id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2502.21036 | Qi Dai | Qi Dai, Beixiong Zheng, Qiyao Wang, Xue Xiong, Xiaodan Shao, Lipeng
Zhu, Rui Zhang | A Demo of Radar Sensing Aided Rotatable Antenna for Wireless
Communication System | null | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Rotatable antenna (RA) represents a novel antenna architecture that enhances
wireless communication system performance by independently or collectively
adjusting each antenna's boresight/orientation. In this demonstration, we
develop a prototype of radar sensing-aided rotatable antenna that integrates
radar sensing with dynamic antenna orientation to enhance wireless
communication performance while maintaining low hardware costs. The proposed
prototype consists of a transmitter (TX) module and a receiver (RX) module,
both of which employ universal software radio peripherals (USRPs) for
transmitting and receiving signals. Specifically, the TX utilizes a laser radar
to detect the RX's location and conveys the angle of arrival (AoA) information
to its antenna servo, which enables the RA to align its boresight direction
with the identified RX. Experimental results examine the effectiveness of the
proposed prototype and indicate that the RA significantly outperforms the
traditional fixed-antenna system in terms of increasing received
signal-to-noise ratio (SNR).
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 13:28:29 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 08:24:06 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Dai",
"Qi",
""
],
[
"Zheng",
"Beixiong",
""
],
[
"Wang",
"Qiyao",
""
],
[
"Xiong",
"Xue",
""
],
[
"Shao",
"Xiaodan",
""
],
[
"Zhu",
"Lipeng",
""
],
[
"Zhang",
"Rui",
""
]
] |
2503.00170 | Roi Bar-Zur | Roi Bar-Zur and Ittay Eyal | Elastic Restaking Networks | null | null | null | null | cs.GT cs.DC | http://creativecommons.org/licenses/by/4.0/ | Many blockchain-based decentralized services require their validators
(operators) to deposit stake (collateral), which is forfeited (slashed) if they
misbehave. Restaking networks let validators secure multiple services by
reusing stake. These networks have quickly gained traction, leveraging
over~\$20 billion in stake. However, restaking introduces a new attack vector
where validators can coordinate to misbehave across multiple services
simultaneously, extracting digital assets while forfeiting their stake only
once.
Previous work focused either on preventing coordinated misbehavior or on
protecting services if all other services are Byzantine and might unjustly
cause slashing due to bugs or malice. The first model overlooks how a single
Byzantine service can collapse the network, while the second ignores
shared-stake benefits.
To bridge the gap, we analyze the system as a strategic game of coordinated
misbehavior, when a given fraction of the services are Byzantine. We introduce
elastic restaking networks, where validators can allocate portions of their
stake that may cumulatively exceed their total stake, and when allocations are
lost, the remaining stake stretches to cover remaining allocations. We show
that elastic networks exhibit superior robustness compared to previous
approaches, and demonstrate a synergistic effect where an elastic restaking
network enhances its blockchain's security, contrary to community concerns of
an opposite effect in existing networks. We then design incentives for tuning
validators' allocations.
Our elastic restaking system and incentive design have immediate practical
implications for deployed restaking networks.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 20:33:17 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Mar 2025 10:15:29 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 18:54:25 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Bar-Zur",
"Roi",
""
],
[
"Eyal",
"Ittay",
""
]
] |
2503.00595 | Alexander Smith Clark | Alexander Smith Clark, Michael Zingale | Multidimensional Nova Simulations with an Extended Buffer and Lower
Initial Mixing Temperatures | 11 pages, 10 figures | The Open Journal of Astrophysics, volume 8 (2025) | 10.33232/001c.136890 | null | astro-ph.SR astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | A classical nova is a thermonuclear runaway initiated on a white dwarf
accreting solar-like material from its stellar companion. Once the white dwarf
accretes enough mass, the pressure at the base of the accreted layer reaches a
critical point, leading to the ignition of the hydrogen fuel at their
interface. This paper presents a set of two-dimensional CO classical nova
simulations with an extended buffer zone of a fixed low density and temperature
between the top of the accreted layer and the upper boundary, allowing us to
capture the thermonuclear outburst in the domain. Our domain reduces the role
of the upper-outflow boundary condition that has affected previous simulations
and allows us to explore the nucleosynthesis evolution in detail. We also study
the effects of the initial temperature perturbation and buffer size to explore
their sensitivity in our simulations. Finally, we start our simulations with a
lower temperature at the base of the accreted layer ($7\times
10^7\,\mathrm{K}$) than previous work, allowing us to capture mixing earlier in
the evolution, reducing the effects of the mixing-length-theory assumptions.
This allows for a more realistic description of convective transport in our
models.
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 19:33:10 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:11:45 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Clark",
"Alexander Smith",
""
],
[
"Zingale",
"Michael",
""
]
] |
2503.02323 | YuSen An | Hai-Yang Zhang, Ya-Peng Hu, Yu-Sen An | Curled orbit and epicyclic oscillation of charged particles around the
weakly magnetized black hole in the presence of Lorentz violation | v2: 15 pages, 6 figures,2 tables, added a section using MCMC
algorithm to give the tighter constraint on Lorentz violation parameter,
reference added | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we investigate the motion of charged particles around the
weakly magnetized Schwarzschild-like bumblebee black hole which has Lorentz
symmetry breaking. Charged particles have curled orbits around the black hole
which can only appear in the presence of external magnetic field. We
investigate the effect of Lorentz violation factor on the curled orbit for both
the case with and without cosmological constant. Furthermore, we investigate
the harmonic oscillation behavior of the charged particles around the stable
circular orbit. By using the epicyclic resonance model, we relate the harmonic
oscillations of charged particles to the twin high frequency quasi-periodic
oscillations observed in micro-quasars. Based on the observations of
quasi-periodic oscillation, we provide a stringent constraint on the Lorentz
violating parameters by using Markov Chain Monte Carlo algorithm. As the black
hole shadow for Schwarzschild-like bumblebee black hole degenerates to the
ordinary Schwarzschild black hole, the constraints we obtained from the
quasi-period oscillation is crucial for further searching for the imprint of
Lorentz symmetry breaking in our universe.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 06:29:37 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 06:35:17 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zhang",
"Hai-Yang",
""
],
[
"Hu",
"Ya-Peng",
""
],
[
"An",
"Yu-Sen",
""
]
] |
2503.04515 | Vasileios A. Letsios | Atsushi Higuchi and Vasileios A. Letsios | Unitary Rigid Supersymmetry for the Chiral Graviton and Chiral Gravitino
in de Sitter Spacetime | 93 pages - main text is 76 pages long. No figures; v2: minor typos
corrected, references added | null | null | null | hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is commonly believed that a unitary supersymmetric quantum field theory
(QFT) involving graviton and gravitino fields on fixed 4-dimensional de Sitter
spacetime ($dS_{4}$) cannot exist due to known challenges associated with
supersymmetry (SUSY) in $dS_{4}$. In this paper, we contradict this expectation
by presenting a new unitary supersymmetric QFT on $dS_{4}$: the free
supersymmetric theory of the chiral graviton and chiral gravitino fields. By
chiral, we mean that the corresponding field strengths are anti-self-dual, and
the gauge potentials are complex, each carrying a single complex propagating
degree of freedom. The global SUSY transformations are generated by the
standard Dirac Killing spinors of $dS_{4}$. The theory overcomes the known
obstacles to unitary global SUSY on $dS_{4}$ by closing the commutator between
two SUSY transformations on $so(4,2) \oplus u(1)$ rather than the de Sitter
algebra $so(4,1)$. Crucially, the $so(4,2)$ symmetry is realised through
unconventional conformal-like transformations. This free theory cannot become
interacting while preserving SUSY in a way that makes the spin-2 sector the
true graviton sector of General Relativity, as the three-graviton coupling
cannot be $u(1)$-invariant. We establish the unitarity of the free
supersymmetric theory in two complementary ways. First, by studying the action
of the superalgebra generators on the space of physical gravitino and graviton
mode solutions. Second, by quantising the fields and explicitly constructing
the complex quantum supercharges $Q_{A}$ and $Q^{A\dagger}$, we show that the
trace $\sum_{A} \{ Q_{A}, Q^{A \dagger} \}$ is positive-definite.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 15:01:13 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 19:59:00 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Higuchi",
"Atsushi",
""
],
[
"Letsios",
"Vasileios A.",
""
]
] |
2503.04546 | Jiahua Tian | Qiang Jia, Ran Luo, Jiahua Tian, Yi-Nan Wang and Yi Zhang | Symmetry Topological Field Theory for Flavor Symmetry | null | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | In this Letter, we demonstrate that the Symmetry Topological Field Theory
(SymTFT) associated to a Quantum Field Theory (QFT) with continuous non-abelian
$G$-flavor symmetry is a $BF$-theory with gauge group $G$. We show that gauging
$G$-symmetry with a flat connection yields a theory with global symmetry
characterized by exchanging the conjugate variables in the quantization of
$BF$-theory. We construct the extended operators that generate the $G$-flavor
symmetry and the $(d-2)$-form $\text{Rep}(G)$-symmetry of the gauged QFT. We
demonstrate that $BF$-theory arises as the theory characterizing $G$-flavor
symmetry of a QFT in the AdS/CFT setup. 't Hooft anomalies of the $G$-flavor
symmetry are realized as extra terms in the action.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 15:33:30 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 01:14:39 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Jia",
"Qiang",
""
],
[
"Luo",
"Ran",
""
],
[
"Tian",
"Jiahua",
""
],
[
"Wang",
"Yi-Nan",
""
],
[
"Zhang",
"Yi",
""
]
] |
2503.04649 | Paul Atzberger | Blaine Quackenbush, Paul J. Atzberger | Transferable Foundation Models for Geometric Tasks on Point Cloud
Representations: Geometric Neural Operators | null | null | null | null | cs.LG cs.CV cs.NA math.NA math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce methods for obtaining pretrained Geometric Neural Operators
(GNPs) that can serve as basal foundation models for use in obtaining geometric
features. These can be used within data processing pipelines for machine
learning tasks and numerical methods. We show how our GNPs can be trained to
learn robust latent representations for the differential geometry of
point-clouds to provide estimates of metric, curvature, and other shape-related
features. We demonstrate how our pre-trained GNPs can be used (i) to estimate
the geometric properties of surfaces of arbitrary shape and topologies with
robustness in the presence of noise, (ii) to approximate solutions of geometric
partial differential equations (PDEs) on manifolds, and (iii) to solve
equations for shape deformations such as curvature driven flows. We release
codes and weights for using GNPs in the package geo_neural_op. This allows for
incorporating our pre-trained GNPs as components for reuse within existing and
new data processing pipelines. The GNPs also can be used as part of numerical
solvers involving geometry or as part of methods for performing inference and
other geometric tasks.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 17:35:37 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 03:47:25 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Quackenbush",
"Blaine",
""
],
[
"Atzberger",
"Paul J.",
""
]
] |
2503.04950 | Nikita Borisov | Nikita Borisov | Monomial stability of Frobenius images | 28 pages, 6 figures | null | null | null | math.CO math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study representation stability in the sense of Church, Ellenberg, and Farb
\cite{FI-module} through the lens of symmetric function theory and the
different symmetric function bases. We show that a sequence, $(F_n)_n$, where
$F_n$ is a homogeneous symmetric function of degree $n$, has stabilizing Schur
coefficients if and only if it has stabilizing monomial coefficients. More
generally, we develop a framework for checking when stabilizing coefficients
transfer from one symmetric function basis to another. We also see how one may
compute representation stable ranges from the monomial expansions of the
$F_n$.\parspace
As applications, we reprove and refine the representation stability of
diagonal coinvariant algebras, $DR_n$. We also observe new representation
stability phenomena of the Garsia-Haiman modules. This establishes certain
stability properties of the modified Macdonald polynomials,
$\tilde{H}_{\mu[n]}[X;q,t]$, and the modified $q,t$-Kostka numbers,
$\tilde{K}_{\mu[n],\nu[n]}(q,t)$. In an upcoming addition to the paper, these
methods will be be applied to \textit{any} sequence
$\tilde{H}_{\mu^{(n)}}[X;q,t]$ with $|\mu^{(n)}|=n$ and $\mu^{(n)}\subseteq
\mu^{(n+1)}$.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 20:27:36 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:58:12 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Borisov",
"Nikita",
""
]
] |
2503.06032 | Franciole Marinho | F. Marinho | APEX: Optimized vertical drift PDS for DUNE FD3 | null | null | null | null | hep-ex physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Deep Underground Neutrino Experiment currently under construction in the
US will be a long-baseline neutrino oscillation experiment dedicated to
determining the neutrino mass ordering and to measure the CP violation phase in
the lepton sector. DUNE will also perform studies of non-beam physics such as
atmospheric neutrinos, bursts from supernovae and nucleon decays in which
photon detection systems will play a major role in triggering and also provide
calorimetric measurements. For the second phase of DUNE, two additional
detector modules will be added in the far detector complex in the Sanford
Underground Research Facility. We present the Aluminum Profiles with Embedded
X-ARAPUCA (APEX) concept as an advanced proposal for the photon detector system
of the third DUNE far detector module. This system aims to have an optical
coverage of approximately 60% made viable by the technology advancement
achieved by the DUNE collaboration on the use of non-conductive optical fibers
for power and signal readout of the photon detector units. Such large coverage
will provide enhanced light collection capabilities at MeV-scale energy deposit
level per interaction and optimal energy reconstruction resolution up to the
GeV scale. The attained electrical isolation of the detector units with low
noise levels allows for a complete instrumentation of the field cage walls with
satisfactory segmentation as the readout scheme envisages a much larger than
typical number of channels to be adopted. We discuss the main features of the
system, first estimates on its expected performances, potential for physics
measurements and prototyping plans for R\&D.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 03:06:05 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 17:08:45 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 10:46:04 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Marinho",
"F.",
""
]
] |
2503.06125 | Xiaohan Shi | Kai Yang, Zijian Bai, Yang Xiao, Xinyu Li, Xiaohan Shi | RGB-Phase Speckle: Cross-Scene Stereo 3D Reconstruction via Wrapped
Pre-Normalization | Submitted to ICCV 2025 | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | 3D reconstruction garners increasing attention alongside the advancement of
high-level image applications, where dense stereo matching (DSM) serves as a
pivotal technique. Previous studies often rely on publicly available datasets
for training, focusing on modifying network architectures or incorporating
specialized modules to extract domain-invariant features and thus improve model
robustness. In contrast, inspired by single-frame structured-light
phase-shifting encoding, this study introduces RGB-Speckle, a cross-scene 3D
reconstruction framework based on an active stereo camera system, designed to
enhance robustness. Specifically, we propose a novel phase pre-normalization
encoding-decoding method: first, we randomly perturb phase-shift maps and embed
them into the three RGB channels to generate color speckle patterns;
subsequently, the camera captures phase-encoded images modulated by objects as
input to a stereo matching network. This technique effectively mitigates
external interference and ensures consistent input data for RGB-Speckle,
thereby bolstering cross-domain 3D reconstruction stability. To validate the
proposed method, we conduct complex experiments: (1) construct a color speckle
dataset for complex scenarios based on the proposed encoding scheme; (2)
evaluate the impact of the phase pre-normalization encoding-decoding technique
on 3D reconstruction accuracy; and (3) further investigate its robustness
across diverse conditions. Experimental results demonstrate that the proposed
RGB-Speckle model offers significant advantages in cross-domain and cross-scene
3D reconstruction tasks, enhancing model generalization and reinforcing
robustness in challenging environments, thus providing a novel solution for
robust 3D reconstruction research.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 08:37:20 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 12:57:37 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yang",
"Kai",
""
],
[
"Bai",
"Zijian",
""
],
[
"Xiao",
"Yang",
""
],
[
"Li",
"Xinyu",
""
],
[
"Shi",
"Xiaohan",
""
]
] |
2503.06341 | Elijah Pelofske | Elijah Pelofske, Vincent Russo | Digital Zero-Noise Extrapolation with Quantum Circuit Unoptimization | null | null | null | null | quant-ph cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum circuit unoptimization is an algorithm that transforms a quantum
circuit into a different circuit that uses more gate operations while
maintaining the same unitary transformation. We demonstrate that this method
can implement digital zero-noise extrapolation (ZNE), a quantum error
mitigation technique. By employing quantum circuit unoptimization as a form of
circuit folding, noise can be systematically amplified. The key advantages of
this approach are twofold. First, its ability to generate an exponentially
increasing number of distinct circuit variants as the noise level is amplified,
which allows noise averaging over many circuit instances with slightly
different circuit structure which mitigates the effect of biased error
propagation because of the significantly altered circuit structure from quantum
circuit unoptimization, or highly biased local noise on a quantum processor.
Second, quantum circuit unoptimization by design resists circuit simplification
back to the original unmodified circuit, making it plausible to use ZNE in
contexts where circuit compiler optimization is applied server-side. We
evaluate the effectiveness of quantum circuit unoptimization as a noise-scaling
method for ZNE in two test cases using depolarizing noise numerical
simulations: random quantum volume circuits, where the observable is the heavy
output probability, and QAOA circuits for the (unweighted) maximum cut problem
on random 3-regular graphs, where the observable is the cut value. We show that
using quantum circuit unoptimization to perform ZNE can approximately recover
signal from noisy quantum simulations.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 21:06:48 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 18:53:44 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Pelofske",
"Elijah",
""
],
[
"Russo",
"Vincent",
""
]
] |
2503.07374 | Sjoerd Dirksen | Simon Hakvoort, Bastien Francois, Kirien Whan, Sjoerd Dirksen | Improving Statistical Postprocessing for Extreme Wind Speeds using Tuned
Weighted Scoring Rules | Typos corrected | null | null | null | stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent statistical postprocessing methods for wind speed forecasts have
incorporated linear models and neural networks to produce more skillful
probabilistic forecasts in the low-to-medium wind speed range. At the same
time, these methods struggle in the high-to-extreme wind speed range. In this
work, we aim to increase the performance in this range by training using a
weighted version of the continuous ranked probability score (wCRPS). We develop
an approach using shifted Gaussian cdf weight functions, whose parameters are
tuned using a multi-objective hyperparameter tuning algorithm that balances
performance on low and high wind speed ranges. We explore this approach for
both linear models and convolutional neural network models combined with
various parametric distributions, namely the truncated normal, log-normal, and
generalized extreme value distributions, as well as adaptive mixtures. We apply
these methods to forecasts from KNMI's deterministic Harmonie-Arome numerical
weather prediction model to obtain probabilistic wind speed forecasts in the
Netherlands for 48 hours ahead. For linear models we observe that even with a
tuned weight function, training using the wCRPS produces a strong body-tail
trade-off, where increased performance on extremes comes at the price of lower
performance for the bulk of the distribution. For the best models using
convolutional neural networks, we find that using a tuned weight function the
performance on extremes can be increased without a significant deterioration in
performance on the bulk. The best-performing weight function is shown to be
model-specific. Finally, the choice of distribution has no significant impact
on the performance of our models.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 14:30:51 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:11:56 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Hakvoort",
"Simon",
""
],
[
"Francois",
"Bastien",
""
],
[
"Whan",
"Kirien",
""
],
[
"Dirksen",
"Sjoerd",
""
]
] |
2503.09093 | Sifan Yu | Sifan Yu, Feng He, Anlan Xie, Luxi Zhao | Efficient Adaptive Bandwidth Allocation for Deadline-Aware Online
Admission Control in Time-Sensitive Networking | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the growing demand for dynamic real-time applications, online admission
control for time-critical event-triggered (ET) traffic in Time-Sensitive
Networking (TSN) has become a critical challenge. The main issue lies in
dynamically allocating bandwidth with real-time guarantees in response to
traffic changes while also meeting the requirements for rapid response,
scalability, and high resource utilization in online scenarios. To address this
challenge, we propose an online admission control method for ET traffic based
on the TSN/ATS+CBS (asynchronous traffic shaper and credit-based shaper)
architecture. This method provides a flexible framework for real-time
guaranteed online admission control, supporting dynamic bandwidth allocation
and reclamation at runtime without requiring global reconfiguration, thus
improving scalability. Within this framework, we further integrate a novel
strategy based on network calculus (NC) theory for efficient and
high-utilization bandwidth reallocation. On the one hand, the strategy focuses
on adaptively balancing residual bandwidth with deadline awareness to prevent
bottleneck egress ports, thereby improving admission capacity. On the other
hand, it employs a non-trivial analytical result to reduce the search space,
accelerating the solving process. Experimental results from both large-scale
synthetic and realistic test cases show that, compared to the state-of-the-art,
our method achieves an average 56% increase in admitted flows and an average
92% reduction in admission time. Additionally, it postpones the occurrence of
bottleneck egress ports and the first rejection of admission requests, thereby
enhancing adaptability.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 06:06:49 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 17:22:43 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yu",
"Sifan",
""
],
[
"He",
"Feng",
""
],
[
"Xie",
"Anlan",
""
],
[
"Zhao",
"Luxi",
""
]
] |
2503.09288 | Mitsuharu Yashima | Mitsuharu Yashima, Nina Seto, Yujiro Oshita, Masataka Kakoi, Hiroya
Sakurai, Yoshihiko Takano, Hidekazu Mukuda | Microscopic evidence for spin-spinless stripe order with reduced Ni
moments within $ab$ plane for bilayer nickelate La$_3$Ni$_2$O$_7$ probed by
$^{139}$La-NQR | 6+2 pages, 4+2 figures | J. Phys. Soc. Jpn. 94, 054704 (2025) | 10.7566/JPSJ.94.054704 | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The intrinsic electronic properties of La$_3$Ni$_2$O$_{7}$ have been
selectively investigated by nuclear quadrupole resonance (NQR) at the La(2)
site outside the NiO$_2$ bilayers. The La(2)$_{\rm a}$ site of the ideal
La$_3$Ni$_2$O$_{7}$ is clearly distinguished from the La(2)$_{\rm b}$ site
close to the local defects. Below 150K, almost half of the intrinsic
La(2)$_{\rm a}$ sites are dominated by a finite internal field within the $ab$
plane, while the other half are dominated by zero internal field. The result is
fully consistent with the single spin-spinless stripe order of
($\cdots\uparrow\circ\downarrow\circ\uparrow\circ\cdots$), where the reduced Ni
magnetic moments are parallel to the $ab$-plane. Even for the La(2)$_{\rm b}$
site, the result is also explained within the same model by considering the
inhomogeneous internal magnetic fields enhanced around the nearby defects such
as oxygen vacancies. These findings provide unambiguous microscopic evidence
for the single spin-spinless stripe order below 150 K at ambient pressure.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:32:24 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 07:23:10 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yashima",
"Mitsuharu",
""
],
[
"Seto",
"Nina",
""
],
[
"Oshita",
"Yujiro",
""
],
[
"Kakoi",
"Masataka",
""
],
[
"Sakurai",
"Hiroya",
""
],
[
"Takano",
"Yoshihiko",
""
],
[
"Mukuda",
"Hidekazu",
""
]
] |
2503.09644 | Fabrizio Tamburini | Fabrizio Tamburini (Rotonium - Quantum Computing, Le Village by CA,
Padova, Italy) | A Majorana Relativistic Quantum Spectral Approach to the Riemann
Hypothesis in (1+1)-Dimensional Rindler Spacetimes | 33 pages, improved presentation | null | null | null | math.GM hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Following the Hilbert-P\'olya approach to the Riemann Hypothesis, we present
an exact spectral realization of the nontrivial zeros of the Riemann zeta
function $\zeta(z)$ with a Mellin-Barnes integral that explicitly contains it.
