text
stringlengths 6
128k
|
|---|
Learning multi-lingual sentence embeddings is a fundamental task in natural
language processing. Recent trends in learning both mono-lingual and
multi-lingual sentence embeddings are mainly based on contrastive learning (CL)
among an anchor, one positive, and multiple negative instances. In this work,
we argue that leveraging multiple positives should be considered for
multi-lingual sentence embeddings because (1) positives in a diverse set of
languages can benefit cross-lingual learning, and (2) transitive similarity
across multiple positives can provide reliable structural information for
learning. In order to investigate the impact of multiple positives in CL, we
propose a novel approach, named MPCL, to effectively utilize multiple positive
instances to improve the learning of multi-lingual sentence embeddings.
Experimental results on various backbone models and downstream tasks
demonstrate that MPCL leads to better retrieval, semantic similarity, and
classification performances compared to conventional CL. We also observe that
in unseen languages, sentence embedding models trained on multiple positives
show better cross-lingual transfer performance than models trained on a single
positive instance.
|
We study the spectrum of one dimensional integral operators in bounded real
intervals of length $2L$, for value of $L$ large. The integral operators are
obtained by linearizing a non local evolution equation for a non conserved
order parameter describing the phases of a fluid. We prove a Perron-Frobenius
theorem showing that there is an isolated, simple minimal eigenvalue strictly
positive for $L$ finite, going to zero exponentially fast in $L$. We lower
bound, uniformly on $L$, the spectral gap by applying a generalization of the
Cheeger's inequality. These results are usefulfor deriving spectral properties
for non local Cahn-Hilliard type of equations in problems of interface
dynamics.
|
Let K be a 3-stranded knot (or link), and let L denote the number of
crossings in K. Let $\epsilon_{1}$ and $\epsilon_{2}$ be two positive real
numbers such that $\epsilon_{2}$ is less than or equal to 1.
In this paper, we create two algorithms for computing the value of the Jones
polynomial of K at all points $t=exp(i\phi)$ of the unit circle in the complex
plane such that the absolute value of $\phi$ is less than or equal to $\pi/3$.
The first algorithm, called the classical 3-stranded braid (3-SB) algorithm,
is a classical deterministic algorithm that has time complexity O(L). The
second, called the quantum 3-SB algorithm, is a quantum algorithm that computes
an estimate of the Jones polynomial of K at $exp(i\phi))$ within a precision of
$\epsilon_{1}$ with a probability of success bounded below by $1-\epsilon_{2}%.
The execution time complexity of this algorithm is O(nL), where n is the
ceiling function of (ln(4/\epsilon_{2}))/(2(\epsilon_{2})^2). The compilation
time complexity, i.e., an asymptotic measure of the amount of time to assemble
the hardware that executes the algorithm, is O(L).
|
Deuterium plays a crucial role in cosmology because the primordial D/H
abundance, in the context of big bang nucleosynthesis (BBN) theory, yields a
precise measure of the cosmic baryon content. Observations of D/H can limit or
measure the true primordial abundance because D is thought to be destroyed by
stars and thus D/H monotonically decreases after BBN. Recently, however, Mullan
& Linsky have pointed out that D arises as a secondary product of neutrons in
stellar flares which then capture on protons via n+p \to d + gamma, and that
this could dominate over direct D production in flares. Mullan & Linsky note
that if this process is sufficiently vigorous in flaring dwarf stars, it could
lead to significant non-BBN D production. We have considered the production of
D in stellar flares, both directly and by n capture. We find that for
reasonable flare spectra, n/d < 10 and (n+d)/6Li < 400, both of which indicate
that the n-capture channel does not allow for Galactic D production at a level
which will reverse the monotonic decline of D. We also calculate the 2.22 MeV
gamma-ray line production associated with n capture, and find that existing
COMPTEL limits also rule out significant D production in the Galaxy today.
Thus, we find flares in particular, and neutron captures in general, are not an
important Galactic source of D. On the other hand, we cannot exclude that flare
production might contribute to recent FUSE observations of large variations in
the local interstellar D/H abundance; we do, however, offer important
constraints on this possibility. Finally, since flare stars should inevitably
produce some n-capture events, a search for diffuse 2.22 MeV gamma-rays by
INTEGRAL can further constrain (or measure!) Galactic deuterium production via
n-capture.
|
We propose a flexible framework that deals with both singer conversion and
singers vocal technique conversion. The proposed model is trained on
non-parallel corpora, accommodates many-to-many conversion, and leverages
recent advances of variational autoencoders. It employs separate encoders to
learn disentangled latent representations of singer identity and vocal
technique separately, with a joint decoder for reconstruction. Conversion is
carried out by simple vector arithmetic in the learned latent spaces. Both a
quantitative analysis as well as a visualization of the converted spectrograms
show that our model is able to disentangle singer identity and vocal technique
and successfully perform conversion of these attributes. To the best of our
knowledge, this is the first work to jointly tackle conversion of singer
identity and vocal technique based on a deep learning approach.
|
Supersymmetric gauge theories, in higher dimensions compactified in an
orbifold, give a natural framework to unify the gauge bosons, Higgs fields and
even the matter fields in a single multiplet of the unifying gauge symmetry.
The extra dimensions and the supersymmetry are the two key ingredients for such
an unification. In this work, we investigate various scenarios for the
unification of the three gauge couplings, and the Yukawa couplings in the
Minimal Supersymmetric Standard Model (MSSM), as well as the trilinear Higgs
couplings \lambda and \kappa of the Non-Minimal Supersymmetric Standard Model
(NMSSM). We present an SU(8) model in six dimensions with N=2 supersymmetry,
compactified in a T^2/Z_6 orbifold which unifies the three gauge couplings with
\lambda and \kappa of NMSSM. Then, we present an SU(9) model in 6D, which, in
addition, includes partial unification of Yukawa couplings, either for the
up-type (top quark and Dirac tau-neutrino) or down-type (bottom quark and tau
lepton). We also study the phenomenological implications of these various
unification scenarios using the appropriate renormalization group equations,
and show that such unification works very well with the measured low energy
values of the couplings. The predicted upper bounds for the lightest neutral
Higgs boson mass in our model is higher than those in MSSM, but lower that
those in the general NMSSM (where the couplings \lambda and \kappa are
arbitrary). Some of the predictions of our models can be tested in the upcoming
Large Hadron Collider.
|
Fix $a \in \mathbb{Z}$, $a\notin \{0,\pm 1\}$. A simple argument shows that
for each $\epsilon > 0$, and almost all (asymptotically 100% of) primes $p$,
the multiplicative order of $a$ modulo $p$ exceeds $p^{\frac12-\epsilon}$. It
is an open problem to show the same result with $\frac12$ replaced by any
larger constant. We show that if $a,b$ are multiplicatively independent, then
for almost all primes $p$, one of $a,b,ab, a^2b, ab^2$ has order exceeding
$p^{\frac{1}{2}+\frac{1}{30}}$. The same method allows one to produce, for each
$\epsilon > 0$, explicit finite sets $\mathcal{A}$ with the property that for
almost all primes $p$, some element of $\mathcal{A}$ has order exceeding
$p^{1-\epsilon}$. Similar results hold for orders modulo general integers $n$
rather than primes $p$.
|
Using a representative sample of 14 star-forming dwarf galaxies in the local
Universe, we show the existence of a spaxel-to-spaxel anti-correlation between
the index N2 (log([NII]6583/Halpha)) and the Halpha flux. These two quantities
are commonly employed as proxies for gas-phase metallicity and star formation
rate (SFR), respectively. Thus, the observed N2 to Halpha relation may reflect
the existence of an anti-correlation between the metallicity of the gas forming
stars and the SFR it induces. Such an anti-correlation is to be expected if
variable external metal-poor gas fuels the star-formation process.
Alternatively, it can result from the contamination of the star-forming gas by
stellar winds and SNe, provided that intense outflows drive most of the metals
out of the star-forming regions. We also explore the possibility that the
observed anti-correlation is due to variations in the physical conditions of
the emitting gas, other than metallicity. Using alternative methods to compute
metallicity, as well as previous observations of HII regions and
photoionization models, we conclude that this possibility is unlikely. The
radial gradient of metallicity characterizing disk galaxies does not produce
the correlation either.
|
Network virtualization techniques allow for the coexistence of many virtual
networks (VNs) jointly sharing the resources of an underlying substrate
network. The Virtual Network Embedding problem (VNE) arises when looking for
the most profitable set of VNs to embed onto the substrate. In this paper, we
address the offline version of the problem. We propose a Mixed-Integer Linear
Programming formulation to solve it to optimality which accounts for acceptance
and rejection of virtual network requests, allowing for both splittable and
unsplittable (single path) routing schemes. Our formulation also considers a
Rent-at-Bulk (RaB) model for the rental of substrate capacities where economies
of scale apply. To better emphasize the importance of RaB, we also compare our
method to a baseline one which only takes RaB into account a posteriori, once a
solution to VNE, oblivious to RaB, has been found. Computational experiments
show the viability of our approach, stressing the relevance of addressing RaB
directly with an exact formulation
|
The experimental evidence for pentaquarks, both old an new, is discussed.
Constraints due to $K^+N$ scattering data from previous decades is first
reviewed, followed by experiments with positive evidence and those with null
results. Finally, the problem of the narrow width of the Theta+ pentaquark is
discussed, along with theoretical implications.
|
We review the concept of infinity as applied to regularization procedures in
Quantum Electrodynamics. A clear distinction that is lacking in current
literature is made between the physical contents of renormalization, and the
mathematical aspects of regularization. Robinson's non-standard analysis is
offered as a means to settle the ambiguities of the theory, in the spirit of
Paul Dirac's well known comments concerning the weak status of the mathematics
used in traditional regularization schemes. As a case study we consider the
Casimir effect
|
We present new analysis of HST images of (47171) 1999 TC36 that confirm it as
a triple system. Fits to the point-spread function consistently show that the
apparent primary is itself composed of two similar-sized components. The two
central components, A1 and A2, can be consistently identified in each of nine
epochs spread over seven years of time. In each instance the component
separation, ranging from 0.023+/-0.002 to 0.031+/-0.003 arcsec, is roughly one
half of the Hubble Space Telescope's diffraction limit at 606 nm. The orbit of
the central pair has a semi-major axis of a~867 km with a period of P~1.9 days.
