Search

Traditional superoscillation focusing typically requires complex optimization of the incident light field. These complexities may limit the practical application of superoscillation. High-order radially polarized Laguerre-Gaussian beams inherently support superoscillation focusing due to their multi-ring amplitude distribution and 0 ~ \pi phase alternation, which align with the necessary destructive interference mechanisms. In this study, we demonstrate that by adjusting the beam mode order together with the incident beam size, we can easily control the full width at half maximum, field of view, and energy distribution of superoscillation focusing. Moreover, high-order azimuthally polarized vortex-phase Laguerre-Gaussian beams can also achieve superoscillation focusing, offering even better super-resolution effects. The distinct focusing behaviors of their circular components present unique opportunities for applications involving circular dichroism materials., Comment: 5 pages, 4 figures

The development of a general purpose service robot for daily life necessitates the robot's ability to deploy a myriad of fundamental behaviors judiciously. Recent advancements in training Large Language Models (LLMs) can be used to generate action sequences directly, given an instruction in natural language with no additional domain information. However, while the outputs of LLMs are semantically correct, the generated task plans may not accurately map to acceptable actions and might encompass various linguistic ambiguities. LLM hallucinations pose another challenge for robot task planning, which results in content that is inconsistent with real-world facts or user inputs. In this paper, we propose a task planning method based on a constrained LLM prompt scheme, which can generate an executable action sequence from a command. An exceptional handling module is further proposed to deal with LLM hallucinations problem. This module can ensure the LLM-generated results are admissible in the current environment. We evaluate our method on the commands generated by the RoboCup@Home Command Generator, observing that the robot demonstrates exceptional performance in both comprehending instructions and executing tasks.

Relation extraction is an efficient way of mining the extraordinary wealth of human knowledge on the Web. Existing methods rely on domain-specific training data or produce noisy outputs. We focus here on extracting targeted relations from semi-structured web pages given only a short description of the relation. We present GraphScholarBERT, an open-domain information extraction method based on a joint graph and language model structure. GraphScholarBERT can generalize to previously unseen domains without additional data or training and produces only clean extraction results matched to the search keyword. Experiments show that GraphScholarBERT can improve extraction F1 scores by as much as 34.8\% compared to previous work in a zero-shot domain and zero-shot website setting., Comment: 7 pages, 2 figures

Arrays of neutral atoms have emerged as promising platforms for quantum computing. Realization of high-fidelity two-qubit gates with robustness is currently a significant important task for large-scale operations. In this paper, we present a convenient approach for implementing a two-qubit controlled-phase gate using Rydberg blockade. We achieve the noncyclic geometric control with a single modulated pulse. As compared with the control scheme by cyclic evolution that determined by dynamical parameters, the robustness of the proposal against systematic errors will be remarkably improved due to the geometric characteristic. Importantly, the noncyclic geometric control reduces the gate time for small rotation angles and will be more insensitive to the decoherence effect. We accelerate the adiabatic control with the aid of shortcuts to adiabaticity to further shorten the operation time. We apply our protocol to the algorithm of quantum Fourier transformation to show the actual acceleration. Therefore, the proposed scheme will provide an analytical waveforms for arbitrary two-qubit gates and may have important use in the experiments of atomic arrays., Comment: 8 pages, 5 figures

The braiding operations of quantum states have attracted substantial attention due to their great potential for realizing topological quantum computations. In this paper, we show that a three-fold degenerate eigen subspace can be obtained in a four-level Hamiltonian which is the minimal physical system. Braiding operations are proposed to apply to dressed states in the subspace. The topology of the braiding diagram can be characterized through physical methods once that the sequential braiding pulses are adopted. We establish an equivalent relationship function between the permutation group and the output states where different output states correspond to different values of the function. The topological transition of the braiding happens when two operations overlap, which is detectable through the measurement of the function. Combined with the phase variation method, we can analyze the wringing pattern of the braiding. Therefore, the experimentally-feasible system provides a platform to investigate braiding dynamics, the SU(3) physics and the qutrit gates., Comment: 10 pages, 6 figures

Large-language models (LLMs) such as GPT-4 caught the interest of many scientists. Recent studies suggested that these models could be useful in chemistry and materials science. To explore these possibilities, we organized a hackathon. This article chronicles the projects built as part of this hackathon. Participants employed LLMs for various applications, including predicting properties of molecules and materials, designing novel interfaces for tools, extracting knowledge from unstructured data, and developing new educational applications. The diverse topics and the fact that working prototypes could be generated in less than two days highlight that LLMs will profoundly impact the future of our fields. The rich collection of ideas and projects also indicates that the applications of LLMs are not limited to materials science and chemistry but offer potential benefits to a wide range of scientific disciplines.

