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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2503.10313 | Manoj Yadav K. | Arpan Kanrar, Charlotte Roelants and Manoj K. Yadav | Central series' and ($n$)-isoclinism of skew left braces | 29 pages, 1 table. Corrected Theorem 4.8, added Corollary 4.9, added
Table 1, added Proposition 7.12 and modified remarks in the last section.
Many other typo corrected | null | null | null | math.RA | http://creativecommons.org/licenses/by/4.0/ | The aim of this article is to advance the knowledge on the theory of skew
left braces. We introduce a subclass of skew left braces, which we denote by
$\mathcal{I}_n$, $n \ge 1$, such that elements of the annihilator and lower
central series' interact `nicely' with respect to commutation. That allows us
to define a concept of $n$-isoclinism of skew left braces in $\mathcal{I}_n$,
by using a concept of brace commutator words, which we have introduced. We
prove results on $1$-isoclinism (isoclinism) of skew left braces analogous to
important results in group theory. For any two symmetric $n$-isoclinic skew
left braces $A$ and $B$, we prove that, there exist skew left braces $C$ and
$R$ such that both $A$ and $B$ are $n$-isoclinic to both $C$ and $R$ and (i)
$A$ and $B$ are quotient skew left braces of $C$; (ii) $A$ and $B$ are sub-skew
left braces of $R$. Connections between a skew left brace and the group which
occurs as a natural semi-direct product of additive and multiplicative groups
of the skew left brace are investigated, and it is proved that $n$-isoclinism
is preserved from braces to groups. We also show that various nilpotency
concepts on skew left braces are invariant under $n$-isoclinism.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 12:47:45 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:24:36 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Kanrar",
"Arpan",
""
],
[
"Roelants",
"Charlotte",
""
],
[
"Yadav",
"Manoj K.",
""
]
] |
2503.10491 | Alexander Osipov | Alexander V. Osipov | The $\Delta_1$-property of $X$ is equivalent to the Choquet property of
$B_1(X)$ | 8 pages. arXiv admin note: text overlap with arXiv:2110.15496,
arXiv:2501.06898 | null | null | null | math.GN | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give a characterization of the $\Delta_1$-property of any Tychonoff space
$X$ in terms of the function space $B_1(X)$ of all Baire-one real-valued
functions on a space $X$ with the topology of pointwise convergence. We
establish that for a Tychonoff space $X$ the $\Delta_1$-property is equivalent
to the Choquet property of $B_1(X)$. Also we construct under $ZFC$ an example
of a separable pseudocompact space $X$ such that $C_p(X)$ is
$\kappa$-Frechet-Urysohn but $X$ fails to be a $\Delta_1$-space. This answers a
question of Kakol-Leiderman-Tkachuk.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 15:56:25 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:55:50 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Osipov",
"Alexander V.",
""
]
] |
2503.10540 | Vadim Abramkin | Vadim Abramkin, George G. Pavlov, Yuriy Shibanov, B. Posselt and Oleg
Kargaltsev | The middle-aged pulsar PSR J1741-2054 and its bow-shock nebula in far
ultraviolet | 16 pages, 9 figures, accepted for publication in Astronomy &
Astrophysics | A&A 696, A121 (2025) | 10.1051/0004-6361/202452211 | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Context. The nearby middle-aged gamma-ray pulsar J1741-2054 and its pulsar
wind nebula (PWN) have been studied in X-rays, and its bow-shock nebula (BSN)
has been investigated in the Balmer lines, but they have never been observed in
far ultraviolet (FUV). Aims. To further study the thermal and magnetospheric
emission from PSR J1741-2054 and the BSN properties, we observed them in the
FUV range with the Hubble Space Telescope (HST). Methods. We imaged the target
in two FUV filters of the HST's ACS/SBC detector. We also re-analyzed previous
optical observations of the pulsar and its BSN. We fit the pulsar's FUV-optical
spectrum separately and together with its X-ray spectrum. Results. We found
that the pulsar's FUV-optical spectrum consists of a thermal and nonthermal
components. A joint fit of the FUV-optical and X-ray spectra with combinations
of a nonthermal and thermal components showed a hard optical nonthermal
spectrum with a photon index $\Gamma_{opt} \approx 1.0-1.2$ and a softer X-ray
component, $\Gamma_X \approx 2.6-2.7$. The thermal emission is dominated by the
cold component with the temperature $kT_{cold}\approx 40-50$ eV and emitting
sphere radius $R_{cold}\approx 8-15$ km, at $d=270$ pc. An additional hot
thermal component, with $kT_{hot}\sim 80$ eV and $R_{hot}\sim 1$ km, is also
possible. Such a spectrum resembles the spectra of other middle-aged pulsars,
but it shows a harder (softer) optical (X-ray) nonthermal spectrum. We detected
the FUV BSN, the first one associated with a middle-aged pulsar. Its
closed-shell morphology is similar to the H$\alpha$ BSN morphology, while its
FUV flux, $\sim10^{-13}$ erg cm$^{-2}$ s$^{-1}$, is a factor of $\sim 4$ higher
than the H$\alpha$ flux. This FUV BSN has a higher surface brightness than the
two previously known ones.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 16:55:33 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Abramkin",
"Vadim",
""
],
[
"Pavlov",
"George G.",
""
],
[
"Shibanov",
"Yuriy",
""
],
[
"Posselt",
"B.",
""
],
[
"Kargaltsev",
"Oleg",
""
]
] |
2503.10612 | Eric Tovar | Bennett Clayton and Eric J. Tovar | Approximation technique for preserving the minimum principle on the
entropy for the compressible Euler Equations | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper is concerned with constructing an invariant-domain preserving
approximation technique for the compressible Euler equations that preserves the
minimum principle on the physical entropy. We show that any numerical method
that can be written as a convex combination of good auxiliary states will
satisfy the minimum principle on the physical entropy provided the equation of
state satisfies some mild assumptions. Furthermore, we derive a wave speed
estimate in an extended Riemann problem necessary for constructing the
auxiliary states with desired properties. Finally, we numerically illustrate
the proposed methodology.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 17:55:05 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:26:01 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Clayton",
"Bennett",
""
],
[
"Tovar",
"Eric J.",
""
]
] |
2503.10650 | Tabia Tanzin Prama | Tabia Tanzin Prama, Jannatul Ferdaws Amrin, Md. Mushfique Anwar, Iqbal
H. Sarker | AI Enabled User-Specific Cyberbullying Severity Detection with
Explainability | null | null | null | null | cs.LG cs.CL cs.CY | http://creativecommons.org/licenses/by/4.0/ | The rise of social media has significantly increased the prevalence of
cyberbullying (CB), posing serious risks to both mental and physical
well-being. Effective detection systems are essential for mitigating its
impact. While several machine learning (ML) models have been developed, few
incorporate victims' psychological, demographic, and behavioral factors
alongside bullying comments to assess severity. In this study, we propose an AI
model intregrating user-specific attributes, including psychological factors
(self-esteem, anxiety, depression), online behavior (internet usage,
disciplinary history), and demographic attributes (race, gender, ethnicity),
along with social media comments. Additionally, we introduce a re-labeling
technique that categorizes social media comments into three severity levels:
Not Bullying, Mild Bullying, and Severe Bullying, considering user-specific
factors.Our LSTM model is trained using 146 features, incorporating emotional,
topical, and word2vec representations of social media comments as well as
user-level attributes and it outperforms existing baseline models, achieving
the highest accuracy of 98\% and an F1-score of 0.97. To identify key factors
influencing the severity of cyberbullying, we employ explainable AI techniques
(SHAP and LIME) to interpret the model's decision-making process. Our findings
reveal that, beyond hate comments, victims belonging to specific racial and
gender groups are more frequently targeted and exhibit higher incidences of
depression, disciplinary issues, and low self-esteem. Additionally, individuals
with a prior history of bullying are at a greater risk of becoming victims of
cyberbullying.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 05:11:42 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 19:01:43 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Prama",
"Tabia Tanzin",
""
],
[
"Amrin",
"Jannatul Ferdaws",
""
],
[
"Anwar",
"Md. Mushfique",
""
],
[
"Sarker",
"Iqbal H.",
""
]
] |
2503.10825 | Pavel Fileviez Perez | Hridoy Debnath, Pavel Fileviez Perez | On Gauge Theories of Neutrino Masses and Dark Matter | null | Physical Review D 111, 075020 (2025) | 10.1103/PhysRevD.111.075020 | null | hep-ph astro-ph.CO hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We discuss the predictions in the simplest theory for neutrino masses based
on the spontaneous breaking of local lepton number. This theory provides a
simple theoretical framework to understand the possible relation between the
origin of neutrino masses and the nature of the dark matter. In this theory,
one of the fields needed for anomaly cancellation is a dark matter candidate
and the local lepton number is broken at the low scale. We discuss in great
detail the dark matter properties showing the allowed parameter space by the
relic density bounds and the predictions for direct detection. The predictions
for gamma and neutrino lines from dark matter annihilation are investigated. In
the case of Dirac neutrinos, the bound on the effective number of relativistic
degrees of freedom plays an important role and the predictions for gamma lines
could be tested in the near future. We discuss the predictions in the case of
Majorana neutrinos where the dark matter candidate has extra annihilation
channels and compare all the predictions to the case with Dirac neutrinos.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 19:15:37 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Debnath",
"Hridoy",
""
],
[
"Perez",
"Pavel Fileviez",
""
]
] |
2503.11121 | Mika Juvela | Mika Juvela | Fast spectral line calculations with the escape probability method and
tests with synthetic observations of interstellar clouds | Accepted for publication in A&A | A&A 696, A144 (2025) | 10.1051/0004-6361/202453409 | null | astro-ph.IM astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radiative transfer effects need to be taken into account when analysing
spectral line observations. When the data are not sufficient for detailed
modelling, simpler methods are needed. The escape probability formalism (EPF)
is one such tool. We wish to quantify the model errors in the EPF analysis of
interstellar clouds and cores. We introduce PEP, a parallel program for
calculating fast EPF parameters quickly. We model full radiative transfer to
generate synthetic observations for various cloud models. These are examined
with the PEP program, and their results are compared to the actual
beam-averaged kinetic temperatures, column densities, and volume densities. PEP
enables the calculations of even millions of parameter combinations in a matter
of seconds. However, the simple assumptions of EPF can lead to significant
errors. In the tests the errors were typically within a factor of two, but
could in some cases reach an order of magnitude. The model errors are thus
similar or even larger than the statistical errors caused by the typical
observational noise. Due to degeneracies, parameter combinations are better
constrained than the individual parameters. The model errors could be reduced
by using full radiative transfer modelling. However, in the absence of full
knowledge of the source structure, the errors are difficult to quantify. We
also present a method for approximate handling of hyperfine structure lines in
EPF calculations. Both the observational statistical errors and the model
errors need to be considered when estimating the reliability of EPF results.
Full radiative transfer modelling is needed to better understand the true
uncertainties.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 06:33:29 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Juvela",
"Mika",
""
]
] |
2503.11266 | Jonas Utz | Jonas Utz, Stefan Vocht, Anne Tjorven Buessen, Dennis Possart, Fabian
Wagner, Mareike Thies, Mingxuan Gu, Stefan Uderhardt, Katharina Breininger | CyclePose -- Leveraging Cycle-Consistency for Annotation-Free Nuclei
Segmentation in Fluorescence Microscopy | under review for MICCAI 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In recent years, numerous neural network architectures specifically designed
for the instance segmentation of nuclei in microscopic images have been
released. These models embed nuclei-specific priors to outperform generic
architectures like U-Nets; however, they require large annotated datasets,
which are often not available. Generative models (GANs, diffusion models) have
been used to compensate for this by synthesizing training data. These two-stage
approaches are computationally expensive, as first a generative model and then
a segmentation model has to be trained. We propose CyclePose, a hybrid
framework integrating synthetic data generation and segmentation training.
CyclePose builds on a CycleGAN architecture, which allows unpaired translation
between microscopy images and segmentation masks. We embed a segmentation model
into CycleGAN and leverage a cycle consistency loss for self-supervision.
Without annotated data, CyclePose outperforms other weakly or unsupervised
methods on two public datasets. Code is available at
https://github.com/jonasutz/CyclePose
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 10:22:26 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:54:10 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Utz",
"Jonas",
""
],
[
"Vocht",
"Stefan",
""
],
[
"Buessen",
"Anne Tjorven",
""
],
[
"Possart",
"Dennis",
""
],
[
"Wagner",
"Fabian",
""
],
[
"Thies",
"Mareike",
""
],
[
"Gu",
"Mingxuan",
""
],
[
"Uderhardt",
"Stefan",
""
],
[
"Breininger",
"Katharina",
""
]
] |
2503.11451 | Jonas M\"uller | Jonas M\"uller, Quentin Copp\'ee, Saskia Hekker | Oscillations of red giant stars with magnetic damping in the core. I.
Dissipation of mode energy in dipole-like magnetic fields | accepted by A&A, 20 pages, 19 figures, 3 tables | A&A 696, A134 (2025) | 10.1051/0004-6361/202553888 | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Strong magnetic fields in the core of red-giant branch stars are expected to
suppress the amplitudes of the multipole modes. This occurs when the strength
of the internal magnetic field approaches the critical field strength, at which
the magnetic forces become comparable to the buoyancy. We performed Hamiltonian
ray tracing simulations of magneto-gravity waves to investigate the suppression
of the multipole modes in the presence of an internal dipole-like magnetic
field. We took into account different stellar masses, metallicities, and ages,
as well as various oscillation frequencies and spherical degrees. In
particular, we estimated the trapped fraction, a measure of multipole mode
suppression, which quantifies the fraction of mode energy in the core that is
dissipated by the interaction with the magnetic field. Our results indicate
that the trapped fraction can be described by a simple expression, which
smoothly connects the regime without multipole mode suppression with the regime
with complete suppression of the multipole modes. Crucially, the trapped
fraction depends only on the ratio between the strength of the internal
magnetic field and the critical field strength. Therefore, our expression for
the trapped fraction provides a flexible tool that can be used, for example, to
estimate the amount of multipole mode suppression as a star ascends the
red-giant branch or to investigate the onset of the suppression in observed
power spectral densities.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 14:40:08 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Müller",
"Jonas",
""
],
[
"Coppée",
"Quentin",
""
],
[
"Hekker",
"Saskia",
""
]
] |
2503.11496 | Shaofeng Liang | Shaofeng Liang and Runwei Guan and Wangwang Lian and Daizong Liu and
Xiaolou Sun and Dongming Wu and Yutao Yue and Weiping Ding and Hui Xiong | Cognitive Disentanglement for Referring Multi-Object Tracking | 26 pages, 11 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As a significant application of multi-source information fusion in
intelligent transportation perception systems, Referring Multi-Object Tracking
(RMOT) involves localizing and tracking specific objects in video sequences
based on language references. However, existing RMOT approaches often treat
language descriptions as holistic embeddings and struggle to effectively
integrate the rich semantic information contained in language expressions with
visual features. This limitation is especially apparent in complex scenes
requiring comprehensive understanding of both static object attributes and
spatial motion information. In this paper, we propose a Cognitive
Disentanglement for Referring Multi-Object Tracking (CDRMT) framework that
addresses these challenges. It adapts the "what" and "where" pathways from the
human visual processing system to RMOT tasks. Specifically, our framework first
establishes cross-modal connections while preserving modality-specific
characteristics. It then disentangles language descriptions and hierarchically
injects them into object queries, refining object understanding from coarse to
fine-grained semantic levels. Finally, we reconstruct language representations
based on visual features, ensuring that tracked objects faithfully reflect the
referring expression. Extensive experiments on different benchmark datasets
demonstrate that CDRMT achieves substantial improvements over state-of-the-art
methods, with average gains of 6.0% in HOTA score on Refer-KITTI and 3.2% on
Refer-KITTI-V2. Our approach advances the state-of-the-art in RMOT while
simultaneously providing new insights into multi-source information fusion.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 15:21:54 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:35:24 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Liang",
"Shaofeng",
""
],
[
"Guan",
"Runwei",
""
],
[
"Lian",
"Wangwang",
""
],
[
"Liu",
"Daizong",
""
],
[
"Sun",
"Xiaolou",
""
],
[
"Wu",
"Dongming",
""
],
[
"Yue",
"Yutao",
""
],
[
"Ding",
"Weiping",
""
],
[
"Xiong",
"Hui",
""
]
] |
2503.11497 | Jonas Erhardt | Jonas Erhardt, Mattia Iannetti, Fernando Dominguez, Ewelina M.