These orbital parameters yield a system mass that is consistent with Msys =
12.75+/-0.06 10^18 kg derived from the orbit of the more distant secondary,
component B. The diameters of the three components are dA1= 286(+45,-38) km,
dA2= 265(+41,-35 km and dB= 139(+22,-18) km. The relative sizes of these
components are more similar than in any other known multiple in the solar
system. Taken together, the diameters and system mass yield a bulk density of
p=542(+317,-211) kg m^-3. HST Photometry shows that component B is variable
with an amplitude of >=0.17+/-0.05 magnitudes. Components A1 and A2 do not show
variability larger than 0.08+/-0.03 magnitudes approximately consistent with
the orientation of the mutual orbit plane and tidally-distorted equilibrium
shapes. The system has high specific angular momentum of J/J'=0.93, comparable
to most of the known Transneptunian binaries.
|
Molecular dynamics simulations are used to study a unique expanded jammed
state. Tension transforms many glassy polymers from a dense glass to a network
of fibrils and voids called a craze. Entanglements between polymers and
interchain friction jam the system after a fixed increase in volume. As in
dense jammed systems, the distribution of forces is exponential, but they are
tensile rather than compressive. The broad distribution of forces has important
implications for fibril breakdown and the ultimate strength of crazes.
|
In this work, we construct a phenomenological effective model for the
heavy-light quark systems, which consist of (u,d,c,b) quarks, i.e. extended
nonlocal chiral-quark model (ExNLChQM) to compute the heavy-meson weak-decay
constants f_D and f_B. ExNLChQM is based on the heavy-quark effective field
theory as well as the dilute instanton-vacuum configuration. In ExNLChQM, a
certain portion of the heavy-meson mass is considered to be generated from the
nontrivial QCD vacuum contribution, being similar to that for the light quarks
in usual instanton approaches. Hence, the effective heavy- and light-quark
masses become momentum-dependent and play the role of a smooth UV regulator.
Employing a generic external-field method applied to the effective action from
ExNLChQM, we obtain f_D = (169.28 ~ 234.57) MeV and f_B = (165.41 ~ 229.21) MeV
from the numerical results, depending on different model parameters. These
values are in relatively good agreement with experimental data and various
theoretical estimations. We also discuss the heavy-quark effects on the QCD
vacuum, and the decay constants f_D^* and f_B^* in terms of the heavy-quark
spin symmetry.
|
We study the occurrence of excitonic superfluidity in electron-hole bilayers
at zero temperature. We not only identify the crossover in the phase diagram
from the BCS limit of overlapping pairs to the BEC limit of non-overlapping
tightly-bound pairs but also, by varying the electron and hole densities
independently, we can analyze a number of phases that occur mainly in the
crossover region. With different electron and hole effective masses, the phase
diagram is asymmetric with respect to excess electron or hole densities. We
propose as the criterion for the onset of superfluidity, the jump of the
electron and hole chemical potentials when their densities cross.
|
A computational study of three-dimensional instability of steady flows in a
helical pipe of arbitrary curvature and torsion is carried out for the first
time. The problem is formulated in Germano coordinates in two equivalent but
different forms of the momentum equation, so that results obtained using both
formulations cross verify each other. An additional formulation in the
cylindrical coordinates is applied for a limiting case of the toroidal pipe.
The calculations are performed by the finite volume and finite difference
methods. Grid independence of the results is established for both steady flows,
the eigenvalues associated with the linear stability problem, and the critical
parameters. The calculated steady flows agree well with experimental
measurements and previous numerical results. The computed critical Reynolds
numbers corresponding to the onset of oscillatory instability agree well with
the most recent experimental results, but disagree with the earlier ones. Novel
results related to the parametric stability study are reported.
|
Over the past decade, the issue of gain degradation at broadside in periodic
leaky-wave antennas (P-LWAs) has been resolved, using a circuit modeling
approach, by introducing proper asymmetry in the unit cell of the antenna
structure. This paper provides a more fundamental and insightful perspective of
the problem by showing, using a simple coupled-mode analysis, that the optimal
level of structural asymmetry corresponds to an exceptional point of the
coupling parameter between the two eigenmodes of the P-LWA. This contribution
represents a key step towards the development of a full electromagnetic
resolution of the broadside issue.
|
In a graph $G$, an {\it efficient dominating set} is a subset $D$ of vertices
such that $D$ is an independent set and each vertex outside $D$ has exactly one
neighbor in $D$. The {\textsc{Efficient Dominating Set}} problem (EDS) asks for
the existence of an efficient dominating set in a given graph $G$. The EDS is
known to be $NP$-complete for $P_7$-free graphs, and is known to be polynomial
time solvable for $P_5$-free graphs. However, the computational complexity of
the EDS problem is unknown for $P_6$-free graphs. In this paper, we show that
the EDS problem can be solved in polynomial time for a subclass of $P_6$-free
graphs, namely ($P_6$, banner)-free graphs.
|
It is shown that the simplest multiplicative random complex matrix model
generalizes the large-N phase structure found in the unitary case: A
perturbative regime is joined to a nonperturbative regime at a point of
nonanalyticity.
|
The following study investigates the thermodynamic reaction barriers during
nitrogen fixation for an inorganic sulfur desorbed photocatalyst Molybdenum
disulfide surface. The design space is investigated using an density functional
theory (DFT) method within a space bound by MMoS$_2$ M=Mo,Fe,Co. The discussion
focuses on Heyrovsky type reactions along both the associative and dissociative
pathway. A key insight into the roles of the inorganic and the balance between
nitrogen and hydrogen affinity, providing evidence for an optimal material that
minimizes the required over-potential. It is found that phases with a higher
concentration of Mo face high reaction barrier involving nitrogen, where phases
with higher concentrations of Fe and Co face high reaction barriers involving
hydrogen species. In the absence of kinetic considerations, the best phase was
predicted to be the 1T phase with a Mo$_{0.75}$Fe$_{0.25}$S$_{2}$ composition.
This phase proved to have a balance of hydrogen and nitrogen affinity and
follows the dissociative pathway, which can be evolved through non-thermal
methods.
|
This article summarizes theoretical predictions for the density and isospin
dependence of the nuclear mean field and the corresponding nuclear equation of
state. We compare predictions from microscopic and phenomenological approaches.
An application to heavy ion reactions requires to incorporate these forces into
the framework of dynamical transport models. Constraints on the nuclear
equation of state derived from finite nuclei and from heavy ion reactions are
discussed.
|
Diderot's \textit{Encyclop\'edie} is a reference work from XVIIIth century in
Europe that aimed at collecting the knowledge of its era. \textit{Wikipedia}
has the same ambition with a much greater scope. However, the lack of digital
connection between the two encyclopedias may hinder their comparison and the
study of how knowledge has evolved. A key element of \textit{Wikipedia} is
Wikidata that backs the articles with a graph of structured data. In this
paper, we describe the annotation of more than 10,300 of the
\textit{Encyclop\'edie} entries with Wikidata identifiers enabling us to
connect these entries to the graph. We considered geographic and human
entities. The \textit{Encyclop\'edie} does not contain biographic entries as
they mostly appear as subentries of locations. We extracted all the geographic
entries and we completely annotated all the entries containing a description of
human entities. This represents more than 2,600 links referring to locations or
human entities. In addition, we annotated more than 9,500 entries having a
geographic content only. We describe the annotation process as well as
application examples. This resource is available at
https://github.com/pnugues/encyclopedie_1751
|
Motivated by a natural problem in online model selection with bandit
information, we introduce and analyze a best arm identification problem in the
rested bandit setting, wherein arm expected losses decrease with the number of
times the arm has been played. The shape of the expected loss functions is
similar across arms, and is assumed to be available up to unknown parameters
that have to be learned on the fly. We define a novel notion of regret for this
problem, where we compare to the policy that always plays the arm having the
smallest expected loss at the end of the game. We analyze an arm elimination
algorithm whose regret vanishes as the time horizon increases. The actual rate
of convergence depends in a detailed way on the postulated functional form of
the expected losses. Unlike known model selection efforts in the recent bandit
literature, our algorithm exploits the specific structure of the problem to
learn the unknown parameters of the expected loss function so as to identify
the best arm as quickly as possible. We complement our analysis with a lower
bound, indicating strengths and limitations of the proposed solution.
|
Purpose: In some proton therapy facilities, patient alignment relies on two
2D orthogonal kV images, taken at fixed, oblique angles, as no 3D on-the-bed
imaging is available. The visibility of the tumor in kV images is limited since
the patient's 3D anatomy is projected onto a 2D plane, especially when the
tumor is behind high-density structures such as bones. This can lead to large
patient setup errors. A solution is to reconstruct the 3D CT image from the kV
images obtained at the treatment isocenter in the treatment position.
Methods: An asymmetric autoencoder-like network built with vision-transformer
blocks was developed. The data was collected from 1 head and neck patient: 2
orthogonal kV images (1024x1024 voxels), 1 3D CT with padding (512x512x512)
acquired from the in-room CT-on-rails before kVs were taken and 2
digitally-reconstructed-radiograph (DRR) images (512x512) based on the CT. We
resampled kV images every 8 voxels and DRR and CT every 4 voxels, thus formed a
dataset consisting of 262,144 samples, in which the images have a dimension of
128 for each direction. In training, both kV and DRR images were utilized, and
the encoder was encouraged to learn the jointed feature map from both kV and
DRR images. In testing, only independent kV images were used. The full-size
synthetic CT (sCT) was achieved by concatenating the sCTs generated by the
model according to their spatial information. The image quality of the
synthetic CT (sCT) was evaluated using mean absolute error (MAE) and
per-voxel-absolute-CT-number-difference volume histogram (CDVH).