Non-Abelian gauge field (NAGF) plays a central role in understanding the geometrical and topological phenomena in physics. Here we experimentally induce a NAGF in the degenerate eigen subspace of a double-$\Lambda$ four-level atomic system. The non-Abelian nature of the gauge field is detected through the measurement of the non-commutativity of two successive evolution loops. Then we theoretically propose and experimentally demonstrate a novel scheme to measure the NAGF through multi-loop evolution and robust holonomic quantum gates. The demonstrated scheme offers the advantage of detecting the NAGF with amplification through multi-loop evolution. Our results pave the way for an experimentally-feasible approach to achieving high-resolution and high-precision measurements of the gauge fields., Comment: 11 pages, 8 figures

Transformer-based masked language models such as BERT, trained on general corpora, have shown impressive performance on downstream tasks. It has also been demonstrated that the downstream task performance of such models can be improved by pretraining larger models for longer on more data. In this work, we empirically evaluate the extent to which these results extend to tasks in science. We use 14 domain-specific transformer-based models (including ScholarBERT, a new 770M-parameter science-focused masked language model pretrained on up to 225B tokens) to evaluate the impact of training data, model size, pretraining and finetuning time on 12 downstream scientific tasks. Interestingly, we find that increasing model sizes, training data, or compute time does not always lead to significant improvements (i.e., >1% F1), if at all, in scientific information extraction tasks and offered possible explanations for the surprising performance differences., Comment: 12 pages. 3 figures. 5 tables. Accepted to the Findings of ACL 2023

In view of the fact that most invisibility devices focus on linear polarization cloaking and that the characteristics of mid infrared cloaking are rarely studied, we propose a cross circularly polarized invisibility carpet cloaking device in the mid infrared band. Based on the Pancharatnam Berry phase principle, the unit cells with the cross circular polarization gradient phase were carefully designed and constructed into a metasurface. In order to achieve tunable cross circular polarization carpet cloaks, a phase change material is introduced into the design of the unit structure. When the phase change material is in amorphous and crystalline states, the proposed metasurface unit cells can achieve high efficiency cross polarization conversion and reflection intensity can be tuned. According to the phase compensation principle of carpet cloaking, we construct a metasurface cloaking device with a phase gradient using the designed unit structure. From the near and far field distributions, the cross circular polarization cloaking property is confirmed in the broadband wavelength range. The proposed cloaking device can effectively resist detection of cross-circular polarization.

The topology of quantum systems has become a topic of great interest since the discovery of topological insulators. However, as a hallmark of the topological insulators, the spin Chern number has not yet been experimentally detected. The challenge to directly measure this topological invariant lies in the fact that this spin Chern number is defined based on artificially constructed wavefunctions. Here we experimentally mimic the celebrated Bernevig-Hughes-Zhang model with cold atoms, and then measure the spin Chern number with the linear response theory. We observe that, although the Chern number for each spin component is ill defined, the spin Chern number measured by their difference is still well defined when both energy and spin gaps are non-vanished., Comment: 12 pages, 9 figures

Researchers worldwide are seeking to repurpose existing drugs or discover new drugs to counter the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A promising source of candidates for such studies is molecules that have been reported in the scientific literature to be drug-like in the context of coronavirus research. We report here on a project that leverages both human and artificial intelligence to detect references to drug-like molecules in free text. We engage non-expert humans to create a corpus of labeled text, use this labeled corpus to train a named entity recognition model, and employ the trained model to extract 10912 drug-like molecules from the COVID-19 Open Research Dataset Challenge (CORD-19) corpus of 198875 papers. Performance analyses show that our automated extraction model can achieve performance on par with that of non-expert humans., Comment: 17 single-column pages, 6 figures, and 6 tables

To enhance transmission efficiency of Pancharatnam-Berry (PB) phase metasurfaces, multilayer split-ring resonators were proposed to develop encoding sequences. As per the generalized Snell law, the deflection angle of the PB phase encoding metasurfaces depends on the metasurface period size. Therefore, it is impossible to design an infinitesimal metasurface unit.Consequently, the continuous transmission scattering angle cannot be obtained. In digital signal processing, this study introduces the Fourier convolution principle on encoding metasurface sequences to freely control the transmitted scattering angles. Both addition and subtraction operations between two different encoding sequences were then performed to achieve the continuous variation of the scattering angle. Furthermore, we established that the Fourier convolution principle can be applied to the checkerboard coded metasurfaces.