Hankiewicz, Bj\"orn Trauzettel, Gianni Profeta, Domenico Di Sante, Giorgio
Sangiovanni, Simon Moser, Ralph Claessen | Backscattering in Topological Edge States Despite Time-Reversal Symmetry | null | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Spin-momentum-locked edge states of quantum spin Hall insulators (QSHIs)
provide a compelling platform for spintronic applications, owing to their
intrinsic protection against backscattering from non-magnetic disorder. This
protection emerges from time-reversal symmetry, which pairs Kramers partners of
helical edge modes with opposite spin and momentum, thereby strictly forbidding
elastic single-particle backscattering within the pair. Yet, contrary to the
idealized notion of linear edge bands, the non-monotonic dispersions of
realistic materials can host multiple Kramers pairs, reintroducing
backscattering channels between them without violating time-reversal symmetry.
Here, we investigate inter-Kramers pair backscattering in the non-linear edge
bands of the QSHI indenene, highlighting a critical aspect of edge-state
stability. Using quasiparticle interference in scanning tunneling spectroscopy
-- a direct probe of backscattering -- we observe pairwise coupling between
energy-degenerate Kramers pairs, while energy regions with only a single
Kramers pair remain strictly protected. Supported by theoretical analysis, our
findings provide an unprecedented experimental demonstration of edge state
backscattering fully consistent with their underlying topological protection.
This insight has profound implications for numerous QSHI candidates,
emphasizing that the mere presence of gap-traversing edge modes does not
inherently guarantee their protection against backscattering.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 15:21:56 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 22:55:19 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Erhardt",
"Jonas",
""
],
[
"Iannetti",
"Mattia",
""
],
[
"Dominguez",
"Fernando",
""
],
[
"Hankiewicz",
"Ewelina M.",
""
],
[
"Trauzettel",
"Björn",
""
],
[
"Profeta",
"Gianni",
""
],
[
"Di Sante",
"Domenico",
""
],
[
"Sangiovanni",
"Giorgio",
""
],
[
"Moser",
"Simon",
""
],
[
"Claessen",
"Ralph",
""
]
] |
2503.11595 | Yin Chen | Yin Chen, Shan Ren, Jiawen Shan, and Runxuan Zhang | Generalized derivations of $\omega$-Lie algebras | To appear in J. Algebra Appl. (2026) 2650206 (16 pages) | null | 10.1142/S0219498826502063 | null | math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This article explores the structure theory of compatible generalized
derivations of finite-dimensional $\omega$-Lie algebras over a field
$\mathbb{K}$. We prove that any compatible quasiderivation of an $\omega$-Lie
algebra can be embedded as a compatible derivation into a larger $\omega$-Lie
algebra, refining the general result established by Leger and Luks in 2000 for
finite-dimensional nonassociative algebras. We also provide an approach to
explicitly compute (compatible) generalized derivations and quasiderivations
for all $3$-dimensional non-Lie complex $\omega$-Lie algebras.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 17:04:20 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 02:38:43 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Yin",
""
],
[
"Ren",
"Shan",
""
],
[
"Shan",
"Jiawen",
""
],
[
"Zhang",
"Runxuan",
""
]
] |
2503.11689 | Mehulkumar Rajkumar Lakhadive | Lakhadive Mehulkumar R, Anshu Sharma, and Basuraj Bhowmik | Investigating dimensionally-reduced highly-damped systems with
multivariate variational mode decomposition: An experimental approach | We have identified significant methodological and mathematical
inconsistencies in the current formulation of the paper, particularly in the
absence of the optimization framework that is crucial to the core analysis.
These issues may lead to misinterpretation of the proposed approach. We are
currently working on a revised and complete version that accurately reflects
the intended contributions | null | null | null | stat.AP | http://creativecommons.org/licenses/by/4.0/ | Structural health monitoring (SHM) is an essential engineering field aimed at
ensuring the safety and reliability of civil infrastructures. This study
proposes a methodology using multivariate variational mode decomposition (MVMD)
for damage detection and modal identification. MVMD decomposes multi-sensor
vibration responses into intrinsic modal components, facilitating the
extraction of natural frequencies and damping ratios by analyzing amplitude
decay in the identified modes. Mode shapes are determined through
peak-normalization of Fourier spectra corresponding to each mode. The
methodology is further applied to detect damage by identifying changes in the
extracted modal parameters and spatial features of the structure. The proposed
approach enables damage detection by tracking variations in modal parameters
and spatial structural characteristics. To validate its efficacy, the
methodology is applied to a benchmark eight-degree-of-freedom (8-DOF) system
from Los Alamos National Laboratory (LANL), demonstrating its robustness in
identifying structural damage under non-stationary excitation and narrowband
frequency content. The results confirm that MVMD provides a reliable and
adaptable framework for modal analysis and damage assessment in complex
infrastructure systems, addressing key challenges such as environmental
variability and practical scenarios.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 14:02:09 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:59:29 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"R",
"Lakhadive Mehulkumar",
""
],
[
"Sharma",
"Anshu",
""
],
[
"Bhowmik",
"Basuraj",
""
]
] |
2503.12297 | Gagan Khandate | Gagan Khandate, Boxuan Wang, Sarah Park, Weizhe Ni, Jaoquin Palacious,
Kate Lampo, Philippe Wu, Rosh Ho, Eric Chang, Matei Ciocarlie | Train Robots in a JIF: Joint Inverse and Forward Dynamics with Human and
Robot Demonstrations | 9 pages, 8 figures, submission to RSS 2025 | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pre-training on large datasets of robot demonstrations is a powerful
technique for learning diverse manipulation skills but is often limited by the
high cost and complexity of collecting robot-centric data, especially for tasks
requiring tactile feedback. This work addresses these challenges by introducing
a novel method for pre-training with multi-modal human demonstrations. Our
approach jointly learns inverse and forward dynamics to extract latent state
representations, towards learning manipulation specific representations. This
enables efficient fine-tuning with only a small number of robot demonstrations,
significantly improving data efficiency. Furthermore, our method allows for the
use of multi-modal data, such as combination of vision and touch for
manipulation. By leveraging latent dynamics modeling and tactile sensing, this
approach paves the way for scalable robot manipulation learning based on human
demonstrations.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 23:37:15 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 17:42:16 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Khandate",
"Gagan",
""
],
[
"Wang",
"Boxuan",
""
],
[
"Park",
"Sarah",
""
],
[
"Ni",
"Weizhe",
""
],
[
"Palacious",
"Jaoquin",
""
],
[
"Lampo",
"Kate",
""
],
[
"Wu",
"Philippe",
""
],
[
"Ho",
"Rosh",
""
],
[
"Chang",
"Eric",
""
],
[
"Ciocarlie",
"Matei",
""
]
] |
2503.12936 | Tong Lei | Tong Lei, Qinwen Hu, Ziyao Lin, Andong Li, Rilin Chen, Meng Yu, Dong
Yu and Jing Lu | FNSE-SBGAN: Far-field Speech Enhancement with Schrodinger Bridge and
Generative Adversarial Networks | 13 pages, 6 figures | null | null | null | eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The prevailing method for neural speech enhancement predominantly utilizes
fully-supervised deep learning with simulated pairs of far-field
noisy-reverberant speech and clean speech. Nonetheless, these models frequently
demonstrate restricted generalizability to mixtures recorded in real-world
conditions. To address this issue, this study investigates training enhancement
models directly on real mixtures. Specifically, we revisit the single-channel
far-field to near-field speech enhancement (FNSE) task, focusing on real-world
data characterized by low signal-to-noise ratio (SNR), high reverberation, and
mid-to-high frequency attenuation. We propose FNSE-SBGAN, a framework that
integrates a Schrodinger Bridge (SB)-based diffusion model with generative
adversarial networks (GANs). Our approach achieves state-of-the-art performance
across various metrics and subjective evaluations, significantly reducing the
character error rate (CER) by up to 14.58% compared to far-field signals.
Experimental results demonstrate that FNSE-SBGAN preserves superior subjective
quality and establishes a new benchmark for real-world far-field speech
enhancement. Additionally, we introduce an evaluation framework leveraging
matrix rank analysis in the time-frequency domain, providing systematic
insights into model performance and revealing the strengths and weaknesses of
different generative methods.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 08:51:03 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Apr 2025 17:47:04 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 16:31:23 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Lei",
"Tong",
""
],
[
"Hu",
"Qinwen",
""
],
[
"Lin",
"Ziyao",
""
],
[
"Li",
"Andong",
""
],
[
"Chen",
"Rilin",
""
],
[
"Yu",
"Meng",
""
],
[
"Yu",
"Dong",
""
],
[
"Lu",
"Jing",
""
]
] |
2503.13042 | Hans A. Weidenm\"uller | Hans A. Weidenm\"uller | The rise of stochasticity in physics | null | Eur. Phys. J. Plus 140:304 (2025) | null | null | physics.hist-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In the last 175 years, the physical understanding of
nature has seen a revolutionary change. Until about 1850, Newton's
theory and the mechanical world view derived from it provided the
dominant view of the physical world, later supplemented by Maxwell's
theory of the electromagnetic field. That approach was entirely
deterministic and free of probabilistic concepts. In contrast to
that conceptual edifice, today many fields of physics are governed
by probabilistic concepts. Statistical mechanics in its classical or
quantum version and random-matrix theory are obvious
examples. Quantum mechanics is an intrinsically statistical
theory. Classical chaos and its quantum manifestations also require
a stochastic approach. The paper describes how a combination of
discoveries and conceptual problems undermined the mechanical world
view, led to novel concepts, and shaped the modern understanding of
physics.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 10:46:42 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Weidenmüller",
"Hans A.",
""
]
] |
2503.13534 | Wonjun Yi | Wonjun Yi, Yong-Hwa Park | Multi-output Classification for Compound Fault Diagnosis in Motor under
Partially Labeled Target Domain | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This study presents a novel multi-output classification (MOC) framework
designed for domain adaptation in fault diagnosis, addressing challenges posed
by partially labeled (PL) target domain dataset and coexisting faults in
rotating machinery. Unlike conventional multi-class classification (MCC)
approaches, the MOC framework independently classifies the severity of each
fault, enhancing diagnostic accuracy. By integrating multi-kernel maximum mean
discrepancy loss (MKMMD) and entropy minimization loss (EM), the proposed
method improves feature transferability between source and target domains,
while frequency layer normalization (FLN) effectively handles stationary
vibration signals by leveraging mechanical characteristics. Experimental
evaluations across six domain adaptation cases, encompassing partially labeled
(PL) scenarios, demonstrate the superior performance of the MOC approach over
baseline methods in terms of macro F1 score.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 14:15:10 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Yi",
"Wonjun",
""
],
[
"Park",
"Yong-Hwa",
""
]
] |
2503.14420 | Marc Levine | Marc Levine and Anna M. Viergever | Quadratic Donaldson-Thomas invariants for $(\mathbb{P}^1)^3$ and some
other smooth proper toric threefolds | In the introduction, we altered our main conjecture slightly,
deleting a condition for the precise value of the constant
$\epsilon\in\{\pm1\}$, and added comments about the vanishing of the odd
quadratic DT invariants | null | null | null | math.AG math.AT | http://creativecommons.org/licenses/by/4.0/ | Using virtual localization in Witt sheaf cohomology, we show that the
generating series of quadratic Donaldson-Thomas invariants of
$(\mathbb{P}^1)^3$, valued in the Witt ring of $\mathbb{R}$,
$W(\mathbb{R})\cong \mathbb{Z}$, is equal to $M(q^2)^{-8}$, where $M(q)$ is the
MacMahon function. This confirms a modified version of a conjecture of
Viergever. We also show that a localized version of this conjecture holds for
certain iterated blow-ups of $(\mathbb{P}^1)^3$ and other related smooth proper
toric varieties.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 16:54:22 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 12:23:16 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Levine",
"Marc",
""
],
[
"Viergever",
"Anna M.",
""
]
] |
2503.14675 | Guilherme Xavier | Guilherme B. Xavier, Jan-{\AA}ke Larsson, Paolo Villoresi, Giuseppe
Vallone, Ad\'an Cabello | Energy-time and time-bin entanglement: past, present and future | Review paper. 27 pages, 9 figures, 250+ references. New references
added. Comments welcome! | null | null | null | quant-ph physics.optics | http://creativecommons.org/licenses/by/4.0/ | Entanglement is a key resource in many quantum information tasks. From a
fundamental perspective entanglement is at the forefront of major philosophical
discussions advancing our understanding of nature. An experimental scheme was
proposed in 1989 by Franson that exploited the unpredictability in the
generation time of a photon pair in order to produce a then new form of quantum
entanglement, known as energy-time entanglement. A later modification gave rise
to the very popular time-bin entanglement, an important cornerstone in many
real-world quantum communication applications. Both forms of entanglement have
radically pushed forward our understanding of quantum mechanics throughout the
1990s and 2000s. A decade later modifications to the original proposals were
proposed and demonstrated, which opens the path for the highly sought-after
device-independence capability for entanglement certification, with a goal of
ultra-secure quantum communication. In this review we cover the beginnings of
energy-time and time-bin entanglement, many key experiments that expanded our
understanding of what was achievable in quantum information experiments all the
way down to modern demonstrations based on new technological advances. We will
then point out to the future discussing the important place that energy-time
and time-bin entanglement will have in upcoming quantum networks and novel
protocols based on nonlocality.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 19:30:33 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 18:40:04 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Xavier",
"Guilherme B.",
""
],
[
"Larsson",
"Jan-Åke",
""
],
[
"Villoresi",
"Paolo",
""
],
[
"Vallone",
"Giuseppe",
""
],
[
"Cabello",
"Adán",
""
]
] |
2503.14904 | Andrei Linde | John Joseph Carrasco, Renata Kallosh, Andrei Linde, and Diederik Roest | Axion Stabilization in Modular Cosmology | 19 pages, 10 figures, a new section is added on $SL(2,\mathbb{Z})$
invariant potentials in disk variables | null | null | null | hep-th astro-ph.CO gr-qc hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The $SL(2,\mathbb{Z})$ invariant $\alpha$-attractor models have plateau
potentials with respect to the inflaton and axion fields. The potential in the
axion direction is almost exactly flat during inflation, hence, the axion field
remains nearly massless. In this paper, we develop a generalized class of such
models, where the $SL(2,\mathbb{Z})$ symmetry is preserved, but the axion
acquires a large mass and becomes strongly stabilized during inflation, which
eliminates isocurvature perturbations in this scenario. Inflation in such
two-field models occurs as in the single-field $\alpha$-attractors and leads to
the same cosmological predictions.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 05:10:21 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:45:57 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Carrasco",
"John Joseph",
""
],
[
"Kallosh",
"Renata",
""
],
[
"Linde",
"Andrei",
""
],
[
"Roest",
"Diederik",
""
]
] |
2503.14936 | Yifan Zhang | Yifan Zhang, Chen Huang, Zachary Karas, Dung Thuy Nguyen, Kevin Leach,
Yu Huang | Enhancing Code LLM Training with Programmer Attention | null | null | 10.1145/3696630.3728510 | null | cs.SE cs.HC cs.LG | http://creativecommons.org/licenses/by/4.0/ | Human attention provides valuable yet underexploited signals for code LLM
training, offering a perspective beyond purely machine-driven attention.