This integral defines the spectrum of the real-valued energy eigenvalues $E_n$
of a Majorana particle in a $(1+1)$-dimensional Rindler spacetime or equivalent
Kaluza-Klein reductions of $(n+1)$-dimensional geometries. We show that the
Hamiltonian $H_M$ describing the particle is hermitian and the spectrum of
energy eigenvalues $\{E_n\}_{n \in \mathbb{N}}$ is countably infinite in number
in a bijective correspondence with the imaginary part of the nontrivial zeros
of $\zeta(z)$ having the same cardinality as required by Hardy-Littlewood's
theorem from number theory. The correspondence between the two spectra with the
essential self-adjointness of $H_M$, confirmed with deficiency index analysis,
boundary triplet theory and Krein's extension theorem, imply that all
nontrivial zeros have real part $\Re ( z )=1/2$, i.e., lie on the ``critical
line''. In the framework of noncommutative geometry, $H_M$ is interpreted as a
Dirac operator $D$ in a spectral triple $(\mathcal{A}, \mathcal{H}, D)$,
linking these results to Connes' program for the Riemann Hypothesis. The
algebra $\mathcal{A}$ encodes the modular symmetries underlying the spectral
realization of $\zeta (z)$ in the Hilbert space $\mathcal{H}$ of Majorana
wavefunctions, integrating concepts from quantum mechanics, general relativity,
and number theory. This analysis offers a promising Hilbert-P\'olya-inspired
path to prove the Riemann Hypothesis.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 06:35:32 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 05:32:26 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Tamburini",
"Fabrizio",
"",
"Rotonium - Quantum Computing, Le Village by CA,\n Padova, Italy"
]
] |
2503.09956 | Loc Nguyen | Yu Qiao, Phuong-Nam Tran, Ji Su Yoon, Loc X. Nguyen, Eui-Nam Huh,
Dusit Niyato and Choong Seon Hong | DeepSeek-Inspired Exploration of RL-based LLMs and Synergy with Wireless
Networks: A Survey | 45 pages, 12 figures | null | null | null | cs.LG cs.AI cs.CV cs.ET | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reinforcement learning (RL)-based large language models (LLMs), such as
ChatGPT, DeepSeek, and Grok-3, have gained significant attention for their
exceptional capabilities in natural language processing and multimodal data
understanding. Meanwhile, the rapid expansion of information services has
driven the growing need for intelligence, efficient, and adaptable wireless
networks. Wireless networks require the empowerment of RL-based LLMs while
these models also benefit from wireless networks to broaden their application
scenarios. Specifically, RL-based LLMs can enhance wireless communication
systems through intelligent resource allocation, adaptive network optimization,
and real-time decision-making. Conversely, wireless networks provide a vital
infrastructure for the efficient training, deployment, and distributed
inference of RL-based LLMs, especially in decentralized and edge computing
environments. This mutual empowerment highlights the need for a deeper
exploration of the interplay between these two domains. We first review recent
advancements in wireless communications, highlighting the associated challenges
and potential solutions. We then discuss the progress of RL-based LLMs,
focusing on key technologies for LLM training, challenges, and potential
solutions. Subsequently, we explore the mutual empowerment between these two
fields, highlighting key motivations, open challenges, and potential solutions.
Finally, we provide insights into future directions, applications, and their
societal impact to further explore this intersection, paving the way for
next-generation intelligent communication systems. Overall, this survey
provides a comprehensive overview of the relationship between RL-based LLMs and
wireless networks, offering a vision where these domains empower each other to
drive innovations.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 01:59:11 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Mar 2025 01:32:33 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 01:30:17 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Qiao",
"Yu",
""
],
[
"Tran",
"Phuong-Nam",
""
],
[
"Yoon",
"Ji Su",
""
],
[
"Nguyen",
"Loc X.",
""
],
[
"Huh",
"Eui-Nam",
""
],
[
"Niyato",
"Dusit",
""
],
[
"Hong",
"Choong Seon",
""
]
] |
2503.10014 | Yirui Zheng | Yirui Zheng, Juntai Shen, Xufen Wu, and Bin-Hui Chen | Comparison of bar formation mechanisms II: does a tidally induced bar
grow faster than an internally developed bar? | Accepted by ApJ, revised slightly to match the accepted version. No
major changes. 13 pages, 7 figures + 1 appendix | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Bar structures can form internally due to the instability of their host
galaxies or externally due to perturbations from other galaxies. We
systematically quantify the growth timescales ($\tau_\mathrm{bar}$) of bars
formed through these two mechanisms with a series of controlled $N$-body
simulations. In galaxies susceptible to bar instability, tidally forced bars
display $\tau_\mathrm{bar}$ values comparable to those of internally developed
bars within the same disk. Tidal perturbations promote/delay bar formation by
advancing/postponing its onset, but the growth rate of the bar structure
remains largely unchanged. In these interaction scenarios, the bar formation is
still driven primarily by the galaxy's internal nature that remains unaffected
by tidal perturbations. As the external perturbation wave reaches the galaxy's
center, it evokes a "seed bar" that is then swing amplified. In this scenario,
the onset of bar formation is advanced. Conversely, bar formation may be
delayed if the external perturbation wave is out of phase with the pre-existing
spontaneously developed seed bar, which causes destructive interference and
limits the bar growth. In the hot disk model that resists bar formation in
isolation, $\tau_\mathrm{bar}$ of the tidally forced bar correlates with the
strength of the perturbation. The bar growth in this model deviates from the
exponential profile and is better described by a linear function. The vaired
$\tau_\mathrm{bar}$ and the preference for linear growth contrast with bars
formed in galaxies inherently susceptible to bar instability. These tidally
forced bars may not adhere to the swing amplification mechanism that predicts
an exponential bar growth.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 03:50:33 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:33:37 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zheng",
"Yirui",
""
],
[
"Shen",
"Juntai",
""
],
[
"Wu",
"Xufen",
""
],
[
"Chen",
"Bin-Hui",
""
]
] |
2503.10074 | Taehun Kim | Taehun Kim, Hyerean Jang, Youngjoo Shin | Demoting Security via Exploitation of Cache Demote Operation in Intel's
Latest ISA Extension | The modified version of this preprint has been submitted to ESORICS
2025 | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | ISA extensions are increasingly adopted to boost the performance of
specialized workloads without requiring an entire architectural redesign.
However, these enhancements can inadvertently expose new attack surfaces in the
microarchitecture. In this paper, we investigate Intel's recently introduced
cldemote extension, which promotes efficient data sharing by transferring cache
lines from upper-level caches to the Last Level Cache (LLC). Despite its
performance benefits, we uncover critical properties-unprivileged access,
inter-cache state transition, and fault suppression-that render cldemote
exploitable for microarchitectural attacks. We propose two new attack
primitives, Flush+Demote and Demote+Time, built on our analysis. Flush+Demote
constructs a covert channel with a bandwidth of 2.84 Mbps and a bit error rate
of 0.018%, while Demote+Time derandomizes the kernel base address in 2.49 ms on
Linux. Furthermore, we show that leveraging cldemote accelerates eviction set
construction in non-inclusive LLC designs by obviating the need for helper
threads or extensive cache conflicts, thereby reducing construction time by 36%
yet retaining comparable success rates. Finally, we examine how ISA extensions
contribute to broader microarchitectural attacks, identifying five key
exploitable characteristics and categorizing four distinct attack types. We
also discuss potential countermeasures, highlighting the far-reaching security
implications of emerging ISA extensions.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 05:43:14 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 00:38:24 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Kim",
"Taehun",
""
],
[
"Jang",
"Hyerean",
""
],
[
"Shin",
"Youngjoo",
""
]
] |
2503.10505 | Srivatsav Kunnawalkam Elayavalli | Srivatsav Kunnawalkam Elayavalli and Christopher Schafhauser | Negative resolution to the $C^*$-algebraic Tarski problem | Comments welcome. In v2, minor edits including adding references,
fixed typos | null | null | null | math.OA math.GR math.LO | http://creativecommons.org/publicdomain/zero/1.0/ | We compute the $K_1$-group of ultraproducts of unital, simple $C^*$-algebras
with unique trace and strict comparison. As an application, we prove that the
reduced free group $C^*$-algebras $C^*_r(F_m)$ and $C^*_r(F_n)$ are
elementarily equivalent (i.e., have isomorphic ultrapowers) if and only if $m =
n$. This settles in the negative the $C^*$-algebraic analogue of Tarski's 1945
problem for groups.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 16:08:40 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 17:47:21 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Elayavalli",
"Srivatsav Kunnawalkam",
""
],
[
"Schafhauser",
"Christopher",
""
]
] |
2503.10673 | Haidar Khan | Hisham A. Alyahya, Haidar Khan, Yazeed Alnumay, M Saiful Bari,
B\"ulent Yener | ZeroSumEval: An Extensible Framework For Scaling LLM Evaluation with
Inter-Model Competition | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | We introduce ZeroSumEval, a dynamic, competition-based, and evolving
evaluation framework for Large Language Models (LLMs) that leverages
competitive games. ZeroSumEval encompasses a diverse suite of games, including
security challenges (Capture the Flag), classic board games (chess), and
knowledge tests (MathQuiz). These games are designed to evaluate a range of
capabilities such as strategic reasoning, planning, knowledge application,
safety, and adaptability. Building upon recent studies that highlight the
effectiveness of game-based evaluations for LLMs, ZeroSumEval enhances these
approaches by providing a standardized and extensible framework for easily
implementing games and leverages DSPy to provide a better abstraction for LLM
player strategies.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 16:54:27 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:14:21 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Alyahya",
"Hisham A.",
""
],
[
"Khan",
"Haidar",
""
],
[
"Alnumay",
"Yazeed",
""
],
[
"Bari",
"M Saiful",
""
],
[
"Yener",
"Bülent",
""
]
] |
2503.12296 | Vu Thi Hue | Vu Thi Hue | Sharp estimates for Lyapunov exponents of Milstein approximation of
stochastic differential systems | null | null | null | null | math.PR math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Milstein approximation with step size $\Delta t>0$ of the solution $(X,
Y)$ to a two-by-two system of linear stochastic differential equations is
considered. It is proved that when the solution of the underlying model is
exponentially stable or exponentially blowing up at infinite time, these
behaviours are preserved at the level of the Milstein approximate solution
$\{(X_n, Y_n)\}$ in both the mean-square and almost-sure senses, provided
sufficiently small step size $\Delta t$. This result is based on sharp
estimates, from both above and below, of the discrete Lyapunov exponent. This
type of sharp estimate for approximate solutions to stochastic differential
equations seems to be first studied in this work. In particular, the proposed
method covers the setting for linear stochastic differential equations as well
as the $\theta$-Milstein scheme's setting.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 23:35:18 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:15:32 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Hue",
"Vu Thi",
""
]
] |
2503.12825 | Jian Zhai | Jian Zhai | Determination of the density in the linear elastic wave equation | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the inverse boundary value problem for the linear elastic wave
equation in three-dimensional isotropic medium. We show that both the Lam\'e
parameters and the density can be uniquely recovered from the boundary
measurements under the strictly convex foliation condition.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 05:11:33 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 00:02:16 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zhai",
"Jian",
""
]
] |
2503.12888 | Yang Guo | Siyuan Yao, Yang Guo, Yanyang Yan, Wenqi Ren, Xiaochun Cao | UncTrack: Reliable Visual Object Tracking with Uncertainty-Aware
Prototype Memory Network | 14 pages,11 figures,references added | null | 10.1109/TIP.2025.3559796 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformer-based trackers have achieved promising success and become the
dominant tracking paradigm due to their accuracy and efficiency. Despite the
substantial progress, most of the existing approaches tackle object tracking as
a deterministic coordinate regression problem, while the target localization
uncertainty has been greatly overlooked, which hampers trackers' ability to
maintain reliable target state prediction in challenging scenarios. To address
this issue, we propose UncTrack, a novel uncertainty-aware transformer tracker
that predicts the target localization uncertainty and incorporates this
uncertainty information for accurate target state inference. Specifically,
UncTrack utilizes a transformer encoder to perform feature interaction between
template and search images. The output features are passed into an
uncertainty-aware localization decoder (ULD) to coarsely predict the
corner-based localization and the corresponding localization uncertainty. Then
the localization uncertainty is sent into a prototype memory network (PMN) to
excavate valuable historical information to identify whether the target state
prediction is reliable or not. To enhance the template representation, the
samples with high confidence are fed back into the prototype memory bank for
memory updating, making the tracker more robust to challenging appearance
variations. Extensive experiments demonstrate that our method outperforms other
state-of-the-art methods. Our code is available at
https://github.com/ManOfStory/UncTrack.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 07:33:16 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yao",
"Siyuan",
""
],
[
"Guo",
"Yang",
""
],
[
"Yan",
"Yanyang",
""
],
[
"Ren",
"Wenqi",
""
],
[
"Cao",
"Xiaochun",
""
]
] |
2503.13145 | Entao Yang | Entao Yang, Xiaotian Zhang, Yue Shang, and Ge Zhang | High-entropy Advantage in Neural Networks' Generalizability | null | null | null | null | cs.LG cond-mat.stat-mech | http://creativecommons.org/licenses/by-nc-nd/4.0/ | One of the central challenges in modern machine learning is understanding how
neural networks generalize knowledge learned from training data to unseen test
data. While numerous empirical techniques have been proposed to improve
generalization, a theoretical understanding of the mechanism of generalization
remains elusive. Here we introduce the concept of Boltzmann entropy into neural
networks by re-conceptualizing such networks as hypothetical molecular systems
where weights and biases are atomic coordinates, and the loss function is the
potential energy. By employing molecular simulation algorithms, we compute
entropy landscapes as functions of both training loss and test accuracy (or
test loss), on networks with up to 1 million parameters, across four distinct
machine learning tasks: arithmetic question, real-world tabular data, image
recognition, and language modeling. Our results reveal the existence of
high-entropy advantage, wherein high-entropy network states generally
outperform those reached via conventional training techniques like stochastic
gradient descent. This entropy advantage provides a thermodynamic explanation
for neural network generalizability: the generalizable states occupy a larger
part of the parameter space than its non-generalizable analog at low train
loss. Furthermore, we find this advantage more pronounced in narrower neural
networks, indicating a need for different training optimizers tailored to
different sizes of networks.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 13:16:25 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 03:28:45 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yang",
"Entao",
""
],
[
"Zhang",
"Xiaotian",
""
],
[
"Shang",
"Yue",
""
],
[
"Zhang",
"Ge",
""
]
] |
2503.13379 | Mil\'an Mosonyi | P\'eter E. Frenkel, Mil\'an Mosonyi, P\'eter Vrana, Mih\'aly Weiner | Error bounds for composite quantum hypothesis testing and a new
characterization of the weighted Kubo-Ando geometric means | 36 pages. v3: Added explicit example with strict improvement in the
strong converse exponent using geometric means | null | null | null | quant-ph cs.IT math-ph math.FA math.IT math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The optimal error exponents of binary composite i.i.d. state discrimination
are trivially bounded by the worst-case pairwise exponents of discriminating
individual elements of the sets representing the two hypotheses, and in the
finite-dimensional classical case, these bounds in fact give exact single-copy
expressions for the error exponents. In contrast, in the non-commutative case,
the optimal exponents are only known to be expressible in terms of regularized
divergences, resulting in formulas that, while conceptually relevant,
practically not very useful. In this paper, we develop further an approach
initiated in [Mosonyi, Szil\'agyi, Weiner, IEEE Trans. Inf. Th.
68(2):1032--1067, 2022] to give improved single-copy bounds on the error
exponents by comparing not only individual states from the two hypotheses, but
also various unnormalized positive semi-definite operators associated to them.
Here, we show a number of equivalent characterizations of such operators giving
valid bounds, and show that in the commutative case, considering weighted
geometric means of the states, and in the case of two states per hypothesis,
considering weighted Kubo-Ando geometric means, are optimal for this approach.