Researchers across the globe are seeking to rapidly repurpose existing drugs or discover new drugs to counter the the novel coronavirus disease (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). One promising approach is to train machine learning (ML) and artificial intelligence (AI) tools to screen large numbers of small molecules. As a contribution to that effort, we are aggregating numerous small molecules from a variety of sources, using high-performance computing (HPC) to computer diverse properties of those molecules, using the computed properties to train ML/AI models, and then using the resulting models for screening. In this first data release, we make available 23 datasets collected from community sources representing over 4.2 B molecules enriched with pre-computed: 1) molecular fingerprints to aid similarity searches, 2) 2D images of molecules to enable exploration and application of image-based deep learning methods, and 3) 2D and 3D molecular descriptors to speed development of machine learning models. This data release encompasses structural information on the 4.2 B molecules and 60 TB of pre-computed data. Future releases will expand the data to include more detailed molecular simulations, computed models, and other products., Comment: 11 pages, 5 figures

We investigated numerically the electromagnetically induced transparency (EIT) behavior with the maximum Q factor is up to 28000 in a metal-dielectric metamaterial. It composed of two layers, a metallic strip as bright mode stacked above a dielectric rod as another bright mode. The coherent coupling between the broad electric dipole induced by the metal strip and the narrow toroidal dipole induced by the dielectric rod leads to the destructive interference, then introduces a transparency window. Interestingly, the EIT window could be effectively influenced by varying parameters of the two layers and the thickness of middle spacer. This process can be explained by using two-particle model. Furthermore, the radiation loss is greatly suppressed within the EIT windows and the slow light effect is improved a lot. These excellent properties also provide an effective platform for the filters and sensitive chemical and biological sensors in the optical range.

Fast and robust quantum gates is the cornerstone of fault-tolerance quantum computation. In this paper, we propose to achieve quantum gates based on non-cyclic geometric evolution. Dynamical phase during the evolution is cancelled by spin-echo process and the adiabatic control can be sped up through shortcut to adiabatic manner. Different from geometric gates based on cyclic evolution, the superiority of non-cyclic scheme is that the operation time is proportional to the rotation angle (but not the geometric phase) of quantum state. Therefore, the non-cyclic scheme becomes fairly fast in the case of quantum gates with small rotation angle which will be more insensitive to the decoherence and leakage to the states outside the computational basis. The proposed scheme is also robust against random noise due to the geometric characteristic of projective Hilbert space. Since the refined proposed scheme is fast and robust, it is an particularly suitable way to manipulate the physical systems with weak nonlinearity, such as superconucting systems., Comment: 6 pages, 4figures

Virtual Observatory (VO) is a data-intensively online astronomical research and education environment, which takes advantages of advanced information technologies to achieve seamless and global access to astronomical information. AstroCloud is a cyber-infrastructure for astronomy research initiated by Chinese Virtual Observatory (China-VO) project, and also a kind of physical distributed platform which integrates lots of tasks such as telescope access proposal management, data archiving, data quality control, data release and open access, cloud based data processing and analysis. It consists of five application channels, i.e. observation, data, tools, cloud and public and is acting as a full lifecycle management system and gateway for astronomical data and telescopes. Physically, the platform is hosted in six cities currently, i.e. Beijing, Nanjing, Shanghai, Kunming, Lijiang and Urumqi, and serving more than 17 thousand users. Achievements from international Virtual Observatories and Cloud Computing are adopted heavily. In the paper, backgrounds of the project, architecture, Cloud Computing environment, key features of the system, current status and future plans are introduced., Comment: 5 pages, 8 figures, to appear in the Proceedings of "The 3rd International Conference on Wireless Communication and Sensor Network (WCSN 2016)"