Despite the complexity and cost of collecting eye-tracking data, there has also
been limited progress in systematically using these signals for code LLM
training. To address both issues, we propose a cohesive pipeline spanning
augmentation and reward-based fine-tuning. Specifically, we introduce (1) an
eye-tracking path augmentation method to expand programmer attention datasets,
(2) a pattern abstraction step that refines raw fixations into learnable
attention motifs, and (3) a reward-guided strategy for integrating these
insights directly into a CodeT5 supervised fine-tuning process. Our experiments
yield +7.16 in CodeBLEU on the CodeXGlue benchmark for code summarization,
underscoring how uniting human and machine attention can boost code
intelligence. We hope this work encourages broader exploration of human-centric
methods in next-generation AI4SE.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 06:44:29 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 07:53:50 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Zhang",
"Yifan",
""
],
[
"Huang",
"Chen",
""
],
[
"Karas",
"Zachary",
""
],
[
"Nguyen",
"Dung Thuy",
""
],
[
"Leach",
"Kevin",
""
],
[
"Huang",
"Yu",
""
]
] |
2503.15055 | Arina Razmyslovich | Arina Razmyslovich, Kseniia Murasheva, Sofia Sedlova, Julien
Capitaine, Eugene Dmitriev | ELTEX: A Framework for Domain-Driven Synthetic Data Generation | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | We introduce Efficient LLM Token Extraction (ELTEX), a framework addressing
the critical challenge of LLM domain specialization by systematically
extracting and integrating domain indicators throughout synthetic data
generation. Unlike approaches relying on implicit knowledge transfer, ELTEX
explicitly leverages domain signals to maintain specialized knowledge
integrity. In our cybersecurity case study, ELTEX-enhanced data enables a
fine-tuned Gemma-2B model to achieve performance competitive with GPT-4o on
blockchain cyberattack classification while reducing computational
requirements. Our Google Sheets implementation makes ELTEX accessible to
non-technical users. Our contributions include: (1) the ELTEX framework; (2)
Google Sheets Add-on implementation; (3) empirical validation showing how ELTEX
bridges performance gaps between small and large models; and (4) a synthetic
dataset of 11,448 texts for blockchain cyberattack detection.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 09:46:54 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:14:25 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Razmyslovich",
"Arina",
""
],
[
"Murasheva",
"Kseniia",
""
],
[
"Sedlova",
"Sofia",
""
],
[
"Capitaine",
"Julien",
""
],
[
"Dmitriev",
"Eugene",
""
]
] |
2503.15475 | Kangle Deng | Foundation AI Team Roblox: Kiran Bhat, Nishchaie Khanna, Karun Channa,
Tinghui Zhou, Yiheng Zhu, Xiaoxia Sun, Charles Shang, Anirudh Sudarshan,
Maurice Chu, Daiqing Li, Kangle Deng, Jean-Philippe Fauconnier, Tijmen
Verhulsdonck, Maneesh Agrawala, Kayvon Fatahalian, Alexander Weiss, Christian
Reiser, Ravi Kiran Chirravuri, Ravali Kandur, Alejandro Pelaez, Akash Garg,
Michael Palleschi, Jessica Wang, Skylar Litz, Leon Liu, Anying Li, David
Harmon, Derek Liu, Liangjun Feng, Denis Goupil, Lukas Kuczynski, Jihyun Yoon,
Naveen Marri, Peiye Zhuang, Yinan Zhang, Brian Yin, Haomiao Jiang, Marcel van
Workum, Thomas Lane, Bryce Erickson, Salil Pathare, Kyle Price, Steve Han,
Yiqing Wang, Anupam Singh, David Baszucki | Cube: A Roblox View of 3D Intelligence | Our code and model weights can be found at:
https://github.com/Roblox/cube | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Foundation models trained on vast amounts of data have demonstrated
remarkable reasoning and generation capabilities in the domains of text,
images, audio and video. Our goal at Roblox is to build such a foundation model
for 3D intelligence, a model that can support developers in producing all
aspects of a Roblox experience, from generating 3D objects and scenes to
rigging characters for animation to producing programmatic scripts describing
object behaviors. We discuss three key design requirements for such a 3D
foundation model and then present our first step towards building such a model.
We expect that 3D geometric shapes will be a core data type and describe our
solution for 3D shape tokenizer. We show how our tokenization scheme can be
used in applications for text-to-shape generation, shape-to-text generation and
text-to-scene generation. We demonstrate how these applications can collaborate
with existing large language models (LLMs) to perform scene analysis and
reasoning. We conclude with a discussion outlining our path to building a fully
unified foundation model for 3D intelligence.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 17:52:17 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 19:52:06 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Foundation AI Team",
"",
""
],
[
"Bhat",
"Kiran",
""
],
[
"Khanna",
"Nishchaie",
""
],
[
"Channa",
"Karun",
""
],
[
"Zhou",
"Tinghui",
""
],
[
"Zhu",
"Yiheng",
""
],
[
"Sun",
"Xiaoxia",
""
],
[
"Shang",
"Charles",
""
],
[
"Sudarshan",
"Anirudh",
""
],
[
"Chu",
"Maurice",
""
],
[
"Li",
"Daiqing",
""
],
[
"Deng",
"Kangle",
""
],
[
"Fauconnier",
"Jean-Philippe",
""
],
[
"Verhulsdonck",
"Tijmen",
""
],
[
"Agrawala",
"Maneesh",
""
],
[
"Fatahalian",
"Kayvon",
""
],
[
"Weiss",
"Alexander",
""
],
[
"Reiser",
"Christian",
""
],
[
"Chirravuri",
"Ravi Kiran",
""
],
[
"Kandur",
"Ravali",
""
],
[
"Pelaez",
"Alejandro",
""
],
[
"Garg",
"Akash",
""
],
[
"Palleschi",
"Michael",
""
],
[
"Wang",
"Jessica",
""
],
[
"Litz",
"Skylar",
""
],
[
"Liu",
"Leon",
""
],
[
"Li",
"Anying",
""
],
[
"Harmon",
"David",
""
],
[
"Liu",
"Derek",
""
],
[
"Feng",
"Liangjun",
""
],
[
"Goupil",
"Denis",
""
],
[
"Kuczynski",
"Lukas",
""
],
[
"Yoon",
"Jihyun",
""
],
[
"Marri",
"Naveen",
""
],
[
"Zhuang",
"Peiye",
""
],
[
"Zhang",
"Yinan",
""
],
[
"Yin",
"Brian",
""
],
[
"Jiang",
"Haomiao",
""
],
[
"van Workum",
"Marcel",
""
],
[
"Lane",
"Thomas",
""
],
[
"Erickson",
"Bryce",
""
],
[
"Pathare",
"Salil",
""
],
[
"Price",
"Kyle",
""
],
[
"Han",
"Steve",
""
],
[
"Wang",
"Yiqing",
""
],
[
"Singh",
"Anupam",
""
],
[
"Baszucki",
"David",
""
]
] |
2503.15792 | Christopher Berg Smiet | Christopher Berg Smiet, Ludovic Rais, Joaquim Loizu and Robert Davies | Turnstile area as a measure for chaotic transport in magnetic
confinement fusion devices | Submitted to Chaos | null | null | null | physics.plasm-ph | http://creativecommons.org/licenses/by-sa/4.0/ | We analyze stochasticity in the magnetic fields of magnetic confinement
fusion reactors by calculating the lobe areas of turnstiles - a method
developed for characterizing transport into and out of resonance zones in
Hamiltonian dynamical systems. We develop an efficient algorithm based on an
action principle to calculate this quantity directly from the magnetic field,
including stellarator magnetic fields which are sourced by a complicated set of
three-dimensional coils. In the analyzed devices, the turnstile area on the
inboard (plasma-facing) manifolds is much smaller than the turnstile area on
the outboard (wall-facing) manifolds. The application of the turnstile area
calculation for the design of future reactors will be discussed.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 02:14:16 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Apr 2025 15:58:55 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 10:14:03 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Smiet",
"Christopher Berg",
""
],
[
"Rais",
"Ludovic",
""
],
[
"Loizu",
"Joaquim",
""
],
[
"Davies",
"Robert",
""
]
] |
2503.16092 | Lassi Paunonen | Anthony Hastir and Lassi Paunonen | Well-Posedness and Stability of Infinite-Dimensional Systems Under
Monotone Feedback | 41 pages, 1 figure. Submitted. In this version, the stability results
in Section 4 were generalised to system nodes which are impedance passive but
not necessarily well-posed | null | null | null | math.OC math.AP math.FA | http://creativecommons.org/licenses/by/4.0/ | We study the well-posedness and stability of an impedance passive
infinite-dimensional linear system under nonlinear feedback of the form
$u(t)=\phi(v(t)-y(t))$, where $\phi$ is a monotone function. Our first main
result introduces conditions guaranteeing the existence of classical and
generalised solutions in a situation where the original linear system is
well-posed. In the absence of the external input $v$ we establish the existence
of strong and generalised solutions under strictly weaker conditions. Finally,
we introduce conditions guaranteeing that the origin is a globally
asymptotically stable equilibrium point of the closed-loop system. Motivated by
the analysis of partial differential equations with nonlinear boundary
conditions, we use our results to investigate the well-posedness and stablility
of abstract boundary control systems, port-Hamiltonian systems, a Timoshenko
beam model, and a two-dimensional boundary controlled heat equation.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 12:33:45 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 12:54:33 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Hastir",
"Anthony",
""
],
[
"Paunonen",
"Lassi",
""
]
] |
2503.16188 | Ming Li | Ming Li, Jike Zhong, Shitian Zhao, Yuxiang Lai, Kaipeng Zhang | Think or Not Think: A Study of Explicit Thinking in Rule-Based Visual
Reinforcement Fine-Tuning | Preprint, work in progress. Add results on math, cvbench, and puzzle | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper investigates the thinking process in rule-based reinforcement
learning fine-tuning (RFT) for multi-modal large language models (MLLMs). We
first propose CLS-RL for classification, using verifiable rewards to encourage
MLLM thinking. Experiments show CLS-RL significantly outperforms SFT and yields
a 'free-lunch' generalization effect (improving performance on unseen datasets
after training on one dataset). We then question if this explicit thinking is
always necessary for RFT. Challenging convention that explicit thinking is
crucial for RFT, we introduce No-Thinking-RL, minimizing thinking via a simple
equality accuracy reward. Experiments show No-Thinking-RL surpasses CLS-RL in
in-domain and generalization abilities, with significantly less fine-tuning
time. This suggests reducing thinking can improve MLLM fine-tuning efficiency
and effectiveness for certain visual tasks. We hypothesize explicit thinking
negatively impacts reward convergence during RFT. To test this, we propose the
Think-After-Answerwer method to let models first output the answer and then
generate thinking process to alliviate the negative impact of thinking. We
further test No-Thinking-RL on diverse tasks (including math, spatial, puzzles)
with 2B and 7B models. For 2B models, No-Thinking-RL outperforms thinking-based
RFT for all tasks, even on math, with Think-After-Answerwer performing
intermediately. For 7B models, performance is comparable on simple visual
tasks, but RFT with thinking excels on complex reasoning (math). This implies
when dealing with complex math problems, smaller models struggle with
generating effective reasoning, hurting performance on complex tasks.
Conversely, for simple visual tasks, thinking is not indispensable, and its
removal can boost performance and reduce training time. We hope our findings
offer insights for better understanding the effect of the thinking process in
RFT.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 14:37:45 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 09:52:37 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 15:31:31 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Li",
"Ming",
""
],
[
"Zhong",
"Jike",
""
],
[
"Zhao",
"Shitian",
""
],
[
"Lai",
"Yuxiang",
""
],
[
"Zhang",
"Kaipeng",
""
]
] |
2503.16231 | Elizabeth Gasparim | Elizabeth Gasparim | Intrinsic Mirrors for Minimal Adjoint Orbit (ICM G&T 2022) | Section 6 of this talk describes the results of arXiv:2108.02471 | Cheltsov, I., & Martinez-Garcia, J. (Eds.). (2025). ZAG Handbook
of Modern Algebraic Geometry (1st ed.). Chapman and Hall/CRC | 10.1201/9781003286363 | null | math.AG math.SG | http://creativecommons.org/licenses/by/4.0/ | This text is contribution 77 to the ZAG Handbook of Modern Algebraic
Geometry, edited by I. Cheltsov and J. Martinez-Garcia, and summarises the
Short Communication I presented at the Geometry and Topology Session of the
International Congress of Mathematicians which took place at the University of
Copenhagen in 2022.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 15:28:39 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Gasparim",
"Elizabeth",
""
]
] |
2503.16304 | Peiran Gu | Peiran Gu, Fuhao Duan, Wenhao Li, Bochen Xu, Ying Cai, Teng Yao,
Chenxun Zhuo, Tianming Liu and Bao Ge | Bridging Technology and Humanities: Evaluating the Impact of Large
Language Models on Social Sciences Research with DeepSeek-R1 | 52 pages, 19 figures | null | null | null | cs.CY cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, the development of Large Language Models (LLMs) has made
significant breakthroughs in the field of natural language processing and has
gradually been applied to the field of humanities and social sciences research.
LLMs have a wide range of application value in the field of humanities and
social sciences because of its strong text understanding, generation and
reasoning capabilities. In humanities and social sciences research, LLMs can
analyze large-scale text data and make inferences.
This article analyzes the large language model DeepSeek-R1 from seven
aspects: low-resource language translation, educational question-answering,
student writing improvement in higher education, logical reasoning, educational
measurement and psychometrics, public health policy analysis, and art education
. Then we compare the answers given by DeepSeek-R1 in the seven aspects with
the answers given by o1-preview. DeepSeek-R1 performs well in the humanities
and social sciences, answering most questions correctly and logically, and can
give reasonable analysis processes and explanations. Compared with o1-preview,
it can automatically generate reasoning processes and provide more detailed
explanations, which is suitable for beginners or people who need to have a
detailed understanding of this knowledge, while o1-preview is more suitable for
quick reading.
Through analysis, it is found that LLM has broad application potential in the
field of humanities and social sciences, and shows great advantages in
improving text analysis efficiency, language communication and other fields.
LLM's powerful language understanding and generation capabilities enable it to
deeply explore complex problems in the field of humanities and social sciences,
and provide innovative tools for academic research and practical applications.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 16:25:24 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Mar 2025 16:34:40 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 15:09:24 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Gu",
"Peiran",
""
],
[
"Duan",
"Fuhao",
""
],
[
"Li",
"Wenhao",
""
],
[
"Xu",
"Bochen",
""
],
[
"Cai",
"Ying",
""
],
[
"Yao",
"Teng",
""
],
[
"Zhuo",
"Chenxun",
""
],
[
"Liu",
"Tianming",
""
],
[
"Ge",
"Bao",
""
]
] |
2503.16713 | Shi Feng | Shi Feng | Optimal Matching Problem on the Boolean Cube | 26 pages, 0 figure | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish upper and lower bounds for the expected Wasserstein distance
between the random empirical measure and the uniform measure on the Boolean
cube. Our analysis leverages techniques from Fourier analysis, following the
framework introduced in \cite{bobkov2021simple}, as well as methods from large
deviations theory.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 21:14:57 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 17:50:39 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Feng",
"Shi",
""
]
] |
2503.16769 | Snigdha Ghosh | Snigdha Ghosh, Nilanjan Chaudhuri, Pradip Roy and Sourav Sarkar | Shear and bulk viscous coefficients of a hot and chirally imbalanced
quark matter using NJL model | Version published in Physical Review D | Physical Review D 111, 076012 (2025) | 10.1103/PhysRevD.111.076012 | null | hep-ph nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The shear $\eta$ and bulk $\zeta$ viscous coefficients have been calculated
in a hot and chirally asymmetric quark matter quantified in terms of a chiral
chemical potential (CCP) using the two-flavor Nambu-Jona--Lasinio (NJL) model.