Results: The model achieved a speed of 2.1s and a MAE of <40HU. The CDVH
showed that <5% of the voxels had a per-voxel-absolute-CT-number-difference
larger than 185 HU.
Conclusion: A patient-specific vision-transformer-based network was developed
and shown to be accurate and efficient to reconstruct 3D CT images from kV
images.
|
We formulate problems of tight closure theory in terms of projective bundles
and subbundles. This provides a geometric interpretation of such problems and
allows us to apply intersection theory to them. This yields new results
concerning the tight closure of a primary ideal in a two-dimensional graded
domain.
|
The potential of photon-magnon hybrid systems as building blocks for quantum
information science has been widely demonstrated, and it is still the focus of
much research. We leverage the strengths of this unique heterogeneous physical
system in the field of precision physics beyond the standard model, where the
sensitivity to the so-called "invisibles" is currently being boosted by quantum
technologies. Here, we demonstrate that quanta of spin waves, driven by
tiniest, effective magnetic field, can be detected in a large frequency band
using a hybrid system as transducer. This result can be applied to the search
of cosmological signals related, for example, to cold Dark Matter, which may
directly interact with magnons. Our model of the transducer is based on a
second-quantisation two-oscillators hybrid system, it matches the observations,
and can be easily extended to thoroughly describe future large-scale
ferromagnetic haloscopes.
|
We present near-infrared imaging of a sample of 57 relatively large, Northern
spiral galaxies with low inclination. After describing the selection criteria
and some of the basic properties of the sample, we give a detailed description
of the data collection and reduction procedures. The K_s lambda=2.2 micron
images cover most of the disk for all galaxies, with a field of view of at
least 4.2 arcmin. The spatial resolution is better than an arcsec for most
images. We fit bulge and exponential disk components to radial profiles of the
light distribution. We then derive the basic parameters of these components, as
well as the bulge/disk ratio, and explore correlations of these parameters with
several galaxy parameters.
|
We discuss gamma gamma partial widths of pseudoscalar/isoscalar mesons eta(M)
in the mass region M = 1000-1500 MeV. The transition amplitudes eta(1295) ->
gamma gamma and eta(1440) -> gamma gamma are studied within an assumption that
the decaying mesons are the members of the first radial excitation nonet 2^1S_0
q\bar q. The calculations show that partial widths being of the order of 0.1
keV are dominantly due to the n\bar n meson component while the contribution of
the s\bar s component is small.
|
Selecting safe landing sites in non-cooperative environments is a key step
towards the full autonomy of UAVs. However, the existing methods have the
common problems of poor generalization ability and robustness. Their
performance in unknown environments is significantly degraded and the error
cannot be self-detected and corrected. In this paper, we construct a UAV system
equipped with low-cost LiDAR and binocular cameras to realize autonomous
landing in non-cooperative environments by detecting the flat and safe ground
area. Taking advantage of the non-repetitive scanning and high FOV coverage
characteristics of LiDAR, we come up with a dynamic time depth completion
algorithm. In conjunction with the proposed self-evaluation method of the depth
map, our model can dynamically select the LiDAR accumulation time at the
inference phase to ensure an accurate prediction result. Based on the depth
map, the high-level terrain information such as slope, roughness, and the size
of the safe area are derived. We have conducted extensive autonomous landing
experiments in a variety of familiar or completely unknown environments,
verifying that our model can adaptively balance the accuracy and speed, and the
UAV can robustly select a safe landing site.
|
Given a closed smooth manifold M which carries a positive scalar curvature
metric, one can associate an abelian group P(M) to the space of positive scalar
curvature metrics on this manifold. The group of all diffeomorphisms of the
manifold naturally acts on P(M). The moduli group of positive scalar curvature
metrics is defined to be the quotient abelian group of this action, i.e. the
coinvariant of the action. The moduli group measures the size of the moduli
space of positive scalar curvature metrics on M. In this paper, we use the
higher rho invariant and the finite part of the K-theory of the group
C*-algebra of the fundamental group of M to give a lower bound of the rank of
the moduli group.
|
The bandgap constitutes a challenging problem in density functional theory
(DFT) methodologies. It is known that the energy gap values calculated by
common DFT approaches are underestimated. The bandgap was also found to be
related to the derivative discontinuity (DD) of the exchange-correlation
potential in the Kohn-Sham formulation of DFT. Several reports have shown that
DD appears as a step on the potential curve. The step structure is a mandatory
structure for aligning the KS energy levels in the ionization potentials in a
dissociated molecule in both fragments and is a function of electron
localisation. Reproducing the step in the DFT framework gives the charge
transfer process and the correct energy gap and describes the source of
dissociation. This step phenomenon has not yet been studied in the non-additive
kinetic potential $v^{\text{NAD}}[\rho_A,\rho_B](\textbf{r})$, a key quantity
used in embedding theories. While $v^{\text{NAD}}[\rho_A,\rho_B](\textbf{r})$
is known to be difficult to approximate, in this work, we explain how an
accurate energy gap can be produced from the analytically inverted
$v^{\text{NAD}}[\rho_A,\rho_B](\textbf{r})$, even if we use the input densities
calculated by the local and semi-local functionals. We used the precisely
calculated $v^{\text{NAD}}[\rho_A,\rho_B](\textbf{r})$ reported in our previous
publication [Phys. Rev. A 106, 042812 (2022)] to produce the energy gap for
some model systems and report in this work the promising accuracy of our
results through the comparison with the results obtained from one of the most
accurate calculations, OEP theory with the KLI local approximation.
|
Thermally-treating amorphous indium-oxide films is used in various basic
studies as a means of tuning the system disorder. In this process the
resistance of a given sample decreases while its amorphous structure and
chemical composition is preserved. The main effect of the process is an
increase in the system density which in turn leads to improved interatomic
overlap which is easily detected as improved conductivity. A similar effect has
been observed in studies of other amorphous systems that were subjected to
pressure. In the current work we show that the Raman spectra of amorphous
indium-oxide change in response to thermal-treatment in a similar way as in
pressure experiments performed on other disordered and amorphous systems. We
present a study of how thermal-treatment changes the system dynamics by
monitoring the resistance versus time of indium-oxide films following various
stages of thermal-treatment. The time dependence of the sample resistance fits
the stretched exponential law with parameters that change systematically with
further annealing. Implication of these results to slow dynamics phenomena that
are governed by the Kohlrausch's law are discussed.
|
We present the homogeneous abundance analysis for a combined sample of 185
giants in the bulge globular cluster (GC) NGC 6388. Our results are used to
describe the multiple stellar populations and differences or analogies with
bulge field stars. Proton-capture elements indicate that a single class of
first-generation polluters is sufficient to reproduce both the extreme and
intermediate parts of the anti-correlations among light elements O, Na, Mg, and
Al, which is at odds with our previous results based on a much smaller sample.
The abundance pattern of other species in NGC 6388 closely tracks the trends
observed in bulge field stars. In particular, the alpha-elements, including Si,
rule out an accreted origin for NGC 6388, confirming our previous results based
on iron-peak elements, chemo-dynamical analysis, and the age-metallicity
relation. The neutron-capture elements are generally uniform, although the
[Zr/Fe] ratio shows an intrinsic scatter, correlated to Na and Al abundances.
Instead, we do not find enhancement in neutron-capture elements for stars whose
photometric properties would classify NGC 6388 as a type II GC. Together with
the homogeneity in [Fe/H] we found in a previous paper, this indicates we need
to better understand the criteria to separate classes of GCs, coupling
photometry, and spectroscopy. These results are based on abundances of 22
species (O, Na, Mg, Al, Si, Ca, Ti, Sc, V, Cr, Mn, Fe, Co, Ni, Zn, Y, Zr, Ba,
La, Ce, Nd, and Eu) from UVES spectra sampling proton-, alpha-, neutron-capture
elements, and Fe-peak elements. For 12 species, we also obtain abundances in a
large number of giants (up to 150) from GIRAFFE spectra.
|
In this article, we calculate the friction between two counter-flowing
bosonic and fermionic super-fluids. In the limit where the boson-boson and
boson-fermion interactions can be treated within the mean-field approximation,
we show that the force can be related to the dynamical structure factor of the
fermionic component. Finally, we provide asymptotic expressions for weakly and
strongly attractive fermions and show that the damping rate obeys simple
scaling laws close to the critical velocity.
|
In this paper, we explore the limit structure of a sequence of Riemannian
manifolds with Bakry-\'Emery Ricci curvature bounded below in the
Gromov-Hausdorff topology. By extending the techniques established by
Cheeger-Cloding for Riemannian manifolds with Ricci curvature bounded below, we
prove that each tangent space at a point of the limit space is a metric cone.
