arXiv subjects

Shuyi Li

Publications and source records attributed to Shuyi Li.

32 records · Page 2Linked to original sources

arXiv2024

SememeLM: A Sememe Knowledge Enhanced Method for Long-tail Relation Representation

Recognizing relations between two words is a fundamental task with the broad applications. Different from extracting relations from text, it is difficult to identify relations among words without their contexts. Especially for long-tail relations, it becomes more difficult due to inadequate semantic features. Existing approaches based on language models (LMs) utilize rich knowledge of LMs to enhance the semantic features of relations. However, they capture uncommon relations while overlooking less frequent but meaningful ones since knowledge of LMs seriously relies on trained data where often represents common relations. On the other hand, long-tail relations are often uncommon in training data. It is interesting but not trivial to use external knowledge to enrich LMs due to collecting corpus containing long-tail relationships is hardly feasible. In this paper, we propose a sememe knowledge enhanced method (SememeLM) to enhance the representation of long-tail relations, in which sememes can break the contextual constraints between wors. Firstly, we present a sememe relation graph and propose a graph encoding method. Moreover, since external knowledge base possibly consisting of massive irrelevant knowledge, the noise is introduced. We propose a consistency alignment module, which aligns the introduced knowledge with LMs, reduces the noise and integrates the knowledge into the language model. Finally, we conducted experiments on word analogy datasets, which evaluates the ability to distinguish relation representations subtle differences, including long-tail relations. Extensive experiments show that our approach outperforms some state-of-the-art methods.

cs.CL↗

arXiv2024

Kitaev physics in the two-dimensional magnet NiPSe$_3$

The Kitaev interaction, found in candidate materials such as $α$-RuCl$_3$, occurs through the metal ($M$)-ligand ($X$)-metal ($M$) paths of the edge-sharing octahedra because the large spin-orbit coupling (SOC) on the metal atoms activates directional spin interactions. Here, we show that even in $3d$ transition-metal compounds, where the SOC of the metal atom is negligible, heavy ligands can induce bond-dependent Kitaev interactions. In this work, we take as an example the $3d$ transition-metal chalcogenophosphate NiPSe$_3$ and show that the key is found in the presence of a sizable SOC on the Se $p$ orbital, one which mediates the super-exchange between the nearest-neighbor Ni sites. Our study provides a pathway for engineering enhanced Kitaev interactions through the interplay of SOC strength, lattice distortions, and chemical substitutions.

cond-mat.str-el↗

arXiv2024

Field induced non-BEC transitions in frustrated magnets

Frustrated spin-systems have traditionally proven challenging to understand, owing to a scarcity of controlled methods for their analyses. By contrast, under strong magnetic fields, certain aspects of spin systems admit simpler and universal description in terms of hardcore bosons. The bosonic formalism is anchored by the phenomenon of Bose-Einstein condensation (BEC), which has helped explain the behaviors of a wide range of magnetic compounds under applied magnetic fields. Here, we focus on the interplay between frustration and externally applied magnetic field to identify instances where the BEC paradigm is no longer applicable. As a representative example, we consider the antiferromagnetic $J_1 - J_2 - J_3$ model on the square lattice in the presence of a uniform external magnetic field, and demonstrate that the frustration-driven suppression of the Néel order leads to a Lifshitz transition for the hardcore bosons. In the vicinity of the Lifshitz point, the physics becomes unmoored from the BEC paradigm, and the behavior of the system, both at and below the saturation field, is controlled by a Lifshitz multicritical point. We obtain the resultant universal scaling behaviors, and provide strong evidence for the existence of a frustration and magnetic-field driven correlated bosonic liquid state along the entire phase boundary separating the Néel phase from other magnetically ordered states.