This is done by employing the one-loop Green-Kubo formalism where the viscous
coefficients have been extracted from the long-wavelength limit of the
in-medium spectral function corresponding to the energy momentum tensor (EMT)
current correlator calculated using the real time formalism of finite
temperature field theory. The momentum dependent thermal width of the
quark/antiquark that enters into the expression of the viscosities as a
dynamical input containing interactions, has been obtained from the $2\to2$
scattering processes mediated via the collective mesonic modes in scalar and
pseudoscalar chanels encoded in respective in-medium polarization functions
having explicit temperature and CCP dependence. Several thermodynamic
quantities such as pressure, energy density, entropy density $(s)$, specific
heat and isentropic speed of sound have also been calculated at finite CCP. The
temperature and CCP dependence of the viscosity to entropy density ratios
$\eta/s$ and $\zeta/s$ have also been studied.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 00:53:52 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:58:14 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ghosh",
"Snigdha",
""
],
[
"Chaudhuri",
"Nilanjan",
""
],
[
"Roy",
"Pradip",
""
],
[
"Sarkar",
"Sourav",
""
]
] |
2503.17609 | Ning Jiang | Ning Jiang, Zhen Pan | Embers of Active Galactic Nuclei: Tidal Disruption Events and
Quasiperiodic Eruptions | Publication in ApJL (983,L18). 2 Tables and 1 Figure | null | 10.3847/2041-8213/adc456 | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent observations have confirmed the direct association between tidal
disruption events (TDEs) and quasiperiodic eruptions (QPEs). In addition, TDE
hosts and QPE hosts are statistically found to be similar in their
morphological properties and in the strong overrepresentation of poststarburst
galaxies. Particularly, both of them show an intriguing preference for
extending emission line regions (EELRs), which are indicative of recently faded
active galactic nuclei (AGNs). This further suggests that QPEs might be
produced following TDEs involving supermassive black holes at a particular
stage, when the AGN activity has recently ceased. Moreover, in the framework of
"QPEs=extreme mass ratio inspiral (EMRI) + accretion disk" model, a large
fraction of QPE EMRIs are inferred to be quasi-circular from the QPE timing,
indicating that they are wet EMRIs that were formed in the AGN disk during a
previous AGN phase. Based on these facts, we propose a unified scenario that
connects these three phenomena: AGN activities boost both the TDE rate and the
formation rate of low-eccentricity EMRIs, consequently TDEs are preferentially
found in recently faded AGNs instead of in ongoing AGNs due to selection
effects, and QPEs are also preferentially found in recently faded AGNs where
TDEs frequently feed a misaligned accretion disk to the EMRI.
| [
{
"version": "v1",
"created": "Sat, 22 Mar 2025 01:42:55 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 00:55:16 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Jiang",
"Ning",
""
],
[
"Pan",
"Zhen",
""
]
] |
2503.17751 | Chenhao Zhang | Yi Ouyang, Qimin Song and Chenhao Zhang | Finiteness of Pythagoras numbers of finitely generated real algebras | 19 pages, comments welcome! | null | null | null | math.NT math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we establish two finiteness results and propose a conjecture
concerning the Pythagoras number $P(A)$ of a finitely generated real algebra
$A$. Let $X \hookrightarrow \mathbb{P}^n$ be an integral projective surface
over $\mathbb{R}$, and let $s \in \Gamma(X,\mathcal{O}_X(1))$ be a nonzero
section such that the boundary $\partial_{X(\mathbb{C})}\bigl(X_{s\neq
0}(\mathbb{R})\bigr)$, taken in the standard topology, is Zariski dense in
$X_{s=0}$. We prove $P\bigl(\Gamma(X_{s\neq 0})\bigr)=\infty$. As a corollary,
the Pythagoras numbers of integral smooth affine curves over $\mathbb{R}$ are
shown to be unbounded. For any finitely generated $\mathbb{R}$-algebra $A$, if
the Zariski closure of the real points of $\mathrm{Spec}(A)$ has dimension less
than two, we demonstrate $P(A)<\infty$.
| [
{
"version": "v1",
"created": "Sat, 22 Mar 2025 12:27:49 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Mar 2025 12:37:22 GMT"
},
{
"version": "v3",
"created": "Wed, 2 Apr 2025 04:23:15 GMT"
},
{
"version": "v4",
"created": "Tue, 15 Apr 2025 09:08:02 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ouyang",
"Yi",
""
],
[
"Song",
"Qimin",
""
],
[
"Zhang",
"Chenhao",
""
]
] |
2503.18035 | Zhenyu Li | Tianyi Shang, Zhenyu Li, Pengjie Xu, Zhaojun Deng, Ruirui Zhang | Text-Driven 3D Lidar Place Recognition for Autonomous Driving | 13 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Environment description-based localization in large-scale point cloud maps
constructed through remote sensing is critically significant for the
advancement of large-scale autonomous systems, such as delivery robots
operating in the last mile. However, current approaches encounter challenges
due to the inability of point cloud encoders to effectively capture local
details and long-range spatial relationships, as well as a significant modality
gap between text and point cloud representations. To address these challenges,
we present Des4Pos, a novel two-stage text-driven remote sensing localization
framework. In the coarse stage, the point-cloud encoder utilizes the
Multi-scale Fusion Attention Mechanism (MFAM) to enhance local geometric
features, followed by a bidirectional Long Short-Term Memory (LSTM) module to
strengthen global spatial relationships. Concurrently, the Stepped Text Encoder
(STE) integrates cross-modal prior knowledge from CLIP [1] and aligns text and
point-cloud features using this prior knowledge, effectively bridging modality
discrepancies. In the fine stage, we introduce a Cascaded Residual Attention
(CRA) module to fuse cross-modal features and predict relative localization
offsets, thereby achieving greater localization precision. Experiments on the
KITTI360Pose test set demonstrate that Des4Pos achieves state-of-the-art
performance in text-to-point-cloud place recognition. Specifically, it attains
a top-1 accuracy of 40% and a top-10 accuracy of 77% under a 5-meter radius
threshold, surpassing the best existing methods by 7% and 7%, respectively.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 11:36:19 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:22:14 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Shang",
"Tianyi",
""
],
[
"Li",
"Zhenyu",
""
],
[
"Xu",
"Pengjie",
""
],
[
"Deng",
"Zhaojun",
""
],
[
"Zhang",
"Ruirui",
""
]
] |
2503.18127 | Maximilian Sommer | Max Sommer, Mark Wyatt, Yinuo Han | A PR drag origin for the Fomalhaut disk's pervasive inner dust:
constraints on collisional strengths, icy composition, and embedded planets | null | Monthly Notices of the Royal Astronomical Society, v. 539-1
(2025), pp. 439-456 | 10.1093/mnras/staf494 | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Recent JWST observations of the Fomalhaut debris disk have revealed a
significant abundance of dust interior to the outer planetesimal belt, raising
questions about its origin and maintenance. In this study, we apply an
analytical model to the Fomalhaut system, that simulates the dust distribution
interior to a planetesimal belt, as collisional fragments across a range of
sizes are dragged inward under Poynting-Robertson (PR) drag. We generate
spectral energy distributions and synthetic JWST/MIRI images of the model
disks, and perform an extensive grid search for particle parameters --
pertaining to composition and collisional strength -- that best match the
observations. We find that a sound fit can be found for particle properties
that involve a substantial water ice component, around 50%--80% by total
volume, and a catastrophic disruption threshold, $Q_D^\star$, at a particle
size of $D\!\approx\!30\,$um of 2--4$\,\times\,10^6\,$erg/g. Based on the
expected dynamical depletion of migrating dust by an intervening planet we
discount planets with masses $>1\,M_\mathrm{Saturn}$ beyond $\sim50\,$au in the
extended disk, though a planet shepherding the inner edge of the outer belt of
up to $\sim2\,M_\mathrm{Saturn}$ is reconcilable with the PR-drag-maintained
disk scenario, contingent upon higher collisional strengths. These results
indicate that PR drag transport from the outer belt alone can account for the
high interior dust contents seen in the Fomalhaut system, which may thus
constitute a common phenomenon in other belt-bearing systems. This establishes
a framework for interpreting mid-planetary system dust around other stars, with
our results for Fomalhaut providing a valuable calibration of the models.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 16:16:58 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:22:46 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Sommer",
"Max",
""
],
[
"Wyatt",
"Mark",
""
],
[
"Han",
"Yinuo",
""
]
] |
2503.18172 | Zixin Chen | Zixin Chen, Sicheng Song, Kashun Shum, Yanna Lin, Rui Sheng, Huamin Qu | Unmasking Deceptive Visuals: Benchmarking Multimodal Large Language
Models on Misleading Chart Question Answering | 31 pages in total. Under Review | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Misleading chart visualizations, which intentionally manipulate data
representations to support specific claims, can distort perceptions and lead to
incorrect conclusions. Despite decades of research, misleading visualizations
remain a widespread and pressing issue. Recent advances in multimodal large
language models (MLLMs) have demonstrated strong chart comprehension
capabilities, yet no existing work has systematically evaluated their ability
to detect and interpret misleading charts. This paper introduces the Misleading
Chart Question Answering (Misleading ChartQA) Benchmark, a large-scale
multimodal dataset designed to assess MLLMs in identifying and reasoning about
misleading charts. It contains over 3,000 curated examples, covering 21 types
of misleaders and 10 chart types. Each example includes standardized chart
code, CSV data, and multiple-choice questions with labeled explanations,
validated through multi-round MLLM checks and exhausted expert human review. We
benchmark 16 state-of-the-art MLLMs on our dataset, revealing their limitations
in identifying visually deceptive practices. We also propose a novel pipeline
that detects and localizes misleaders, enhancing MLLMs' accuracy in misleading
chart interpretation. Our work establishes a foundation for advancing
MLLM-driven misleading chart comprehension. We publicly release the sample
dataset to support further research in this critical area.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 18:56:33 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Mar 2025 17:24:41 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 15:48:57 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Zixin",
""
],
[
"Song",
"Sicheng",
""
],
[
"Shum",
"Kashun",
""
],
[
"Lin",
"Yanna",
""
],
[
"Sheng",
"Rui",
""
],
[
"Qu",
"Huamin",
""
]
] |
2503.18653 | Tong Chen | Tong Chen, Yu Gao, Shijie Liu, Ying-nan Mao, Kechen Wang | Limiting on ALP-photon Coupling through GRB221009A | We request the withdrawal of our manuscript as it remains incomplete.
The co-authors have not yet reached a consensus on certain details, and
further analysis is needed regarding the anomalous dip observed in the
high-energy parameter space. Since the current version lacks conclusive
interpretations, we believe withdrawal is appropriate at this stage | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates the constraints on ALPs parameters through the
photon-ALP oscillation model, based on the high-energy photon propagation
characteristics of the gamma-ray burst GRB 221009A. We briefly describe the
Primakoff process and use it to derive the oscillation probability. Numerical
simulations incorporate the GMF, IGMF, and EBL, utilizing the open-source code
gammaALPs to simulate photon propagation from the source to Earth. By employing
the chi2 statistical method and segmented spectral fitting of low-energy and
high-energy observational data from HXMT-GECAM, the dependence of photon
survival probability on ALP mass (ma) and coupling strength (ga) is analyzed.
Results show that for the ALP mass range 10^-7 < ma < 10^2 neV, the upper limit
on the coupling strength is ga < 0.27 * 10^-11 GeV^-1, improving constraints by
one order of magnitude compared to the CAST experiment (6.6 * 10^-11 GeV^-1).
Notably, high-energy data exhibit significantly stronger constraining power.
This work provides a novel theoretical framework and observational basis for
indirectly probing ALPs through extreme astrophysical phenomena.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 13:17:44 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:12:15 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Tong",
""
],
[
"Gao",
"Yu",
""
],
[
"Liu",
"Shijie",
""
],
[
"Mao",
"Ying-nan",
""
],
[
"Wang",
"Kechen",
""
]
] |
2503.18740 | Sichao Qu | Sichao Qu, Min Yang, Sibo Huang, Shuohan Liu, Erqian Dong, Helios Y.
Li, Ping Sheng, I. David Abrahams, and Nicholas X. Fang | Duality Symmetry in Causality Constraints for Enhanced Acoustic
Absorption | 14 Pages, 4 Figures, 5 Extended Data Items | null | null | null | physics.app-ph physics.class-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We derive a generalized causality constraint for acoustic reflection and
transmission for a flat slab with finite thickness, via the duality
transformation. It is known that achieving the upper limit of the causality
constraint necessitates a critical coupling condition to optimize absorption
bandwidth within a specified material thickness. However, the importance of
duality symmetry has been overlooked in this context. Our analytical model
demonstrates that optimal absorption in a 2-port setup not only relies on the
well-established critical coupling but also requires duality symmetry, defined
as the invariance under duality transformation. To verify our theoretical
prediction, we have experimentally realized customized metamaterials that
exhibit quasi-duality symmetry. This was achieved by inducing global degeneracy
between the first-order monopole and dipole resonances, consequently realizing
an exceptionally large sound absorption capacity as permitted by the proposed
causality constraint. Our findings elucidate the intrinsic connection between
duality symmetry and scattering causality, and they facilitate the exploitation
of the untapped potential in existing passive absorbers for wave transport
control.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 14:50:01 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:38:36 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Qu",
"Sichao",
""
],
[
"Yang",
"Min",
""
],
[
"Huang",
"Sibo",
""
],
[
"Liu",
"Shuohan",
""
],
[
"Dong",
"Erqian",
""
],
[
"Li",
"Helios Y.",
""
],
[
"Sheng",
"Ping",
""
],
[
"Abrahams",
"I. David",
""
],
[
"Fang",
"Nicholas X.",
""
]
] |
2503.19694 | Brendon Rhoades | Jaeseong Oh and Brendon Rhoades | Zigzags, contingency tables, and quotient rings | 38 pages, 1 figure | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $\mathbf{x}_{k \times p}$ be a $k \times p$ matrix of variables and let
$\mathbb{F}[\mathbf{x}_{k \times p}]$ be the polynomial ring in these
variables. Given two weak compositions $\alpha,\beta \models_0 n$ of lengths
$\ell(\alpha) = k$ and $\ell(\beta) = p$, we study the ideal $I_{\alpha,\beta}
\subseteq \mathbb{F}[\mathbf{x}_{k \times \ell}]$ generated by row sums, column
sums, monomials in row $i$ of degree $> \alpha_i$, and monomials in column $j$
of degree $> \beta_j$. We prove results connecting algebraic properties of the
quotient ring $R_{\alpha,\beta} := \mathbb{F}[\mathbf{x}_{k \times
\ell}]/I_{\alpha,\beta}$ with the set $C_{\alpha,\beta}$ of
$\alpha,\beta$-contingency tables. The standard monomial basis of
$R_{\alpha,\beta}$ with respect to a diagonal term order is encoded by the
matrix-ball avatar of the RSK correspondence. We describe the Hilbert series of
$R_{\alpha,\beta}$ in terms of a zigzag statistic on contingency tables. The
ring $R_{\alpha,\beta}$ carries a graded action of the product
$\mathrm{Stab}(\alpha) \times \mathrm{Stab}(\beta)$ of symmetry groups of the
sequences $\alpha = (\alpha_1,\dots,\alpha_k)$ and $\beta =
(\beta_1,\dots,\beta_p)$; we describe how to calculate the isomorphism type of
this graded action. Our analysis regards the set $C_{\alpha,\beta}$ as a locus
in the affine space $\mathrm{Mat}_{k \times p}(\mathbb{F})$ and applies orbit
harmonics to this locus.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 14:20:39 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 21:55:15 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Oh",
"Jaeseong",
""
],
[
"Rhoades",
"Brendon",
""
]
] |
2503.19821 | Matthew Greenig | Matthew Greenig, Haowen Zhao, Vladimir Radenkovic, Aubin Ramon, Pietro
Sormanni | IgCraft: A versatile sequence generation framework for antibody
discovery and engineering | null | null | null | null | q-bio.BM cs.LG q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Designing antibody sequences to better resemble those observed in natural
human repertoires is a key challenge in biologics development. We introduce
IgCraft: a multi-purpose model for paired human antibody sequence generation,
built on Bayesian Flow Networks. IgCraft presents one of the first unified
generative modeling frameworks capable of addressing multiple antibody sequence
design tasks with a single model, including unconditional sampling, sequence
inpainting, inverse folding, and CDR motif scaffolding. Our approach achieves
competitive results across the full spectrum of these tasks while constraining
generation to the space of human antibody sequences, exhibiting particular
strengths in CDR motif scaffolding (grafting) where we achieve state-of-the-art
performance in terms of humanness and preservation of structural properties. By
integrating previously separate tasks into a single scalable generative model,
IgCraft provides a versatile platform for sampling human antibody sequences
under a variety of contexts relevant to antibody discovery and engineering.