As a result, we give a new characterization of the weighted Kubo-Ando geometric
means as the only $2$-variable operator geometric means that are block
additive, tensor multiplicative, and satisfy the arithmetic-geometric mean
inequality. We also extend our results to composite quantum channel
discrimination, and show an analogous optimality property of the weighted
Kubo-Ando geometric means of two quantum channels, a notion that seems to be
new. We extend this concept to defining the notion of superoperator perspective
function and establish some of its basic properties, which may be of
independent interest.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:06:27 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Mar 2025 07:41:23 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 13:32:39 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Frenkel",
"Péter E.",
""
],
[
"Mosonyi",
"Milán",
""
],
[
"Vrana",
"Péter",
""
],
[
"Weiner",
"Mihály",
""
]
] |
2503.15433 | Alexandra Artusio-Glimpse | Alexandra B. Artusio-Glimpse, Adil Meraki, Hunter Shillingburg, Guy
Lavallee, Miao Liu, Chad Eichfeld, Matthew T. Simons, Glenn Holland,
Christopher L. Holloway, Vladimir A. Aksyuk, Daniel Lopez | Wafer-level fabrication of all-dielectric vapor cells enabling optically
addressed Rydberg atom electrometry | null | null | null | null | physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | Rydberg-atom electrometry enables highly sensitive electric-field
measurements by exploiting the extreme polarizability of Rydberg states in
alkali atoms. Millimeter-scale atomic vapor cells can be accurately and
economically batch-fabricated by anodically bonding silicon and glass wafers,
enabling the large-volume manufacturing of miniature atomic clocks and quantum
sensors. However, silicon is not always an ideal constitutive material for
electric-field sensing because of its high dielectric constant and conductive
losses at millimeter wave frequencies. A broader selection of low-loss
all-dielectric alternatives may be beneficial for specific applications. Here,
we present an all-glass wafer-level microfabrication process that eliminates
silicon, creating hermetically sealed vapor cells that are stable over long
timelines with embedded cesium dispensers. Femtosecond laser machining
precisely defines the cell geometry, and laser-activated alkali loading ensures
reliable filling. We demonstrate long-term vacuum stability and robust Rydberg
excitation through electromagnetically induced transparency measurements of
several Rydberg states. We then use these cells to measure a 34 GHz millimeter
wave field resonant with the 58D$_{5/2}\rightarrow$60P$_{3/2}$ transition using
Autler-Townes splitting showing expected linear dependence with field strength.
This work demonstrates that the all-glass approach offers a highly durable
low-loss cell alternative for miniaturized millimeter wave and microwave
quantum sensing, with the potential to flexibly incorporate a range of other
dielectric and semiconductor materials and integrated photonic and electronic
technologies.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 17:17:27 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 17:32:26 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Artusio-Glimpse",
"Alexandra B.",
""
],
[
"Meraki",
"Adil",
""
],
[
"Shillingburg",
"Hunter",
""
],
[
"Lavallee",
"Guy",
""
],
[
"Liu",
"Miao",
""
],
[
"Eichfeld",
"Chad",
""
],
[
"Simons",
"Matthew T.",
""
],
[
"Holland",
"Glenn",
""
],
[
"Holloway",
"Christopher L.",
""
],
[
"Aksyuk",
"Vladimir A.",
""
],
[
"Lopez",
"Daniel",
""
]
] |
2503.15466 | Brice Coffer | Brice Coffer, Matthew Parker, Michael Coniglio, Cameron Homeyer | Supercell environments using GridRad-Severe and the HRRR: Addressing
discrepancies between prior tornado datasets | null | null | null | null | physics.ao-ph | http://creativecommons.org/licenses/by/4.0/ | Storm-relative helicity (SRH) is an important ingredient in supercell
development, as well as mesocyclone intensity, and is linked to tornadogenesis
and tornado potential. Derived from the storm-relative wind profile, SRH is
composed of both the vertical wind shear and storm-relative flow. Recent
studies have come to conflicting findings regarding whether shallower or deeper
layers of SRH have more skill in tornado forecasting. Possible causes of this
discrepancy include the use of observed versus model-based proximity soundings,
as well as whether the storm-relative wind profile is determined via observed
versus estimated storm motions. This study uses a new dataset of objectively
identified supercells, with observed storm motions, paired with high-resolution
model analyses to address the discrepancies among prior studies. Unlike in
previous model-based tornado environmental datasets, the present approach
reveals substantive differences in storm-relative flow, vertical wind shear,
and SRH within the low-to-mid-levels between nontornadic and tornadic
supercells. Using observed storm motions for storm-relative variables further
magnifies differences in the low-to-mid-level storm-relative winds between
nontornadic and tornadic supercells, ultimately leading to deeper layers of SRH
having more forecast skill than near-ground SRH. Thus, the combination of a
higher-resolution model analysis, which better represents the near-storm
environment, with observed storm motions appears to explain why many past
tornado climatologies using model-based environmental analyses have failed to
find significant differences in the storm-relative wind profile. These results
help bridge the gap between previous studies that employed coarser model-based
analyses and those that aggregated observed soundings from field projects.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 17:44:36 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:40:16 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Coffer",
"Brice",
""
],
[
"Parker",
"Matthew",
""
],
[
"Coniglio",
"Michael",
""
],
[
"Homeyer",
"Cameron",
""
]
] |
2503.15601 | David Prieto | Thomas W. Grimm, David Prieto and Mick van Vliet | Tame Embeddings, Volume Growth, and Complexity of Moduli Spaces | 9 pages | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | Quantum gravity is expected to impose constraints on the moduli spaces of
massless fields that can arise in effective quantum field theories. A recent
proposal asserts that the asymptotic volume growth of these spaces is severely
restricted, and related to the existence of duality symmetries. In this work we
link this proposal to a tameness criterion, by suggesting that any consistent
moduli space should admit a tame isometric embedding into Euclidean space. This
allows us to promote the volume growth constraint to a local condition, and
give the growth coefficient a geometric interpretation in terms of complexity.
We study the implications of this proposal for the emergence of dualities, as
well as for the curvature and infinite distance limits of moduli spaces.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 18:00:01 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 10:37:28 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Grimm",
"Thomas W.",
""
],
[
"Prieto",
"David",
""
],
[
"van Vliet",
"Mick",
""
]
] |
2503.16923 | Akito Daido | Akito Daido, Youichi Yanase, and Kam Tuen Law | Nonreciprocal Current-induced Zero-Resistance State in Valley-Polarized
Superconductors | 11 pages, 5 figures | null | null | null | cond-mat.supr-con | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recently observed nonreciprocal current-induced zero-resistance state
(CIZRS) in twisted trilayer graphene/WSe$_2$ heterostructure has posed a
significant theoretical challenge. Therein, the system shows a zero-resistance
state only when a sufficiently large current is applied in a particular
direction, while stays in an incipient superconducting state with small
resistance when the current is small or in the opposite direction. In this
Letter, we provide a theory for this phenomenon using a model that encapsulates
valley polarization and trigonal warping inherent to the trilayer
graphene/WSe$_2$ heterostructure. We elucidate that threefold-degenerate
Fulde-Ferrell (FF) states are stabilized, and establish the thermodynamic phase
diagram in electric current that manifests the switching behavior of different
FF states. We thereby propose a scenario that predicts CIZRS in the presence of
domains of different FF states: The incipient superconducting state in small
current is naturally understood as a multi-domain state, wherein the
inter-domain supercurrent is difficult to flow due to the tiny Josephson
coupling caused by the mismatch of Cooper-pair momenta. Nevertheless, a
sufficiently large current in a particular direction can selectively populate a
certain FF state and create mono-domain pathways with zero resistance.
Crucially, due to the threefold symmetry of the system, a current flowing in
the opposite direction fails to generate these zero-resistance pathways, thus
giving rise to the observed nonreciprocity. Finally, we suggest that the
long-sought-after triangular finite-momentum state can also be realized in
valley-polarized superconductors.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 07:36:21 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 08:07:42 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Daido",
"Akito",
""
],
[
"Yanase",
"Youichi",
""
],
[
"Law",
"Kam Tuen",
""
]
] |
2503.17192 | Teoman Toprak | Teoman Toprak, Michael Loibl, Guilherme Teixeira, Irina Shiskina, Chen
Miao, Josef Kiendl, Benjamin Marussig, Florian Kummer | Employing Continuous Integration inspired workflows for benchmarking of
scientific software -- a use case on numerical cut cell quadrature | 22 pages, 8 figures, pre-print (not submitted) | null | null | null | cs.SE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Scientific software often offers numerous (open or closed-source)
alternatives for a given problem. A user needs to make an informed choice by
selecting the best option based on specific metrics. However, setting up
benchmarks ad-hoc can become overwhelming as the parameter space expands
rapidly. Very often, the design of the benchmark is also not fully set at the
start of some project. For instance, adding new libraries, adapting metrics, or
introducing new benchmark cases during the project can significantly increase
complexity and necessitate laborious re-evaluation of previous results. This
paper presents a proven approach that utilizes established Continuous
Integration tools and practices to achieve high automation of benchmark
execution and reporting. Our use case is the numerical integration (quadrature)
on arbitrary domains, which are bounded by implicitly or parametrically defined
curves or surfaces in 2D or 3D.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 14:42:24 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 07:44:50 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Toprak",
"Teoman",
""
],
[
"Loibl",
"Michael",
""
],
[
"Teixeira",
"Guilherme",
""
],
[
"Shiskina",
"Irina",
""
],
[
"Miao",
"Chen",
""
],
[
"Kiendl",
"Josef",
""
],
[
"Marussig",
"Benjamin",
""
],
[
"Kummer",
"Florian",
""
]
] |
2503.18041 | Yang Zheng | Weiquan Chen, Zhao Dong, Yang Zheng | Non-uniqueness of Leray-Hopf Solutions to Forced Stochastic
Hyperdissipative Navier-Stokes Equations up to Lions Index | null | null | null | null | math.PR math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show non-uniqueness of local strong solutions to stochastic fractional
Navier-Stokes equations with linear multiplicative noise and some certain
deterministic force. Such non-uniqueness holds true even if we perturb such
deterministic force in appropriate sense.This is closely related to a critical
condition on force under which Leray-Hopf solution to the stochastic equations
is locally unique. Meanwhile, by a new idea, we show that for some stochastic
force the system admits two different global Leray-Hopf solutions smooth on any
compact subset of $(0,\infty) \times \mathbb{R}^d$.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 12:05:37 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 08:22:25 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Chen",
"Weiquan",
""
],
[
"Dong",
"Zhao",
""
],
[
"Zheng",
"Yang",
""
]
] |
2503.18672 | Juncen Guo | Juncen Guo, Yang Liu, Xiaoguang Zhu, Lianlong Sun, Liangyu Teng,
Jingyi Wu, Di Li, Wei Zhou, Liang Song | Feature Calibration enhanced Parameter Synthesis for CLIP-based
Class-incremental Learning | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Class-Incremental Learning (CIL) enables models to continuously learn new
class knowledge while retaining previous classes, facilitating adaptation and
evolution in dynamic, real-world environments. Traditional CIL methods
primarily rely on visual features, which limits their effectiveness in complex,
multimodal scenarios. In contrast, VLMs show promising potential for enhancing
CIL by leveraging pre-trained knowledge and integrating multi-modal semantic
cues such as text and vision. However, existing approaches struggle to mitigate
catastrophic forgetting while preserving the generalization strengths of VLMs
across diverse modalities. To address these challenges, we propose a Feature
Calibration Enhanced Parameter Synthesis (FCPS) framework. Specifically, FCPS
introduces a dynamic parameter adjustment mechanism that iteratively calibrates
the contribution of original visual features to the final class decision, thus
preserving the model's intrinsic generalization capability across modalities.
Simultaneously, parameter integration enables effective knowledge transfer,
maintaining a balance between acquiring new class representations and
preserving old knowledge. Experimental results on popular benchmarks (e.g.,
CIFAR100 and ImageNet100) validate the superiority of the proposed method.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 13:44:12 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Mar 2025 10:00:27 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 13:10:16 GMT"
},
{
"version": "v4",
"created": "Thu, 17 Apr 2025 12:26:16 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Guo",
"Juncen",
""
],
[
"Liu",
"Yang",
""
],
[
"Zhu",
"Xiaoguang",
""
],
[
"Sun",
"Lianlong",
""
],
[
"Teng",
"Liangyu",
""
],
[
"Wu",
"Jingyi",
""
],
[
"Li",
"Di",
""
],
[
"Zhou",
"Wei",
""
],
[
"Song",
"Liang",
""
]
] |
2503.18855 | Kai Xu | Kai Xu, Fei He | Boundary Effects in Biological Planar Networks: Pentagons Dominate
Marginal Cells | null | null | null | null | q-bio.OT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The topological and geometrical features at the boundary zone of planar
polygonal networks remain poorly understood. Based on observations and
mathematical proofs, we propose that marginal cells in Pyropia haitanensis
thalli, a two-dimensional (2D) biological polygonal network, have an average
edge number of exactly five. We demonstrate that this number is maintained by
specific division patterns. Furthermore, we reveal significant limitations of
Lewis law and Aboav-Weaire law by comparing the topological and geometrical
parameters of marginal cells and inner cells. We find strong boundary effects
that are manifested in the distinct distributions of interior angles and edge
lengths in marginal cells. Similar to inner cells, cell division tend to occur
in marginal cells with large sizes. Our findings suggest that inner cells
should be strictly defined based on their positional relationship to the
marginal cells.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 16:32:25 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 05:45:21 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Xu",
"Kai",
""
],
[
"He",
"Fei",
""
]
] |
2503.19185 | Francesco Calabr\'o | Davide Elia De Falco, Enrico Schiassi, Francesco Calabr\`o | Least Squares with Equality constraints Extreme Learning Machines for
the resolution of PDEs | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | In this paper, we investigate the use of single hidden-layer neural networks
as a family of ansatz functions for the resolution of partial differential
equations (PDEs). In particular, we train the network via Extreme Learning
Machines (ELMs) on the residual of the equation collocated on -- eventually
randomly chosen -- points. Because the approximation is done directly in the
formulation, such a method falls into the framework of Physically Informed
Neural Networks (PINNs) and has been named PIELM. Since its first introduction,
the method has been refined variously, and one successful variant is the
Extreme Theory of Functional Connections (XTFC). However, XTFC strongly takes
advantage of the description of the domain as a tensor product. Our aim is to
extend XTFC to domains with general shapes. The novelty of the procedure
proposed in the present paper is related to the treatment of boundary
conditions via constrained imposition, so that our method is named Least
Squares with Equality constraints ELM (LSEELM). An in-depth analysis and
comparison with the cited methods is performed, again with the analysis of the
convergence of the method in various scenarios. We show the efficiency of the
procedure both in terms of computational cost and in terms of overall accuracy.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 22:22:24 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 13:46:26 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"De Falco",
"Davide Elia",
""
],
[
"Schiassi",
"Enrico",
""
],
[
"Calabrò",
"Francesco",
""
]
] |
2503.19325 | Yuchao Gu | Yuchao Gu, Weijia Mao, Mike Zheng Shou | Long-Context Autoregressive Video Modeling with Next-Frame Prediction | Project page at https://farlongctx.github.io/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Long-context autoregressive modeling has significantly advanced language
generation, but video generation still struggles to fully utilize extended
temporal contexts. To investigate long-context video modeling, we introduce
Frame AutoRegressive (FAR), a strong baseline for video autoregressive
modeling. Just as language models learn causal dependencies between tokens
(i.e., Token AR), FAR models temporal causal dependencies between continuous
frames, achieving better convergence than Token AR and video diffusion
transformers. Building on FAR, we observe that long-context video modeling
faces challenges due to visual redundancy. Training on long videos is
computationally expensive, as vision tokens grow much faster than language
tokens. To tackle this issue, we propose balancing locality and long-range
dependency through long short-term context modeling. A high-resolution
short-term context window ensures fine-grained temporal consistency, while an
unlimited long-term context window encodes long-range information using fewer
tokens. With this approach, we can train on long video sequences with a
manageable token context length, thereby significantly reducing training time
and memory usage. Furthermore, we propose a multi-level KV cache designed to
support the long short-term context modeling, which accelerating inference on
long video sequences. We demonstrate that FAR achieves state-of-the-art
performance in both short- and long-video generation, providing a simple yet
effective baseline for video autoregressive modeling. The code is released at
https://github.com/showlab/FAR.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 03:38:06 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 15:26:04 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Gu",
"Yuchao",
""
],
[
"Mao",
"Weijia",
""
],
[
"Shou",
"Mike Zheng",
""
]
] |
2503.19492 | Taichi Tsukamoto | Shigeki Sugimoto, Taichi Tsukamoto | Energy-Momentum Tensor and D-term of Baryons in Top-down Holographic QCD | 19 pages, 6 figures, references added | null | null | KUNS-3037 | hep-th hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the energy-momentum tensor of a baryon in a top-down holographic
QCD. In holographic QCD, the baryons are represented as solitons in a
5-dimensional gauge theory. We obtain the soliton solution by solving the
equations of motion numerically. Using this result, the energy-momentum tensor
and related quantities such as the mass, mean square radii, and the D-term
(druck term) are computed. The evaluated D-term is about -2.05, whose absolute
value is significantly larger than that in the previous work arXiv:2206.06578.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 09:28:34 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:01:13 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Sugimoto",
"Shigeki",
""
],
[
"Tsukamoto",
"Taichi",
""
]
] |
2503.19563 | Jakob Reiffenstein | Jakob Reiffenstein | Nevanlinna matrix estimates without regularity conditions | 30 pages | null | null | null | math.SP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The Nevanlinna matrix of a half-line Jacobi operator coincides, up to
multiplication with a constant matrix, with the monodromy matrix of an
associated canonical system. This canonical system is discrete in a certain
sense, and is determined by two sequences, called "lengths" and "angles". We
derive new lower and upper estimates for the norm of the monodromy matrix in
terms of the lengths and angles, without imposing any restrictions on these
sequences. Returning to the Jacobi setting, we show that the order of the
Nevanlinna matrix is always greater than or equal to the convergence exponent
of the off-diagonal sequence of Jacobi parameters, and obtain a generalisation
of a classical theorem of Berezanskii.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 11:28:15 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:46:35 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Reiffenstein",
"Jakob",
""
]
] |
2503.22512 | Wenqiang Luo | Wenqiang Luo, Jacky Wai Keung, Boyang Yang, Jacques Klein, Tegawende
F. Bissyande, Haoye Tian, Bach Le | Unlocking LLM Repair Capabilities in Low-Resource Programming Languages
Through Cross-Language Translation and Multi-Agent Refinement | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in leveraging LLMs for APR have demonstrated impressive
capabilities in fixing software defects. However, current LLM-based approaches
predominantly focus on mainstream programming languages like Java and Python,
neglecting less prevalent but emerging languages such as Rust due to expensive
training resources, limited datasets, and insufficient community support. This
narrow focus creates a significant gap in repair capabilities across the
programming language spectrum, where the full potential of LLMs for
comprehensive multilingual program repair remains largely unexplored. To
address this limitation, we introduce a novel cross-language program repair
approach LANTERN that leverages LLMs' differential proficiency across languages
through a multi-agent iterative repair paradigm. Our technique strategically
translates defective code from languages where LLMs exhibit weaker repair
capabilities to languages where they demonstrate stronger performance, without
requiring additional training. A key innovation of our approach is an LLM-based
decision-making system that dynamically selects optimal target languages based
on bug characteristics and continuously incorporates feedback from previous
repair attempts. We evaluate our method on xCodeEval, a comprehensive
multilingual benchmark comprising 5,068 bugs across 11 programming languages.