We also analyze the singular structure of the limit space analogous to a work
of Cheeger-Colding-Tian. Our results will be applied to study the limit space
of a sequence of K\"ahler metrics arising from solutions of certain complex
Monge-Amp\`ere equations for the existence of K\"ahler-Ricci solitons on a Fano
manifold via the continuity method.
|
In this paper, we study the mathematical structure and numerical
approximation of elliptic problems posed in a (3D) domain $\Omega$ when the
right-hand side is a (1D) line source $\Lambda$. The analysis and approximation
of such problems is known to be non-standard as the line source causes the
solution to be singular. Our main result is a splitting theorem for the
solution; we show that the solution admits a split into an explicit, low
regularity term capturing the singularity, and a high-regularity correction
term $w$ being the solution of a suitable elliptic equation. The splitting
theorem states the mathematical structure of the solution; in particular, we
find that the solution has anisotropic regularity. More precisely, the solution
fails to belong to $H^1$ in the neighbourhood of $\Lambda$, but exhibits
piecewise $H^2$-regularity parallel to $\Lambda$. The splitting theorem can
further be used to formulate a numerical method in which the solution is
approximated via its correction function $w$. This approach has several
benefits. Firstly, it recasts the problem as a 3D elliptic problem with a 3D
right-hand side belonging to $L^2$, a problem for which the discretizations and
solvers are readily available. Secondly, it makes the numerical approximation
independent of the discretization of $\Lambda$; thirdly, it improves the
approximation properties of the numerical method. We consider here the Galerkin
finite element method, and show that the singularity subtraction then recovers
optimal convergence rates on uniform meshes, i.e., without needing to refine
the mesh around each line segment. The numerical method presented in this paper
is therefore well-suited for applications involving a large number of line
segments. We illustrate this by treating a dataset (consisting of $\sim 3000$
line segments) describing the vascular system of the brain.
|
Phosphorene, a monolayer of black phosphorus, is promising for nanoelectronic
applications not only because it is a natural p-type semiconductor but also it
possesses a layer-number dependent direct bandgap (in the range of 0.3 eV~1.5
eV). On basis of the density functional theory calculations, we investigate
electronic properties of the bilayer phosphorene with different stacking
orders. We find that the direct bandgap of the bilayers can vary from 0.78 -
1.04 eV with three different stacking orders. In addition, a vertical electric
field can further reduce the bandgap down to 0.56 eV (at the field strength 0.5
V/{\AA}). More importantly, we find that when a monolayer of MoS_2 is
superimposed with the p-type AA- or AB-stacked bilayer phosphorene, the
combined tri-layer can be an effective solar-cell material with type-II
heterojunction alignment. The power conversion efficiency is predicted to be
~18% or 16% with AA- or AB-stacked bilayer phosphorene, higher than reported
efficiencies of the state-of-the-art trilayer graphene/transition metal
dichalcogenide solar cells.
|
We study the role of adiabatic index in determining the critical points in
the transonic low angular momentum accretion flow onto a black hole. We present
the general relativistic 2D hydrodynamic simulations of axisymmetric, inviscid
accretion flows in a fixed Kerr black hole gravitational field. A relativistic
fluid where its bulk velocity is comparable to the speed of light, flowing in
the accretion disk very close to the horizon can be described by an adiabatic
index of 4/3 < {\gamma} < 5/3. The time dependent evolution of the shock
position and respective effects on mass accretion rate and oscillation
frequency with varying adiabatic index is discussed in the context of the
observed microquasars.
|
We consider a one-Laplace equation perturbed by $p$-Laplacian with
$1<p<\infty$. We prove that a weak solution is continuously differentiable
($C^{1}$) if it is convex. Note that similar result fails to hold for the
unperturbed one-Laplace equation. The main difficulty is to show
$C^{1}$-regularity of the solution at the boundary of a facet where the
gradient of the solution vanishes. For this purpose we blow-up the solution and
prove that its limit is a constant function by establishing a Liouville-type
result, which is proved by showing a strong maximum principle. Our argument is
rather elementary since we assume that the solution is convex. A few
generalization is also discussed.
|
The dynamics of infectious diseases propagating in populations depends both
on human interaction patterns, the contagion process and the pathogenesis
within hosts. The immune system follows a circadian rhythm and, consequently,
the chance of getting infected varies with the time of day an individual is
exposed to the pathogen. The movement and interaction of people also follow
24-hour cycles, which couples these two phenomena. We use a stochastic
metapopulation model informed by hourly mobility data for two medium-sized
Chinese cities. By this setup, we investigate how the epidemic risk depends on
the difference of the clocks governing the population movement and the immune
systems. In most of the scenarios we test, we observe circadian rhythms would
constrain the pace and extent of disease emergence. The three measures
(strength, outward transmission risk and introduction risk) are highly
correlated with each other. For example of the Yushu City, outward transmission
risk and introduction risk are correlated with a Pearson's correlation
coefficient of 0.83, and the risks correlate to strength with coefficients of
$-0.85$ and $-0.75$, respectively (all have $p<0.05$), in simulations with no
circadian effect and $R_0=1.5$. The relation between the circadian rhythms of
the immune system and daily routines in human mobility can affect the pace and
extent of infectious disease spreading. Shifting commuting times could mitigate
the emergence of outbreaks.
|
A simple model of the driven motion of interacting particles in a two
dimensional random medium is analyzed, focusing on the critical behavior near
to the threshold that separates a static phase from a flowing phase with a
steady-state current. The critical behavior is found to be surprisingly robust,
being independent of whether the driving force is increased suddenly or
adiabatically. Just above threshold, the flow is concentrated on a sparse
network of channels, but the time scale for convergence to this fixed network
diverges with a larger exponent that that for convergence of the current
density to its steady-state value. This is argued to be caused by the
``dangerous irrelevance'' of dynamic particle collisions at the critical point.
Possible applications to vortex motion near to the critical current in dirty
thin film superconductors are discussed briefly.
|
We investigate the effect of coupling between translational and internal
degrees of freedom of composite quantum particles on their localization in a
random potential. We show that entanglement between the two degrees of freedom
weakens localization due to the upper bound imposed on the inverse
participation ratio by purity of a quantum state. We perform numerical
calculations for a two-particle system bound by a harmonic force in a 1D
disordered lattice and a rigid rotor in a 2D disordered lattice. We illustrate
that the coupling has a dramatic effect on localization properties, even with a
small number of internal states participating in quantum dynamics.
|
GomalizingFlow.jl: is a package to generate configurations for quantum field
theory on the lattice using the flow based sampling algorithm in Julia
programming language. This software serves two main purposes: to accelerate
research of lattice QCD with machine learning with easy prototyping, and to
provide an independent implementation to an existing public Jupyter notebook in
Python/PyTorch. GomalizingFlow.jl implements, the flow based sampling
algorithm, namely, RealNVP and Metropolis-Hastings test for two dimension and
three dimensional scalar field, which can be switched by a parameter file. HMC
for that theory also implemented for comparison. This package has Docker image,
which reduces effort for environment construction. This code works both on CPU
and NVIDIA GPU.
|
We make a theoretical study of the three-body system composed of ${\bar
K^*}{\bar B^*}{\bar B^*}$ to look for possible bound states, which could be
associated to mesonic resonances of very exotic nature, containing open strange
and double-bottom flavours. The three-body interaction is evaluated by using
the fixed center approach to the Faddeev equations where the ${\bar B^*}{\bar
B^*}$ is bound forming an $I(J^P)=0(1^+)$ state, as it was found in previous
works, and the third particle, the ${\bar B^*}$, of much smaller mass,
interacts with the components of the cluster. We obtain bound states for all
the channels considered: spin $J=0$, 1 and 2, all of them with isospin $I=1/2$
and negative parity.
|
In this work, we investigate the relationship between the geometrical
properties, the photon sphere, the shadow, and the eikonal quasinormal modes of
electrically charged black holes in 4D Einstein-Gauss-Bonnet gravity.
Quasinormal modes are complex frequency oscillations that are dependent on the
geometry of spacetime and have significant applications in studying black hole
properties and testing alternative theories of gravity. Here, we focus on the
eikonal limit for high frequency quasinormal modes and their connection to the
black holes geometric characteristics. To study the photon sphere, quasinormal
modes, and black hole shadow, we employ various techniques such as the WKB
method in various orders of approximation, the Poschl-Teller potential method,
and Churilova's analytical formulas. Our results indicate that the real part of
the eikonal quasinormal mode frequencies of test fields are linked to the
unstable circular null geodesic and are correlated with the shadow radius for
an Charged Einstein-Gauss-Bonnet 4D black hole. Furthermore, we found that the
real part of quasinormal modes, the photon sphere and shadow radius have a
lower value for charged black holes in 4D Einstein-Gauss-Bonnet gravity
compared to black holes without electric charge and those of static black holes
in general relativity. Additionally, we explore various analytical formulas for
the photon spheres and shadows, deducing an Churilova's approximate formula for
the black hole shadow radius of the Charged Einstein-Gauss-Bonnet 4D black
hole, which arises from its connection with the eikonal quasinormal modes.
|
Original quantum repeater protocols based on single-photon interference
suffer from phase noise of the channel, which makes the long-distance quantum
communication infeasible. Fortunately, two-photon interference type quantum
repeaters can be immune to phase noise of the channel. However, this type
quantum repeaters may still suffer from polarization disturbance of the
channel. Here we propose a quantum repeaters protocol which is free of
polarization disturbance of the channel based on the invariance of the
anti-symmetric Bell state $|\psi^->=(|H>|V>-|V>|H>)/\sqrt{2}$ under collective
noise. Our protocol is also immune to phase noise with the Sagnac
interferometer configuration. Through single-atom cavity-QED technology and
linear optics, this scheme can be implemented easily.
|
High-quality K-band spectra of point sources, deeply embedded in massive
star-forming regions, have revealed a population of 20 young massive stars
showing no photospheric absorption lines, but only emission lines. The K-band
spectra exhibit one or more features commonly associated with massive Young
Stellar Objects surrounded by circumstellar material: a very red color (J-K) =
2, CO bandhead emission, hydrogen emission lines (sometimes doubly peaked), and
FeII and/or MgII emission lines. The CO emission comes from a relatively dense
(~10^10 cm^(-3)) and hot (T ~ 2000-5000 K) region, sufficiently shielded from
the intense UV radiation field of the young massive star. Modeling of the
CO-first overtone emission shows that the CO gas is located within 5 AU of the
star. The hydrogen emission is produced in an ionized medium exposed to UV
radiation. The best geometrical configuration is a dense and neutral
circumstellar disk causing the CO bandhead emission, and an ionized upper layer
where the hydrogen lines are produced. We argue that the circumstellar disk is
likely a remnant of the accretion via a circumstellar disk.
|
In this paper, we initiate a systematic study of graph resilience. The
(local) resilience of a graph G with respect to a property P measures how much
one has to change G (locally) in order to destroy P. Estimating the resilience
leads to many new and challenging problems. Here we focus on random and
pseudo-random graphs and prove several sharp results.
|
Developing robust models against adversarial perturbations has been an active
area of research and many algorithms have been proposed to train individual
robust models. Taking these pretrained robust models, we aim to study whether
it is possible to create an ensemble to further improve robustness. Several
previous attempts tackled this problem by ensembling the soft-label prediction
and have been proved vulnerable based on the latest attack methods. In this
paper, we show that if the robust training loss is diverse enough, a simple
hard-label based voting ensemble can boost the robust error over each
individual model. Furthermore, given a pool of robust models, we develop a
principled way to select which models to ensemble. Finally, to verify the
improved robustness, we conduct extensive experiments to study how to attack a
voting-based ensemble and develop several new white-box attacks. On CIFAR-10
dataset, by ensembling several state-of-the-art pre-trained defense models, our
method can achieve a 59.8% robust accuracy, outperforming all the existing
defensive models without using additional data.
|
We discuss a linear seesaw model with as minimum field content as possible,
introducing a modular $S_4$ with the help of gauged $U(1)_{B-L}$ symmetries.
