arXiv subjects
Publications and source records attributed to Yu Shi.
We consider a wave packet of a spin-1/2 particle in a gravitational field, the effect of which can be described in terms of a succession of local inertial frames. It is shown that integrating out of the momentum yields a spin mixed state, with the entropy dependent on the deviation of metric from the flat spacetime. The decoherence occurs even if the particle is static in the gravitational field.
The direct observation of time reversal symmetry violation (TV) is important for the test of $CPT$ conservation and the Standard Model. In this paper, we study both time-dependent and time-independent genuine TV signals in entangled $D^0-\bar{D}^0$ pairs. A possible $CPT$-violation effect called the $ω$ effect is also investigated. In the $C=-1$ entangled state, the asymmetries due to TV are calculated to be of the order of $10^{-5}$ to $10^{-4}$ within the Standard Model, but the modification due to the $ω$ effect in the $C=-1$ states is found to be about $10\%-30\%$ when $|ω|\sim 10^{-4}$. This result is consistent with our Monte Carlo simulation, which implies that with $10^9$ to $10^{10}$ events, TV signals can be observed in the entangled $D^0-\bar{D}^0$ pairs, and the bound of $ω\sim 10^{-3}$ can be reached. The time-dependent and the time-independent asymmetries in the $C=-1$ $D^0-\bar{D}^0$ system provides a window to detect new physics such as the $ω$ effect, although they are not easily observable.
In the presence of an external magnetic field, the surface plasmon polariton that exists at the metal-dielectric interface is believed to support a unidirectional frequency range near the surface plasmon frequency, where the surface plasmon polariton propagates along one but not the opposite direction. Recent works have pointed to some of the paradoxical consequences of such a unidirectional range, including in particular the violation of the time-bandwidth product constraint that should otherwise apply in general in static systems. Here we show that such a unidirectional frequency range is nonphysical, using both a general thermodynamic argument, and a detailed calculation based on a nonlocal hydrodynamic Drude model for the metal permittivity. Our calculation reveals that the surface plasmon-polariton remains bidirectional for all frequencies. This work overturns a long-held belief in nonreciprocal photonics, and highlights the importance of quantum plasmonic concepts for the understanding of nonreciprocal plasmonic effects.
We compute two-particle production in p+A collisions and extract azimuthal harmonics, using the dilute-dense formalism in the Color Glass Condensate framework. The multiple scatterings of the partons inside the projectile proton on the dense gluons inside the target nucleus are expressed in terms of Wilson lines. They generate interesting correlations, which can be partly responsible for the signals of collectivity measured at RHIC and at the LHC. Most notably, while gluon Wilson loops yield vanishing odd harmonics, quark Wilson loops can generate sizable odd harmonics for two particle correlations. By taking both quark and gluon channels into account, we find that the overall second and third harmonics lie rather close to the recent PHENIX data at RHIC.
Heterogeneous information networks (HINs) are ubiquitous in real-world applications. In the meantime, network embedding has emerged as a convenient tool to mine and learn from networked data. As a result, it is of interest to develop HIN embedding methods. However, the heterogeneity in HINs introduces not only rich information but also potentially incompatible semantics, which poses special challenges to embedding learning in HINs. With the intention to preserve the rich yet potentially incompatible information in HIN embedding, we propose to study the problem of comprehensive transcription of heterogeneous information networks. The comprehensive transcription of HINs also provides an easy-to-use approach to unleash the power of HINs, since it requires no additional supervision, expertise, or feature engineering. To cope with the challenges in the comprehensive transcription of HINs, we propose the HEER algorithm, which embeds HINs via edge representations that are further coupled with properly-learned heterogeneous metrics. To corroborate the efficacy of HEER, we conducted experiments on two large-scale real-words datasets with an edge reconstruction task and multiple case studies. Experiment results demonstrate the effectiveness of the proposed HEER model and the utility of edge representations and heterogeneous metrics. The code and data are available at https://github.com/GentleZhu/HEER.
We consider a magnetic skyrmion adiabatically driven by a spin-polarized electrical current periodic in both space and time and asymmetric in space, and also subject to a random magnetic field representing the thermal fluctuations. We show that when the random magnetic field is low enough, while the time variation of the driving current is slow enough, the skyrmion flow is an integer multiply of the ratio between the space and time periods, the integer being a topological invariant called Chern number. This result is also demonstrated by numerically solving the stochastic Landau-Lifshitz-Gilbert (sLLG) and Langevin equations. Our work suggests a novel method of manipulating skyrmions with topological stability.
We show that a dynamic gain-loss modulation in an optical structure can lead to a direction-dependent parity-time (PT) phase transition. The phase transition can be made thresholdless in the forward direction, and yet remains with a non-zero threshold in the backward direction. As a result, non-reciprocal directional amplification can be realized.
Recently, integrated optics has gained interest as a hardware platform for implementing machine learning algorithms. Of particular interest are artificial neural networks, since matrix-vector multi- plications, which are used heavily in artificial neural networks, can be done efficiently in photonic circuits. The training of an artificial neural network is a crucial step in its application. However, currently on the integrated photonics platform there is no efficient protocol for the training of these networks. In this work, we introduce a method that enables highly efficient, in situ training of a photonic neural network. We use adjoint variable methods to derive the photonic analogue of the backpropagation algorithm, which is the standard method for computing gradients of conventional neural networks. We further show how these gradients may be obtained exactly by performing intensity measurements within the device. As an application, we demonstrate the training of a numerically simulated photonic artificial neural network. Beyond the training of photonic machine learning implementations, our method may also be of broad interest to experimental sensitivity analysis of photonic systems and the optimization of reconfigurable optics platforms.