cond-mat.str-el↗

arXiv2023

Bayesian Learning via Q-Exponential Process

Regularization is one of the most fundamental topics in optimization, statistics and machine learning. To get sparsity in estimating a parameter $u\in\mathbb{R}^d$, an $\ell_q$ penalty term, $\Vert u\Vert_q$, is usually added to the objective function. What is the probabilistic distribution corresponding to such $\ell_q$ penalty? What is the correct stochastic process corresponding to $\Vert u\Vert_q$ when we model functions $u\in L^q$? This is important for statistically modeling large dimensional objects, e.g. images, with penalty to preserve certainty properties, e.g. edges in the image. In this work, we generalize the $q$-exponential distribution (with density proportional to) $\exp{(- \frac{1}{2}|u|^q)}$ to a stochastic process named $Q$-exponential (Q-EP) process that corresponds to the $L_q$ regularization of functions. The key step is to specify consistent multivariate $q$-exponential distributions by choosing from a large family of elliptic contour distributions. The work is closely related to Besov process which is usually defined by the expanded series. Q-EP can be regarded as a definition of Besov process with explicit probabilistic formulation and direct control on the correlation length. From the Bayesian perspective, Q-EP provides a flexible prior on functions with sharper penalty ($q<2$) than the commonly used Gaussian process (GP). We compare GP, Besov and Q-EP in modeling functional data, reconstructing images, and solving inverse problems and demonstrate the advantage of our proposed methodology.

stat.ME↗

arXiv2023

Complete Suppression of Phase Segregation in Mixed-Halide Perovskite Nanocrystals under Periodic Heating

Under continuous light illumination, it is known that localized domains with segregated halide compositions form in semiconducting mixed-halide perovskites, thus severely limiting their optoelectronic applications due to the negative changes in bandgap energies and charge-carrier characteristics. Here we deposit mixed-halide perovskite CsPbBr1.2I1.8 nanocrystals onto an indium tin oxide substrate, whose temperature can be rapidly changed by ~10 degree in a few seconds by applying or removing an external voltage. Such a sudden temperature change induces a temporary transition of CsPbBr1.2I1.8 nanocrystals from the segregated phase to the mixed phase, the latter of which can be permanently maintained when the light illumination is coupled with periodic heating cycles. These findings mark the emergence of a practical solution to the detrimental phase-segregation problem, given that a small temperature modulation is readily available in various fundamental studies and practical devices using mixed-halide perovskites.

cond-mat.mtrl-sci↗

arXiv2023

Diffusive Excitonic Bands from Frustrated Triangular Sublattice in a Singlet-Ground-State System

Magnetic order in most materials occurs when magnetic ions with finite moments in a crystalline lattice arrange in a particular pattern below the ordering temperature determined by exchange interactions between the ions. However, when the crystal electric field (CEF) effect results in a spin-singlet ground state on individual magnetic sites, the collective ground state of the system can either remain non-magnetic, or more intriguingly, the exchange interactions between neighboring ions, provided they are sufficiently strong, can admix the excited CEF levels, resulting in a magnetically ordered ground state. The collective magnetic excitations in such a state are so-called spin excitons that describe the CEF transitions propagating through the lattice. In most cases, spin excitons originating from CEF levels of a localized single ion are dispersion-less in momentum (reciprocal) space and well-defined in both the magnetically ordered and paramagnetic states. Here we use thermodynamic and neutron scattering experiments to study stoichiometric Ni2Mo3O8 without site disorder, where Ni2+ ions form a bipartite honeycomb lattice comprised of two triangular lattices, with ions subject to the tetrahedral and octahedral crystalline environment, respectively. We find that in both types of ions, the CEF excitations have nonmagnetic singlet ground states, yet the material has long-range magnetic order. Furthermore, CEF spin excitons from the triangular-lattice arrangement of tetrahedral sites form, in both the antiferromagnetic and paramagnetic states, a dispersive diffusive pattern around the Brillouin zone boundary in reciprocal space. The present work thus demonstrates that spin excitons in an ideal triangular lattice magnet can have dispersive excitations, irrespective of the existence of static magnetic order, and this phenomenon is most likely due to spin entanglement and geometric frustrations.

cond-mat.str-el↗

arXiv2023

Magnons, Phonons, and Thermal Hall Effect in Candidate Kitaev Magnet $α$-RuCl$_3$