Results demonstrate significant enhancement in repair effectiveness,
particularly for underrepresented languages, with Rust showing a 22.09%
improvement in Pass@10 metrics. Our research provides the first empirical
evidence that cross-language translation significantly expands the repair
capabilities of LLMs and effectively bridges the performance gap between
programming languages with different levels of popularity, opening new avenues
for truly language-agnostic automated program repair.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 15:15:56 GMT"
},
{
"version": "v2",
"created": "Thu, 3 Apr 2025 06:56:58 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 17:00:56 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Luo",
"Wenqiang",
""
],
[
"Keung",
"Jacky Wai",
""
],
[
"Yang",
"Boyang",
""
],
[
"Klein",
"Jacques",
""
],
[
"Bissyande",
"Tegawende F.",
""
],
[
"Tian",
"Haoye",
""
],
[
"Le",
"Bach",
""
]
] |
2503.22522 | Shuqing Liu | Shuqing Liu, Rong Su, Karl H.Johansson | A Centralized Planning and Distributed Execution Method for Shape
Filling with Homogeneous Mobile Robots | null | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The pattern formation task is commonly seen in a multi-robot system. In this
paper, we study the problem of forming complex shapes with functionally limited
mobile robots, which have to rely on other robots to precisely locate
themselves. The goal is to decide whether a given shape can be filled by a
given set of robots; in case the answer is yes, to complete a shape formation
process as fast as possible with a minimum amount of communication. Traditional
approaches either require global coordinates for each robot or are prone to
failure when attempting to form complex shapes beyond the capability of given
approaches - the latter calls for a decision procedure that can tell whether a
target shape can be formed before the actual shape-forming process starts. In
this paper, we develop a method that does not require global coordinate
information during the execution process and can effectively decide whether it
is feasible to form the desired shape. The latter is achieved via a planning
procedure that is capable of handling a variety of complex shapes, in
particular, those with holes, and assigning a simple piece of scheduling
information to each robot, facilitating subsequent distributed execution, which
does not rely on the coordinates of all robots but only those of neighboring
ones. The effectiveness of our shape-forming approach is vividly illustrated in
several simulation case studies.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 15:28:04 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 14:51:14 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 02:30:00 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Liu",
"Shuqing",
""
],
[
"Su",
"Rong",
""
],
[
"Johansson",
"Karl H.",
""
]
] |
2503.22714 | Sergio Torres Aguilar | Sergio Torres Aguilar | TRIDIS: A Comprehensive Medieval and Early Modern Corpus for HTR and NER | 6 pages, 3 figures, 2 tables | null | null | null | cs.CL cs.AI cs.CV cs.DL | http://creativecommons.org/licenses/by-sa/4.0/ | This paper introduces TRIDIS (Tria Digita Scribunt), an open-source corpus of
medieval and early modern manuscripts. TRIDIS aggregates multiple legacy
collections (all published under open licenses) and incorporates large metadata
descriptions. While prior publications referenced some portions of this corpus,
here we provide a unified overview with a stronger focus on its constitution.
We describe (i) the narrative, chronological, and editorial background of each
major sub-corpus, (ii) its semi-diplomatic transcription rules (expansion,
normalization, punctuation), (iii) a strategy for challenging out-of-domain
test splits driven by outlier detection in a joint embedding space, and (iv)
preliminary baseline experiments using TrOCR and MiniCPM2.5 comparing random
and outlier-based test partitions. Overall, TRIDIS is designed to stimulate
joint robust Handwritten Text Recognition (HTR) and Named Entity Recognition
(NER) research across medieval and early modern textual heritage.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 03:44:11 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Aguilar",
"Sergio Torres",
""
]
] |
2503.23230 | Lucas Gerken Starepravo | Lucas Gerken Starepravo, Georgios Fourtakas, Steven Lind, Ajay Harish,
Jack R. C. King | Can Neural Networks Bridge the Gap Between Lagrangian Mesh-Free Methods
and High-Order Interpolants? | null | null | null | null | physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Mesh-free numerical methods offer flexibility in discretising complex
geometries, showing potential where mesh-based methods struggle. While
high-order approximations can be obtained via consistency correction using
linear systems, they remain prohibitively expensive in Lagrangian formulations,
which often exhibit low-order convergence. Here, we explore the use of machine
learning (ML) to bridge the gap between mesh-free Lagrangian simulations and
high-order approximations. We develop strategies to couple data-driven models,
in particular multilayer perceptrons and residual MLPs with the Local
Anisotropic Basis Function Method (LABFM), as an exemplar high-order mesh-free
method. In the first strategy, we use neural networks to surrogate the
high-order kernel; in the second, we develop surrogate models for computing the
solutions of dense, low-rank linear systems present in high-order mesh-free
methods. Results from networks aimed at predicting support nodal weights yield
a qualitative match with validation data, but fall short in eliminating
lower-order errors due to inaccuracies in the ML-computed weights, and thus
leading to divergent behaviour. Regarding the second strategy, the ML-computed
solution vector generates residuals with mean absolute errors of
$\mathcal{O}(10^{-5})$. However, convergence studies reveal this level of
accuracy to be insufficient, causing derivative operators to diverge at a lower
resolution and achieve a lower accuracy than LABFM theoretically allows.
Furthermore, there is marginal computational gain when computing the solution
vector with neural networks compared to LU factorisation. These findings
indicate that insufficient accuracy challenges both using neural networks as
surrogates for high-order kernels and solve ill-conditioned linear systems,
while the additional high computational cost systems further limits the
latter's practicality.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 21:29:01 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 12:52:36 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Starepravo",
"Lucas Gerken",
""
],
[
"Fourtakas",
"Georgios",
""
],
[
"Lind",
"Steven",
""
],
[
"Harish",
"Ajay",
""
],
[
"King",
"Jack R. C.",
""
]
] |
2503.23752 | Jin Zhou | Jin Zhou, Yi Zhou, Pengfei Xu and Hui Huang | StrokeFusion: Vector Sketch Generation via Joint Stroke-UDF Encoding and
Latent Sequence Diffusion | null | null | null | null | cs.GR cs.CV | http://creativecommons.org/licenses/by/4.0/ | In the field of sketch generation, raster-format trained models often produce
non-stroke artifacts, while vector-format trained models typically lack a
holistic understanding of sketches, leading to compromised recognizability.
Moreover, existing methods struggle to extract common features from similar
elements (e.g., eyes of animals) appearing at varying positions across
sketches. To address these challenges, we propose StrokeFusion, a two-stage
framework for vector sketch generation. It contains a dual-modal sketch feature
learning network that maps strokes into a high-quality latent space. This
network decomposes sketches into normalized strokes and jointly encodes stroke
sequences with Unsigned Distance Function (UDF) maps, representing sketches as
sets of stroke feature vectors. Building upon this representation, our
framework exploits a stroke-level latent diffusion model that simultaneously
adjusts stroke position, scale, and trajectory during generation. This enables
high-fidelity sketch generation while supporting stroke interpolation editing.
Extensive experiments on the QuickDraw dataset demonstrate that our framework
outperforms state-of-the-art techniques, validating its effectiveness in
preserving structural integrity and semantic features. Code and models will be
made publicly available upon publication.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 06:03:03 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 03:17:42 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zhou",
"Jin",
""
],
[
"Zhou",
"Yi",
""
],
[
"Xu",
"Pengfei",
""
],
[
"Huang",
"Hui",
""
]
] |
2503.24047 | Shuo Ren | Shuo Ren, Pu Jian, Zhenjiang Ren, Chunlin Leng, Can Xie, Jiajun Zhang | Towards Scientific Intelligence: A Survey of LLM-based Scientific Agents | 34 pages, 10 figures | null | null | null | cs.AI cs.MA | http://creativecommons.org/licenses/by/4.0/ | As scientific research becomes increasingly complex, innovative tools are
needed to manage vast data, facilitate interdisciplinary collaboration, and
accelerate discovery. Large language models (LLMs) are now evolving into
LLM-based scientific agents that automate critical tasks, ranging from
hypothesis generation and experiment design to data analysis and simulation.
Unlike general-purpose LLMs, these specialized agents integrate domain-specific
knowledge, advanced tool sets, and robust validation mechanisms, enabling them
to handle complex data types, ensure reproducibility, and drive scientific
breakthroughs. This survey provides a focused review of the architectures,
design, benchmarks, applications, and ethical considerations surrounding
LLM-based scientific agents. We highlight why they differ from general agents
and the ways in which they advance research across various scientific fields.
By examining their development and challenges, this survey offers a
comprehensive roadmap for researchers and practitioners to harness these agents
for more efficient, reliable, and ethically sound scientific discovery.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 13:11:28 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 07:26:34 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Ren",
"Shuo",
""
],
[
"Jian",
"Pu",
""
],
[
"Ren",
"Zhenjiang",
""
],
[
"Leng",
"Chunlin",
""
],
[
"Xie",
"Can",
""
],
[
"Zhang",
"Jiajun",
""
]
] |
2503.24375 | Howard Milchberg | N. Tripathi, S.W. Hancock, and H.M. Milchberg | Transverse orbital angular momentum: setting the record straight | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | The nature of the transverse orbital angular momentum (tOAM) associated with
spatiotemporal optical vortex (STOV) pulses has been the subject of recent
debate. We demonstrate that the approaches to tOAM presented in several recent
papers are incorrect and lead to unphysical results, including erroneous claims
of zero total tOAM. We emphasize the importance of calculating the OAM of any
extended physical object at a common instant of time, and reemphasize the
special status of the centre of energy as a reference point for all OAM
calculations. The theory presented in [Phys. Rev. Lett. 127, 193901 (2021)] is
the only correct classical field-based framework that both agrees with
experiments and provides a self consistent understanding of transverse OAM in
spatiotemporal light fields.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 17:53:59 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 14:47:03 GMT"
},
{
"version": "v3",
"created": "Thu, 10 Apr 2025 15:25:41 GMT"
},
{
"version": "v4",
"created": "Thu, 17 Apr 2025 16:03:50 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Tripathi",
"N.",
""
],
[
"Hancock",
"S. W.",
""
],
[
"Milchberg",
"H. M.",
""
]
] |
2504.00046 | Fabrizio Marozzo | Loris Belcastro, Cristian Cosentino, Fabrizio Marozzo, Merve
G\"und\"uz-C\"ure, Sule \"Ozt\"urk-Birim | Multi-Stakeholder Disaster Insights from Social Media Using Large
Language Models | null | null | null | null | cs.CL cs.AI cs.ET cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, social media has emerged as a primary channel for users to
promptly share feedback and issues during disasters and emergencies, playing a
key role in crisis management. While significant progress has been made in
collecting and analyzing social media content, there remains a pressing need to
enhance the automation, aggregation, and customization of this data to deliver
actionable insights tailored to diverse stakeholders, including the press,
police, EMS, and firefighters. This effort is essential for improving the
coordination of activities such as relief efforts, resource distribution, and
media communication. This paper presents a methodology that leverages the
capabilities of LLMs to enhance disaster response and management. Our approach
combines classification techniques with generative AI to bridge the gap between
raw user feedback and stakeholder-specific reports. Social media posts shared
during catastrophic events are analyzed with a focus on user-reported issues,
service interruptions, and encountered challenges. We employ full-spectrum
LLMs, using analytical models like BERT for precise, multi-dimensional
classification of content type, sentiment, emotion, geolocation, and topic.
Generative models such as ChatGPT are then used to produce human-readable,
informative reports tailored to distinct audiences, synthesizing insights
derived from detailed classifications. We compare standard approaches, which
analyze posts directly using prompts in ChatGPT, to our advanced method, which
incorporates multi-dimensional classification, sub-event selection, and
tailored report generation. Our methodology demonstrates superior performance
in both quantitative metrics, such as text coherence scores and latent
representations, and qualitative assessments by automated tools and field
experts, delivering precise insights for diverse disaster response
stakeholders.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 22:53:52 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 11:29:06 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Belcastro",
"Loris",
""
],
[
"Cosentino",
"Cristian",
""
],
[
"Marozzo",
"Fabrizio",
""
],
[
"Gündüz-Cüre",
"Merve",
""
],
[
"Öztürk-Birim",
"Sule",
""
]
] |
2504.00399 | Akira Okumura | Ryu Kawarasaki and Akira Okumura and Kazuhiro Furuta and Hiroyasu
Tajima | Multimodal Shifts in the Dark Current of Silicon Photomultipliers | 7 pages, 3 figures, accepted for JINST | null | 10.1088/1748-0221/20/04/C04031 | null | astro-ph.IM hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The dark count rate is one of the key properties of avalanche photodiodes
(APDs) and silicon photomultipliers (SiPMs). Previous studies have reported
discrete shifts in the dark count rate on short timescales ~10 ms to ~100 s
with small increases (up to ~1 kHz per APD). In this study, we report a similar
yet distinct phenomenon in the dark current of SiPMs designed for gamma-ray
telescopes. Long-term stability tests (>100 days) revealed bimodal or
multimodal dark current shifts of the order of 0.1 uA on timescales of days in
48 out of 128 SiPM channels. In addition, optical emission was observed from an
APD surface when the dark current was in a high state. These findings suggest
that multimodal dark current behavior is a common property of SiPMs, which may
be due to defects in the silicon.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 03:45:54 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Kawarasaki",
"Ryu",
""
],
[
"Okumura",
"Akira",
""
],
[
"Furuta",
"Kazuhiro",
""
],
[
"Tajima",
"Hiroyasu",
""
]
] |
2504.00638 | Alireza Aghabagherloo | Alireza Aghabagherloo, Aydin Abadi, Sumanta Sarkar, Vishnu Asutosh
Dasu and Bart Preneel | Impact of Data Duplication on Deep Neural Network-Based Image
Classifiers: Robust vs. Standard Models | null | null | null | null | cs.LG cs.AI eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The accuracy and robustness of machine learning models against adversarial
attacks are significantly influenced by factors such as training data quality,
model architecture, the training process, and the deployment environment. In
recent years, duplicated data in training sets, especially in language models,
has attracted considerable attention. It has been shown that deduplication
enhances both training performance and model accuracy in language models. While
the importance of data quality in training image classifier Deep Neural
Networks (DNNs) is widely recognized, the impact of duplicated images in the
training set on model generalization and performance has received little
attention.
In this paper, we address this gap and provide a comprehensive study on the
effect of duplicates in image classification. Our analysis indicates that the
presence of duplicated images in the training set not only negatively affects
the efficiency of model training but also may result in lower accuracy of the
image classifier. This negative impact of duplication on accuracy is
particularly evident when duplicated data is non-uniform across classes or when
duplication, whether uniform or non-uniform, occurs in the training set of an
adversarially trained model. Even when duplicated samples are selected in a
uniform way, increasing the amount of duplication does not lead to a
significant improvement in accuracy.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 10:48:00 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 16:01:23 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Aghabagherloo",
"Alireza",
""
],
[
"Abadi",
"Aydin",
""
],
[
"Sarkar",
"Sumanta",
""
],
[
"Dasu",
"Vishnu Asutosh",
""
],
[
"Preneel",
"Bart",
""
]
] |
2504.01632 | Giulia Marchiori Pietrosanti Dr. | Giulia Marchiori Pietrosanti, Giulio Rossolini, Alessandro Biondi,
Giorgio Buttazzo | Benchmarking the Spatial Robustness of DNNs via Natural and Adversarial
Localized Corruptions | Under review | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The robustness of DNNs is a crucial factor in safety-critical applications,
particularly in complex and dynamic environments where localized corruptions
can arise. While previous studies have evaluated the robustness of semantic
segmentation (SS) models under whole-image natural or adversarial corruptions,
a comprehensive investigation into the spatial robustness of dense vision
models under localized corruptions remained underexplored. This paper fills
this gap by introducing specialized metrics for benchmarking the spatial
robustness of segmentation models, alongside with an evaluation framework to
assess the impact of localized corruptions. Furthermore, we uncover the
inherent complexity of characterizing worst-case robustness using a single
localized adversarial perturbation. To address this, we propose region-aware
multi-attack adversarial analysis, a method that enables a deeper understanding
of model robustness against adversarial perturbations applied to specific
regions. The proposed metrics and analysis were exploited to evaluate 14
segmentation models in driving scenarios, uncovering key insights into the
effects of localized corruption in both natural and adversarial forms. The
results reveal that models respond to these two types of threats differently;
for instance, transformer-based segmentation models demonstrate notable
robustness to localized natural corruptions but are highly vulnerable to
adversarial ones and vice-versa for CNN-based models. Consequently, we also
address the challenge of balancing robustness to both natural and adversarial
localized corruptions by means of ensemble models, thereby achieving a broader
threat coverage and improved reliability for dense vision tasks.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 11:37:39 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 16:43:20 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Pietrosanti",
"Giulia Marchiori",
""
],
[
"Rossolini",
"Giulio",
""
],
[
"Biondi",
"Alessandro",
""
],
[
"Buttazzo",
"Giorgio",
""
]
] |
2504.01797 | Hanqi Su | Jay Lee and Hanqi Su | Rethinking industrial artificial intelligence: a unified foundation
framework | The paper submitted to IJAMD, the International Journal of AI for
Materials and Design, has been accepted | null | 10.36922/IJAMD025080006 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in industrial artificial intelligence (AI) are reshaping
the industry by driving smarter manufacturing, predictive maintenance, and
intelligent decision-making. However, existing approaches often focus primarily
on algorithms and models while overlooking the importance of systematically
integrating domain knowledge, data, and models to develop more comprehensive
and effective AI solutions. Therefore, the effective development and deployment
of industrial AI require a more comprehensive and systematic approach. To
address this gap, this paper reviews previous research, rethinks the role of
industrial AI, and proposes a unified industrial AI foundation framework
comprising three core modules: the knowledge module, data module, and model
module. These modules help to extend and enhance the industrial AI methodology
platform, supporting various industrial applications. In addition, a case study
on rotating machinery diagnosis is presented to demonstrate the effectiveness
of the proposed framework, and several future directions are highlighted for
the development of the industrial AI foundation framework.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 15:05:32 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:54:57 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Lee",
"Jay",
""
],
[
"Su",
"Hanqi",
""
]
] |
2504.02025 | Riccardo Gonzo | Dogan Akpinar, Vittorio del Duca, Riccardo Gonzo | The spinning self-force EFT: 1SF waveform recursion relation and Compton
scattering | 15 pages + appendices, 4 figures; v2: minor improvements, references
added | null | null | null | hep-th gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Building on recent approaches, we develop an effective field theory for the
interaction of spinning particles modeling Kerr black holes within the
gravitational self-force expansion. To incorporate dimensional regularization
into this framework, we analyze the Myers-Perry black hole and its particle
description, obtaining a novel form of the corresponding linearized stress
tensor. We then derive the 1SF self-force effective action up to quadratic
order in the spin expansion, identifying a new type of spinning recoil term
that arises from integrating out the heavy dynamics. Next, we study the 1SF
metric perturbation both from the traditional self-force perspective and
through the diagrammatic background field expansion, making contact with the
radiative waveform. This leads us to consider a novel recursion relation for
the curved space 1SF Compton amplitude, which we study up to one-loop in the
wave regime and compare with the flat space one-loop Compton amplitude for Kerr
up to quadratic order in spin. Finally, we investigate the 1SF spinning Compton
amplitude in the eikonal regime, clarifying how strong-field effects -- such as
the location of the separatrix -- emerge from the resummation of the
perturbative weak-field expansion.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 18:00:00 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 17:10:42 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Akpinar",
"Dogan",
""
],
[
"del Duca",
"Vittorio",
""
],
[
"Gonzo",
"Riccardo",
""
]
] |
2504.02431 | Mario Kahlhofer | Mario Kahlhofer, Matteo Golinelli, Stefan Rass | Koney: A Cyber Deception Orchestration Framework for Kubernetes | camera-ready version; to be published in the 4th Workshop on Active
Defense and Deception (ADnD 2025) co-located with IEEE EuroS&P, source code
available at https://github.com/dynatrace-oss/koney | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | System operators responsible for protecting software applications remain
hesitant to implement cyber deception technology, including methods that place
traps to catch attackers, despite its proven benefits. Overcoming their
concerns removes a barrier that currently hinders industry adoption of
deception technology. Our work introduces deception policy documents to
describe deception technology "as code" and pairs them with Koney, a Kubernetes
operator, which facilitates the setup, rotation, monitoring, and removal of
traps in Kubernetes. We leverage cloud-native technologies, such as service
meshes and eBPF, to automatically add traps to containerized software
applications, without having access to the source code. We focus specifically
on operational properties, such as maintainability, scalability, and
simplicity, which we consider essential to accelerate the adoption of cyber
deception technology and to facilitate further research on cyber deception.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 09:37:14 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 11:12:49 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Kahlhofer",
"Mario",
""
],
[
"Golinelli",
"Matteo",
""
],
[
"Rass",
"Stefan",
""
]
] |
2504.02497 | Abhiram Soori | Sachchidanand Das and Abhiram Soori | An all-electrical scheme for valley polarization in graphene | 7 pages, 7 captioned figures | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We propose an all-electrical setup to generate valley polarization in
graphene. A finite graphene sheet is connected to two normal metal electrodes
along its zigzag edges, with armchair edges left free. When a bias is applied
at one terminal and the others are grounded, valley polarization arises from