Model code and weights are publicly available at
https://github.com/mgreenig/IgCraft.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 16:32:03 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 04:24:18 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Greenig",
"Matthew",
""
],
[
"Zhao",
"Haowen",
""
],
[
"Radenkovic",
"Vladimir",
""
],
[
"Ramon",
"Aubin",
""
],
[
"Sormanni",
"Pietro",
""
]
] |
2503.19822 | Love Pettersson | Love Pettersson, Anders S. S{\o}rensen | Long-distance quantum communication using concatenated ring graph codes | Equation typos fixed in text and code. Conclusion remains the same | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To realize long-distance quantum communication, it is crucial to design
quantum repeater architectures that can deal with transmission losses and
operational errors. Code concatenation of photonic graph codes is a promising
way to achieve this; however, existing concatenated codes that can correct both
transmission losses and operational errors are extremely hardware-demanding. We
propose a one-way quantum repeater architecture based on concatenated ring
graph codes and linear optical Bell-state measurements. We construct a scheme
to generate the concatenated ring graph codes using quantum emitters, where the
number of matter qubits scales linearly with concatenation depth. Furthermore,
we devise a measurement strategy at each repeater station with a simple
experimental setup where photons are measured in the order that they are
created and show that entanglement swapping is fault-tolerant to both
transmission losses and operational errors. This allows for long-distance
quantum communication ($> 10^4$ km) at a kHZ rate even in the presence of
single qubit error rates $\epsilon > 10^{-3}$.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 16:32:18 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 07:45:47 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Pettersson",
"Love",
""
],
[
"Sørensen",
"Anders S.",
""
]
] |
2503.20385 | Adam Stevenson | A. T. Stevenson (1), C. A. Haswell (1), J. P. Faria (2), J. R. Barnes
(1), J. K. Barstow (1), H. Dickinson (1), M. R. Standing (3) ((1) School of
Physical Sciences, The Open University, UK, (2) Observatoire Astronomique de
l'Universit\'e de Gen\'eve, Switzerland, (3) European Space Agency, European
Space Astronomy Centre, Spain) | RV-exoplanet eccentricities: Good, Beta, and Best | 28 pages, 22 figures, 13 tables. Accepted for publication in MNRAS | null | 10.1093/mnras/staf502 | null | astro-ph.EP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We examine the eccentricity distribution(s) of radial velocity detected
exoplanets. Previously, the eccentricity distribution was found to be described
well by a Beta distribution with shape parameters $a=0.867, b=3.03$. Increasing
the sample size by a factor of 2.25, we find that the CDF regression method now
prefers a mixture model of Rayleigh + Exponential distributions over the Beta
distribution, with an increase in Bayesian evidence of $\Delta\ln{Z}\sim 77$
($12.6\,\sigma$). Using PDF regression, the eccentricity distribution is best
described by a Gamma distribution, with a Rayleigh + Exponential mixture a
close second. The mixture model parameters, $\alpha = 0.68\pm0.05,
\lambda=3.32\pm0.25, \sigma =0.11\pm0.01$, are consistent between methods. We
corroborate findings that exoplanet eccentricities are drawn from independent
parent distributions when splitting the sample by period, mass, and
multiplicity. Systems with a known outer massive companion provide no positive
evidence for an eccentricity distribution distinct from those without. We
quantitatively show M-dwarf hosted planets share a common eccentricity
distribution with those orbiting FGK-type stars. We release our python code,
eccentriciPy, which allows bespoke tailoring of the input archive to create
more relevant priors for particular problems in RV planet discovery and
characterisation. We re-characterised example planets using either traditional
Beta, or updated priors, finding differences for recovery of low-amplitude
multi-signal systems. We explore the effects of a variety of prior choices. The
accurate determination of small but non-zero eccentricity values has
wide-ranging implications for modelling the structure and evolution of planets
and their atmospheres due to the energy dissipated by tidal flexing.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 10:06:13 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:28:53 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Stevenson",
"A. T.",
""
],
[
"Haswell",
"C. A.",
""
],
[
"Faria",
"J. P.",
""
],
[
"Barnes",
"J. R.",
""
],
[
"Barstow",
"J. K.",
""
],
[
"Dickinson",
"H.",
""
],
[
"Standing",
"M. R.",
""
]
] |
2503.20761 | Fan Yang | Zhuqing Wang, Ruochen Gao, Xiaoling Wu, Berislav Bu\v{c}a, Klaus
M{\o}lmer, Li You, and Fan Yang | Boundary Time Crystals Induced by Local Dissipation and Long-Range
Interactions | 7+6 pages, 5+3 figures | null | null | null | quant-ph cond-mat.quant-gas physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Driven-dissipative many-body system supports nontrivial quantum phases absent
in equilibrium. As a prominent example, the interplay between coherent driving
and collective dissipation can lead to a dynamical quantum phase that
spontaneously breaks time-translation symmetry. This so-called boundary time
crystal (BTC) is fragile in the presence of local dissipation, which can easily
relax the system to a stationary state. In this work, we demonstrate a robust
BTC that is intrinsically induced by local dissipation. We provide extensive
numerical evidences to support existence of the BTC and study its behaviors in
different regimes. In particular, with decreasing interaction range, we
identify a transition from classical limit cycles to quantum BTCs featuring
sizable spatial correlations. Our studies significantly broaden the scope of
nonequilibrium phases and shed new light on experimental search for dynamical
quantum matter.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 17:49:59 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:35:45 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Wang",
"Zhuqing",
""
],
[
"Gao",
"Ruochen",
""
],
[
"Wu",
"Xiaoling",
""
],
[
"Buča",
"Berislav",
""
],
[
"Mølmer",
"Klaus",
""
],
[
"You",
"Li",
""
],
[
"Yang",
"Fan",
""
]
] |
2503.20826 | Zhiwei Yang | Zhiwei Yang, Yucong Meng, Kexue Fu, Feilong Tang, Shuo Wang, Zhijian
Song | Exploring CLIP's Dense Knowledge for Weakly Supervised Semantic
Segmentation | CVPR2025 | null | null | null | cs.CV cs.CL cs.LG eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Weakly Supervised Semantic Segmentation (WSSS) with image-level labels aims
to achieve pixel-level predictions using Class Activation Maps (CAMs).
Recently, Contrastive Language-Image Pre-training (CLIP) has been introduced in
WSSS. However, recent methods primarily focus on image-text alignment for CAM
generation, while CLIP's potential in patch-text alignment remains unexplored.
In this work, we propose ExCEL to explore CLIP's dense knowledge via a novel
patch-text alignment paradigm for WSSS. Specifically, we propose Text Semantic
Enrichment (TSE) and Visual Calibration (VC) modules to improve the dense
alignment across both text and vision modalities. To make text embeddings
semantically informative, our TSE module applies Large Language Models (LLMs)
to build a dataset-wide knowledge base and enriches the text representations
with an implicit attribute-hunting process. To mine fine-grained knowledge from
visual features, our VC module first proposes Static Visual Calibration (SVC)
to propagate fine-grained knowledge in a non-parametric manner. Then Learnable
Visual Calibration (LVC) is further proposed to dynamically shift the frozen
features towards distributions with diverse semantics. With these enhancements,
ExCEL not only retains CLIP's training-free advantages but also significantly
outperforms other state-of-the-art methods with much less training cost on
PASCAL VOC and MS COCO.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 02:00:49 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Yang",
"Zhiwei",
""
],
[
"Meng",
"Yucong",
""
],
[
"Fu",
"Kexue",
""
],
[
"Tang",
"Feilong",
""
],
[
"Wang",
"Shuo",
""
],
[
"Song",
"Zhijian",
""
]
] |
2503.21136 | Daniel R. Johnston | Shashi Chourasiya and Daniel R. Johnston | Power-free palindromes and reversed primes | 14 pages, 1 figure | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove new results related to the digital reverse $\overleftarrow{n}$ of a
positive integer $n$ in a fixed base $b$. First we show that for $b\geq 26000$,
there exists infinitely many primes $p$ such that $\overleftarrow{p}$ is
square-free. Further, we show that for $b\geq 2$ there are infinitely many
palindromes (with $n=\overleftarrow{n}$) that are cube-free. We also give
asymptotic expressions for the counting functions corresponding to these
results. The main tools we use are recent bounds from the literature on
reversed primes and palindromes in arithmetic progressions.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 03:55:53 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 06:46:49 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chourasiya",
"Shashi",
""
],
[
"Johnston",
"Daniel R.",
""
]
] |
2503.21188 | Aixin Sun | Aixin Sun | Are We Solving a Well-Defined Problem? A Task-Centric Perspective on
Recommendation Tasks | Work in progress | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Recommender systems (RecSys) leverage user interaction history to predict and
suggest relevant items, shaping user experiences across various domains. While
many studies adopt a general problem definition, i.e., to recommend preferred
items to users based on past interactions, such abstraction often lacks the
domain-specific nuances necessary for practical deployment. However, models are
frequently evaluated using datasets from online recommender platforms, which
inherently reflect these specificities. In this paper, we analyze RecSys task
formulations, emphasizing key components such as input-output structures,
temporal dynamics, and candidate item selection. All these factors directly
impact offline evaluation. We further examine the complexities of user-item
interactions, including decision-making costs, multi-step engagements, and
unobservable interactions, which may influence model design and loss functions.
Additionally, we explore the balance between task specificity and model
generalizability, highlighting how well-defined task formulations serve as the
foundation for robust evaluation and effective solution development. By
clarifying task definitions and their implications, this work provides a
structured perspective on RecSys research. The goal is to help researchers
better navigate the field, particularly in understanding specificities of the
RecSys tasks and ensuring fair and meaningful evaluations.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 06:10:22 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:19:42 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Sun",
"Aixin",
""
]
] |
2503.21334 | Mathieu Gerber | Mathieu Gerber | Safety of particle filters: Some results on the time evolution of
particle filter estimates | 27 pages | null | null | null | math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Particle filters (PFs) is a class of Monte Carlo algorithms that propagate
over time a set of $N\in\mathbb{N}$ particles which can be used to estimate, in
an online fashion, the sequence of filtering distributions
$(\hat{\eta}_t)_{t\geq 1}$ defined by a state-space model. Despite the
popularity of PFs, the study of the time evolution of their estimates has
received barely attention in the literature. Denoting by
$(\hat{\eta}_t^N)_{t\geq 1}$ the PF estimate of $(\hat{\eta}_t)_{t\geq 1}$ and
letting $\kappa\in (0,1)$, in this work we first show that for any number of
particles $N$ it holds that, with probability one, we have $\|\hat{\eta}_t^N-
\hat{\eta}_t\|\geq \kappa$ for infinitely many $t\geq 1$, with $\|\cdot\|$ a
measure of distance between probability distributions. Considering a simple
filtering problem we then provide reassuring results concerning the ability of
PFs to estimate jointly a finite set $\{\hat{\eta}_t\}_{t=1}^T$ of filtering
distributions by studying
$\mathbb{P}(\sup_{t\in\{1,\dots,T\}}\|\hat{\eta}_t^{N}-\hat{\eta}_t\|\geq
\kappa)$. Finally, on the same toy filtering problem, we prove that sequential
quasi-Monte Carlo, a randomized quasi-Monte Carlo version of PF algorithms,
offers greater safety guarantees than PFs in the sense that, for this
algorithm, it holds that $\lim_{N\rightarrow\infty}\sup_{t\geq
1}\|\hat{\eta}_t^N-\hat{\eta}_t\|=0$ with probability one.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 10:13:00 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 11:35:54 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 12:25:48 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Gerber",
"Mathieu",
""
]
] |
2503.21882 | Matthew Zaremsky | James Belk, Francesco Fournier-Facio, James Hyde, Matthew C. B.
