As4F2, both of which contain separate double Fe2As2 layers that are self hole-doped in the stoichiometric composition. Bulk superconductivity at [Formula: see text] K is demonstrated by the measurements of electrical resistivity, dc magnetic susceptibility and heat capacity. An exceptionally high value of the initial slope of the upper critical field ([Formula: see text]d[Formula: see text]/d[Formula: see text] [Formula: see text] T K-1) is measured for the polycrystalline sample.

[Application of SF-36 scale on pulmonary tuberculosis patients in Yunnan province of China and southern Thailand].

OBJECTIVE: To explore the effect of SF-36 scale being applied in different countries under different culture and to describe the quality of life of pulmonary tuberculosis patients in China and Thailand. METHODS: SF-36 scale was applied to pulmonary tuberculosis patients in both countries using face to face interview. RESULTS: Many coefficients among domains were greater than 0.5 when quality of life of tuberculosis patient in both countries was measured. Cronbach's coefficient of all domains were greater than 0.7 for tuberculosis patients in China while cronbach's Coefficient of most domains were equal or greater than 0.7 for tuberculosis patients in Thailand except for vitality and social domains. The score of social domain for patients in Thailand was greater than that of China. CONCLUSION: Structure validity was not good for tuberculosis patients in both countries since there were some items overlapped in different domains. However, the reliability was good for measuring quality of life of tuberculosis patients both in China and in Thailand.