transverse momentum matching between graphene and the normal metal. Significant
polarization occurs when the Fermi wave vector in the metal exceeds half the \(
K \)--\( K' \) valley separation. We analyze the dependence of conductance and
valley polarization on geometric and electronic parameters, and show that while
increased width enhances both, increased length leads to Fabry--P\'erot
oscillations and suppresses polarization due to intervalley mixing. Disorder
near the Dirac point enhances conductance but reduces polarization. Our
findings offer a route to electrical control of valley degrees of freedom in
graphene-based devices.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 11:20:41 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 03:13:22 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Das",
"Sachchidanand",
""
],
[
"Soori",
"Abhiram",
""
]
] |
2504.02546 | Xiangxiang Chu | Xiangxiang Chu and Hailang Huang and Xiao Zhang and Fei Wei and Yong
Wang | GPG: A Simple and Strong Reinforcement Learning Baseline for Model
Reasoning | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reinforcement Learning (RL) can directly enhance the reasoning capabilities
of large language models without extensive reliance on Supervised Fine-Tuning
(SFT). In this work, we revisit the traditional Policy Gradient (PG) mechanism
and propose a minimalist RL approach termed Group Policy Gradient (GPG). Unlike
conventional methods, GPG directly optimize the original RL objective, thus
obviating the need for surrogate loss functions. By eliminating the critic and
reference models, avoiding KL divergence constraints, and addressing the
advantage and gradient estimation bias, our approach significantly simplifies
the training process compared to Group Relative Policy Optimization (GRPO). Our
approach achieves superior performance without relying on auxiliary techniques
or adjustments. As illustrated in Figure 1, extensive experiments demonstrate
that our method not only reduces computational costs but also consistently
outperforms GRPO across various unimodal and multimodal tasks. Our code is
available at https://github.com/AMAP-ML/GPG.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 12:53:41 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 15:53:07 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Chu",
"Xiangxiang",
""
],
[
"Huang",
"Hailang",
""
],
[
"Zhang",
"Xiao",
""
],
[
"Wei",
"Fei",
""
],
[
"Wang",
"Yong",
""
]
] |
2504.02634 | Guy Wilkinson | Alain Blondel, Christophe Grojean, Patrick Janot, Srini Rajagopalan
and Guy Wilkinson | The FCC integrated programme: a physics manifesto | Contribution to the European Strategy for Particle Physics Update
2025-2026. This document is submitted on behalf of the Physics, Experiments
and Detectors pillar of the FCC Project, including those who signed the FCC
Feasibility Study Report, and those who contributed to the FCC documents
referenced herein. (v2 - correct author affiliation, and implement minor
changes to appendix.) | null | null | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | The FCC integrated programme comprises an $\rm e^+e^-$ high-luminosity
circular collider that will produce very large samples of data in an energy
range $88 \le \sqrt{s} \le 365$ GeV, followed by a high-energy $\rm pp$ machine
that, with the current baseline plan, will operate at a collision energy of
around 85 TeV and deliver datasets an order of magnitude larger than those of
the HL-LHC. This visionary project will allow for transformative measurements
across a very broad range of topics, which in almost all cases will exceed in
sensitivity the projections of any other proposed facility, and simultaneously
provide the best possible opportunity for discovering physics beyond the
Standard Model. The highlights of the physics programme are presented, together
with discussion on the key attributes of the integrated project that enable the
physics reach. It is noted that the baseline programme of FCC-ee, in
particular, is both flexible and extendable, and also that the synergy and
complementarity of the electron and proton machines, and the sharing of a
common infrastructure, provides a remarkably efficient, timely and
cost-effective approach to addressing the most pressing open questions in
elementary particle physics.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 14:31:24 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 11:29:35 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Blondel",
"Alain",
""
],
[
"Grojean",
"Christophe",
""
],
[
"Janot",
"Patrick",
""
],
[
"Rajagopalan",
"Srini",
""
],
[
"Wilkinson",
"Guy",
""
]
] |
2504.02786 | Pedro Meert | Pedro Meert, Andrea Giusti and Roberto Casadio | Quantum maximally symmetric space-times | null | null | null | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that 4-dimensional maximally symmetric spacetimes can be obtained
from a coherent state quantisation of gravity, always resulting in geometries
that approach the Minkowski vacuum exponentially away from the radius of
curvature. A possible connection with the central charge in the AdS/CFT
correspondence is also noted.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 17:30:31 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 18:49:17 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Meert",
"Pedro",
""
],
[
"Giusti",
"Andrea",
""
],
[
"Casadio",
"Roberto",
""
]
] |
2504.02792 | Chuning Zhu | Chuning Zhu, Raymond Yu, Siyuan Feng, Benjamin Burchfiel, Paarth Shah,
and Abhishek Gupta | Unified World Models: Coupling Video and Action Diffusion for
Pretraining on Large Robotic Datasets | null | null | null | null | cs.RO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Imitation learning has emerged as a promising approach towards building
generalist robots. However, scaling imitation learning for large robot
foundation models remains challenging due to its reliance on high-quality
expert demonstrations. Meanwhile, large amounts of video data depicting a wide
range of environments and diverse behaviors are readily available. This data
provides a rich source of information about real-world dynamics and
agent-environment interactions. Leveraging this data directly for imitation
learning, however, has proven difficult due to the lack of action annotation
required for most contemporary methods. In this work, we present Unified World
Models (UWM), a framework that allows for leveraging both video and action data
for policy learning. Specifically, a UWM integrates an action diffusion process
and a video diffusion process within a unified transformer architecture, where
independent diffusion timesteps govern each modality. By simply controlling
each diffusion timestep, UWM can flexibly represent a policy, a forward
dynamics, an inverse dynamics, and a video generator. Through simulated and
real-world experiments, we show that: (1) UWM enables effective pretraining on
large-scale multitask robot datasets with both dynamics and action predictions,
resulting in more generalizable and robust policies than imitation learning,
(2) UWM naturally facilitates learning from action-free video data through
independent control of modality-specific diffusion timesteps, further improving
the performance of finetuned policies. Our results suggest that UWM offers a
promising step toward harnessing large, heterogeneous datasets for scalable
robot learning, and provides a simple unification between the often disparate
paradigms of imitation learning and world modeling. Videos and code are
available at https://weirdlabuw.github.io/uwm/.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 17:38:59 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 20:27:54 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zhu",
"Chuning",
""
],
[
"Yu",
"Raymond",
""
],
[
"Feng",
"Siyuan",
""
],
[
"Burchfiel",
"Benjamin",
""
],
[
"Shah",
"Paarth",
""
],
[
"Gupta",
"Abhishek",
""
]
] |
2504.03160 | Yuxiang Zheng | Yuxiang Zheng, Dayuan Fu, Xiangkun Hu, Xiaojie Cai, Lyumanshan Ye,
Pengrui Lu, Pengfei Liu | DeepResearcher: Scaling Deep Research via Reinforcement Learning in
Real-world Environments | null | null | null | null | cs.AI cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) equipped with web search capabilities have
demonstrated impressive potential for deep research tasks. However, current
approaches predominantly rely on either manually engineered prompts (prompt
engineering-based) with brittle performance or reinforcement learning within
controlled Retrieval-Augmented Generation (RAG) environments (RAG-based) that
fail to capture the complexities of real-world interaction. In this paper, we
introduce DeepResearcher, the first comprehensive framework for end-to-end
training of LLM-based deep research agents through scaling reinforcement
learning (RL) in real-world environments with authentic web search
interactions. Unlike RAG-based approaches that assume all necessary information
exists within a fixed corpus, our method trains agents to navigate the noisy,
unstructured, and dynamic nature of the open web. We implement a specialized
multi-agent architecture where browsing agents extract relevant information
from various webpage structures and overcoming significant technical
challenges. Extensive experiments on open-domain research tasks demonstrate
that DeepResearcher achieves substantial improvements of up to 28.9 points over
prompt engineering-based baselines and up to 7.2 points over RAG-based RL
agents. Our qualitative analysis reveals emergent cognitive behaviors from
end-to-end RL training, including the ability to formulate plans,
cross-validate information from multiple sources, engage in self-reflection to
redirect research, and maintain honesty when unable to find definitive answers.
Our results highlight that end-to-end training in real-world web environments
is not merely an implementation detail but a fundamental requirement for
developing robust research capabilities aligned with real-world applications.
We release DeepResearcher at https://github.com/GAIR-NLP/DeepResearcher.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 04:41:28 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Apr 2025 10:45:47 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 02:57:20 GMT"
},
{
"version": "v4",
"created": "Thu, 17 Apr 2025 04:46:08 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zheng",
"Yuxiang",
""
],
[
"Fu",
"Dayuan",
""
],
[
"Hu",
"Xiangkun",
""
],
[
"Cai",
"Xiaojie",
""
],
[
"Ye",
"Lyumanshan",
""
],
[
"Lu",
"Pengrui",
""
],
[
"Liu",
"Pengfei",
""
]
] |
2504.03619 | Haozhou Hu | Haozhou Hu, Harpreet S. Dhillon, R. Michael Buehrer | A New Statistical Approach to Calibration-Free Localization Using
Unlabeled Crowdsourced Data | 15 pages | null | null | null | eess.SP stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fingerprinting-based indoor localization methods typically require
labor-intensive site surveys to collect signal measurements at known reference
locations and frequent recalibration, which limits their scalability. This
paper addresses these challenges by presenting a novel approach for indoor
localization that utilizes crowdsourced data without location labels. We
leverage the statistical information of crowdsourced data and propose a
cumulative distribution function (CDF) based distance estimation method that
maps received signal strength (RSS) to distances from access points. This
approach overcomes the limitations of conventional distance estimation based on
the empirical path loss model by efficiently capturing the impacts of shadow
fading and multipath. Compared to fingerprinting, our unsupervised statistical
approach eliminates the need for signal measurements at known reference
locations. The estimated distances are then integrated into a three-step
framework to determine the target location. The localization performance of our
proposed method is evaluated using RSS data generated from ray-tracing
simulations. Our results demonstrate significant improvements in localization
accuracy compared to methods based on the empirical path loss model.
Furthermore, our statistical approach, which relies on unlabeled data, achieves
localization accuracy comparable to that of the supervised approach, the
$k$-Nearest Neighbor ($k$NN) algorithm, which requires fingerprints with
location labels. For reproducibility and future research, we make the
ray-tracing dataset publicly available at [2].
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 17:37:30 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 15:34:35 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Hu",
"Haozhou",
""
],
[
"Dhillon",
"Harpreet S.",
""
],
[
"Buehrer",
"R. Michael",
""
]
] |
2504.03879 | Jiho Kim | Jiho Kim, Cong Hao | RealProbe: An Automated and Lightweight Performance Profiler for In-FPGA
Execution of High-Level Synthesis Designs | Accepted at FCCM 2025. Artifact evaluated | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | High-level synthesis (HLS) accelerates FPGA design by rapidly generating
diverse implementations using optimization directives. However, even with
cycle-accurate C/RTL co-simulation, the reported clock cycles often differ
significantly from actual FPGA performance. This discrepancy hampers accurate
bottleneck identification, leading to suboptimal design choices. Existing
in-FPGA profiling tools, such as the Integrated Logic Analyzer (ILA), require
tedious inspection of HLS-generated RTL and manual signal monitoring, reducing
productivity. To address these challenges, we introduce RealProbe, the first
fully automated, lightweight in-FPGA profiling tool for HLS designs. With a
single directive--#pragma HLS RealProbe--the tool automatically generates all
necessary code to profile cycle counts across the full function hierarchy,
including submodules and loops. RealProbe extracts, records, and visualizes
cycle counts with high precision, providing actionable insights into on-board
performance. RealProbe is non-intrusive, implemented as independent logic to
ensure minimal impact on kernel functionality or timing. It also supports
automated design space exploration (DSE), optimizing resource allocation based
on FPGA constraints and module complexity. By leveraging incremental synthesis
and implementation, DSE runs independently of the original HLS kernel.
Evaluated across 28 diverse test cases, including a large-scale design,
RealProbe achieves 100% accuracy in capturing cycle counts with minimal logic
overhead-just 16.98% LUTs, 43.15% FFs, and 0% BRAM usage. The tool, with full
documentation and examples, is available on GitHub at
https://github.com/sharc-lab/RealProbe .
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 19:03:56 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 18:23:31 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Kim",
"Jiho",
""
],
[
"Hao",
"Cong",
""
]
] |
2504.03946 | Onofre Rojas | Onofre Rojas, Moises Rojas, and S. M. de Souza | Quantum Otto engine mimicking Carnot near pseudotransitions in the
one-dimensional extended Hubbard model in the atomic limit | 12 pages, 9 figures | Phys. Rev. E 111 044121 (2025) | 10.1103/PhysRevE.111.044121 | null | cond-mat.str-el cond-mat.mes-hall | http://creativecommons.org/publicdomain/zero/1.0/ | The one-dimensional extended Hubbard model (EHM) in the atomic limit has
recently been found to exhibit a curious thermal pseudo-transition behavior,
which closely resembles first and second-order thermal phase transitions. This
phenomenon, occurring at half-filling, is influenced by the quantum phase
transition between the alternating pair (AP) and paramagnetic (PM) phases at
zero temperature. In this study, we leverage this anomalous behavior to
investigate the performance of quantum many-body machines, using the EHM as the
working substance. Our analysis reveals that the quantum Otto engine, when
operating in the anomalous region, closely mimics the ideal Carnot engine. In
this region, both the work output and thermal efficiency of the Otto engine
increase, approaching the performance of a Carnot engine. This highlights the
potential of many-body systems, such as the EHM, in enhancing quantum
thermodynamic performance. Our findings demonstrate that, although the second
law of thermodynamics prevents engines from surpassing Carnot efficiency, the
Otto engine can operate remarkably close to this limit in the anomalous region,
offering insights into new directions for future research on quantum
thermodynamic cycles and working substances.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 21:27:06 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 17:33:24 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Rojas",
"Onofre",
""
],
[
"Rojas",
"Moises",
""
],
[
"de Souza",
"S. M.",
""
]
] |
2504.03982 | Shreya Khisa | Shreya Khisa, Ali Amhaz, Mohamed Elhattab, Chadi Assi, Sanaa
Sharafeddine | Meta-Learning Driven Movable-Antenna-assisted Full-Duplex RSMA for
Multi-User Communication: Performance and Optimization | null | null | null | null | eess.SY cs.SY eess.SP | http://creativecommons.org/licenses/by/4.0/ | Full-duplex (FD) radios at base station (BS) have gained significant interest
because of their ability to simultaneously transmit and receive signals on the
same frequency band. However, FD communication is hindered by self-interference
(SI) and intra-cell interference caused by simultaneous uplink (UL)
transmissions affecting downlink (DL) reception. These interferences
significantly limit the ability to fully exploit FD's potential. Recently,
movable antenna (MA) technology has emerged as a groundbreaking innovation,
offering an effective way to mitigate interference by adjusting the position of
each MA within the transmitter or receiver region. This dynamic repositioning
allows MAs to move away from high-interference zones to areas with minimal
interference, thereby enhancing multiplexing gain and improving spectral
efficiency (SE). In light of this, in this paper, we investigate an FD
communication system by integrating it with MAs to evaluate and investigate its
effectiveness in handling SI and intra-cell interference. Moreover, we utilize
rate-splitting multiple access (RSMA) as our multiple access technique in both
UL and DL transmission. To achieve the full potential of the system, we
evaluated three different scenarios with FD-BS-RSMA with MAs where our goal is
to maximize the total sum rate of the system by jointly optimizing the
transmitting and receiving beamforming vectors, UL user equipment (UE)
transmission power, MA positions, and common stream split ratio of RSMA while
satisfying the minimum data rate requirements of all UEs, common stream
constraint, power budget requirements of BS and UL UEs, and inter-MA distance.
The formulated optimization problem is highly non-convex in nature, and hence,
we propose a gradient-based meta-learning (GML) approach which can handle the
non-convexity in a discrete manner by optimizing each variable in a different
neural network.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 22:54:05 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 15:39:03 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Khisa",
"Shreya",
""
],
[
"Amhaz",
"Ali",
""
],
[
"Elhattab",
"Mohamed",
""
],
[
"Assi",
"Chadi",
""
],
[
"Sharafeddine",
"Sanaa",
""
]
] |
2504.04035 | Ting-Tung Wang | Ting-Tung Wang, Xiaoxue Ran, and Zi Yang Meng | Symmetrizing the Constraints -- Density Matrix Renormalization Group for
Constrained Lattice Models | 12 pages, 8 figures | null | null | null | cond-mat.str-el quant-ph | http://creativecommons.org/licenses/by/4.0/ | We develop a density matrix renormalization group (DMRG) algorithm for
constrained quantum lattice models that successfully {\it{implements the local
constraints as symmetries in the contraction of the matrix product states and
matrix product operators}}. Such an implementation allows us to investigate a
quantum dimer model in DMRG for any lattice geometry wrapped around a cylinder
with substantial circumference. We have thence computed the ground state phase
diagram of the quantum dimer model on triangular lattice, with the
symmetry-breaking characteristics of the columnar solid phase and
$\sqrt{12}\times\sqrt{12}$ valence bond solid phase fully captured, as well as
the topological entanglement entropy of the $\mathbb{Z}_2$ quantum spin liquid
phase that extends to the RK point on non-bipartite lattice accurately
revealed. Our DMRG algorithm on constrained quantum lattice models opens new
opportunities for matrix and tensor-based algorithms for these systems that
have immediate relevance towards the frustrated quantum magnets and synthetic
quantum simulators.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 03:05:11 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 09:51:31 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 08:44:37 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Wang",
"Ting-Tung",
""
],
[
"Ran",
"Xiaoxue",
""
],
[
"Meng",
"Zi Yang",
""
]
] |
2504.04049 | Tian-Xiao He | Tian-Xiao He | The Multiple Riordan Group and the Multiple Riordan Semigroup | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | In this paper, we define Riordan type arrays and the Riordan semigroup and
extend them to multiple settings. Multiple Riordan type arrays and the multiple
Riordan semigroup are related to multiple Riordan arrays and the multiple
Riordan group. We give a comprehensive discussion of the latter and
characterize them by an $A$-sequence and multiple $Z$-sequences. Applications
of Riordan type arrays in the construction of identities are given. In
addition, we give compressions of multiple Riordan arrays and multiple Riordan
type arrays and their sequence characterizations.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 03:58:45 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 04:11:41 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"He",
"Tian-Xiao",
""
]
] |
2504.04165 | Wadim Gerner | Ugo Boscain and Wadim Gerner | Charged particle motion in a strong magnetic field: Applications to
plasma confinement | 23 pages. Some additional remarks and an explicit example with a
figure have been added. Minor typos were corrected. The main
results/conclusions remained unchanged | null | null | null | math-ph math.CA math.MP physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We derive the zero order approximation of a charged particle under the
influence of a strong magnetic field in a mathematically rigorous manner and
clarify in which sense this approximation is valid. We use this to further
rigorously derive a displacement formula for the pressure of plasma equilibria
and compare our findings to results in the physics literature. The main novelty
of our results is a qualitative estimate of the confinement time for optimised
plasma equilibria with respect to the gyro frequency. These results are of
interest in the context of plasma fusion confinement.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 13:01:30 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 13:15:00 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Boscain",
"Ugo",
""
],
[
"Gerner",
"Wadim",
""
]
] |
2504.04312 | Xiaodong Shao | Xiaodong Shao, Haoyang Yang, Haoran Li, Zongyu Zuo, Jose Guadalupe
Romero, Qinglei Hu | Prescribed-Time Boresight Control of Spacecraft Under Pointing
Constraints | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | This article proposes an integrated boresight guidance and control (IBGC)
scheme to address the boresight reorientation problem of spacecraft under
temporal and pointing constraints. A $C^1$ continuous, saturated
prescribed-time adjustment (PPTA) function is presented, along with the
establishment of a practical prescribed-time stability criterion. Utilizing the
time scale transformation technique and the PPTA function, we propose a
prescribed-time guidance law that guides the boresight vector from almost any
initial orientation in free space to a small neighborhood of the goal
orientation within a preassigned time, while avoiding all forbidden zones
augmented with safety margins. Subsequently, a prescribed-time disturbance
observer (PTDO) is derived to reconstruct the external disturbances. By
leveraging barrier and PPTA functions, a PTDO-based reduced-attitude tracking
controller is developed, which ensures prescribed-time boresight tracking
within a ``safe tube''. By judiciously setting the safety margins, settling
times, and safe tube for the guidance and control laws, the proposed IBGC
scheme achieves pointing-constrained boresight reorientation within a required
task completion time. Simulation and experimental results demonstrate the
efficacy of the proposed IBGC scheme.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 01:00:12 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Apr 2025 05:08:56 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 03:18:07 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Shao",
"Xiaodong",
""
],
[
"Yang",
"Haoyang",
""
],
[
"Li",
"Haoran",
""
],
[
"Zuo",
"Zongyu",
""
],
[
"Romero",
"Jose Guadalupe",
""
],
[
"Hu",
"Qinglei",
""
]
] |
2504.04581 | Xufei Fang Dr. | Alexander Frisch, Chukwudalu Okafor, Oliver Preu{\ss}, Jiawen Zhang,
Katsuyuki Matsunaga, Atsutomo Nakamura, Wenjun Lu, Xufei Fang | Room-temperature dislocation plasticity in ceramics: Methods, Materials,
and Mechanisms | null | Journal of the American Ceramic Society, 2025 | 10.1111/jace.20575 | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Dislocation-mediated plastic deformation in ceramic materials has sparked
renewed research interest due to the technological potential of dislocations.