Zaremsky | Boone-Higman embeddings of $\mathrm{Aut}(F_n)$ and mapping class groups
of punctured surfaces | 19 pages. V2: Minor edits, submitted version | null | null | null | math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove that the groups $\mathrm{Aut}(F_n)$ satisfy the Boone-Higman
conjecture for all $n$, meaning each $\mathrm{Aut}(F_n)$ embeds in a finitely
presented simple group. In fact, we prove that each $\mathrm{Aut}(F_n)$
satisfies the "permutational" Boone-Higman conjecture, which means the simple
group in question can be taken to be a twisted Brin-Thompson group. A
far-reaching consequence of our approach is that finitely presented twisted
Brin-Thompson groups are universal among finitely presented simple groups that
are highly transitive. This is evidence toward the Boone-Higman conjecture
being equivalent to its permutational version. Proving the conjecture for
$\mathrm{Aut}(F_n)$ also confirms the conjecture for all groups (virtually)
embedding into some $\mathrm{Aut}(F_n)$, such as mapping class groups of
non-closed surfaces, braid groups, loop braid groups, ribbon braid groups and
certain Artin groups. This answers several questions of the first and fourth
authors with Bleak and Matucci. Yet another consequence of our approach is that
satisfying the permutational Boone-Higman conjecture is closed under free
products.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 18:01:03 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 19:37:41 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Belk",
"James",
""
],
[
"Fournier-Facio",
"Francesco",
""
],
[
"Hyde",
"James",
""
],
[
"Zaremsky",
"Matthew C. B.",
""
]
] |
2503.21958 | Kibon Ku | Kibon Ku, Talukder Z Jubery, Elijah Rodriguez, Aditya Balu, Soumik
Sarkar, Adarsh Krishnamurthy, and Baskar Ganapathysubramanian | SC-NeRF: NeRF-based Point Cloud Reconstruction using a Stationary Camera
for Agricultural Applications | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | This paper presents a NeRF-based framework for point cloud (PCD)
reconstruction, specifically designed for indoor high-throughput plant
phenotyping facilities. Traditional NeRF-based reconstruction methods require
cameras to move around stationary objects, but this approach is impractical for
high-throughput environments where objects are rapidly imaged while moving on
conveyors or rotating pedestals. To address this limitation, we develop a
variant of NeRF-based PCD reconstruction that uses a single stationary camera
to capture images as the object rotates on a pedestal. Our workflow comprises
COLMAP-based pose estimation, a straightforward pose transformation to simulate
camera movement, and subsequent standard NeRF training. A defined Region of
Interest (ROI) excludes irrelevant scene data, enabling the generation of
high-resolution point clouds (10M points). Experimental results demonstrate
excellent reconstruction fidelity, with precision-recall analyses yielding an
F-score close to 100.00 across all evaluated plant objects. Although pose
estimation remains computationally intensive with a stationary camera setup,
overall training and reconstruction times are competitive, validating the
method's feasibility for practical high-throughput indoor phenotyping
applications. Our findings indicate that high-quality NeRF-based 3D
reconstructions are achievable using a stationary camera, eliminating the need
for complex camera motion or costly imaging equipment. This approach is
especially beneficial when employing expensive and delicate instruments, such
as hyperspectral cameras, for 3D plant phenotyping. Future work will focus on
optimizing pose estimation techniques and further streamlining the methodology
to facilitate seamless integration into automated, high-throughput 3D
phenotyping pipelines.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 20:14:45 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 16:57:59 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ku",
"Kibon",
""
],
[
"Jubery",
"Talukder Z",
""
],
[
"Rodriguez",
"Elijah",
""
],
[
"Balu",
"Aditya",
""
],
[
"Sarkar",
"Soumik",
""
],
[
"Krishnamurthy",
"Adarsh",
""
],
[
"Ganapathysubramanian",
"Baskar",
""
]
] |
2503.22044 | Shurui Li | Shurui Li, Puneet Gupta | CIMPool: Scalable Neural Network Acceleration for Compute-In-Memory
using Weight Pools | 12 pages, 11 figures | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | Compute-in-memory (CIM) based neural network accelerators offer a promising
solution to the Von Neumann bottleneck by computing directly within memory
arrays. However, SRAM CIM faces limitations in executing larger models due to
its cell size and on-chip memory constraints. This work proposes CIMPool, a
CIM-aware compression and acceleration framework that counters this limitation
through a weight sharing-based compression technique, aptly named `Weight
Pool,' enabling significantly larger neural networks to be accommodated within
on-chip memory constraints. This method minimizes the accuracy trade-off
typically associated with parameter compression, allowing CIMPool to achieve a
significantly larger compression ratio compared to the traditional quantization
method with iso-accuracy.
Furthermore, CIMPool co-optimizes the compression algorithm, hardware, and
dataflow to efficiently implement the hardware permutation required by weight
pool compression, with negligible area and throughput overhead. Empirical
results demonstrate that CIMPool can achieve 8-bit level accuracy with an
effective 0.5-bit precision, reduce chip area by 62.3% for ResNet-18, and
enable the execution of an order of magnitude larger models for a given area
budget in SRAM CIMs. When DRAM is used to store weights, CIMPool can reduce the
total energy by 3.24x compared to iso-accuracy traditional CIMs.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 23:38:03 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 07:41:36 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Li",
"Shurui",
""
],
[
"Gupta",
"Puneet",
""
]
] |
2503.22303 | Magdalena Kaiser | Magdalena Kaiser and Gerhard Weikum | Preference-based Learning with Retrieval Augmented Generation for
Conversational Question Answering | WWW 2025 Short Paper, 5 pages | null | 10.1145/3701716.3715544 | null | cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conversational Question Answering (ConvQA) involves multiple subtasks, i) to
understand incomplete questions in their context, ii) to retrieve relevant
information, and iii) to generate answers. This work presents PRAISE, a
pipeline-based approach for ConvQA that trains LLM adapters for each of the
three subtasks. As labeled training data for individual subtasks is unavailable
in practice, PRAISE learns from its own generations using the final answering
performance as feedback signal without human intervention and treats
intermediate information, like relevant evidence, as weakly labeled data. We
apply Direct Preference Optimization by contrasting successful and unsuccessful
samples for each subtask. In our experiments, we show the effectiveness of this
training paradigm: PRAISE shows improvements per subtask and achieves new
state-of-the-art performance on a popular ConvQA benchmark, by gaining 15.5
percentage points increase in precision over baselines.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 10:26:49 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:10:39 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Kaiser",
"Magdalena",
""
],
[
"Weikum",
"Gerhard",
""
]
] |
2503.22348 | Christos Charmousis | Christos Charmousis, Simon Iteanu, David Langlois, Karim Noui | Axial perturbations of black holes with primary scalar hair | Regular article, 32 pages, numerous figures, references added | null | null | CERN-TH-2025-058 | gr-qc astro-ph.HE hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | We study axial perturbations of static black holes with primary hair in a
family of degenerate higher-order scalar-tensor (DHOST) theories. These
solutions possess a scalar charge, fully independent of the mass, leading to a
continuous one-parameter deformation of the standard Schwarzschild black hole.
Starting from these solutions, we also construct new black holes, solutions of
other DHOST theories, obtained via disformal transformations of the metric. In
particular, we investigate two specific types of disformal transformations: the
first leading to a theory where gravitational waves propagate at the speed of
light, the second to a Horndeski theory, where the equations of motion remain
second order. The dynamics of axial perturbations can be formally related to
the general relativistic equations of motion of axial perturbations in an
effective metric. The causal structure of the effective metric differs from
that of the background metric, leading to distinct gravitational and luminous
horizons. Using a WKB approximation, we compute the quasi-normal modes for the
Schr\"odinger-like equation associated with the effective metric outside the
gravitational horizon.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 11:44:25 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:54:37 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Charmousis",
"Christos",
""
],
[
"Iteanu",
"Simon",
""
],
[
"Langlois",
"David",
""
],
[
"Noui",
"Karim",
""
]
] |
2503.22400 | Owidiusz Makuta | Owidiusz Makuta, B{\l}a\.zej Kuzaka, Remigiusz Augusiak | Frustration graph formalism for qudit observables | 19 pages, We added a few references missing from the first version as
well as a short discussion on their impact on our paper | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | The incompatibility of measurements is the key feature of quantum theory that
distinguishes it from the classical description of nature. Here, we consider
groups of d-outcome quantum observables with prime d represented by
non-Hermitian unitary operators whose eigenvalues are d'th roots of unity. We
additionally assume that these observables mutually commute up to a scalar
factor being one of the d'th roots of unity. By representing commutation
relations of these observables via a frustration graph, we show that for such a
group, there exists a single unitary transforming them into a tensor product of
generalized Pauli matrices and some ancillary mutually commuting operators.
Building on this result, we derive upper bounds on the sum of the squares of
the absolute values and the sum of the expected values of the observables
forming a group. We finally utilize these bounds to compute the generalized
geometric measure of entanglement for qudit stabilizer subspaces.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 13:09:13 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 17:03:34 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Makuta",
"Owidiusz",
""
],
[
"Kuzaka",
"Błażej",
""
],
[
"Augusiak",
"Remigiusz",
""
]
] |
2503.22477 | Max Hirschberger | Ryo Misawa, Markus Kriener, Rinsuke Yamada, Ryota Nakano, Milena
Jovanovic, Leslie M. Schoop, Max Hirschberger | Chemical enhancement of superconductivity in LaRu3Si2 with
mode-selective coupling between kagome phonons and flat bands | 25 pages, 4 figures, 4 supplemental figures | null | null | null | cond-mat.supr-con cond-mat.mtrl-sci cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | In kagome metals, flat electronic bands induced by frustrated hopping are a
platform for strong electron correlations. In particular, a selective coupling
of flat band states to certain kagome phonon modes is proposed as a universal
origin of superconductivity in this material class. Here, we investigate the
superconductivity in the kagome system LaRu$_3$(Si$_{1-x}$Ge$_x$)$_2$ by
chemical pressure tuning while preserving the Ru-$4d$ states that constitute
the kagome flat bands. We observe a sizable enhancement in the density of
states up to $x = 0.07$, as determined by the specific heat, with a concomitant
increase in the superconducting transition temperature $T_\mathrm{c}$.
Ge-dopants induce a uniaxial lattice expansion along the $c$-axis. Our
first-principles calculations suggest that this mitigates the detrimental
effect of hybridization between kagome layers and reduces the dispersion of the
Ru-$4d_{x^2-y^2}$ flat band. The calculated chemical potential moves closer to
the maximum in the energy-dependent density of states. Our result is consistent
with a theoretical prediction of tunable flat band superconductivity in
LaRu$_3$Si$_2$ by mode-selective coupling between specific kagome phonons and
the Ru-$d_{x^2-y^2}$ orbitals.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 14:38:28 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 00:10:41 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Misawa",
"Ryo",
""
],
[
"Kriener",
"Markus",
""
],
[
"Yamada",
"Rinsuke",
""
],
[
"Nakano",
"Ryota",
""
],
[
"Jovanovic",
"Milena",
""
],
[
"Schoop",
"Leslie M.",
""
],
[
"Hirschberger",
"Max",
""
]
] |
2503.23001 | Bin Han | Bin Han, Di Feng, Jie Wang, and Hans D. Schotten | Buyer-Initiated Auction Mechanism for Data Redemption in Machine
Unlearning | Submitted to IEEE GLOBECOM 2025 | null | null | null | cs.LG cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of artificial intelligence (AI) has raised privacy concerns
over user data, leading to regulations like the General Data Protection
Regulation (GDPR) and the California Consumer Privacy Act (CCPA). With the
essential toolbox provided by machine unlearning, AI service providers are now
able to remove user data from their trained models as well as the training
datasets, so as to comply with such regulations. However, extensive data
redemption can be costly and degrade model accuracy. To balance the cost of
unlearning and the privacy protection, we propose a buyer-initiated auction
mechanism for data redemption, enabling the service provider to purchase data
from willing users with appropriate compensation. This approach does not
require the server to have any a priori knowledge about the users' privacy
preference, and provides an efficient solution for maximizing the social
welfare in the investigated problem.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 07:44:34 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 04:25:31 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 09:43:59 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Han",
"Bin",
""
],
[
"Feng",
"Di",
""
],
[
"Wang",
"Jie",
""
],
[
"Schotten",
"Hans D.",
""
]
] |
2503.23020 | Tatsuya Yoshida | Tatsuya Yoshida and Eric Gaidos | Water-Cooled (sub)-Neptunes Get Better Gas Mileage | 11 pages, 4 figures | A&A 696, L13 (2025) | 10.1051/0004-6361/202553667 | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | The demographics of sub-Jovian planets around low-mass stars is dominated by
populations of ``sub-Neptunes" and ``super-Earths", distinguished by the
presence or absence of envelopes of low-molecular weight volatiles, i.e., H2,
He, and H2O. The current paradigm is that sub-Neptunes on close-in orbits
evolve into super-Earths via atmospheric escape driven by high-energy stellar
irradiation. We use an integrated hydrodynamic-radiation-chemical network model
of outflow to demonstrate that this escape is modulated by the abundance of
H2O, an efficient infrared coolant. Increasing H2O/H2 at the base of the flow
induces an order-of-magnitude decline in escape rate, with definitive
consequences for retention of envelopes over Gyr. We show that saturation
limits on H2O in the upper atmospheres of temperate sub-Neptunes could explain
the paradoxical observations that these objects disappear more rapidly than
their counterparts closer to their host stars. We also propose that the
scarcity of sub-Neptunes around very low mass stars could be related to the
water-poor chemistry of their antecedent protoplanetary disks. Observations of
atmospheric H2O by JWST as well as searches for atmospheric escape from younger
planets using H and He lines could test these predictions.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 08:58:12 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Yoshida",
"Tatsuya",
""
],
[
"Gaidos",
"Eric",
""
]
] |
2503.23425 | Rasim Y{\i}lmaz | Rasim Y{\i}lmaz | Towards the Double Copy Formulation for the Abelian Sector of Heterotic
Double Field Theory | 7 pages | null | null | null | hep-th gr-qc | http://creativecommons.org/licenses/by/4.0/ | It has been shown that double field theory (DFT) action can be constructed in
a perturbative manner up to cubic order by using double copy of Yang-Mills.
This construction can also be extended by starting with higher-derivative
Yang-Mills theory to obtain higher-derivative DFT action. In this work, I try
to use this classical double copy formulation for Abelian subsector of
heterotic DFT, starting with a Yang-Mills theory including a charged scalar
field, namely (Y M + \phi^3) theory. Although the methods I used do not give an
exact match, they give some promising results.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 12:57:38 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:47:58 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Yılmaz",
"Rasim",
""
]
] |
2503.23520 | Christian Grussler | Christian Grussler | Tractable Characterization of Discrete-Time Periodic Monotonicity
Preserving Systems | null | null | null | null | math.OC math.DS | http://creativecommons.org/publicdomain/zero/1.0/ | This paper studies three classes of discrete-time linear time-invariant
systems, which differ by the set of periodic signals that they leave invariant.
The first class preserves the property of periodic monotonicity, i.e.,
period-wise unimodality. The second class is invariant to signals with at most
two sign changes per period, and the third class results from the second by
additionally requiring that periodic signals with zero sign-changes are mapped
to the same kind. We provide tractable characterizations for each system class
by the use and extension of total positivity theory and combination with its
geometric interpretations. In particular, central to our results is the
characterization of sequentially convex contours.
Moreover, as many static non-linearities, e.g., ideal relay, saturation,
sigmoid function, quantizer, etc. also preserve these signal sets, our
invariance characterizations also apply to the loop gain of Lur'e feedback
systems. Thus, potentially forming the base for new developments of
signal-based fixed-point theorems towards the prediction of self-sustained
oscillations. In particular, our examples provide first indications for how the
property of periodic monotonicity preservation is valuable to the study of
relay feedback systems.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 17:03:38 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 20:31:09 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Grussler",
"Christian",
""
]
] |
2503.23565 | Friedrich Bauermeister | Friedrich Bauermeister | Topological consequences of null-geodesic refocusing and applications to
$Z^x$ manifolds | 21 pages. version 2: made small changes to sections 3 and 6, added
references, fixed typos | null | null | null | math.DG gr-qc math-ph math.GT math.MP math.SG | http://creativecommons.org/licenses/by/4.0/ | Let $(M,h)$ be a connected, complete Riemannian manifold, let $x\in M$ and
$l>0$. Then $M$ is called a $Z^x$ manifold if all geodesics starting at $x$
return to $x$ and it is called a $Y^x_l$ manifold if every unit-speed geodesic
starting at $x$ returns to $x$ at time $l$. It is unknown whether there are
$Z^x$ manifolds that are not $Y^x_l$-manifolds for some $l>0$. By the
B\'erard-Bergery theorem, any $Y^x_l$ manifold of dimension at least $2$ is
compact with finite fundamental group. We prove the same result for $Z^x$
manifolds $M$ for which all unit-speed geodesics starting at $x$ return to $x$
in uniformly bounded time. We also prove that any $Z^x$ manifold $(M,h)$ with
$h$ analytic is a $Y^x_l$ manifold for some $l>0$. We start by defining a class
of globally hyperbolic spacetimes (called observer-refocusing) such that any
$Z^x$ manifold is the Cauchy surface of some observer-refocusing spacetime. We
then prove that under suitable conditions the Cauchy surfaces of
observer-refocusing spacetimes are compact with finite fundamental group and
show that analytic observer-refocusing spacetimes of dimension at least $3$ are
strongly refocusing. We end by stating a contact-theoretic conjecture analogous
to our results in Riemannian and Lorentzian geometry.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 19:09:19 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 23:31:08 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Bauermeister",
"Friedrich",
""
]
] |
2503.23720 | Yiyang Zhang | Jiaqi Chen, Li Lin Yang, Yiyang Zhang | On an approach to canonicalizing elliptic Feynman integrals | 45 pages, 2 figures, 1 ancillary file; v2: typos fixed and text
improved | null | null | null | hep-th hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a systematic approach to systematically constructing canonical
bases for univariate elliptic Feynman integrals with multiple kinematic scales,
which frequently arise in phenomenologically relevant scattering processes. The
construction is performed in the Legendre normal form of elliptic curves, where
the geometric properties of the curves are simple and explicit, and further
kinematic singularities are present as marked points. The simplicity of the
normal form facilitates the construction of universal canonical bases with an
arbitrary number of marked points, which can be mapped to generic univariate
elliptic integral families via an appropriate M\"obius transformation.