Due to rank two neutrino mass matrix, we have a vanishing neutrino mass
eigenvalue, and only the normal mass hierarchy of neutrinos is favored through
the modular $S_4$ symmetry. In our numerical $\Delta \chi^2$ analysis, we
especially find rather sharp prediction on sum of neutrino masses to be around
$60$ meV in addition to the other predictions.
|
We present exact results for the electronic transport properties of graphene
sheets connected to two metallic electrodes. Our results, obtained by
transfer-matrix methods, are valid for all sheet widths and lengths. In the
limit of large width-to-length ratio relevant to recent experiments, we find a
Dirac-point conductivity of $2e^2/\sqrt{3}h$ and a sub-Poissonian Fano factor
of $2 - 3\sqrt{3}/\pi \simeq 0.346$ for armchair graphene; for the zigzag
geometry these are respectively 0 and 1. Our results reflect essential effects
from both the topology of graphene and the electronic structure of the leads,
giving a complete microscopic understanding of the unique intrinsic transport
in graphene.
|
The paper evaluates three variants of the Gated Recurrent Unit (GRU) in
recurrent neural networks (RNN) by reducing parameters in the update and reset
gates. We evaluate the three variant GRU models on MNIST and IMDB datasets and
show that these GRU-RNN variant models perform as well as the original GRU RNN
model while reducing the computational expense.
|
By using the infinite time-evolving block decimation, we study quantum
fidelity and entanglement entropy in the spin-1/2 Heisenberg alternating chain
under an external magnetic field. The effects of the magnetic field on the
fidelity are investigated, and its relation with the quantum phase transition
(QPT) is analyzed. The phase diagram of the model is given accordingly, which
supports the Haldane phase, the singlet-dimer phase, the Luttinger liquid phase
and the paramagnetic phase. The scaling of entanglement entropy in the gapless
Luttinger liquid phase is studied, and the central charge c = 1 is obtained. We
also study the relationship between the quantum coherence, string order
parameter and QPTs. Results obtained from these quantum information
observations are consistent with the previous reports.
|
Higher order quark effective interactions are found for SU(2) flavor by
departing from a non local quark-quark interaction. By integrating out a
component of the quark field, the determinant is expanded in chirally symmetric
and symmetry breaking effective interactions up to the fifh order in the quark
bilinears. The resulting coupling constants are resolved in the leading order
of the longwavelength limit and exact numerical ratios between several of these
coupling constants are obtained in the large quark mass limit. In this level,
chiral invariant interactions only show up in even powers of the quark
bilinears, i.e. ${\cal O}(\bar{\psi} \psi)^{2 n}$ ($n=1,2,3,..$), whereas
(explicit) chiral symmetry breaking terms emerge as ${\cal O}(\bar{\psi}
\psi)^{n}$ being always proportional to some power of the Lagrangian quark
mass.
|
In this work, we investigate the local density of states (LDOS) near a
magnetic impurity in single-layer FeSe superconductors. The two-orbital model
with spin-orbit coupling proposed in Ref. [{\emph{Phys. Rev. Lett.}
\textbf{119}, 267001 (2017)}] is used to describe the FeSe superconductor. In
the strong coupling regime, two impurity resonance peaks appear with opposite
resonance energies in the LDOS spectral function. For a strong spin-orbit
coupling, the superconducting gap in this model is $d$-wave symmetric with
nodes, the spatial distributions of the LDOS at the two resonance energies are
fourfold symmetric, which reveals typical characteristic of $d$-wave pairing.
When the spin-orbit coupling is not strong enough to close the superconducting
gap, we find that the spatial distribution of the LDOS at one of the resonance
energies manifests $s$-wave symmetry, while the pairing potential preserves
$d$-wave symmetry. This result is consistent with previous experimental
investigations.
|
This paper deals with the fault detection and isolation (FDI) problem for
linear structured systems in which the system matrices are given by
zero/nonzero/arbitrary pattern matrices. In this paper, we follow a geometric
approach to verify solvability of the FDI problem for such systems. To do so,
we first develop a necessary and sufficient condition under which the FDI
problem for a given particular linear time-invariant system is solvable. Next,
we establish a necessary condition for solvability of the FDI problem for
linear structured systems. In addition, we develop a sufficient algebraic
condition for solvability of the FDI problem in terms of a rank test on an
associated pattern matrix. To illustrate that this condition is not necessary,
we provide a counterexample in which the FDI problem is solvable while the
condition is not satisfied. Finally, we develop a graph-theoretic condition for
the full rank property of a given pattern matrix, which leads to a
graph-theoretic condition for solvability of the FDI problem.
|
Using new and published photometric observations of mu1 Sco (HR 6247),
spanning 70 years, a period of 1.4462700(5) days was determined. It was found
that the epoch of primary minimum suggested by Shobbrook at HJD 2449534.178
requires an adjustment to HJD 2449534.17700(9) to align all the available
photometric datasets. Using the resulting combined-data light-curve, radial
velocities derived from IUE data and the modelling software PHOEBE, a new
system solution for this binary was obtained. It appears that the secondary is
close to, or just filling, its Roche-lobe.
|
In this work, we consider the case where a source with bursty traffic can
adjust the transmission duration in order to increase the reliability. The
source is equipped with a queue in order to store the arriving packets. We
model the system with a discrete time Markov Chain, and we characterize the
performance in terms of service probability and average delay per packet. The
accuracy of the theoretical results is validated through simulations. This work
serves as an initial step in order to provide a framework for systems with
arbitrary arrivals and variable transmission durations and it can be utilized
for the derivation of the delay distribution and the delay violation
probability.
|
This paper presents a new robust data-driven predictive control scheme for
unknown linear time-invariant systems by using input-state-output or
input-output data based on whether the state is measurable. To remove the need
for the persistently exciting (PE) condition of a sufficiently high order on
pre-collected data, a set containing all systems capable of generating such
data is constructed. Then, at each time step, an upper bound of a given
objective function is derived for all systems in the set, and a feedback
controller is designed to minimize this bound. The optimal control gain at each
time step is determined by solving a set of linear matrix inequalities. We
prove that if the synthesis problem is feasible at the initial time step, it
remains feasible for all future time steps. Unlike current data-driven
predictive control schemes based on behavioral system theory, our approach
requires less stringent conditions for the pre-collected data, facilitating
easier implementation. Further, the proposed predictive control scheme features
an infinite prediction horizon, potentially resulting in superior overall
control performance compared to existing methods with finite prediction
horizons. The effectiveness of our proposed methods is demonstrated through
application to an unknown and unstable batch reactor.
|
We present luminosity functions derived from a spectroscopic survey of AGN
selected from Spitzer Space Telescope imaging surveys. Selection in the
mid-infrared is significantly less affected by dust obscuration. We can thus
compare the luminosity functions of the obscured and unobscured AGN in a more
reliable fashion than by using optical or X-ray data alone. We find that the
AGN luminosity function can be well described by a broken power-law model in
which the break luminosity decreases with redshift. At high redshifts
($z>1.6$), we find significantly more AGN at a given bolometric luminosity than
found by either optical quasar surveys or hard X-ray surveys. The fraction of
obscured AGN decreases rapidly with increasing AGN luminosity, but, at least at
high redshifts, appears to remain at $\approx 50$\% even at bolometric
luminosities $\sim 10^{14}L_{\odot}$. The data support a picture in which the
obscured and unobscured populations evolve differently, with some evidence that
high luminosity obscured quasars peak in space density at a higher redshift
than their unobscured counterparts. The amount of accretion energy in the
Universe estimated from this work suggests that AGN contribute about 12\% to
the total radiation intensity of the Universe, and a high radiative accretion
efficiency $\approx 0.18^{+0.12}_{-0.07}$ is required to match current
estimates of the local mass density in black holes.
|
We investigate open charm meson production in fixed-target LHCb experiment at
$\sqrt{s}=86.6$ GeV in $p+^4\!\mathrm{He}$ collisions. Theoretical calculations
of charm cross section are done in the framework of the $k_{T}$-factorization
approach. Its application in the kinematical range never examined before is
carefully discussed. We consider different schemes for the calculations
relevant for different unintegrated (transverse momentum dependent) parton
densities in a proton. We include in the analysis both CCFM- and DGLAP-based
models of unintegrated parton distributions appropriate for the considered
kinematics. Integrated as well as differential cross sections as a function of
$D^{0}$ meson rapidity and transverse momentum are shown and compared with the
experimental data. As a reference point, predictions of next-to-leading order
collinear approach are also presented and discussed. A very good agreement
between the experimental data and the $k_{T}$-factorization predictions was
obtained. Both the CCFM and the DGLAP-based frameworks for parton distributions
in a proton are successfully used to explain the LHCb fixed-target open charm
cross section.