Heterogeneous information networks (HINs) are ubiquitous in real-world applications. Due to the heterogeneity in HINs, the typed edges may not fully align with each other. In order to capture the semantic subtlety, we propose the concept of aspects with each aspect being a unit representing one underlying semantic facet. Meanwhile, network embedding has emerged as a powerful method for learning network representation, where the learned embedding can be used as features in various downstream applications. Therefore, we are motivated to propose a novel embedding learning framework---AspEm---to preserve the semantic information in HINs based on multiple aspects. Instead of preserving information of the network in one semantic space, AspEm encapsulates information regarding each aspect individually. In order to select aspects for embedding purpose, we further devise a solution for AspEm based on dataset-wide statistics. To corroborate the efficacy of AspEm, we conducted experiments on two real-words datasets with two types of applications---classification and link prediction. Experiment results demonstrate that AspEm can outperform baseline network embedding learning methods by considering multiple aspects, where the aspects can be selected from the given HIN in an unsupervised manner.
We introduce an implementation of a Laplace differentiator based on a photonic crystal slab that operates at transmission mode. We show that the Laplace differentiator can be implemented provided that the guided resonances near the $Γ$ point exhibit an isotropic band structure. Such a device may facilitate nanophotonics-based optical analog computing for image processing.
We evaluate the octupole in the large-$N_c$ limit in the McLerran-Venugopalan model, and derive a general expression of the 2n-point correlator, which can be applied in analytical studies of the multi-particle production in the scatterings between hard probes and dense targets.
We describe analytically and numerically the geometric phase arising from nonlinear frequency conversion and show that such a phase can be made non-reciprocal by momentum-dependent photonic transition. Such non-reciprocity is immune to the shortcomings imposed by dynamic reciprocity in Kerr and Kerr-like devices. We propose a simple and practical implementation, requiring only a single waveguide and one pump, while the geometric phase is controllable by the pump and promises robustness against fabrication errors.
We provide an analytic solution to the coupled-mode equations describing the steady-state of a single periodically-modulated optical resonator driven by a monochromatic input. The phenomenology of this system was qualitatively understood only in the adiabatic limit, i.e. for low modulation speed. However, both in and out of this regime, we find highly non-trivial effects for specific parameters of the modulation. For example, we show complete suppression of the transmission even with zero detuning between the input and the static resonator frequency. We also demonstrate the possibility for complete, lossless frequency conversion of the input into the side-band frequencies, as well as for optimizing the transmitted signal towards a given target temporal waveform. The analytic results are validated by first-principle simulations.
In 2012, Chen Ning Yang received a 90th birthday gift in the form of a black cube inscribed with his 13 most important contributions, which cover four major areas of physics: statistical mechanics, condensed matter physics, particle physics and field theory. We briefly describe these 13 contributions and make general comments about Yang's distinctive style as a trailblazing leader in research.
Quantum entanglement is the characteristic quantum correlation. Here we use this concept to analyze the quantum entanglement generated by Schwinger production of particle-antiparticle pairs in an electric field, as well as the change of mode entanglement as a consequence of the electric field effect on an entangled pair of particles. The system is partitioned by using momentum modes. Various kinds of pairwise mode entanglement are calculated as functions of the electric field. Both constant and pulsed electric fields are considered. The use of entanglement exposes information beyond that in particle number distributions.
We consider quantum entanglement of three accelerating qubits, each of which is locally coupled with a real scalar field, without causal influence among the qubits or among the fields. The initial states are assumed to be the GHZ and W states, which are the two representative three-partite entangled states. For each initial state, we study how various kinds of entanglement depend on the accelerations of the three qubits. All kinds of entanglement eventually suddenly die if at least two of three qubits have large enough accelerations. This result implies the eventual sudden death of all kinds of entanglement among three particles coupled with scalar fields when they are sufficiently close to the horizon of a black hole.
We consider Unruh-Wald qubit detector model adopted for the far future region of an exactly solvable 1+1 dimensional scalar field theory in a Robertson-Walker expanding spacetime. It is shown that the expansion of the universe in its history enhances the decoherence of the qubit coupled with a scalar field. Moreover, we consider two entangled qubits, each locally coupled a scalar field. The expansion of the universe in its history degrades the entanglement between the qubits, and can lead to entanglement sudden death if the initial entanglement is small enough. The details depend on the parameters characterizing the expansion of the universe. This work, albeit on a toy model, suggests that the history of the universe might be probed through the coherent and entanglement behavior of future detectors of quantum fields.
We consider two entangled accelerating qubits coupled with real scalar fields, each described by the Unruh-Wald model. It is demonstrated that because of the Unruh effect, the bipartite entanglement of the two qubits suddenly dies when the acceleration of one or more qubits are large enough. We also consider three entangled accelerating qubits in GHZ state and in W state, with equal acceleration-frequency ratio, and found that in either state, the tripartite entanglement suddenly dies at a certain value of acceleration-frequency ratio. The equivalence between the Rindler metric and the Schwarzchild metric in the vicinity of the horizon of a black hole implies that for the two entangled qubits outside a black hole, the entanglement suddenly dies when one or both of the qubits are close enough to the horizon, while for the three entangled qubits in GHZ or W state, the tripartite entanglement suddenly dies when these qubits are close enough to the horizon.