Therefore, our result can be readily applied to these cases straightforwardly.
As a demonstration, we discuss the application of our method to several
concrete examples, where we show that the $\varepsilon$-factorization of
sub-sector dependence can also be simply done.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 04:42:54 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:23:50 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Jiaqi",
""
],
[
"Yang",
"Li Lin",
""
],
[
"Zhang",
"Yiyang",
""
]
] |
2503.23921 | Hidenori Fukaya | Shoto Aoki, Hidenori Fukaya, Mikio Furuta, Shinichiroh Matsuo, Tetsuya
Onogi, and Satoshi Yamaguchi | $K$-theoretic computation of the Atiyah(-Patodi)-Singer index of lattice
Dirac operators | 19 pages, 6 figures, minor corrections and references added | null | null | OU-HET-1267 | hep-th hep-lat math.KT | http://creativecommons.org/licenses/by/4.0/ | We show that the Wilson Dirac operator in lattice gauge theory can be
identified as a mathematical object in $K$-theory and that its associated
spectral flow is equal to the index. In comparison to the standard lattice
Dirac operator index, our formulation does not require the Ginsparg-Wilson
relation and has broader applicability to systems with boundaries and to the
mod-two version of the indices in general dimensions. We numerically verify
that the $K$ and $KO$ group formulas reproduce the known index theorems in
continuum theory. We examine the Atiyah-Singer index on a flat two-dimensional
torus and, for the first time, demonstrate that the Atiyah-Patodi-Singer index
with nontrivial curved boundaries, as well as the mod-two versions, can be
computed on a lattice.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 10:11:44 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 04:46:42 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Aoki",
"Shoto",
""
],
[
"Fukaya",
"Hidenori",
""
],
[
"Furuta",
"Mikio",
""
],
[
"Matsuo",
"Shinichiroh",
""
],
[
"Onogi",
"Tetsuya",
""
],
[
"Yamaguchi",
"Satoshi",
""
]
] |
2503.24043 | Zhenkai Qin | Jiahui Lu, Shuang Wu, Zhenkai Qin, Guifang Yang | Frequency-Aware Attention-LSTM for PM$_{2.5}$ Time Series Forecasting | null | null | null | null | cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | To enhance the accuracy and robustness of PM$_{2.5}$ concentration
forecasting, this paper introduces FALNet, a Frequency-Aware LSTM Network that
integrates frequency-domain decomposition, temporal modeling, and
attention-based refinement. The model first applies STL and FFT to extract
trend, seasonal, and denoised residual components, effectively filtering out
high-frequency noise. The filtered residuals are then fed into a stacked LSTM
to capture long-term dependencies, followed by a multi-head attention mechanism
that dynamically focuses on key time steps. Experiments conducted on real-world
urban air quality datasets demonstrate that FALNet consistently outperforms
conventional models across standard metrics such as MAE, RMSE, and $R^2$. The
model shows strong adaptability in capturing sharp fluctuations during
pollution peaks and non-stationary conditions. These results validate the
effectiveness and generalizability of FALNet for real-time air pollution
prediction, environmental risk assessment, and decision-making support.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 13:07:33 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 13:59:14 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 15:16:00 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Lu",
"Jiahui",
""
],
[
"Wu",
"Shuang",
""
],
[
"Qin",
"Zhenkai",
""
],
[
"Yang",
"Guifang",
""
]
] |
2503.24318 | Walter Krawec | Walter O. Krawec | Quantum Conference Key Agreement with Classical Advantage Distillation | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we prove security of a quantum conference key agreement (QCKA)
protocol augmented with a classical advantage distillation (CAD) protocol. We
derive a proof of security, in the finite key setting, that is able to bound
the secure key rate for any general, coherent, attack. We evaluate the
performance of the system, showing our result can improve the noise tolerance
of the protocol in some, but not all, scenarios.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 17:07:36 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:41:07 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Krawec",
"Walter O.",
""
]
] |
2504.00762 | Jianhao Chen | Jianhao Chen, Zishuo Xun, Bocheng Zhou, Han Qi, Qiaosheng Zhang, Yang
Chen, Wei Hu, Yuzhong Qu, Wanli Ouyang, Shuyue Hu | Do We Truly Need So Many Samples? Multi-LLM Repeated Sampling
Efficiently Scales Test-Time Compute | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper presents a simple, effective, and cost-efficient strategy to
improve LLM performance by scaling test-time compute. Our strategy builds upon
the repeated-sampling-then-voting framework, with a novel twist: incorporating
multiple models, even weaker ones, to leverage their complementary strengths
that potentially arise from diverse training data and paradigms. By using
consistency as a signal, our strategy dynamically switches between models.
Theoretical analysis highlights the efficiency and performance advantages of
our strategy. Extensive experiments on six datasets demonstrate that our
strategy not only outperforms self-consistency and state-of-the-art multi-agent
debate approaches, but also significantly reduces inference costs.
Additionally, ModelSwitch requires only a few comparable LLMs to achieve
optimal performance and can be extended with verification methods,
demonstrating the potential of leveraging multiple LLMs in the
generation-verification paradigm.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 13:13:43 GMT"
},
{
"version": "v2",
"created": "Wed, 2 Apr 2025 08:55:04 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 06:58:14 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Jianhao",
""
],
[
"Xun",
"Zishuo",
""
],
[
"Zhou",
"Bocheng",
""
],
[
"Qi",
"Han",
""
],
[
"Zhang",
"Qiaosheng",
""
],
[
"Chen",
"Yang",
""
],
[
"Hu",
"Wei",
""
],
[
"Qu",
"Yuzhong",
""
],
[
"Ouyang",
"Wanli",
""
],
[
"Hu",
"Shuyue",
""
]
] |
2504.00846 | Paul Smith | Chutian Ma, Giacinto Paolo Saggese, Paul Smith | The effect of latency on optimal order execution policy | 15 figures, 2 tables; additional references added in v2 | null | null | null | q-fin.MF math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Market participants regularly send bid and ask quotes to exchange-operated
limit order books. This creates an optimization challenge where their potential
profit is determined by their quoted price and how often their orders are
successfully executed. The expected profit from successful execution at a
favorable limit price needs to be balanced against two key risks: (1) the
possibility that orders will remain unfilled, which hinders the trading agenda
and leads to greater price uncertainty, and (2) the danger that limit orders
will be executed as market orders, particularly in the presence of order
submission latency, which in turn results in higher transaction costs. In this
paper, we consider a stochastic optimal control problem where a risk-averse
trader attempts to maximize profit while balancing risk. The market is modeled
using Brownian motion to represent the price uncertainty. We analyze the
relationship between fill probability, limit price, and order submission
latency. We derive closed-form approximations of these quantities that perform
well in the practical regime of interest. Then, we utilize a mean-variance
method where our total reward function features a risk-tolerance parameter to
quantify the combined risk and profit.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 14:33:23 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:24:24 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ma",
"Chutian",
""
],
[
"Saggese",
"Giacinto Paolo",
""
],
[
"Smith",
"Paul",
""
]
] |
2504.00883 | Zhenyi Liao | Zhenyi Liao, Qingsong Xie, Yanhao Zhang, Zijian Kong, Haonan Lu,
Zhenyu Yang, Zhijie Deng | Improved Visual-Spatial Reasoning via R1-Zero-Like Training | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Increasing attention has been placed on improving the reasoning capacities of
multi-modal large language models (MLLMs). As the cornerstone for AI agents
that function in the physical realm, video-based visual-spatial intelligence
(VSI) emerges as one of the most pivotal reasoning capabilities of MLLMs. This
work conducts a first, in-depth study on improving the visual-spatial reasoning
of MLLMs via R1-Zero-like training. Technically, we first identify that the
visual-spatial reasoning capacities of small- to medium-sized Qwen2-VL models
cannot be activated via Chain of Thought (CoT) prompts. We then incorporate
GRPO training for improved visual-spatial reasoning, using the carefully
curated VSI-100k dataset, following DeepSeek-R1-Zero. During the investigation,
we identify the necessity to keep the KL penalty (even with a small value) in
GRPO. With just 120 GPU hours, our vsGRPO-2B model, fine-tuned from
Qwen2-VL-2B, can outperform the base model by 12.1% and surpass GPT-4o.
Moreover, our vsGRPO-7B model, fine-tuned from Qwen2-VL-7B, achieves
performance comparable to that of the best open-source model
LLaVA-NeXT-Video-72B. Additionally, we compare vsGRPO to supervised fine-tuning
and direct preference optimization baselines and observe strong performance
superiority. The code and dataset will be available soon.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 15:11:11 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 20:12:57 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Liao",
"Zhenyi",
""
],
[
"Xie",
"Qingsong",
""
],
[
"Zhang",
"Yanhao",
""
],
[
"Kong",
"Zijian",
""
],
[
"Lu",
"Haonan",
""
],
[
"Yang",
"Zhenyu",
""
],
[
"Deng",
"Zhijie",
""
]
] |
2504.01222 | Israel Campero Jurado | Israel Campero Jurado, Pieter Gijsbers, Joaquin Vanschoren | AutoML Benchmark with shorter time constraints and early stopping | Workshop on the Future of Machine Learning Data Practices and
Repositories, ICLR 2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated Machine Learning (AutoML) automatically builds machine learning
(ML) models on data. The de facto standard for evaluating new AutoML frameworks
for tabular data is the AutoML Benchmark (AMLB). AMLB proposed to evaluate
AutoML frameworks using 1- and 4-hour time budgets across 104 tasks. We argue
that shorter time constraints should be considered for the benchmark because of
their practical value, such as when models need to be retrained with high
frequency, and to make AMLB more accessible. This work considers two ways in
which to reduce the overall computation used in the benchmark: smaller time
constraints and the use of early stopping. We conduct evaluations of 11 AutoML
frameworks on 104 tasks with different time constraints and find the relative
ranking of AutoML frameworks is fairly consistent across time constraints, but
that using early-stopping leads to a greater variety in model performance.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 22:22:15 GMT"
},
{
"version": "v2",
"created": "Fri, 4 Apr 2025 15:38:37 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Apr 2025 23:15:34 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Jurado",
"Israel Campero",
""
],
[
"Gijsbers",
"Pieter",
""
],
[
"Vanschoren",
"Joaquin",
""
]
] |
2504.01832 | Ali J. Ghandour | Khalil Al Salahat, Mohamad El Moussawi, Veera Ganesh Yalla, Ali J.
Ghandour | Quantum Meets SAR: A Novel Range-Doppler Algorithm for Next-Gen Earth
Observation | null | null | null | null | quant-ph eess.SP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Synthetic Aperture Radar (SAR) plays a vital role in remote sensing due to
its ability to capture high-resolution images regardless of weather conditions
or daylight. However, to transform the raw SAR signals into interpretable
imagery, advanced data processing techniques are essential. A widely used
technique for this purpose is the Range Doppler Algorithm (RDA), which takes
advantage of Fast Fourier Transform (FFT) to convert signals into the frequency
domain for further processing. However, the computational cost of this approach
becomes significant when dealing with large datasets. This paper presents a
Quantum Range Doppler Algorithm (QRDA) that utilizes the Quantum Fourier
Transform (QFT) to accelerate processing compared to the classical FFT.
Furthermore, it introduces a quantum implementation of the Range Cell Migration
Correction (RCMC) in the Fourier domain, a critical step in the RDA pipeline,
and evaluates its performance relative to its classical counterpart.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 15:40:12 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 10:06:02 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Salahat",
"Khalil Al",
""
],
[
"Moussawi",
"Mohamad El",
""
],
[
"Yalla",
"Veera Ganesh",
""
],
[
"Ghandour",
"Ali J.",
""
]
] |
2504.01880 | Sriram Shastry | B Sriram Shastry | Partition function zeros of quantum many-body systems | 32 pages, 4 figures | null | null | null | cond-mat.stat-mech cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We present a method for calculating the Yang-Lee partition function zeros of
a translationally invariant model of lattice fermions, exemplified by the
Hubbard model. The method rests on a theorem involving the single electron
self-energy $\Sigma_\sigma(k, i \omega_n)$ in the imaginary time Matsubara
formulation. The theorem maps the Yang-Lee zeros to a set of wavevector and
spin labeled virtual energies $\xi_{k \sigma}$. These, thermodynamically
derived virtual energies, are solutions of equations involving the self-energy
at corresponding $k\sigma$'s. Examples of the method in simplified situations
are provided.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 16:39:15 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 16:59:03 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Shastry",
"B Sriram",
""
]
] |
2504.02153 | Nathan TeBlunthuis PhD | Nathan TeBlunthuis | Niche Dynamics in Complex Online Community Ecosystems | 12 pages, 4 figures, conditionally accepted to ICWSM 2025 | null | null | null | cs.SI cs.CY cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Online communities are important organizational forms where members socialize
and share information. Curiously, different online communities often overlap
considerably in topic and membership. Recent research has investigated
competition and mutualism among overlapping online communities through the lens
of organizational ecology; however, it has not accounted for how the nonlinear
dynamics of online attention may lead to episodic competition and mutualism.
