From Brain Science to Artificial Intelligence

Reviewing the history of the development of artificial intelligence(AI)clearly reveals that brain science has resulted in breakthroughs in AI,such as deep learning.At present,although the developmental trend in AI and its applications has surpassed expectations,an insurmountable gap remains between AI and human intelligence.It is urgent to establish a bridge between brain science and AI research,including a link from brain science to AI,and a connection from knowing the brain to simulating the brain.The first steps toward this goal are to explore the secrets of brain science by studying new brain-imaging technology;to establish a dynamic connection diagram of the brain;and to integrate neuroscience experiments with theory,models,and statistics.Based on these steps,a new generation of AI theory and methods can be studied,and a subversive model and working mode from machine perception and learning to machine thinking and decision-making can be established.This article discusses the opportunities and challenges of adapting brain science to AI.

Characterization of topological phase of superlattices in superconducting circuits

The recent experimental observation of topological magnon insulator states in a superconducting circuit chain marks a breakthrough for topological physics with qubits, in which a dimerized qubit chain has been realized. Here, we extend such a dimer lattice to superlattice with arbitrary number of qubits in each unit cell in superconducting circuits, which exhibits rich topological properties. Specifically, by considering a quadrimeric superlattice, we show that the topological invariant(winding number) can be effectively characterized by the dynamics of the single-excitation quantum state through time-dependent quantities. Moreover, we explore the appearance and detection of the topological protected edge states in such a multiband qubit system. Finally, we also demonstrate the stable Bloch-like-oscillation of multiple interface states induced by the interference of them. Our proposal can be readily realized in experiment and may pave the way towards the investigation of topological quantum phases and topologically protected quantum information processing.

Topological phases and type-II edge state in two-leg-coupled Su-Schrieffer-Heeger chains

Topological quantum states have attracted great attention both theoretically and experimentally.Here,we show that the momentum-space lattice allows us to couple two Su-Schrieffer-Heeger(SSH)chains with opposite dimerizations and staggered interleg hoppings.The coupled SSH chain is a four-band model which has sublattice symmetry similar to the SSH4.Interestingly,the topological edge states occupy two sublattices at the same time,which can be regarded as a one-dimension analogue of the type-II corner state.The analytical expressions of the edge states are also obtained by solving the eigenequations.Finally,we provide a possible experimental scheme to detect the topological winding number and corresponding edge states.

Biosorption of Cd from Aqueous Solution by Live and Dead Biomass of Fusarium oxysporum Isolated from the Sludge of Abandoned Mine

Owing to exploring the biosorption mechanism of Fusarium oxysporum to Cd,the adsorption capacities of live and dead biomass of F.oxysporum strain KF2 were detected.The result showed both the live and dead biomass of strain KF2 could tolerate the Cd concentration up to 200 mg/L,and had the largest adsorption capacities for Cd,at 2.21 and 1.86 mg/g,respectively,with the inoculation amount of 3 g,the pH at 6,and the initial Cd concentration of 100 mg/L.The pseudo-second-order kinetic model(live biomass r^(2)>0.99,dead biomass r^(2)>0.90)was more suitable for describing the adsorption process of strain KF2.It indicated that the physicochemical adsorption on the cell surface might be the main pattern for Cd removal.Furthermore,the FTIR results showed that the main functional groups for cell wall to bind Cd were carboxyl,hydroxyl,amino,and phosphate.

In situ optical backpropagation training of diffractive optical neural networks

Training an artificial neural network with backpropagation algorithms to perform advanced machine learning tasks requires an extensive computational process.This paper proposes to implement the backpropagation algorithm optically for in situ training of both linear and nonlinear diffractive optical neural networlks,which enables the acceleration of training speed and improvement in energy efficiency on core computing modules.We demonstrate that the gradient of a loss function with respect to the weights of diffractive layers can be accurately calculated by measuring the forward and backward propagated optical fields based on light reciprocity and phase conjunction principles.The diffractive modulation weights are updated by programming a high-speed spatial light modulator to minimize the error between prediction and target output and perform inference tasks at the speed of light.We numerically validate the effectiveness of our approach on simulated networks for various applications.The proposed in situ optical learning architecture achieves accuracy comparable to in silico training with an electronic computer on the tasks of object dlassification and matrix-vector multiplication,which further allows the diffractive optical neural network to adapt to system imperfections.Also,the self-adaptive property of our approach facilitates the novel application of the network for all-optical imaging through scattering media.The proposed approach paves the way for robust implementation of large-scale difractive neural networks to perform distinctive tasks all-optically.

In situ optical backpropagation training of diffractive optical neural networks:publisher’s note

This publisher’s note corrects the authors’affiliations in Photon.Res.8,940(2020).

Superfluid-superradiant mixed phase of the interacting degenerate Fermi gas in an optical cavity

We investigate the ground-state properties of an attractively interacting degenerate Fermi gas coupling with a high-finesse optical cavity. We predict a new mixed phase with both the superfluid and superradiant properties for the intermediate fermion-fermion interaction and fermion-photon coupling strengths. Moreover, in this mixed phase a relatively large ratio of the scaled polarization to the dimensionless mean-field gap, which is in contrast to that in the conventional superfluid regime can be obtained. We also figure out rich phase diagrams depending crucially on the atomic resonant frequency(effective Zeeman field) and address briefly the experimental detection of our predicted quantum phases.