Despite the long research history of dislocations as one-dimensional lattice
defects in crystalline solids, the understanding of plastically deformable
ceramics at room temperature seems lacking. The conventional view holds that
ceramics are brittle, difficult to deform at room temperature and exhibit no
dislocation plasticity except in small-scale testing such as nanoindentation
and nano-micropillar compression. In this review, we attempt to gather the
evidence and reports of room-temperature dislocation plasticity in ceramics
beyond the nano-/microscale, with a focus on meso-macroscale plasticity. First,
we present a mechanical deformation toolbox covering various experimental
approaches for assessing the dislocation plasticity, with a focus on bulk
plasticity. Second, we provide a materials toolbox listing 44 ceramic compounds
that have been reported to exhibit dislocation plasticity at meso-/macroscale
under ambient conditions. Finally, we discuss the mechanics of dislocations in
ceramics, aiming to establish a foundation for predicting and discovering
additional ceramics capable of room-temperature plastic deformation, thereby
advancing the development of prospective dislocation-based technologies.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 18:44:24 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Frisch",
"Alexander",
""
],
[
"Okafor",
"Chukwudalu",
""
],
[
"Preuß",
"Oliver",
""
],
[
"Zhang",
"Jiawen",
""
],
[
"Matsunaga",
"Katsuyuki",
""
],
[
"Nakamura",
"Atsutomo",
""
],
[
"Lu",
"Wenjun",
""
],
[
"Fang",
"Xufei",
""
]
] |
2504.04904 | Zilong Li | Zilong Li, Haotian Chen, Zhe Gao and Wei Chen | A kinetic CMA diagram | null | null | 10.1088/1741-4326/adca44 | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a kinetic Clemmow-Mullaly-Allis (CMA) diagram by systematically
analysing the kinetic effects on the wave propagation in a homogeneous thermal
plasma. The differences between the cold and kinetic CMA diagrams are outlined.
It is found that new boundaries for weakly damped left- and right-handed
circularly polarized waves are located above the ion and electron cyclotron
frequency lines in the kinetic CMA diagram. Additionally, Langmuir waves in the
kinetic CMA diagram occupy a specific region between the new Langmuir wave
boundary and the plasma frequency line, while in the cold CMA diagram, they
exist on the plasma frequency line. The extraordinary-Bernstein mode
transformation frequency lines in the kinetic CMA diagram replace the hybrid
resonant frequency lines of the cold CMA diagram, with discontinuities between
different cyclotron harmonics. These new boundaries partition the parameter
space in the kinetic CMA diagram differently, leading to new inverse wave
normal surfaces in the regions bounded by new boundaries. The kinetic CMA
diagram not only contributes to a basic understanding of wave properties in
thermal plasmas, but also can provide a powerful tool to explore new possible
propagation paths.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 10:22:29 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Li",
"Zilong",
""
],
[
"Chen",
"Haotian",
""
],
[
"Gao",
"Zhe",
""
],
[
"Chen",
"Wei",
""
]
] |
2504.05378 | Kevin Liguori | Mitrajyoti Ghosh, Kevin Liguori, Takemichi Okui, Kohsaku Tobioka | Neutrino properties from muonium-antimuonium mixing | Minor change to introduction, no change to analysis or conclusions | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | The nature of neutrino mass -- whether neutrinos are Dirac or Majorana
particles -- remains one of the central open questions in particle physics.
While observation of neutrinoless double beta decay would confirm the Majorana
case, the absence of a signal offers no definitive insight. In light of this,
we investigate muonium-antimuonium mixing -- proposed for further study at the
MACE experiment -- as an alternative probe of neutrino properties. We compute
the mixing amplitude in the Standard Model minimally extended to include
massive Dirac or Majorana neutrinos, and correct previous calculations by
properly treating the relevant infrared scales. As the GIM mechanism strongly
suppresses the Dirac contribution, we explore whether relaxing unitarity of the
PMNS matrix can enhance the mixing without obscuring neutrino properties.
Surprisingly, the answer to this question is negative. We also examine the
pseudo-Dirac case -- predominantly Dirac neutrinos with small Majorana masses
-- and find that this scenario can significantly enhance the mixing compared to
the pure Dirac case, especially for normal mass ordering.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 18:00:01 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 19:14:56 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Ghosh",
"Mitrajyoti",
""
],
[
"Liguori",
"Kevin",
""
],
[
"Okui",
"Takemichi",
""
],
[
"Tobioka",
"Kohsaku",
""
]
] |
2504.05521 | Andrei Neagu | Andrei Neagu, Fr\'ed\'eric Godin, Leila Kosseim | Deep Reinforcement Learning Algorithms for Option Hedging | null | null | null | null | q-fin.CP cs.AI cs.CE | http://creativecommons.org/licenses/by/4.0/ | Dynamic hedging is a financial strategy that consists in periodically
transacting one or multiple financial assets to offset the risk associated with
a correlated liability. Deep Reinforcement Learning (DRL) algorithms have been
used to find optimal solutions to dynamic hedging problems by framing them as
sequential decision-making problems. However, most previous work assesses the
performance of only one or two DRL algorithms, making an objective comparison
across algorithms difficult. In this paper, we compare the performance of eight
DRL algorithms in the context of dynamic hedging; Monte Carlo Policy Gradient
(MCPG), Proximal Policy Optimization (PPO), along with four variants of Deep
Q-Learning (DQL) and two variants of Deep Deterministic Policy Gradient (DDPG).
Two of these variants represent a novel application to the task of dynamic
hedging. In our experiments, we use the Black-Scholes delta hedge as a baseline
and simulate the dataset using a GJR-GARCH(1,1) model. Results show that MCPG,
followed by PPO, obtain the best performance in terms of the root
semi-quadratic penalty. Moreover, MCPG is the only algorithm to outperform the
Black-Scholes delta hedge baseline with the allotted computational budget,
possibly due to the sparsity of rewards in our environment.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 21:32:14 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 00:36:34 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Neagu",
"Andrei",
""
],
[
"Godin",
"Frédéric",
""
],
[
"Kosseim",
"Leila",
""
]
] |
2504.05554 | Karl Schwede | Neil Epstein, Peter M. McDonald, Rebecca R.G., Karl Schwede | Closure operations induced via resolutions of singularities in
characteristic zero | 45 pages. Numerous minor changes, corrections, clarifications, and
improvements. Notably, various equidimensional hypotheses added. Comments
welcome | null | null | null | math.AC math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Using the fact that the structure sheaf of a resolution of singularities, or
regular alteration, pushes forward to a Cohen-Macaulay complex in
characteristic zero with a differential graded algebra structure, we introduce
a tight-closure-like operation on ideals in characteristic zero using the
Koszul complex, which we call KH closure (Koszul-Hironaka). We prove it
satisfies various strong colon capturing properties and a version of the
Brian\c{c}on-Skoda theorem, and it behaves well under finite extensions. It
detects rational singularities and is tighter than characteristic zero tight
closure. Furthermore, its formation commutes with localization and it can be
computed effectively. On the other hand, the product of the KH closures of
ideals is not always contained in the KH closure of the product, as one might
expect.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 23:05:39 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:53:44 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Epstein",
"Neil",
""
],
[
"McDonald",
"Peter M.",
""
],
[
"G.",
"Rebecca R.",
""
],
[
"Schwede",
"Karl",
""
]
] |
2504.05675 | Saman Zare | William D. Hutchins, Saman Zare, Daniel Hirt, Elizabeth Golightly,
Patrick E. Hopkins | Infrared Phonon Thermoreflectance in Polar Dielectrics | null | null | null | null | physics.optics cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | In this work, we investigate dielectric materials for thermoreflectance-based
thermal sensing by extracting key optical parameters using
temperature-dependent spectroscopic ellipsometry in the mid-infrared regime.
Leveraging optical phonon resonances, we demonstrate that the thermoreflectance
coefficients in polar dielectrics rival, and in some cases exceed by an order
of magnitude, those observed in commonly used metals that are typically used as
temperature transducers in thermoreflectance measurements. We introduce a
transducer figure of merit (FOM) that combines pump absorption and probe
reflectance modulation at different wavelengths. Our findings reveal that
materials such as sapphire and aluminum nitride can outperform metals by up to
two orders of magnitude. These results position dielectric materials as
compelling candidates for next-generation thermal metrology, broadening the
design space for optical thermometry, with strong implications for
high-resolution thermal mapping and characterization of layered device
structures based on phonon probing.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 04:36:01 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 18:49:00 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Hutchins",
"William D.",
""
],
[
"Zare",
"Saman",
""
],
[
"Hirt",
"Daniel",
""
],
[
"Golightly",
"Elizabeth",
""
],
[
"Hopkins",
"Patrick E.",
""
]
] |
2504.06235 | Shahryar Zehtabi | Shahryar Zehtabi, Dong-Jun Han, Seyyedali Hosseinalipour, Christopher
G. Brinton | Decentralized Federated Domain Generalization with Style Sharing: A
Formal Modeling and Convergence Analysis | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Much of the federated learning (FL) literature focuses on settings where
local dataset statistics remain the same between training and testing time.
Recent advances in domain generalization (DG) aim to use data from source
(training) domains to train a model that generalizes well to data from unseen
target (testing) domains. In this paper, we are motivated by two major gaps in
existing work on FL and DG: (1) the lack of formal mathematical analysis of DG
objectives and training processes; and (2) DG research in FL being limited to
the conventional star-topology architecture. Addressing the second gap, we
develop $\textit{Decentralized Federated Domain Generalization with Style
Sharing}$ ($\texttt{StyleDDG}$), a fully decentralized DG algorithm designed to
allow devices in a peer-to-peer network to achieve DG based on sharing style
information inferred from their datasets. Additionally, we fill the first gap
by providing the first systematic approach to mathematically analyzing
style-based DG training optimization. We cast existing centralized DG
algorithms within our framework, and employ their formalisms to model
$\texttt{StyleDDG}$. Based on this, we obtain analytical conditions under which
a sub-linear convergence rate of $\texttt{StyleDDG}$ can be obtained. Through
experiments on two popular DG datasets, we demonstrate that $\texttt{StyleDDG}$
can obtain significant improvements in accuracy across target domains with
minimal added communication overhead compared to decentralized gradient methods
that do not employ style sharing.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 17:32:56 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 08:52:03 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zehtabi",
"Shahryar",
""
],
[
"Han",
"Dong-Jun",
""
],
[
"Hosseinalipour",
"Seyyedali",
""
],
[
"Brinton",
"Christopher G.",
""
]
] |
2504.06345 | Adam Finley Dr | Adam J. Finley, Allan Sacha Brun, Antoine Strugarek and Barbara Perri | A prolific solar flare factory. Near-continuous monitoring of an active
region nest with Solar Orbiter | Accepted to A&A. 12 Pages + Appendix. 11 Figures + 5 Appendix Figures | null | null | null | astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The properties of active region nests, which are locations on the Sun with
recurring flux emergence, are poorly constrained by observations from Earth
alone. ESA's Solar Orbiter now monitors the far side of the Sun for extended
periods of time. This facilitates observations of the entire Sun. We combined
observations from near-Earth satellites and Solar Orbiter to evaluate the
contribution of a long-lived active region nest to the global flaring activity
of the Sun. We identified a location in Carrington coordinates with episodic
bursts of flux emergence throughout 2022. The combined observations allowed a
nearly continuous monitoring of this region from April to October, that is,
during its most active period. GOES and Solar Orbiter/STIX were used to compare
its flaring activity to that of the entire Sun. The region morphology was
extracted from SDO/AIA and Solar Orbiter/EUI extreme-ultraviolet images and was
combined with magnetic field measurements from SDO/HMI and Solar Orbiter/PHI to
assess its unsigned magnetic flux. The active region nest grew in complexity
from January to May due to repeated flux emergence events. The peak unsigned
magnetic flux was $5\times10^{22}$Mx. The region caused 40% of the observed
solar flares in 2022, including five months in which it produced 50-70% of all
flares over the entire Sun (in the nearly continuous monitoring window). Of the
17 complex flaring NOAA active regions in 2022, this region contained 10, but
they occupied less than 20% of the area in the active latitudes. Active region
nests can maintain a high rate of flaring activity for several solar rotations
and are more likely to produce complex active regions that can trigger X-class
solar flares. Improving the identification and monitoring of long-lived active
region nests would benefit space weather forecasts in the short to medium term.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 18:00:05 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:57:08 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Finley",
"Adam J.",
""
],
[
"Brun",
"Allan Sacha",
""
],
[
"Strugarek",
"Antoine",
""
],
[
"Perri",
"Barbara",
""
]
] |
2504.06469 | Sergio Lopez-Caceres | Sergio Lopez-Caceres and Daniel Santiago-Gonzalez | AI-Assisted Transport of Radioactive Ion Beams | 6 pages, 6 figures; Section headings added for clarity.
Implementation and Results sections expanded. Minor revisions to Abstract and
to Summary and Conclusion | null | null | null | physics.acc-ph cs.AI nucl-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Beams of radioactive heavy ions allow researchers to study rare and unstable
atomic nuclei, shedding light into the internal structure of exotic nuclei and
on how chemical elements are formed in stars. However, the extraction and
transport of radioactive beams rely on time-consuming expert-driven tuning
methods, where hundreds of parameters are manually optimized. Here, we
introduce a system that employs Artificial Intelligence (AI), specifically
utilizing Bayesian Optimization, to assist in the transport process of
radioactive beams. We apply our methodology to real-life scenarios showing
advantages when compared with standard tuning methods. This AI-assisted
approach can be extended to other radioactive beam facilities around the world
to improve operational efficiency and enhance scientific output.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 22:21:54 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 00:25:50 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Lopez-Caceres",
"Sergio",
""
],
[
"Santiago-Gonzalez",
"Daniel",
""
]
] |
2504.06562 | Longguang Zhong | Longguang Zhong, Fanqi Wan, Ziyi Yang, Guosheng Liang, Tianyuan Shi,
Xiaojun Quan | FuseRL: Dense Preference Optimization for Heterogeneous Model Fusion | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Heterogeneous model fusion enhances the performance of LLMs by integrating
the knowledge and capabilities of multiple structurally diverse models.
However, existing approaches often rely solely on selecting the best output for
each prompt from source models, which underutilizes their full potential due to
limited source knowledge and results in sparse optimization signals. To address
this limitation, we propose FuseRL, a novel two-stage framework comprising
FuseSFT and FusePO to maximize the utilization of source LLMs. FuseSFT
establishes a robust initialization by integrating the strengths of
heterogeneous source models through weighted supervised fine-tuning (SFT) on
diverse outputs for each prompt. FusePO optimizes weighted preferences based on
the outputs of multiple source models to enable superior alignment performance.
Extensive experiments demonstrate the effectiveness of our framework across
various preference alignment methods, including RLOO, DPO, and SimPO. Using
Llama-3.1-8B-Instruct as the target model, our approach achieves
state-of-the-art performance among 8B LLMs on the AlpacaEval-2 and Arena-Hard
benchmarks. Further analysis suggests that FuseSFT regularizes the training
process to reduce overfitting, while FusePO introduces dense and diverse
signals for preference optimization.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 03:51:53 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:49:43 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Zhong",
"Longguang",
""
],
[
"Wan",
"Fanqi",
""
],
[
"Yang",
"Ziyi",
""
],
[
"Liang",
"Guosheng",
""
],
[
"Shi",
"Tianyuan",
""
],
[
"Quan",
"Xiaojun",
""
]
] |
2504.06599 | Morimitsu Tanimoto | Morimitsu Tanimoto and Tsutomu T. Yanagida | Axionless Solution to the Strong CP Problem -- two-zeros textures of the
quark and lepton mass matrices and neutrino CP violation -- | Some numerical results are added | null | null | null | hep-ph | http://creativecommons.org/publicdomain/zero/1.0/ | CP invariance is a very attractive solution to the strong CP problem in QCD.
This solution requires the vanishing ${\rm arg}\,[{\rm det}\, M_d\, {\rm det}
M_u]$, where the $M_d$ and $M_u$ are the mass matrices for the down- and
up-type quarks. It happens if we have several zeros in the quark mass matrices.