As the cyber-infrastructure for Astronomical research from Chinese Virtual Observatory (China-VO) project, AstroCloud has been archived solid progresses during the last one year. Proposal management system and data access system are re-designed. Several new sub-systems are developed, including China-VO PaperData, AstroCloud Statics and Public channel. More data sets and application environments are integrated into the platform. LAMOST DR1, the largest astronomical spectrum archive was released to the public using the platform. The latest progresses will be introduced., Comment: 4 pages, 4 figures, ADASS XXV Proceedings

AstroCloud is a cyber-Infrastructure for Astronomy Research initiated by Chinese Virtual Observatory (China-VO) under funding support from NDRC (National Development and Reform commission) and CAS (Chinese Academy of Sciences). The ultimate goal of this project is to provide a comprehensive end-to-end astronomy research environment where several independent systems seamlessly collaborate to support the full lifecycle of the modern observational astronomy based on big data, from proposal submission, to data archiving, data release, and to in-situ data analysis and processing. In this paper, the architecture and key designs of the AstroCloud platform are introduced, including data access middleware, access control and security framework, extendible proposal workflow, and system integration mechanism., Comment: 4 pages, 4 figures, to be published in ADASS XXIV Proceedings. 5-9 October 2014 in Calgary, Canada

Data access and interoperability module connects the observation proposals, data, virtual machines and software. According to the unique identifier of PI (principal investigator), an email address or an internal ID, data can be collected by PI's proposals, or by the search interfaces, e.g. conesearch. Files associated with the searched results could be easily transported to cloud storages, including the storage with virtual machines, or several commercial platforms like Dropbox. Benefitted from the standards of IVOA (International Observatories Alliance), VOTable formatted searching result could be sent to kinds of VO software. Latter endeavor will try to integrate more data and connect archives and some other astronomical resources., Comment: 4 pages, 4 figures, to be published in ADASS XXIV Proceedings. 5-9 October 2014 in Calgary, Canada

Strain engineering has been recently recognized as an effective way to tailor the electrical properties of graphene. In the optical domain, effects such as strain-induced anisotropic absorption add an appealing functionality to graphene, opening the prospect for atomically thin optical elements. Indeed, graphene is currently one of the notable players in the intense drive towards bendable, thin, and portable electronic displays, where its intrinsically metallic, optically transparent, and mechanically robust nature are major advantages. Given that the intrinsic transparency of a graphene monolayer is 97.7 %, any small, reproducible, controllable, and potentially reversible modulation of transparency can have a significant impact for graphene as a viable transparent conducting electrode. Even more so, if the degree of modulation is polarization dependent. Here we show that the transparency in the visible range of graphene pre-strained on a Polyethylene terephthalate (PET) substrate exhibits a periodic modulation (0.1 %) as a function of polarization direction, which we interpret as strain-induced optical anisotropy. The degree of anisotropy is varied by reversible external manipulation of the level of pre-strain. The magnitude of strain is monitored independently by optical absorption and Raman spectroscopy, and the experimental observations are consistent with the theoretically expected modification of the optical conductivity of graphene arising from the strain-induced changes in the electronic dispersion of graphene. The strain sensitivity of the optical response of graphene demonstrated in this study can be potentially utilized towards novel ultra-thin optical devices and strain sensing applications., Comment: 18 pages, 4 figures

We discuss the matrix model in a class of 11D time dependent supersymmetric backgrounds as obtained in hep-th/0508191 . We construct the matrix model action through the matrix regularization of the membrane action in the background. We show that the action is exact to all order of fermionic coordinates. Furthermore We discuss the fuzzy sphere solutions in this background., Comment: 16 pages, 4 figures; references added, modifications of some comments in introduction, version to appear in PLB

This paper is a supplement of our earlier work JHEP 0410 (2004) 011[gr-qc/0409107].We map the vector potential of charged black holes into GEMS and find that its effect on the thermal spectrum is the same as that on the black hole side, i.e., it will induce a chemical potential in the thermal spectrum which is the same as that in the charged black holes.We also argue that the generalization of GEMS approach to non-stationary motions is not possible., Comment: Latex file, 10 pages, no figure; v2: minor corrections; v3: revised with title changed, one new section added and references added

We generalize the work of Deser and Levin on the unified description of Hawking radiation and Unruh effect to general stationary motions in spherically symmetric black holes. We have also matched the chemical potential term of the thermal spectrum of the two sides for uncharged black holes., Comment: Latex file, 12 pages, no figure; v2: typos fixed; v3: minor corrections, final version published in JHEP