|
We consider an original problem that arises from the issue of security
analysis of a power system and that we name optimal discovery with
probabilistic expert advice. We address it with an algorithm based on the
optimistic paradigm and on the Good-Turing missing mass estimator. We prove two
different regret bounds on the performance of this algorithm under weak
assumptions on the probabilistic experts. Under more restrictive hypotheses, we
also prove a macroscopic optimality result, comparing the algorithm both with
an oracle strategy and with uniform sampling. Finally, we provide numerical
experiments illustrating these theoretical findings.
|
Interior structure of non-Abelian black holes is shown to exhibit in a
general case either an oscillating mass-inflationary behavior, or power-law
behavior with a divergent mass function. In both cases no Cauchy horizon forms.
|
This literature review gives an overview of current approaches to perform
domain adaptation in a low-resource and approaches to perform multilingual
semantic search in a low-resource setting. We developed a new typology to
cluster domain adaptation approaches based on the part of dense textual
information retrieval systems, which they adapt, focusing on how to combine
them efficiently. We also explore the possibilities of combining multilingual
semantic search with domain adaptation approaches for dense retrievers in a
low-resource setting.
|
We present the first internal motion measurement of the 6.7-GHz methanol
maser within S269, a small HII region in the outer Galaxy, which was carried
out in 2006 and 2011 using the Japanese VLBI Network (JVN). Several maser
groups and weak isolated spots were detected in an area spanning by ~200 mas
(1000 AU). Three remarkable maser groups are aligned at a position angle of 80
degree. Two of three maser groups were also detected by a previous observation
in 1998, which allowed us to study a long-term position variation of maser
spots from 1998 to 2011. The angular separation between the two groups
increased ~10 mas, which corresponds to an expansion velocity of ~10 km s^{-1}.
Some velocity gradient (~10^{-2} km s^{-1} mas^{-1}) in the overall
distribution was found. The internal motion between the maser groups support
the hypothesis that the methanol masers in S269 could trace a bipolar outflow.
|
COVID-19 clinical trial design is a critical task in developing therapeutics
for the prevention and treatment of COVID-19. In this study, we apply a deep
learning approach to extract eligibility criteria variables from COVID-19
trials to enable quantitative analysis of trial design and optimization.
Specifically, we train attention-based bidirectional Long Short-Term Memory
(Att-BiLSTM) models and use the optimal model to extract entities (i.e.,
variables) from the eligibility criteria of COVID-19 trials. We compare the
performance of Att-BiLSTM with traditional ontology-based method. The result on
a benchmark dataset shows that Att-BiLSTM outperforms the ontology model.
Att-BiLSTM achieves a precision of 0.942, recall of 0.810, and F1 of 0.871,
while the ontology model only achieves a precision of 0.715, recall of 0.659,
and F1 of 0.686. Our analyses demonstrate that Att-BiLSTM is an effective
approach for characterizing patient populations in COVID-19 clinical trials.
|
The implicit bias towards solutions with favorable properties is believed to
be a key reason why neural networks trained by gradient-based optimization can
generalize well. While the implicit bias of gradient flow has been widely
studied for homogeneous neural networks (including ReLU and leaky ReLU
networks), the implicit bias of gradient descent is currently only understood
for smooth neural networks. Therefore, implicit bias in non-smooth neural
networks trained by gradient descent remains an open question. In this paper,
we aim to answer this question by studying the implicit bias of gradient
descent for training two-layer fully connected (leaky) ReLU neural networks. We
showed that when the training data are nearly-orthogonal, for leaky ReLU
activation function, gradient descent will find a network with a stable rank
that converges to $1$, whereas for ReLU activation function, gradient descent
will find a neural network with a stable rank that is upper bounded by a
constant. Additionally, we show that gradient descent will find a neural
network such that all the training data points have the same normalized margin
asymptotically. Experiments on both synthetic and real data backup our
theoretical findings.
|
We present new observations of the [Ne II] emission from the ionized gas in
Sgr A West with improved resolution and sensitivity. About half of the emission
comes from gas with kinematics indicating it is orbiting in a plane tipped
about 25\degree\ from the Galactic plane. This plane is consistent with that
derived previously for the circumnuclear molecular disk and the northern arm
and western arc ionized features. However, unlike most previous studies, we
conclude that the ionized gas is not moving along the ionized features, but on
more nearly circular paths. The observed speeds are close to, but probably
somewhat less than expected for orbital motions in the potential of the central
black hole and stars and have a small inward component. The spatial
distribution of the emission is well fitted by a spiral pattern. We discuss
possible physical explanations for the spatial distribution and kinematics of
the ionized gas, and conclude that both may be best explained by a one-armed
spiral density wave, which also accounts for both the observed low velocities
and the inward velocity component. We suggest that a density wave may result
from the precession of elliptical orbits in the potential of the black hole and
stellar mass distribution.
|
The paper is concerned with the geometrically non-linear theory of
6-parametric elastic shells with drilling degrees of freedom. This theory
establishes a general model for shells, which is characterized by two
independent kinematic fields: the translation vector and the rotation tensor.
Thus, the kinematical structure of 6-parameter shells is identical to that of
Cosserat shells. We show the existence of global minimizers for the
geometrically non-linear 2D equations of elastic shells. The proof of the
existence theorem is based on the direct methods of the calculus of variations
using essentially the convexity of the energy in the strain and curvature
measures. Since our result is valid for general anisotropic shells, we analyze
separately the particular cases of isotropic shells, orthotropic shells, and
composite shells.
|
Developing a novel experimental technique, we applied photon correlation
spectroscopy using infrared radiation in liquid Sulphur around $T_\lambda$,
i.e. in the temperature range where an abrupt increase in viscosity by four
orders of magnitude is observed upon heating within few degrees. This allowed
us - overcoming photo-induced and absorption effects at visible wavelengths -
to reveal a chain relaxation process with characteristic time in the ms range.
These results do rehabilitate the validity of the Maxwell relation in Sulphur
from an apparent failure, allowing rationalizing the mechanical and
thermodynamic behavior of this system within a viscoelastic scenario.
|
We explore the feasibility of using high conversion-efficiency
periodically-poled crystals to produce photon pairs for photon-counting
detector calibrations at 1550 nm. The goal is the development of an appropriate
parametric down-conversion (PDC) source at telecom wavelengths meeting the
requirements of high-efficiency pair production and collection in single
spectral and spatial modes (single-mode fibers). We propose a protocol to
optimize the photon collection, noise levels and the uncertainty evaluation.
This study ties together the results of our efforts to model the single-mode
heralding efficiency of a two-photon PDC source and to estimate the heralding
uncertainty of such a source.
|
The celebrated Erdos-Hajnal conjecture states that for every $n$-vertex
undirected graph $H$ there exists $\eps(H)>0$ such that every graph $G$ that
does not contain $H$ as an induced subgraph contains a clique or an independent
set of size at least $n^{\eps(H)}$. A weaker version of the conjecture states
that the polynomial-size clique/independent set phenomenon occurs if one
excludes both $H$ and its complement $H^{\compl}$. We show that the weaker
conjecture holds if $H$ is any path with a pendant edge at its third vertex;
thus we give a new infinite family of graphs for which the conjecture holds.
|
We calculate a two-pion wave function for the I=2 $S$-wave two-pion system
with a finite scattering momentum and estimate the interaction range between
two pions, which allows us to examine the validity of a necessary condition for
the finite size formula presented by Rummukainen and Gottlieb. We work in the
quenched approximation employing the plaquette gauge action for gluons and the
improved Wilson action for quarks at $1/a=1.63 {\rm GeV}$ on $32^3\times 120$
lattice. The quark masses are chosen to give $m_\pi = 0.420$, 0.488 and $0.587
{\rm GeV}$. We find that the energy dependence of the interaction range is
small and the necessary condition is satisfied for our range of the quark mass
and the scattering momentum, $k \le 0.16 {\rm GeV}$. We also find that the
scattering phase shift can be obtained with a smaller statistical error from
the two-pion wave function than from the two-pion time correlator.
|
We consider a first-order logic for the integers with addition. This logic
extends classical first-order logic by modulo-counting, threshold-counting and
exact-counting quantifiers, all applied to tuples of variables (here, residues
are given as terms while moduli and thresholds are given explicitly). Our main
result shows that satisfaction for this logic is decidable in two-fold
exponential space. If only threshold- and exact-counting quantifiers are
allowed, we prove an upper bound of alternating two-fold exponential time with
linearly many alternations. This latter result almost matches Berman's exact
complexity of first-order logic without counting quantifiers.
To obtain these results, we first translate threshold- and exact-counting
quantifiers into classical first-order logic in polynomial time (which already
proves the second result). To handle the remaining modulo-counting quantifiers
for tuples, we first reduce them in doubly exponential time to modulo-counting
quantifiers for single elements. For these quantifiers, we provide a quantifier
elimination procedure similar to Reddy and Loveland's procedure for first-order
logic and analyse the growth of coefficients, constants, and moduli appearing
in this process. The bounds obtained this way allow to restrict quantification
in the original formula to integers of bounded size which then implies the
first result mentioned above.
Our logic is incomparable with the logic considered by Chistikov et al. in
2022. They allow more general counting operations in quantifiers, but only
unary quantifiers. The move from unary to non-unary quantifiers is non-trivial,
since, e.g., the non-unary version of the H\"artig quantifier results in an
undecidable theory.
|
We reveal the existence of a new form of spontaneously scalarized black-hole
configurations. In particular, it is proved that Reissner-Nordstr\"om black
holes in the highly charged regime $Q/M>(Q/M)_{\text{crit}}=\sqrt{21}/5$ can
support {\it thin} matter shells that are made of massive scalar fields with a
non-minimal coupling to the Gauss-Bonnet invariant of the curved spacetime.