We proceed a systematic construction, in this paper, of two zeros textures for
the down-type quark mass matrix while the mass matrix for the up-type quarks is
always diagonal. We find only three types of the mass matrices can explain the
observed CKM matrix, the masses of the quarks and the charged leptons and the
small enough vacuum angle $\theta < 10^{-10}$. We extend the mass construction
to the neutrino sector and derive predictions on the CP violating parameter
$\delta_{CP}$ in the neutrino oscillation and the mass parameter
$m_{\beta\beta}$. It is extremely remarkable that the normal (NH) and inverted
(IH) hierarchies in the neutrino masses are equally possible in the case where
we introduce only two right-handed neutrinos $N$s. Furthermore, we have a
strict prediction on the $\delta_{CP} \simeq 200^\circ$ or $250^\circ$ in the
NH case. If it is the case we can naturally explain the positive sign of the
baryon asymmetry in the present universe.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 05:52:11 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 05:01:09 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Tanimoto",
"Morimitsu",
""
],
[
"Yanagida",
"Tsutomu T.",
""
]
] |
2504.06755 | Zhihui Li | Li Yu, Zhihui Li, Jimin Xiao, and Moncef Gabbouj | FANeRV: Frequency Separation and Augmentation based Neural
Representation for Video | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural representations for video (NeRV) have gained considerable attention
for their strong performance across various video tasks. However, existing NeRV
methods often struggle to capture fine spatial details, resulting in vague
reconstructions. In this paper, we present a Frequency Separation and
Augmentation based Neural Representation for video (FANeRV), which addresses
these limitations with its core Wavelet Frequency Upgrade Block. This block
explicitly separates input frames into high and low-frequency components using
discrete wavelet transform, followed by targeted enhancement using specialized
modules. Finally, a specially designed gated network effectively fuses these
frequency components for optimal reconstruction. Additionally, convolutional
residual enhancement blocks are integrated into the later stages of the network
to balance parameter distribution and improve the restoration of high-frequency
details. Experimental results demonstrate that FANeRV significantly improves
reconstruction performance and excels in multiple tasks, including video
compression, inpainting, and interpolation, outperforming existing NeRV
methods.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 10:19:35 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:49:23 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yu",
"Li",
""
],
[
"Li",
"Zhihui",
""
],
[
"Xiao",
"Jimin",
""
],
[
"Gabbouj",
"Moncef",
""
]
] |
2504.06778 | Michael Finkelson | Roi Benita, Michael Finkelson, Tavi Halperin, Gleb Sterkin, Yossi Adi | CAFA: a Controllable Automatic Foley Artist | Renamed paper to "CAFA: a Controllable Automatic Foley Artist" from
"Controllable Automatic Foley Artist". Updated link to demo page | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Foley is a key element in video production, refers to the process of adding
an audio signal to a silent video while ensuring semantic and temporal
alignment. In recent years, the rise of personalized content creation and
advancements in automatic video-to-audio models have increased the demand for
greater user control in the process. One possible approach is to incorporate
text to guide audio generation. While supported by existing methods, challenges
remain in ensuring compatibility between modalities, particularly when the text
introduces additional information or contradicts the sounds naturally inferred
from the visuals. In this work, we introduce CAFA (Controllable Automatic Foley
Artist) a video-and-text-to-audio model that generates semantically and
temporally aligned audio for a given video, guided by text input. CAFA is built
upon a text-to-audio model and integrates video information through a modality
adapter mechanism. By incorporating text, users can refine semantic details and
introduce creative variations, guiding the audio synthesis beyond the expected
video contextual cues. Experiments show that besides its superior quality in
terms of semantic alignment and audio-visual synchronization the proposed
method enable high textual controllability as demonstrated in subjective and
objective evaluations.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 10:58:54 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:09:20 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 12:36:56 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Benita",
"Roi",
""
],
[
"Finkelson",
"Michael",
""
],
[
"Halperin",
"Tavi",
""
],
[
"Sterkin",
"Gleb",
""
],
[
"Adi",
"Yossi",
""
]
] |
2504.06852 | Juncen Li | Jun-Cen Li, An Du | Quantum controlling and the topological properties of the magnon
photo-transport in two-dimensional collinear ferromagnet | 21 pages, 8 figures | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In our work, we study magnon transport induced by light through
Aharonov-Casher (AC) effect, including magnon spin photocurrent (MSPC) and
magnon energy photocurrent (MEPC). Firstly, we regard the effect of the
electric field on the magnon through the AC effect as a perturbation. Then we
derived the expressions of MSPC and MEPC in two-dimensional collinear
ferromagnetic system. And we apply our theory to the two-dimension
ferromagnetic Hexagonal and Kagome lattice. We find that the optical frequency
and the relaxation time of the material can be used to control the
photo-transport of magnons. In addition, under the condition of low light
frequncy and infinite relaxation time, the longitudinal magnon photo-transport
is related to the topological property of the magnon system.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 13:06:39 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 08:13:08 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 12:33:48 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Li",
"Jun-Cen",
""
],
[
"Du",
"An",
""
]
] |
2504.06932 | David Schaurecker | David Schaurecker, David Wozabal, Nils L\"ohndorf, Thorsten Staake | Maximizing Battery Storage Profits via High-Frequency Intraday Trading | null | null | null | null | q-fin.TR cs.SY eess.SY math.OC | http://creativecommons.org/licenses/by/4.0/ | Maximizing revenue for grid-scale battery energy storage systems in
continuous intraday electricity markets requires strategies that are able to
seize trading opportunities as soon as new information arrives. This paper
introduces and evaluates an automated high-frequency trading strategy for
battery energy storage systems trading on the intraday market for power while
explicitly considering the dynamics of the limit order book, market rules, and
technical parameters. The standard rolling intrinsic strategy is adapted for
continuous intraday electricity markets and solved using a dynamic programming
approximation that is two to three orders of magnitude faster than an exact
mixed-integer linear programming solution. A detailed backtest over a full year
of German order book data demonstrates that the proposed dynamic programming
formulation does not reduce trading profits and enables the policy to react to
every relevant order book update, enabling realistic rapid backtesting. Our
results show the significant revenue potential of high-frequency trading: our
policy earns 58% more than when re-optimizing only once every hour and 14% more
than when re-optimizing once per minute, highlighting that profits critically
depend on trading speed. Furthermore, we leverage the speed of our algorithm to
train a parametric extension of the rolling intrinsic, increasing yearly
revenue by 8.4% out of sample.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 14:38:09 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 07:04:13 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Schaurecker",
"David",
""
],
[
"Wozabal",
"David",
""
],
[
"Löhndorf",
"Nils",
""
],
[
"Staake",
"Thorsten",
""
]
] |
2504.06998 | J\"orn Zimmerling | Vladimir Druskin and J\"orn Zimmerling | A Krylov projection algorithm for large symmetric matrices with dense
spectra | Block Lanczos, Quadrature, Transfer function, Kre\u{i}n-Nudelman,
Hermite-Pad\'e | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the approximation of $B^T (A+sI)^{-1} B$ for large s.p.d.
$A\in\mathbb{R}^{n\times n}$ with dense spectrum and $B\in\mathbb{R}^{n\times
p}$, $p\ll n$. We target the computations of Multiple-Input Multiple-Output
(MIMO) transfer functions for large-scale discretizations of problems with
continuous spectral measures, such as linear time-invariant (LTI) PDEs on
unbounded domains. Traditional Krylov methods, such as the Lanczos or CG
algorithm, are known to be optimal for the computation of $(A+sI)^{-1}B$ with
real positive $s$, resulting in an adaptation to the distinctively discrete and
nonuniform spectra. However, the adaptation is damped for matrices with dense
spectra. It was demonstrated in [Zimmerling, Druskin, Simoncini, Journal of
Scientific Computing 103(1), 5 (2025)] that averaging Gau{\ss} and Gau\ss
-Radau quadratures computed using the block-Lanczos method significantly
reduces approximation errors for such problems. Here, we introduce an adaptive
Kre\u{i}n-Nudelman extension to the (block) Lanczos recursions, allowing
further acceleration at negligible $o(n)$ cost. Similar to the Gau\ss -Radau
quadrature, a low-rank modification is applied to the (block) Lanczos matrix.
However, unlike the Gau\ss -Radau quadrature, this modification depends on
$\sqrt{s}$ and can be considered in the framework of the Hermite-Pad\'e
approximants, which are known to be efficient for problems with branch-cuts,
that can be good approximations to dense spectral intervals. Numerical results
for large-scale discretizations of heat-diffusion and quasi-magnetostatic
Maxwell's operators in unbounded domains confirm the efficiency of the proposed
approach.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 16:09:34 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 12:54:50 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Druskin",
"Vladimir",
""
],
[
"Zimmerling",
"Jörn",
""
]
] |
2504.07048 | Avinash Kumar | Avinash Kumar, Meng Wang, Chenxu Liu, Ang Li, Prashant J. Nair,
Poulami Das | Context Switching for Secure Multi-programming of Near-Term Quantum
Computers | null | null | null | null | cs.CR cs.ET | http://creativecommons.org/licenses/by/4.0/ | Multi-programming quantum computers improve device utilization and
throughput. However, crosstalk from concurrent two-qubit CNOT gates poses
security risks, compromising the fidelity and output of co-running victim
programs. We design Zero Knowledge Tampering Attacks (ZKTAs), using which
attackers can exploit crosstalk without knowledge of the hardware error
profile. ZKTAs can alter victim program outputs in 40% of cases on commercial
systems.
We identify that ZKTAs succeed because the attacker's program consistently
runs with the same victim program in a fixed context. To mitigate this, we
propose QONTEXTS: a context-switching technique that defends against ZKTAs by
running programs across multiple contexts, each handling only a subset of
trials. QONTEXTS uses multi-programming with frequent context switching while
identifying a unique set of programs for each context. This helps limit only a
fraction of execution to ZKTAs. We enhance QONTEXTS with attack detection
capabilities that compare the distributions from different contexts against
each other to identify noisy contexts executed with ZKTAs. Our evaluations on
real IBMQ systems show that QONTEXTS increases program resilience by three
orders of magnitude and fidelity by 1.33$\times$ on average. Moreover, QONTEXTS
improves throughput by 2$\times$, advancing security in multi-programmed
environments.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 17:05:16 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 03:48:54 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 04:34:36 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Kumar",
"Avinash",
""
],
[
"Wang",
"Meng",
""
],
[
"Liu",
"Chenxu",
""
],
[
"Li",
"Ang",
""
],
[
"Nair",
"Prashant J.",
""
],
[
"Das",
"Poulami",
""
]
] |
2504.07521 | Yifei Dong | Yuxiang Lin, Jingdong Sun, Zhi-Qi Cheng, Jue Wang, Haomin Liang,
Zebang Cheng, Yifei Dong, Jun-Yan He, Xiaojiang Peng, Xian-Sheng Hua | Why We Feel: Breaking Boundaries in Emotional Reasoning with Multimodal
Large Language Models | Accepted at CVPR Workshop NEXD 2025. 21 pages, Project:
https://github.com/Lum1104/EIBench | null | null | null | cs.AI cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most existing emotion analysis emphasizes which emotion arises (e.g., happy,
sad, angry) but neglects the deeper why. We propose Emotion Interpretation
(EI), focusing on causal factors-whether explicit (e.g., observable objects,
interpersonal interactions) or implicit (e.g., cultural context, off-screen
events)-that drive emotional responses. Unlike traditional emotion recognition,
EI tasks require reasoning about triggers instead of mere labeling. To
facilitate EI research, we present EIBench, a large-scale benchmark
encompassing 1,615 basic EI samples and 50 complex EI samples featuring
multifaceted emotions. Each instance demands rationale-based explanations
rather than straightforward categorization. We further propose a Coarse-to-Fine
Self-Ask (CFSA) annotation pipeline, which guides Vision-Language Models
(VLLMs) through iterative question-answer rounds to yield high-quality labels
at scale. Extensive evaluations on open-source and proprietary large language
models under four experimental settings reveal consistent performance
gaps-especially for more intricate scenarios-underscoring EI's potential to
enrich empathetic, context-aware AI applications. Our benchmark and methods are
publicly available at: https://github.com/Lum1104/EIBench, offering a
foundation for advanced multimodal causal analysis and next-generation
affective computing.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 07:33:49 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:34:26 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Lin",
"Yuxiang",
""
],
[
"Sun",
"Jingdong",
""
],
[
"Cheng",
"Zhi-Qi",
""
],
[
"Wang",
"Jue",
""
],
[
"Liang",
"Haomin",
""
],
[
"Cheng",
"Zebang",
""
],
[
"Dong",
"Yifei",
""
],
[
"He",
"Jun-Yan",
""
],
[
"Peng",
"Xiaojiang",
""
],
[
"Hua",
"Xian-Sheng",
""
]
] |
2504.07717 | Yang Jiao | Yang Jiao, Xiaodong Wang, Kai Yang | PR-Attack: Coordinated Prompt-RAG Attacks on Retrieval-Augmented
Generation in Large Language Models via Bilevel Optimization | Accepted at SIGIR 2025 | null | null | null | cs.CR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) have demonstrated remarkable performance across
a wide range of applications, e.g., medical question-answering, mathematical
sciences, and code generation. However, they also exhibit inherent limitations,
such as outdated knowledge and susceptibility to hallucinations.
Retrieval-Augmented Generation (RAG) has emerged as a promising paradigm to
address these issues, but it also introduces new vulnerabilities. Recent
efforts have focused on the security of RAG-based LLMs, yet existing attack
methods face three critical challenges: (1) their effectiveness declines
sharply when only a limited number of poisoned texts can be injected into the
knowledge database, (2) they lack sufficient stealth, as the attacks are often
detectable by anomaly detection systems, which compromises their effectiveness,
and (3) they rely on heuristic approaches to generate poisoned texts, lacking
formal optimization frameworks and theoretic guarantees, which limits their
effectiveness and applicability. To address these issues, we propose
coordinated Prompt-RAG attack (PR-attack), a novel optimization-driven attack
that introduces a small number of poisoned texts into the knowledge database
while embedding a backdoor trigger within the prompt. When activated, the
trigger causes the LLM to generate pre-designed responses to targeted queries,
while maintaining normal behavior in other contexts. This ensures both high
effectiveness and stealth. We formulate the attack generation process as a
bilevel optimization problem leveraging a principled optimization framework to
develop optimal poisoned texts and triggers. Extensive experiments across
diverse LLMs and datasets demonstrate the effectiveness of PR-Attack, achieving
a high attack success rate even with a limited number of poisoned texts and
significantly improved stealth compared to existing methods.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 13:09:50 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 02:01:42 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Jiao",
"Yang",
""
],
[
"Wang",
"Xiaodong",
""
],
[
"Yang",
"Kai",
""
]
] |
2504.07774 | Xiaoning Wu | Xiaokai He, Xiaoning Wu and Naqing Xie | Bondi Mass, Memory Effect and Balance Law of Polyhomogeneous Spacetime | 36 pages and no figure | null | null | null | gr-qc math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spacetimes with metrics admitting an expansion in terms of a combination of
powers of 1/r and ln r are known as polyhomogeneous spacetimes. The asymptotic
behaviour of the Newman-Penrose quantities for these spacetimes is presented
under certain gauges. The Bondi mass is revisited via the Iyer-Wald formalism.
The memory effect of the gravitational radiation in the polyhomogeneous
spacetimes is also discussed. It is found that the appearance of the
logarithmic terms does not affect the balance law and it remains unchanged as
the one of spacetimes with metrics admitting an expansion in terms of powers of
1/r.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 14:14:25 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:22:27 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"He",
"Xiaokai",
""
],
[
"Wu",
"Xiaoning",
""
],
[
"Xie",
"Naqing",
""
]
] |
2504.07895 | Bryanne McDonough | Bryanne McDonough and Alexander Poulin | Magnetic Fields of Satellite Galaxies Stronger Than Comparable Centrals
in TNG100 | 3 pages, 1 figure; submitted to RNAAS where data behind the figure
will be available; revised version addresses a small code error | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Magnetic fields exist in and around galaxies, but the properties of these
fields have not been fully explored due to the challenges inherent in observing
and modeling them. In this Note, we explore the differences in magnetic field
strength of central and satellite galaxies from the magnetohydrodynamic TNG100
simulation. We find that on average, magnetic fields in satellite galaxies are
roughly an order of magnitude stronger than those of central galaxies with
comparable masses. The difference is greater for satellites that have already
approached within $1 R_{200}$ of their host galaxies. These results indicate
that magnetic fields in satellite galaxies are amplified by environmental
processes as they fall into a host halo.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 16:09:50 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 20:54:46 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"McDonough",
"Bryanne",
""
],
[
"Poulin",
"Alexander",
""
]
] |
2504.07940 | Rundong Luo | Rundong Luo, Matthew Wallingford, Ali Farhadi, Noah Snavely, Wei-Chiu
Ma | Beyond the Frame: Generating 360{\deg} Panoramic Videos from Perspective
Videos | Project page: https://red-fairy.github.io/argus/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | 360{\deg} videos have emerged as a promising medium to represent our dynamic
visual world. Compared to the "tunnel vision" of standard cameras, their
borderless field of view offers a more complete perspective of our
surroundings. While existing video models excel at producing standard videos,
their ability to generate full panoramic videos remains elusive. In this paper,
we investigate the task of video-to-360{\deg} generation: given a perspective
video as input, our goal is to generate a full panoramic video that is
consistent with the original video. Unlike conventional video generation tasks,
the output's field of view is significantly larger, and the model is required
to have a deep understanding of both the spatial layout of the scene and the
dynamics of objects to maintain spatio-temporal consistency. To address these
challenges, we first leverage the abundant 360{\deg} videos available online
and develop a high-quality data filtering pipeline to curate pairwise training
data. We then carefully design a series of geometry- and motion-aware
operations to facilitate the learning process and improve the quality of
360{\deg} video generation. Experimental results demonstrate that our model can
generate realistic and coherent 360{\deg} videos from in-the-wild perspective
video. In addition, we showcase its potential applications, including video
stabilization, camera viewpoint control, and interactive visual question
answering.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 17:51:38 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:35:15 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Luo",
"Rundong",
""
],
[
"Wallingford",
"Matthew",
""
],
[
"Farhadi",
"Ali",
""
],
[
"Snavely",
"Noah",
""
],
[
"Ma",
"Wei-Chiu",
""
]
] |
2504.08141 | Oliver Knitter | Saibal De, Oliver Knitter, Rohan Kodati, Paramsothy Jayakumar, James
Stokes, Shravan Veerapaneni | Variational quantum and neural quantum states algorithms for the linear
complementarity problem | 13 pages, 5 figures, to appear in Philosophical Transactions of the
Royal Society A | null | null | null | cs.CE cs.LG quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Variational quantum algorithms (VQAs) are promising hybrid quantum-classical
methods designed to leverage the computational advantages of quantum computing
while mitigating the limitations of current noisy intermediate-scale quantum
(NISQ) hardware. Although VQAs have been demonstrated as proofs of concept,
their practical utility in solving real-world problems -- and whether
quantum-inspired classical algorithms can match their performance -- remains an
open question. We present a novel application of the variational quantum linear
solver (VQLS) and its classical neural quantum states-based counterpart, the
variational neural linear solver (VNLS), as key components within a minimum map
Newton solver for a complementarity-based rigid body contact model. We
demonstrate using the VNLS that our solver accurately simulates the dynamics of
rigid spherical bodies during collision events. These results suggest that
quantum and quantum-inspired linear algebra algorithms can serve as viable
alternatives to standard linear algebra solvers for modeling certain physical
systems.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 22:03:14 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 10:44:22 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"De",
"Saibal",
""
],
[
"Knitter",
"Oliver",
""
],
[
"Kodati",
"Rohan",
""
],
[
"Jayakumar",
"Paramsothy",
""
],
[
"Stokes",
"James",
""
],
[
"Veerapaneni",
"Shravan",
""
]
] |
2504.08260 | Yonchanok Khaokaew | Yonchanok Khaokaew, Flora D. Salim, Andreas Z\"ufle, Hao Xue, Taylor
Anderson, C. Raina MacIntyre, Matthew Scotch, David J Heslop | Evaluating the Bias in LLMs for Surveying Opinion and Decision Making in
Healthcare | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative agents have been increasingly used to simulate human behaviour in
silico, driven by large language models (LLMs). These simulacra serve as
sandboxes for studying human behaviour without compromising privacy or safety.
However, it remains unclear whether such agents can truly represent real
individuals. This work compares survey data from the Understanding America
Study (UAS) on healthcare decision-making with simulated responses from
generative agents. Using demographic-based prompt engineering, we create
digital twins of survey respondents and analyse how well different LLMs
reproduce real-world behaviours. Our findings show that some LLMs fail to
reflect realistic decision-making, such as predicting universal vaccine
acceptance. However, Llama 3 captures variations across race and Income more
accurately but also introduces biases not present in the UAS data. This study
highlights the potential of generative agents for behavioural research while
underscoring the risks of bias from both LLMs and prompting strategies.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 05:11:40 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 01:36:57 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Khaokaew",
"Yonchanok",
""
],
[
"Salim",
"Flora D.",
""
],
[
"Züfle",
"Andreas",
""
],
[
"Xue",
"Hao",
""
],
[
"Anderson",
"Taylor",
""
],
[
"MacIntyre",
"C. Raina",
""
],
[
"Scotch",
"Matthew",
""
],
[
"Heslop",
"David J",
""
]
] |
2504.08401 | Abdo Abouelrous | Abdo Abouelrous, Laurens Bliek, Adriana F. Gabor, Yaoxin Wu, Yingqian
Zhang | Graph Reduction with Unsupervised Learning in Column Generation: A
Routing Application | 22 pages, 4 figures, 5 tables | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Column Generation (CG) is a popular method dedicated to enhancing
computational efficiency in large scale Combinatorial Optimization (CO)
problems. It reduces the number of decision variables in a problem by solving a
pricing problem. For many CO problems, the pricing problem is an Elementary
Shortest Path Problem with Resource Constraints (ESPPRC). Large ESPPRC
instances are difficult to solve to near-optimality. Consequently, we use a
Graph neural Network (GNN) to reduces the size of the ESPPRC such that it
becomes computationally tractable with standard solving techniques. Our GNN is
trained by Unsupervised Learning and outputs a distribution for the arcs to be
retained in the reduced PP. The reduced PP is solved by a local search that
finds columns with large reduced costs and speeds up convergence. We apply our
method on a set of Capacitated Vehicle Routing Problems with Time Windows and
show significant improvements in convergence compared to simple reduction
techniques from the literature. For a fixed computational budget, we improve
the objective values by over 9\% for larger instances. We also analyze the
performance of our CG algorithm and test the generalization of our method to
different classes of instances than the training data.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:08:38 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 09:01:57 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Abouelrous",
"Abdo",
""
],
[
"Bliek",
"Laurens",
""
],
[
"Gabor",
"Adriana F.",
""
],
[
"Wu",
"Yaoxin",
""
],
[
"Zhang",
"Yingqian",
""
]
] |
2504.08876 | Santiago Nunez-Corrales | Francini Corrales-Garro and Danny Valerio-Ram\'irez and Santiago
N\'u\~nez-Corrales | Is Productivity in Quantum Programming Equivalent to Expressiveness? | 11 pages, 6 figures | null | null | null | quant-ph cs.PL cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The expressiveness of quantum programming languages plays a crucial role in
the efficient and comprehensible representation of quantum algorithms. Unlike
classical programming languages, which offer mature and well-defined
abstraction mechanisms, quantum languages must integrate cognitively
challenging concepts such as superposition, interference and entanglement while
maintaining clarity and usability. However, identifying and characterizing
differences in expressiveness between quantum programming paradigms remains an
open area of study. Our work investigates the landscape of expressiveness
through a comparative analysis of hosted quantum programming languages such as
Qiskit, Cirq, Qrisp, and quAPL, and standalone languages including Q# and Qmod.