We study the nature of the debated thermal Hall effect in the candidate Kitaev material $α$-RuCl$_3$. Without assuming the existence of a gapped spin liquid, we show that a realistic minimal spin model in the canted zigzag phase suffices, at the level of linear spin-wave theory, to qualitatively explain the observed temperature and magnetic field dependence of the non-quantized thermal Hall conductivity $κ_{xy}$, with its origin lying in the Berry curvature of the magnon bands. The magnitude of the effect is however too small compared to the measurement by Czajka et al. [Nat. Mater. 22, 36-41 (2023)], even after scanning a broad range of model parameters so as to maximize $κ_{xy}/T$. Recent experiments suggest that phonons play an important role, which we show couple to the spins, endowing phonons with chirality. The resulting intrinsic contribution, from both magnons and phonons, is however still insufficient to explain the observed magnitude of the Hall signal. After careful analysis of the extrinsic phonon mechanisms, we use the recent experimental data on thermal transport in $α$-RuCl$_3$ by Lefrançois et al. [Phys. Rev. X 12, 021025 (2022)] to determine the phenomenological ratio of the extrinsic and intrinsic contributions $η\equiv κ_{xy}^{E}/κ_{xy}^{I}$. We find $η=1.2\pm 0.5$, which when combined with our computed intrinsic value, explains quantitavely both the magnitude and detailed temperature dependence of the experimental thermal Hall effect in $α$-RuCl$_3$.

cond-mat.str-el↗

arXiv2022

Scaling Up Bayesian Uncertainty Quantification for Inverse Problems using Deep Neural Networks

Due to the importance of uncertainty quantification (UQ), Bayesian approach to inverse problems has recently gained popularity in applied mathematics, physics, and engineering. However, traditional Bayesian inference methods based on Markov Chain Monte Carlo (MCMC) tend to be computationally intensive and inefficient for such high dimensional problems. To address this issue, several methods based on surrogate models have been proposed to speed up the inference process. More specifically, the calibration-emulation-sampling (CES) scheme has been proven to be successful in large dimensional UQ problems. In this work, we propose a novel CES approach for Bayesian inference based on deep neural network models for the emulation phase. The resulting algorithm is computationally more efficient and more robust against variations in the training set. Further, by using an autoencoder (AE) for dimension reduction, we have been able to speed up our Bayesian inference method up to three orders of magnitude. Overall, our method, henceforth called \emph{Dimension-Reduced Emulative Autoencoder Monte Carlo (DREAMC)} algorithm, is able to scale Bayesian UQ up to thousands of dimensions for inverse problems. Using two low-dimensional (linear and nonlinear) inverse problems we illustrate the validity of this approach. Next, we apply our method to two high-dimensional numerical examples (elliptic and advection-diffussion) to demonstrate its computational advantages over existing algorithms.

stat.CO↗

arXiv2022

Bayesian Spatiotemporal Modeling for Inverse Problems

Inverse problems with spatiotemporal observations are ubiquitous in scientific studies and engineering applications. In these spatiotemporal inverse problems, observed multivariate time series are used to infer parameters of physical or biological interests. Traditional solutions for these problems often ignore the spatial or temporal correlations in the data (static model), or simply model the data summarized over time (time-averaged model). In either case, the data information that contains the spatiotemporal interactions is not fully utilized for parameter learning, which leads to insufficient modeling in these problems. In this paper, we apply Bayesian models based on spatiotemporal Gaussian processes (STGP) to the inverse problems with spatiotemporal data and show that the spatial and temporal information provides more effective parameter estimation and uncertainty quantification (UQ). We demonstrate the merit of Bayesian spatiotemporal modeling for inverse problems compared with traditional static and time-averaged approaches using a time-dependent advection-diffusion partial different equation (PDE) and three chaotic ordinary differential equations (ODE). We also provide theoretic justification for the superiority of spatiotemporal modeling to fit the trajectories even it appears cumbersome (e.g. for chaotic dynamics).

stat.ME↗

arXiv2021

Topological Weyl magnons and thermal Hall effect in layered honeycomb ferromagnets