These static scalar shells, which become infinitesimally thin in the
dimensionless large-mass $M\mu\gg1$ regime, hover a finite proper distance
above the black-hole horizon [here $\{M,Q\}$ are respectively the mass and
electric charge of the central supporting black hole, and $\mu$ is the proper
mass of the supported scalar field]. In addition, we derive a remarkably
compact analytical formula for the discrete resonance spectrum
$\{\eta(Q/M,M\mu;n)\}_{n=0}^{n=\infty}$ of the non-trivial coupling parameter
which characterizes the bound-state
charged-black-hole-thin-massive-scalar-shell cloudy configurations of the
composed Einstein-Maxwell-scalar field theory.
|
Data produced by laboratory Einstein-Podolsky-Rosen-Bohm (EPRB) experiments
is tested against the hypothesis that the statistics of this data is given by
quantum theory of this thought experiment. Statistical evidence is presented
that the experimental data, while violating Bell inequalities, does not support
this hypothesis. It is shown that an event-based simulation model, providing a
cause-and-effect description of real EPRB experiments at a level of detail
which is not covered by quantum theory, reproduces the results of quantum
theory of this thought experiment, indicating that there is no fundamental
obstacle for a real EPRB experiment to produce data that can be described by
quantum theory.
|
Domain generalisation (DG) methods address the problem of domain shift, when
there is a mismatch between the distributions of training and target domains.
Data augmentation approaches have emerged as a promising alternative for DG.
However, data augmentation alone is not sufficient to achieve lower
generalisation errors. This project proposes a new method that combines data
augmentation and domain distance minimisation to address the problems
associated with data augmentation and provide a guarantee on the learning
performance, under an existing framework. Empirically, our method outperforms
baseline results on DG benchmarks.
|
Let $\lambda \in (0,1)$ and let $T$ be a $r\times r$ complex matrix with
polar decomposition $T=U|T|$. Then, the $\la$- Aluthge transform is defined by
$$ \Delta_\lambda (T )= |T|^{\lambda} U |T |^{1-\lambda}. $$ Let
$\Delta_\lambda^{n}(T)$ denote the n-times iterated Aluthge transform of $T$,
$n\in\mathbb{N}$. We prove that the sequence
$\{\Delta_\lambda^{n}(T)\}_{n\in\mathbb{N}}$ converges for every $r\times r$
{\bf diagonalizable} matrix $T$. We show regularity results for the two
parameter map $(\la, T) \mapsto \alulit{\infty}{T}$, and we study for which
matrices the map $(0,1)\ni \lambda \mapsto \Delta_\lambda^{\infty}(T)$ is
constant.
|
A new ternary intermetallic compound $\mathrm{Pr_2Rh_2Ga}$ was synthesized by
arc-melting and was characterized by powder X-ray diffraction (PXRD),
magnetization, heat capacity $\mathrm{C}_p(\textit{T})$, and electrical
resistivity $\rho(T)$ measurements. PXRD patterns revealed that
$\mathrm{Pr_2Rh_2Ga}$ crystallizes in the $\rm{La_2Ni_3}$-type of orthorhombic
structure with the space group $Cmca$. The temperature variation of magnetic
susceptibility, $\mathrm {C}_p(\textit{T})$ and $\rho(T)$ confirmed that
$\mathrm{Pr_2Rh_2Ga}$ exhibits a ferromagnetic behavior with the transition
temperature of 18 K. The estimated Sommerfeld coefficient $\gamma$ = 640
mJ/($\mathrm{Pr.mole.K^2}$) from the $\mathrm {C}_p(\textit{T})$ results in the
paramagnetic region just above $T_{C}$~was large in comparison to ordinary
metals. In the paramagnetic region $\rho(T)$ data showed a metallic behavior
characteristic of electron - phonon scattering. The maximum negative
magneto-resistance at high field occurs in the region near the magnetic phase
transition temperature. The maximum value of magnetic entropy change ($\rm
-\Delta \textit{S}_{M}$) and adiabatic temperature change ($\rm \Delta
\textit{T}_{ad}$) are $\rm8.2~J/kg.K$ and $\rm3.6~K$, respectively, around the
transition temperature for the change of magnetic field 0-9 T. The calculated
refrigerant capacity is $\rm70~J/kg$, and $\rm135~J/kg$ for a change of
magnetic field 0-5 T and 0-9 T, respectively. Arrott plot derived from
isothermal magnetization and the universal scaling plot by normalizing $\rm
-\Delta \textit{S}_{M}$ confirm that the compound undergoes a second order
ferromagnetic to paramagnetic phase transition.
|
We develop a generalized covering space theory for a class of uniform spaces
called coverable spaces. Coverable spaces include all geodesic metric spaces,
connected and locally pathwise connected compact topological spaces, in
particular Peano continua, as well as more pathological spaces like the
topologist's sine curve. Each coverable space has a generalized universal cover
with universal and lifting properties. Associated with this generalized
universal cover is a functorial uniform space invariant called the deck group,
which is related to the classical fundamental group by a natural homomorphism.
We obtain some specific results for one-dimensional spaces. Keywords: universal
cover, uniform space, geodesic space, fundamental group
|
Let $\mathfrak{M}$ be a semifinite von Neumann algebra on a Hilbert space
equipped with a faithful normal semifinite trace $\tau$. A closed densely
defined operator $x$ affiliated with $\mathfrak{M}$ is called $\tau$-measurable
if there exists a number $\lambda \geq 0$ such that $\tau
\left(e^{|x|}(\lambda,\infty)\right)<\infty$. A number of useful inequalities,
which are known for the trace on Hilbert space operators, are extended to trace
on $\tau$-measurable operators. In particular, these inequalities imply
Clarkson inequalities for $n$-tuples of $\tau$-measurable operators. A general
parallelogram law for $\tau$-measurable operators are given as well.
|
In this paper, we use the conjugate surface construction to prove the
existence of certain non-periodic symmetric immersed minimal surfaces. These
surfaces have finite total curvature and embedded catenoid ends, and they have
positive genus yet maintain the symmetry of their genus-zero counterparts
constructed by Jorge-Meeks and Xu.
|
We review and discuss recent results on the search for correlations between
astrophysical neutrinos and gamma-ray-detected sources, with many
extra-galactic studies reporting potential associations with different types of
blazars. We investigate possible dependencies on blazar sub-classes by using
the largest catalogues and all the multi-frequency data available. Through the
study of similarities and differences in these sources we conclude that blazars
come in two distinct flavors: LBLs and IHBLs (low-energy-peaked and
intermediate-high-energy-peaked objects). These are distinguished by widely
different properties such as the overall spectral energy distribution shape,
jet speed, cosmological evolution, broad-band spectral variability, and optical
polarization properties. Although blazars of all types have been proposed as
neutrino sources, evidence is accumulating in favor of IHBLs being the
counterparts of astrophysical neutrinos. If this is indeed the case, we argue
that the peculiar observational properties of IHBLs may be indirectly related
to proton acceleration to very high energies.
|
We present X-ray light curves and energy spectra for the persistent accreting
pulsars 4U 1626-67 and GX 301-2 measured by the All-Sky Monitor (ASM) on Ginga
from 1987 March - 1991 October. We compare these with simultaneous and near
simultaneous measurements of spin frequency and flux by other instruments,
principally the Burst and Transient Source Experiment (BATSE) on the Compton
Gamma Ray Observatory (CGRO). A dramatic change in the shape of the X-ray
spectrum and a 20% decrease in the 1-20 keV X-ray flux accompany the 1990
transition from steady spin up to steady spin down in 4U 1626-67. The Ginga ASM
is the only instrument to observe 4U 1626-67 during both spin up and spin down.
We show that the distance to 4U 1626-67 is greater than 5 kpc. If 4U 1626-67 is
a near-equilibrium rotator and if the 0.04 Hz Quasi-Period Oscillations seen
during spin up are magnetospheric beat-frequency oscillations, then the
distance to the source is 5 kpc, assuming a neutron-star mass of 1.4 solar
masses, radius 10 km, and moment of inertia 10^45 g cm^2. The X-ray flux of GX
301-2 measured with the ASM varies with orbital phase. The flux peaks shortly
before periastron, with a secondary maximum near apastron. Such variations were
seen previously in the 20-50 keV pulsed flux with BATSE. The ASM observations
confirm that the 20-50 keV pulsed flux in GX 301-2 is a good tracer of the
bolometric flux. The X-ray flux in GX 301-2 was a factor of 2 larger than
average during the periastron passage prior to an episode of persistent spin up
in 1991 July observed with BATSE that lasted half an orbit and resembled
outbursts seen in transient X-ray pulsars.
|
We provide a framework for part of the homological theory of Z-algebras and
their generalizations, directed towards analogues of the Auslander-Gorenstein
condition and the associated double Ext spectral sequence that are useful for
enveloping algebras of Lie algebras and related rings. As an application, we
prove the equidimensionality of the characteristic variety of an irreducible
representation of the Z-algebra, and for related representations over quantum
symplectic resolutions. In the special case of Cherednik algebras of type A,
this answers a question raised by the authors.
|
Using a result of Fujita on approximate Zariski decompositions and the
singular version of Demailly's holomorphic Morse inequalities as obtained by
Bonavero, we express the volume of a line bundle in terms of the absolutely
continuous parts of all the positive curvature currents on it, with a way to
pick an element among them which is most homogeneous with respect to the
volume. This enables us to introduce the volume of any pseudoeffective class on
a compact Kaehler manifold, and Fujita's theorem is then extended to this
context.