We focused on evaluating how different quantum programming languages support
the implementation of core quantum algorithms -- Deutsch-Jozsa, Simon,
Bernstein-Vazirani, and Grover -- using expressiveness metrics: Lines of Code
(LOC), Cyclomatic Complexity (CC), and Halstead Complexity (HC) metrics as
proxies for developer productivity. Our findings suggest that different quantum
programming paradigms offer distinct trade-offs between expressiveness and
productivity, highlighting the importance of language design in quantum
software development.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:04:16 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 14:17:26 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Corrales-Garro",
"Francini",
""
],
[
"Valerio-Ramírez",
"Danny",
""
],
[
"Núñez-Corrales",
"Santiago",
""
]
] |
2504.08889 | Jillian Rastinejad | J. C. Rastinejad (Northwestern), A. J. Levan, P. G. Jonker, C. D.
Kilpatrick, C. L. Fryer, N. Sarin, B. P. Gompertz, C. Liu, R. A. J.
Eyles-Ferris, W. Fong, E. Burns, J. H. Gillanders, I. Mandel, D. B. Malesani,
P. T. O'Brien, N. R. Tanvir, K. Ackley, A. Aryan, F. E. Bauer, S. Bloemen, T.
de Boer, C. R. Bom, J. A. Chacon, K. Chambers, T.-W. Chen, A. A. Chrimes, J.
N. D. van Dalen, V. D'Elia, M. De Pasquale, M. D. Fulton, P. J. Groot, R.
Gupta, D. H. Hartmann, A. P. C. van Hoof, M. E. Huber, L. Izzo, W.
Jacobson-Galan, P. Jakobsson, A. Kong, T. Laskar, T. B. Lowe, E. A. Magnier,
E. Maiorano, A. Martin-Carrillo, L. Mas-Ribas, D. Mata Sanchez, M. Nicholl,
C. J. Nixon, S. R. Oates, G. Paek, J. Palmerio, D. Paris, D. L. A. Pieterse,
G. Pugliese, J. A. Quirola Vasquez, J. van Roestel, A. Rossi, A. Rouco
Escorial. R. Salvaterra, B. Schneider, S. J. Smartt, K. Smith, I. A. Smith,
S. Srivastav, M. A. P. Torres, C. Ventura, P. Vreeswijk, R. Wainscoat, Y.-J.
Yang and S. Yang | EP 250108a/SN 2025kg: Observations of the most nearby Broad-Line Type Ic
Supernova following an Einstein Probe Fast X-ray Transient | Submitted to ApJL. 28 pages including 10 figures and 3 tables | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | With a small sample of fast X-ray transients (FXTs) with multi-wavelength
counterparts discovered to date, the progenitors of FXTs and their connections
to gamma-ray bursts (GRBs) and supernovae (SNe) remain ambiguous. Here, we
present photometric and spectroscopic observations of SN 2025kg, the supernova
counterpart to the FXT EP 250108a. At $z=0.17641$, this is the closest known SN
discovered following an Einstein Probe (EP) FXT. We show that SN 2025kg's
optical spectra reveal the hallmark features of a broad-lined Type Ic SN. Its
light curve evolution and expansion velocities are also comparable to those of
GRB-SNe, including SN 1998bw, and several past FXT SNe. We present JWST/NIRSpec
spectroscopy taken around SN 2025kg's maximum light, and find weak absorption
due to He I $\lambda 1.0830, \lambda 2.0581$ $\mu$m and a broad, unidentified
feature at $\sim$ 4-4.5 $\mu$m. Further, we observe clear evidence for
broadened H$\alpha$ in optical data at 42.5 days that is not detected at other
epochs, indicating interaction with hydrogen-rich material. From its light
curve, we derive a $^{56}$Ni mass of 0.2 - 0.6 $M_{\odot}$. Together with our
companion paper (Eyles-Ferris et al. 2025), our broadband data of EP 250108a/SN
2025kg are consistent with a trapped or low energy ($\lesssim 10^{51}$ ergs)
jet-driven explosion from a collapsar with a zero-age main sequence mass of
15-30 $M_{\odot}$. Finally, we show that the sample of EP FXT SNe support past
rate estimates that low-luminosity jets seen through FXTs are more common than
successful (GRB) jets, and that similar FXT-like signatures are likely present
in at least a few percent of the brightest Ic-BL SNe.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:08 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:55:46 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 14:55:19 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Rastinejad",
"J. C.",
"",
"Northwestern"
],
[
"Levan",
"A. J.",
""
],
[
"Jonker",
"P. G.",
""
],
[
"Kilpatrick",
"C. D.",
""
],
[
"Fryer",
"C. L.",
""
],
[
"Sarin",
"N.",
""
],
[
"Gompertz",
"B. P.",
""
],
[
"Liu",
"C.",
""
],
[
"Eyles-Ferris",
"R. A. J.",
""
],
[
"Fong",
"W.",
""
],
[
"Burns",
"E.",
""
],
[
"Gillanders",
"J. H.",
""
],
[
"Mandel",
"I.",
""
],
[
"Malesani",
"D. B.",
""
],
[
"O'Brien",
"P. T.",
""
],
[
"Tanvir",
"N. R.",
""
],
[
"Ackley",
"K.",
""
],
[
"Aryan",
"A.",
""
],
[
"Bauer",
"F. E.",
""
],
[
"Bloemen",
"S.",
""
],
[
"de Boer",
"T.",
""
],
[
"Bom",
"C. R.",
""
],
[
"Chacon",
"J. A.",
""
],
[
"Chambers",
"K.",
""
],
[
"Chen",
"T. -W.",
""
],
[
"Chrimes",
"A. A.",
""
],
[
"van Dalen",
"J. N. D.",
""
],
[
"D'Elia",
"V.",
""
],
[
"De Pasquale",
"M.",
""
],
[
"Fulton",
"M. D.",
""
],
[
"Groot",
"P. J.",
""
],
[
"Gupta",
"R.",
""
],
[
"Hartmann",
"D. H.",
""
],
[
"van Hoof",
"A. P. C.",
""
],
[
"Huber",
"M. E.",
""
],
[
"Izzo",
"L.",
""
],
[
"Jacobson-Galan",
"W.",
""
],
[
"Jakobsson",
"P.",
""
],
[
"Kong",
"A.",
""
],
[
"Laskar",
"T.",
""
],
[
"Lowe",
"T. B.",
""
],
[
"Magnier",
"E. A.",
""
],
[
"Maiorano",
"E.",
""
],
[
"Martin-Carrillo",
"A.",
""
],
[
"Mas-Ribas",
"L.",
""
],
[
"Sanchez",
"D. Mata",
""
],
[
"Nicholl",
"M.",
""
],
[
"Nixon",
"C. J.",
""
],
[
"Oates",
"S. R.",
""
],
[
"Paek",
"G.",
""
],
[
"Palmerio",
"J.",
""
],
[
"Paris",
"D.",
""
],
[
"Pieterse",
"D. L. A.",
""
],
[
"Pugliese",
"G.",
""
],
[
"Vasquez",
"J. A. Quirola",
""
],
[
"van Roestel",
"J.",
""
],
[
"Rossi",
"A.",
""
],
[
"Salvaterra",
"A. Rouco Escorial. R.",
""
],
[
"Schneider",
"B.",
""
],
[
"Smartt",
"S. J.",
""
],
[
"Smith",
"K.",
""
],
[
"Smith",
"I. A.",
""
],
[
"Srivastav",
"S.",
""
],
[
"Torres",
"M. A. P.",
""
],
[
"Ventura",
"C.",
""
],
[
"Vreeswijk",
"P.",
""
],
[
"Wainscoat",
"R.",
""
],
[
"Yang",
"Y. -J.",
""
],
[
"Yang",
"S.",
""
]
] |
2504.08937 | Minjie Deng | Minjie Deng, Yan Wei, Hao Zhai, An Wu, Yuncan Ouyang, Qianyao Peng | Rethinking Few-Shot Image Fusion: Granular Ball Priors Enable
General-Purpose Deep Fusion | null | null | null | null | cs.GR cs.CV cs.LG eess.IV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In image fusion tasks, the absence of real fused images as priors presents a
fundamental challenge. Most deep learning-based fusion methods rely on
large-scale paired datasets to extract global weighting features from raw
images, thereby generating fused outputs that approximate real fused images. In
contrast to previous studies, this paper explores few-shot training of neural
networks under the condition of having prior knowledge. We propose a novel
fusion framework named GBFF, and a Granular Ball Significant Extraction
algorithm specifically designed for the few-shot prior setting. All pixel pairs
involved in the fusion process are initially modeled as a Coarse-Grained
Granular Ball. At the local level, Fine-Grained Granular Balls are used to
slide through the brightness space to extract Non-Salient Pixel Pairs, and
perform splitting operations to obtain Salient Pixel Pairs. Pixel-wise weights
are then computed to generate a pseudo-supervised image. At the global level,
pixel pairs with significant contributions to the fusion process are
categorized into the Positive Region, while those whose contributions cannot be
accurately determined are assigned to the Boundary Region. The Granular Ball
performs modality-aware adaptation based on the proportion of the positive
region, thereby adjusting the neural network's loss function and enabling it to
complement the information of the boundary region. Extensive experiments
demonstrate the effectiveness of both the proposed algorithm and the underlying
theory. Compared with state-of-the-art (SOTA) methods, our approach shows
strong competitiveness in terms of both fusion time and image expressiveness.
Our code is publicly available at:
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:33:06 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 15:31:11 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Deng",
"Minjie",
""
],
[
"Wei",
"Yan",
""
],
[
"Zhai",
"Hao",
""
],
[
"Wu",
"An",
""
],
[
"Ouyang",
"Yuncan",
""
],
[
"Peng",
"Qianyao",
""
]
] |
2504.09081 | Prabhat Pandey | Prabhat Pandey, Rupak Vignesh Swaminathan, K V Vijay Girish, Arunasish
Sen, Jian Xie, Grant P. Strimel, Andreas Schwarz | SIFT-50M: A Large-Scale Multilingual Dataset for Speech Instruction
Fine-Tuning | null | null | null | null | eess.AS cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We introduce SIFT (Speech Instruction Fine-Tuning), a 50M-example dataset
designed for instruction fine-tuning and pre-training of speech-text large
language models (LLMs). SIFT-50M is built from publicly available speech
corpora, which collectively contain 14K hours of speech, and leverages LLMs
along with off-the-shelf expert models. The dataset spans five languages,
encompassing a diverse range of speech understanding as well as controllable
speech generation instructions. Using SIFT-50M, we train SIFT-LLM, which
outperforms existing speech-text LLMs on instruction-following benchmarks while
achieving competitive performance on foundational speech tasks. To support
further research, we also introduce EvalSIFT, a benchmark dataset specifically
designed to evaluate the instruction-following capabilities of speech-text
LLMs.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:45:48 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 17:34:21 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Pandey",
"Prabhat",
""
],
[
"Swaminathan",
"Rupak Vignesh",
""
],
[
"Girish",
"K V Vijay",
""
],
[
"Sen",
"Arunasish",
""
],
[
"Xie",
"Jian",
""
],
[
"Strimel",
"Grant P.",
""
],
[
"Schwarz",
"Andreas",
""
]
] |
2504.09115 | Jiongchi Yu | Jiongchi Yu, Xiaofei Xie, Qiang Hu, Bowen Zhang, Ziming Zhao, Yun Lin,
Lei Ma, Ruitao Feng, Frank Liauw | CAShift: Benchmarking Log-Based Cloud Attack Detection under Normality
Shift | Accepted by FSE 2025 | null | null | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | With the rapid advancement of cloud-native computing, securing cloud
environments has become an important task. Log-based Anomaly Detection (LAD) is
the most representative technique used in different systems for attack
detection and safety guarantee, where multiple LAD methods and relevant
datasets have been proposed. However, even though some of these datasets are
specifically prepared for cloud systems, they only cover limited cloud
behaviors and lack information from a whole-system perspective. Besides,
another critical issue to consider is normality shift, which implies the test
distribution could differ from the training distribution and highly affects the
performance of LAD. Unfortunately, existing works only focus on simple shift
types such as chronological changes, while other important and cloud-specific
shift types are ignored, e.g., the distribution shift introduced by different
deployed cloud architectures. Therefore, creating a new dataset that covers
diverse behaviors of cloud systems and normality shift types is necessary.
To fill the gap in evaluating LAD under real-world conditions, we present
CAShift, the first normality shift-aware dataset for cloud systems. CAShift
captures three shift types, including application, version, and cloud
architecture shifts, and includes 20 diverse attack scenarios across various
cloud components. Using CAShift, we conduct an empirical study showing that (1)
all LAD methods are significantly affected by normality shifts, with
performance drops of up to 34%, and (2) continuous learning techniques can
improve F1-scores by up to 27%, depending on data usage and algorithm choice.
Based on our findings, we offer valuable implications for future research in
designing more robust LAD models and methods for LAD shift adaptation.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:58:54 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 07:41:24 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yu",
"Jiongchi",
""
],
[
"Xie",
"Xiaofei",
""
],
[
"Hu",
"Qiang",
""
],
[
"Zhang",
"Bowen",
""
],
[
"Zhao",
"Ziming",
""
],
[
"Lin",
"Yun",
""
],
[
"Ma",
"Lei",
""
],
[
"Feng",
"Ruitao",
""
],
[
"Liauw",
"Frank",
""
]
] |
2504.09229 | Erik DeBenedictis | Erik P. DeBenedictis | Scaling up Reversible Logic with HKI Superconducting Inductors | 6 pages, 5 figures, based on a presentation at the USC4SCE workshop
April 9, 2025 | null | null | ZF013 | cs.ET cs.AR | http://creativecommons.org/licenses/by/4.0/ | Researchers developed about a dozen semiconductor reversible (or adiabatic)
logic chips since the early 1990s, validating circuit designs and proving the
concept--but scale up required a further advance. This document shows that
cryogenic inductors made of a new High Kinetic Inductance (HKI) material
provide the advance. This material can be deposited as an integrated circuit
layer, where it has enough energy recycling capacity to power a reversible
circuit of the same size. This allows a designer to replicate and scale a
complete reversible logic subsystem in accordance with Moore's law.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 14:14:01 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 18:36:48 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"DeBenedictis",
"Erik P.",
""
]
] |
2504.09544 | Yemin Yu | Yemin Yu, Neil Tenenholtz, Lester Mackey, Ying Wei, David
Alvarez-Melis, Ava P. Amini, Alex X. Lu | Causal integration of chemical structures improves representations of
microscopy images for morphological profiling | 24 pages | null | null | null | cs.LG cs.CE cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent advances in self-supervised deep learning have improved our ability to
quantify cellular morphological changes in high-throughput microscopy screens,
a process known as morphological profiling. However, most current methods only
learn from images, despite many screens being inherently multimodal, as they
involve both a chemical or genetic perturbation as well as an image-based
readout. We hypothesized that incorporating chemical compound structure during
self-supervised pre-training could improve learned representations of images in
high-throughput microscopy screens. We introduce a representation learning
framework, MICON (Molecular-Image Contrastive Learning), that models chemical
compounds as treatments that induce counterfactual transformations of cell
phenotypes. MICON significantly outperforms classical hand-crafted features
such as CellProfiler and existing deep-learning-based representation learning
methods in challenging evaluation settings where models must identify
reproducible effects of drugs across independent replicates and data-generating
centers. We demonstrate that incorporating chemical compound information into
the learning process provides consistent improvements in our evaluation setting
and that modeling compounds specifically as treatments in a causal framework
outperforms approaches that directly align images and compounds in a single
representation space. Our findings point to a new direction for representation
learning in morphological profiling, suggesting that methods should explicitly
account for the multimodal nature of microscopy screening data.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 12:27:21 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 19:03:34 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yu",
"Yemin",
""
],
[
"Tenenholtz",
"Neil",
""
],
[
"Mackey",
"Lester",
""
],
[
"Wei",
"Ying",
""
],
[
"Alvarez-Melis",
"David",
""
],
[
"Amini",
"Ava P.",
""
],
[
"Lu",
"Alex X.",
""
]
] |
2504.09593 | Hongwei Yao | Hongwei Yao, Haoran Shi, Yidou Chen, Yixin Jiang, Cong Wang, Zhan Qin | ControlNET: A Firewall for RAG-based LLM System | Project Page: https://ai.zjuicsr.cn/firewall | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Retrieval-Augmented Generation (RAG) has significantly enhanced the factual
accuracy and domain adaptability of Large Language Models (LLMs). This
advancement has enabled their widespread deployment across sensitive domains
such as healthcare, finance, and enterprise applications. RAG mitigates
hallucinations by integrating external knowledge, yet introduces privacy risk
and security risk, notably data breaching risk and data poisoning risk. While
recent studies have explored prompt injection and poisoning attacks, there
remains a significant gap in comprehensive research on controlling inbound and
outbound query flows to mitigate these threats. In this paper, we propose an AI
firewall, ControlNET, designed to safeguard RAG-based LLM systems from these
vulnerabilities. ControlNET controls query flows by leveraging activation shift
phenomena to detect adversarial queries and mitigate their impact through
semantic divergence. We conduct comprehensive experiments on four different
benchmark datasets including Msmarco, HotpotQA, FinQA, and MedicalSys using
state-of-the-art open source LLMs (Llama3, Vicuna, and Mistral). Our results
demonstrate that ControlNET achieves over 0.909 AUROC in detecting and
mitigating security threats while preserving system harmlessness. Overall,
ControlNET offers an effective, robust, harmless defense mechanism, marking a
significant advancement toward the secure deployment of RAG-based LLM systems.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 14:18:35 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 03:50:19 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Yao",
"Hongwei",
""
],
[
"Shi",
"Haoran",
""
],
[
"Chen",
"Yidou",
""
],
[
"Jiang",
"Yixin",
""
],
[
"Wang",
"Cong",
""
],
[
"Qin",
"Zhan",
""
]
] |
2504.09683 | Sa\v{s}a Novakovi\'c | Sa\v{s}a Novakovi\'c | Ulrich complexity and categorical representability dimension | comments welcome! arXiv admin note: text overlap with
arXiv:1807.10919 | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the Ulrich complexity of certain examples of Brauer--Severi
varieties, twisted flags and involution varieties and establish lower and upper
bounds. Furthermore, we relate Ulrich complexity to the categorical
representability dimension of the respective varieties. We also state an idea
why, in general, a relation between Ulrich complexity and categorical
representability dimension may appear.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 18:34:09 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 05:55:53 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Novaković",
"Saša",
""
]
] |
2504.09703 | Trung Chau | Trung Chau, Richie Sheng, Deborah Wooton | Homological invariants of edge ideals of weighted oriented graphs | are welcome!!! 19 pages. Minor revisions | null | null | null | math.AC math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We determine all possible triples of depth, dimension, and regularity of edge
ideals of weighted oriented graphs with a fixed number of vertices. Also, we
compute all the possible Betti table sizes of edge ideals of weighted oriented
trees and bipartite~graphs with a fixed number of vertices.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 19:45:44 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 19:11:11 GMT"
}
] | 2025-04-18T00:00:00 | [
[
"Chau",
"Trung",
""
],
[
"Sheng",
"Richie",
""
],
[
"Wooton",
"Deborah",
""
]
] |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.