In this work, we study the topological properties and magnon Hall effect of a three-dimensional ferromagnet in the ABC stacking honeycomb lattice, motivated by the recent inelastic neutron scattering study of CrI$_3$. We show that the magnon band structure and Chern numbers of the magnon branches are significantly affected by the interlayer coupling $J_c$, which moreover has a qualitatively different effect in the ABC stacking compared to the AA stacking adopted by other authors. The nontrivial Chern number of the lowest magnon band is stabilized by the next-nearest-neighbour Dzyaloshinsky-Moriya interaction in each honeycomb layer, resulting in the hopping term similar to that in the electronic Haldane model for graphene. However, we also find several gapless Weyl points, separating the non-equivalent Chern insulating phases, tuned by the ratio of the interlayer coupling $J_c$ and the third-neighbour Heisenberg interaction $J_3$. We further show that the topological character of magnon bands results in non-zero thermal Hall conductivity, whose sign and magnitude depend on $J_c$ and the intra-layer couplings. Since the interlayer coupling strength $J_c$ can be easily tuned by applying pressure to the quasi-2D material such as CrI$_3$, this provides a potential route to tuning the magnon thermal Hall effect in an experiment.

cond-mat.str-el↗

arXiv2021

Noncollinear Antiferromagnetic Order and Effect of Spin-Orbit Coupling in Spin-1 Honeycomb Lattice

Motivated by the recently synthesized insulating nickelate Ni$_2$Mo$_3$O$_8$, which has been reported to have an unusual non-collinear magnetic order of Ni$^{2+}$ $S=1$ moments with a nontrivial angle between adjacent spins, we construct an effective spin-1 model on the honeycomb lattice, with the exchange parameters determined with the help of first principles electronic structure calculations. The resulting bilinear-biquadratic model, supplemented with the realistic crystal-field induced anisotropy, favors the collinear Néel state. We find that the crucial key to explaining the observed noncollinear spin structure is the inclusion of the Dzyaloshinskii--Moriya (DM) interaction between the neighboring spins. By performing the variational mean-field and linear spin-wave theory (LSWT) calculations, we determine that a realistic value of the DM interaction $D\approx 2.78$ meV is sufficient to quantitatively explain the observed angle between the neighboring spins. We furthermore compute the spectrum of magnetic excitations within the LSWT and random-phase approximation (RPA) which should be compared to future inelastic neutron measurements.

cond-mat.str-el↗

arXiv2018

The CST Bounce Universe model -- a parametric study

A bounce universe model with a scale-invariant and stable spectrum of primordial density perturbations was constructed using a consistent truncation of the D-brane dynamics from Type IIB string theory. A coupling was introduced between the tachyon field and the adjoint Higgs field on the D3-branes to lock the tachyon at the top of its potential hill and to model the bounce process, which is known as the Coupled Scalar and Tachyon Bounce (CSTB) Universe. The CSTB model has been shown to be ghost free, and it fulfils the null energy condition; in addition, it can also solve the Big Bang cosmic singularity problem. In this paper we conduct an extensive follow-up study of the parameter space of the CSTB model. In particular we are interested in the parameter values that can produce a single bounce to arrive at a radiation-dominated universe. We further establish that the CSTB universe is a viable alternative to inflation, as it can naturally produce a sufficient number of e-foldings in the locked inflation epoch and in the post-bounce expansion to overcome the four fundamental limitations of the Big Bang cosmology, which are flatness, horizon, homogeneity and singularity, resulting in a universe of the current size.

gr-qc↗

arXiv2016

Off-shell BCJ Relation in Nonlinear Sigma Model

We investigate relations among tree-level off-shell currents in nonlinear sigma model. Under Cayley parametrization, we propose and prove a general revised BCJ relation for even-point currents. Unlike the on-shell BCJ relation, the off-shell one behaves quite differently from Yang-Mills theory although the algebraic structure is the same. After performing the permutation summation in the revised BCJ relation, the sum is non-vanishing, instead, it equals to the sum of sub-current products with the BCJ coefficients under a specific ordering, which is presented by an explicit formula. Taking on-shell limit, this identity is reduced to the on-shell BCJ relation, and thus provides the full off-shell correspondence of tree-level BCJ relation in nonlinear sigma model.

hep-th↗

arXiv2015

Note on off-shell relations in nonlinear sigma model

In this note, we investigate relations between tree-level off-shell currents in nonlinear sigma model. Under Cayley parametrization, all odd-point currents vanish. We propose and prove a generalized $U(1)$ identity for even-point currents. The off-shell $U(1)$ identity given in [1] is a special case of the generalized identity studied in this note. The on-shell limit of this identity is equivalent with the on-shell KK relation. Thus this relation provides the full off-shell correspondence of tree-level KK relation in nonlinear sigma model.

hep-th↗