|
In this paper, we propose an effective method for fast and accurate scene
parsing called Bidirectional Alignment Network (BiAlignNet). Previously, one
representative work BiSeNet~\cite{bisenet} uses two different paths (Context
Path and Spatial Path) to achieve balanced learning of semantics and details,
respectively. However, the relationship between the two paths is not well
explored. We argue that both paths can benefit each other in a complementary
way. Motivated by this, we propose a novel network by aligning two-path
information into each other through a learned flow field. To avoid the noise
and semantic gaps, we introduce a Gated Flow Alignment Module to align both
features in a bidirectional way. Moreover, to make the Spatial Path learn more
detailed information, we present an edge-guided hard pixel mining loss to
supervise the aligned learning process. Our method achieves 80.1\% and 78.5\%
mIoU in validation and test set of Cityscapes while running at 30 FPS with full
resolution inputs. Code and models will be available at
\url{https://github.com/jojacola/BiAlignNet}.
|
We derive exact expressions for luminosities of massive vector-boson pairs
which can be used to describe the cross sections for processes in hadron
collisions or $e^+e^-$ annihilation which proceed via two-vector-boson
scattering. Our approach correctly takes into account the mutual influence of
the emission of one vector boson on the emission of a second one. We show that
only approximately the exact luminosities can be factorized into convolutions
of single-vector-boson distributions. Numerical results are given and compared
to simplified approaches.
|
In this article, we study the Schur mutiplier of the discrete as well as the
finite Heisenberg groups and their t-variants. We describe the representation
groups of these Heisenberg groups and through these give a construction of
their finite dimensional complex projective irreducible representations.
|
A gyrogroup is a nonassociative group-like structure modelled on the space of
relativistically admissible velocities with a binary operation given by
Einstein's velocity addition law. In this article, we present a few of groups
sitting inside a gyrogroup $G$, including the commutator subgyrogroup, the left
nucleus, and the radical of $G$. The normal closure of the commutator
subgyrogroup, the left nucleus, and the radical of $G$ are in particular normal
subgroups of $G$. We then give a criterion to determine when a subgyrogroup $H$
of a finite gyrogroup $G$, where the index $[G\colon H]$ is the smallest prime
dividing $|G|$, is normal in $G$.
|
Many business web-based applications do not offer applications programming
interfaces (APIs) to enable other applications to access their data and
functions in a programmatic manner. This makes their composition difficult (for
instance to synchronize data between two applications). To address this
challenge, this paper presents Abmash, an approach to facilitate the
integration of such legacy web applications by automatically imitating human
interactions with them. By automatically interacting with the graphical user
interface (GUI) of web applications, the system supports all forms of
integrations including bi-directional interactions and is able to interact with
AJAX-based applications. Furthermore, the integration programs are easy to
write since they deal with end-user, visual user-interface elements. The
integration code is simple enough to be called a "mashup".
|
We present a general framework for classifying partially observed dynamical
systems based on the idea of learning in the model space. In contrast to the
existing approaches using model point estimates to represent individual data
items, we employ posterior distributions over models, thus taking into account
in a principled manner the uncertainty due to both the generative
(observational and/or dynamic noise) and observation (sampling in time)
processes. We evaluate the framework on two testbeds - a biological pathway
model and a stochastic double-well system. Crucially, we show that the
classifier performance is not impaired when the model class used for inferring
posterior distributions is much more simple than the observation-generating
model class, provided the reduced complexity inferential model class captures
the essential characteristics needed for the given classification task.
|
Placing and twisting graphene on transition metal dichalcogenides (TMDC)
forms a van der Waals (vdW) heterostructure. The occurrence of Zeeman splitting
and Rashba spin-orbit coupling (SOC) changes graphene's linear dispersion and
conductivity. Hence, this paper studies the dependence of graphene's
longitudinal optical conductivity on Rashba SOC, the twist-angle and
temperature. At zero temperature, a main conductivity peak exists. When Rashba
SOC increases, a second peak occurs, with both extremes presenting an enhanced
height and width, and the frequencies where the two peaks arise will increase
because the energy gap and the possibility of electron transition increase.
Altering the twist-angle from 0 to 30$^{\circ}$, the conductivity is primarily
affected by chalcogen atoms. Moreover, when temperature increases to room
temperature, besides a Drude peak due to the thermal excitation, a new band
arises in the conductivity owing to the joint effect of the thermal transition
and the photon transition.
|
We propose a simple scheme capable of adiabatically splitting an atomic wave
packet using two independent translating traps. Implemented with optical dipole
traps, our scheme allows a high degree of flexibility for atom interferometry
arrangements and highlights its potential as an efficient and high fidelity
atom optical beam splitter.
|
The 4.9 GHz Micro-Arcsecond Scintillation-Induced Variability (MASIV) Survey
detected a drop in Interstellar Scintillation (ISS) for sources at redshifts z
> 2, indicating an apparent increase in angular diameter or a decrease in flux
density of the most compact components of these sources, relative to their
extended emission. This can result from intrinsic source size effects or
scatter broadening in the Intergalactic Medium (IGM), in excess of the expected
(1+z)^0.5 angular diameter scaling of brightness temperature limited sources
due to cosmological expansion. We report here 4.9 GHz and 8.4 GHz observations
and data analysis for a sample of 140 compact, flat-spectrum sources which may
allow us to determine the origin of this angular diameter-redshift relation by
exploiting their different wavelength dependences. In addition to using ISS as
a cosmological probe, the observations provide additional insight into source
morphologies and the characteristics of ISS. As in the MASIV Survey, the
variability of the sources is found to be significantly correlated with
line-of-sight H-alpha intensities, confirming its link with ISS. For 25
sources, time delays of about 0.15 to 3 days are observed between the
scintillation patterns at both frequencies, interpreted as being caused by a
shift in core positions when probed at different optical depths. Significant
correlation is found between ISS amplitudes and source spectral index; in
particular, a large drop in ISS amplitudes is observed at spectral indices of <
-0.4 confirming that steep spectrum sources scintillate less. We detect a
weakened redshift dependence of ISS at 8.4 GHz over that at 4.9 GHz, with the
mean variance at 4-day timescales reduced by a factor of 1.8 in the z > 2
sources relative to the z < 2 sources, as opposed to the factor of 3 decrease
observed at 4.9 GHz. This suggests scatter broadening in the IGM.
|
Large Vision-Language Models (LVLMs) can understand the world comprehensively
by integrating rich information from different modalities, achieving remarkable
advancements on various multimodal downstream tasks. However, deploying LVLMs
is often problematic due to their massive computational/energy costs and carbon
consumption. Such issues make it infeasible to adopt conventional iterative
global pruning, which is costly due to computing the Hessian matrix of the
entire large model for sparsification. Alternatively, several studies have
recently proposed layer-wise pruning approaches to avoid the expensive
computation of global pruning and efficiently compress model weights according
to their importance within a layer. However, they often suffer from suboptimal
model compression due to their lack of a global perspective. To address this
limitation in recent efficient pruning methods for large models, we propose
Efficient Coarse-to-Fine LayerWise Pruning (ECoFLaP), a two-stage
coarse-to-fine weight pruning approach for LVLMs. We first determine the
sparsity ratios of different layers or blocks by leveraging the global
importance score, which is efficiently computed based on the zeroth-order
approximation of the global model gradients. Then, the model performs local
layer-wise unstructured weight pruning based on globally-informed sparsity
ratios. We validate our proposed method across various multimodal and unimodal
models and datasets, demonstrating significant performance improvements over
prevalent pruning techniques in the high-sparsity regime.
|
I show that the so-called causality paradox of time-dependent density
functional theory arises from an incorrect formulation of the variational
principle for the time evolution of the density. The correct formulation not
only resolves the paradox in real time, but also leads to a new expression for
the causal exchange-correlation kernel in terms of Berry curvature.
Furthermore, I show that all the results that were previously derived from
symmetries of the action functional remain valid in the present formulation.
Finally, I develop a model functional theory which explicitly demonstrates the
workings of the new formulation.
|
In this paper, we investigate thick branes with a nonminimally coupled
background scalar field, whose solution is a single-kink or a double-kink. The
effects of the nonminimal coupling constant $\xi$ on the structure of the thick
branes and the localization of gravity, fermions, scalars and vectors are
discussed. It is shown that each brane will split into two sub-branes as
increasing the nonminimal coupling constant $\xi$. By investigating the tensor
perturbation equations of gravity and the general covariant Dirac equation of
fermions, we find that both the gravity zero mode and left-chiral fermion zero
mode are localized at the center of the single-kink branes and localized
between the two sub-branes generated by the double-kink, which indicates that
the constant $\xi$ does not effect the localization of these zero modes.
However, the zero mode of scalars is localized on each sub-brane (for both
single-kink and double-kink branes) when $\xi$ is larger than its critical
value $\xi_0$. The effects of the nonminimal coupling constant $\xi$ on the
resonances of gravity and fermions with finite lifetime on the branes are also
discussed.
|
We present a tight-binding potential for transition metals, carbon, and
transition metal carbides, which has been optimized through a systematic
fitting procedure. A minimal basis, including the s, p electrons of carbon and
the d electrons of the transition metal, is used to obtain a transferable
tight-binding model of the carbon-carbon, metal-metal and metal-carbon
interactions applicable to binary systems. The Ni-C system is more specifically
discussed. The successful validation of the potential for different atomic
configurations indicates a good transferability of the model and makes it a
good choice for atomistic simulations sampling a large configuration space.
This approach appears to be very efficient to describe interactions in systems
containing carbon and transition metal elements.
|
In this paper we give a negative answer to the question posed in [15, Open
Question 2.1] about possible gains of integrability of determinants of
divergence-free, non-negative definite matrix-fields. We also analyze the case
in which the matrix-field is given by the Hessian of a convex function.
|
Subsets and Splits