Neither has it explored the origins of competition and mutualism in the
processes by which online communities select or adapt to their niches. This
paper presents a large-scale study of 8,806 Reddit communities belonging to
1,919 clusters of high user overlap over a 5-year period. The method uses
nonlinear time series methods to infer bursty, often short-lived ecological
dynamics. Results reveal that mutualism episodes are longer lived and slightly
more frequent than competition episodes. Next, it tests whether online
communities find their niches by specializing to avoid competition using panel
regression models. It finds that competitive ecological interactions lead to
decreasing topic and user overlaps; however, changes that decrease such niche
overlaps do not lead to mutualism. The discussion proposes that future designs
may enable online community ecosystem management by informing online community
leaders to organize "spin-off" communities or via feeds and recommendations.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 21:56:46 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:13:02 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"TeBlunthuis",
"Nathan",
""
]
] |
2504.02348 | Sourav Chatterjee | Sourav Chatterjee | Rigorous results for timelike Liouville field theory | 87 pages. Additional references and minor updates in this revision | null | null | null | math.PR hep-th math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Liouville field theory has long been a cornerstone of two-dimensional quantum
field theory and quantum gravity, which has attracted much recent attention in
the mathematics literature. Timelike Liouville field theory is a version of
Liouville field theory where the kinetic term in the action appears with a
negative sign, which makes it closer to a theory of quantum gravity than
ordinary (spacelike) Liouville field theory. Making sense of this "wrong sign"
requires a theory of Gaussian random variables with negative variance. Such a
theory is developed in this paper, and is used to prove the timelike DOZZ
formula for the $3$-point correlation function when the parameters satisfy the
so-called "charge neutrality condition". Expressions are derived also for the
$k$-point correlation functions for all $k\ge 3$, and it is shown that these
functions approach the correct semiclassical limits as the coupling constant is
sent to zero.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 07:33:00 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Apr 2025 08:07:57 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 14:56:05 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chatterjee",
"Sourav",
""
]
] |
2504.02578 | Andrew Allan | Andrew P. Allan and Aline A. Vidotto | Helium escape signatures are generally strongest during younger ages but
this age dependence is lost in the diversity of observed exoplanets | 14 pages, 8 figures (includes 3 in the appendix), 6 tables (2 in
appendix). Accepted for publication in Monthly Notices of the Royal
Astronomical Society Main Journal | null | 10.1093/mnras/staf566 | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Highly irradiated exoplanets undergo extreme hydrodynamic atmospheric escape,
due to their high level of received XUV flux. Over their lifetime, this escape
varies significantly, making evolution studies essential for interpreting the
growing number of observations of escaping planetary atmospheres. In a previous
work, we modelled this evolving escape, alongside one of its observable
tracers, the helium triplet transit signature at 1083nm. Using hydrodynamic and
ray-tracing models, we demonstrated that atmospheric escape and the
corresponding He 1083nm signature are stronger at younger ages, for a
0.3$~M_\text{J}$ gas-giant. Yet, the current literature includes several young
(<1Gyr) planets with weak or non-detections in He 1083nm. To understand this
apparent discrepancy, we now perform detailed modelling for many of these
systems. The resulting He 1083nm predictions align relatively well with the
observations. From our two studies, we conclude that for any given planet,
stronger atmospheric escape during younger ages produces deeper He 1083nm
absorption. However, for a population of exoplanets, the relation between
younger ages and stronger He absorptions is lost to the broad diversity of
their various other system parameters. Accordingly, for the current sample of
young, 1083nm-observed exoplanets, alternative trends take precedence. One such
trend is that planets with deeper geometrical transits exhibit more favourable
detections. Our modelling also agrees with the strong empirical trend in the
literature between $ EW \cdot R_{*}^{2}$ and $F_{\text{xuv}} \cdot
R_{\text{pl}}^2 / \Phi_{g}$. Additionally, we show that the coupling between
the lower and upper atmospheres is necessary for a robust prediction of the
1083nm signature.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 13:41:16 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 13:58:31 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Allan",
"Andrew P.",
""
],
[
"Vidotto",
"Aline A.",
""
]
] |
2504.02669 | Liang Tao | Tao Liang, Yongsheng Li, Xiaoping Zhai | Stability threshold for two-dimensional Boussinesq systems near-Couette
shear flow in a finite channel | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we investigate the stability threshold problem of the
two-dimensional Navier-Stokes Boussinesq(NSB) equations in a finite channel $
\T \times [-1,1]$, focusing on the stability around the near Couette shear flow
$ (U(y), 0)$, assuming the Navier slip boundary conditions are satisfied. In
particular, when the initial data for the vorticity resides in an anisotropic
Sobolev space of size $ O(\min \{ \mu^{\frac{1}{2}}, \nu^{\frac{1}{2}}\})$, and
the initial perturbation of the temperature resides in an anisotropic Sobolev
space of size $ O(\min \{ \mu, \nu\})$, we derive the nonlinear enhanced
dissipation effect and the inviscid damping effect for the NSB system.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 15:09:57 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 11:43:36 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Liang",
"Tao",
""
],
[
"Li",
"Yongsheng",
""
],
[
"Zhai",
"Xiaoping",
""
]
] |
2504.02705 | Min Jun Jo | Tarek M. Elgindi, Min Jun Jo | Cusp Formation in Vortex Patches | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove instantaneous cusp formation for any initial vortex patch with acute
corners. This was conjectured to occur in the numerical literature.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 15:42:06 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 22:35:09 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Elgindi",
"Tarek M.",
""
],
[
"Jo",
"Min Jun",
""
]
] |
2504.02896 | Jakub Sawicki | Jakub Sawicki | Synchronisation zweier Labialpfeifen | 4 pages, 3 figures, in German | null | null | null | nlin.AO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The synchronization of coupled organ pipes represents a nonlinear phenomenon.
The investigation of synchronization effects in organ pipes using a model of
time-delayed coupled Van der Pol oscillators shows that synchronization
improves pitch stability but can also lead to sound attenuation. The Arnold
tongue illustrates the influence of distance-dependent coupling on stability
and frequency adaptation. The results agree qualitatively well with experiments
and emphasize the relevance of nonlinear coupling mechanisms in organ
acoustics.
-- Die Synchronisation gekoppelter Orgelpfeifen stellt ein nichtlineares
Ph{\"a}nomen dar. Die Untersuchung von Synchronisationseffekten bei
Orgelpfeifen anhand eines Modells zeitverz{\"o}gert gekoppelter
Van-der-Pol-Oszillatoren zeigt, dass Synchronisation die
Tonh{\"o}henstabilit{\"a}t verbessert, aber auch zu Schalld{\"a}mpfung
f{\"u}hren kann. Die Arnold-Zunge verdeutlicht den Einfluss
distanzabh{\"a}ngiger Kopplung auf Stabilit{\"a}t und Frequenzanpassung. Die
Ergebnisse stimmen qualitativ gut mit Experimenten {\"u}berein und betonen die
Relevanz nichtlinearer Kopplungsmechanismen in der Orgelakustik.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 21:41:04 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:12:08 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Sawicki",
"Jakub",
""
]
] |
2504.03193 | Xin Zhang | Xin Zhang and Robby T. Tan | Mamba as a Bridge: Where Vision Foundation Models Meet Vision Language
Models for Domain-Generalized Semantic Segmentation | Accepted to CVPR 2025 (Highlight) | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Vision Foundation Models (VFMs) and Vision-Language Models (VLMs) have gained
traction in Domain Generalized Semantic Segmentation (DGSS) due to their strong
generalization capabilities. However, existing DGSS methods often rely
exclusively on either VFMs or VLMs, overlooking their complementary strengths.
VFMs (e.g., DINOv2) excel at capturing fine-grained features, while VLMs (e.g.,
CLIP) provide robust text alignment but struggle with coarse granularity.
Despite their complementary strengths, effectively integrating VFMs and VLMs
with attention mechanisms is challenging, as the increased patch tokens
complicate long-sequence modeling. To address this, we propose MFuser, a novel
Mamba-based fusion framework that efficiently combines the strengths of VFMs
and VLMs while maintaining linear scalability in sequence length. MFuser
consists of two key components: MVFuser, which acts as a co-adapter to jointly
fine-tune the two models by capturing both sequential and spatial dynamics; and
MTEnhancer, a hybrid attention-Mamba module that refines text embeddings by
incorporating image priors. Our approach achieves precise feature locality and
strong text alignment without incurring significant computational overhead.
Extensive experiments demonstrate that MFuser significantly outperforms
state-of-the-art DGSS methods, achieving 68.20 mIoU on synthetic-to-real and
71.87 mIoU on real-to-real benchmarks. The code is available at
https://github.com/devinxzhang/MFuser.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 05:44:45 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:14:58 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Zhang",
"Xin",
""
],
[
"Tan",
"Robby T.",
""
]
] |
2504.03201 | Jonathan Arnaud | Jonathan S. Arnaud, Xian-Zhu Tang, and Christopher J. McDevitt | A Runaway Electron Avalanche Surrogate for Partially Ionized Plasmas | Submitted to Nuclear Fusion | null | null | null | physics.plasm-ph | http://creativecommons.org/licenses/by/4.0/ | A physics-constrained deep learning surrogate that predicts the exponential
``avalanche'' growth rate of runaway electrons (REs) for a plasma containing
partially ionized impurities is developed. Specifically, a physics-informed
neural network (PINN) that learns the adjoint of the relativistic Fokker-Planck
equation in steady-state is derived, enabling a rapid surrogate of the RE
avalanche for a broad range of plasma parameters, motivating a path towards an
ML-accelerated integrated description of a tokamak disruption. A steady-state
power balance equation together with atomic physics data is embedded directly
into the PINN, thus limiting the PINN to train across physically consistent
temperatures and charge state distributions. This restricted training domain
enables accurate predictions of the PINN while drastically reducing the
computational cost of training the model. In addition, a novel closure for the
relativistic electron population used when evaluating the secondary source of
REs is developed that enables improved accuracy compared to a
Rosenbluth-Putvinski source. The avalanche surrogate is verified against Monte
Carlo simulations, where it is shown to accurately predict the RE avalanche
growth rate across a broad range of plasma parameters encompassing distinct
tokamak disruption scenarios.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 06:12:24 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 21:47:30 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Arnaud",
"Jonathan S.",
""
],
[
"Tang",
"Xian-Zhu",
""
],
[
"McDevitt",
"Christopher J.",
""
]
] |
2504.03278 | Piero Fariselli | Guido Barducci, Ivan Rossi, Francesco Codic\`e, Cesare Rollo, Valeria
Repetto, Corrado Pancotti, Virginia Iannibelli, Tiziana Sanavia and Piero
Fariselli | JanusDDG: A Thermodynamics-Compliant Model for Sequence-Based Protein
Stability via Two-Fronts Multi-Head Attention | 20 pages, 11 figures | null | null | null | q-bio.QM cs.AI cs.LG physics.comp-ph | http://creativecommons.org/licenses/by-sa/4.0/ | Understanding how residue variations affect protein stability is crucial for
designing functional proteins and deciphering the molecular mechanisms
underlying disease-related mutations. Recent advances in protein language
models (PLMs) have revolutionized computational protein analysis, enabling,
among other things, more accurate predictions of mutational effects. In this
work, we introduce JanusDDG, a deep learning framework that leverages
PLM-derived embeddings and a bidirectional cross-attention transformer
architecture to predict $\Delta \Delta G$ of single and multiple-residue
mutations while simultaneously being constrained to respect fundamental
thermodynamic properties, such as antisymmetry and transitivity. Unlike
conventional self-attention, JanusDDG computes queries (Q) and values (V) as
the difference between wild-type and mutant embeddings, while keys (K)
alternate between the two. This cross-interleaved attention mechanism enables
the model to capture mutation-induced perturbations while preserving essential
contextual information. Experimental results show that JanusDDG achieves
state-of-the-art performance in predicting $\Delta \Delta G$ from sequence
alone, matching or exceeding the accuracy of structure-based methods for both
single and multiple mutations. Code
Availability:https://github.com/compbiomed-unito/JanusDDG
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 09:02:32 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 18:11:58 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Barducci",
"Guido",
""
],
[
"Rossi",
"Ivan",
""
],
[
"Codicè",
"Francesco",
""
],
[
"Rollo",
"Cesare",
""
],
[
"Repetto",
"Valeria",
""
],
[
"Pancotti",
"Corrado",
""
],
[
"Iannibelli",
"Virginia",
""
],
[
"Sanavia",
"Tiziana",
""
],
[
"Fariselli",
"Piero",
""
]
] |
2504.03562 | Emil Straube | Emil J. Straube | Regularity in the $\overline{\partial}$--Neumann problem, D'Angelo
forms, and Diederich--Forn\ae ss index | null | null | null | null | math.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This article chronicles a development that started around 1990 with
\cite{BoasStraube91}, where the authors showed that if a smooth bounded
pseudoconvex domain $\Omega$ in $\mathbb{C}^{n}$ admits a defining function
that is plurisubharmonic at points of the boundary, then the
$\overline{\partial}$--Neumann operators on $\Omega$ preserve the Sobolev
spaces $W^{s}_{(0,q)}(\Omega)$, $s\geq 0$. The same authors then proved a
further regularity result and made explicit the role of D'Angelo forms for
regularity (\cite{BoasStraube93}). A few years later, Kohn (\cite{Kohn99})
initiated a quantitative study of the results in \cite{BoasStraube91} by
relating the Sobolev level up to which regularity holds to the
Diederich--Forn\ae ss index of the domain. Many of these ideas were synthesized
and developed further by Harrington
(\cite{Harrington11,Harrington19,Harrington22}). Then, around 2020, Liu
(\cite{Liu19b, Liu19}) and Yum (\cite{Yum21}) discovered that the DF--index is
closely related to certain differential inequalities involving D'Angelo forms.
This relationship in turn led to a recent new result which supports the
conjecture that DF--index one should imply global regularity in the
$\overline{\partial}$--Neumann problem (\cite{LiuStraube22}). Much of the work
described above relies heavily on Kohn's groundbreaking contributions to the
regularity theory of the $\overline{\partial}$--Neumann problem.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 16:11:24 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 19:02:27 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Straube",
"Emil J.",
""
]
] |
2504.03624 | Deepak Narayanan | NVIDIA: Aaron Blakeman, Aarti Basant, Abhinav Khattar, Adithya
Renduchintala, Akhiad Bercovich, Aleksander Ficek, Alexis Bjorlin, Ali
Taghibakhshi, Amala Sanjay Deshmukh, Ameya Sunil Mahabaleshwarkar, Andrew
Tao, Anna Shors, Ashwath Aithal, Ashwin Poojary, Ayush Dattagupta, Balaram
Buddharaju, Bobby Chen, Boris Ginsburg, Boxin Wang, Brandon Norick, Brian
Butterfield, Bryan Catanzaro, Carlo del Mundo, Chengyu Dong, Christine
Harvey, Christopher Parisien, Dan Su, Daniel Korzekwa, Danny Yin, Daria
Gitman, David Mosallanezhad, Deepak Narayanan, Denys Fridman, Dima Rekesh,
Ding Ma, Dmytro Pykhtar, Dong Ahn, Duncan Riach, Dusan Stosic, Eileen Long,
Elad Segal, Ellie Evans, Eric Chung, Erick Galinkin, Evelina Bakhturina, Ewa
Dobrowolska, Fei Jia, Fuxiao Liu, Gargi Prasad, Gerald Shen, Guilin Liu, Guo
Chen, Haifeng Qian, Helen Ngo, Hongbin Liu, Hui Li, Igor Gitman, Ilia
Karmanov, Ivan Moshkov, Izik Golan, Jan Kautz, Jane Polak Scowcroft, Jared
Casper, Jarno Seppanen, Jason Lu, Jason Sewall, Jiaqi Zeng, Jiaxuan You,
Jimmy Zhang, Jing Zhang, Jining Huang, Jinze Xue, Jocelyn Huang, Joey Conway,
John Kamalu, Jon Barker, Jonathan Cohen, Joseph Jennings, Jupinder Parmar,
Karan Sapra, Kari Briski, Kateryna Chumachenko, Katherine Luna, Keshav
Santhanam, Kezhi Kong, Kirthi Sivamani, Krzysztof Pawelec, Kumar Anik, Kunlun
Li, Lawrence McAfee, Leon Derczynski, Lindsey Pavao, Luis Vega, Lukas
Voegtle, Maciej Bala, Maer Rodrigues de Melo, Makesh Narsimhan Sreedhar,
Marcin Chochowski, Markus Kliegl, Marta Stepniewska-Dziubinska, Matthieu Le,
Matvei Novikov, Mehrzad Samadi, Michael Andersch, Michael Evans, Miguel
Martinez, Mike Chrzanowski, Mike Ranzinger, Mikolaj Blaz, Misha Smelyanskiy,
Mohamed Fawzy, Mohammad Shoeybi, Mostofa Patwary, Nayeon Lee, Nima Tajbakhsh,
Ning Xu, Oleg Rybakov, Oleksii Kuchaiev, Olivier Delalleau, Osvald Nitski,
Parth Chadha, Pasha Shamis, Paulius Micikevicius, Pavlo Molchanov, Peter
Dykas, Philipp Fischer, Pierre-Yves Aquilanti, Piotr Bialecki, Prasoon
Varshney, Pritam Gundecha, Przemek Tredak, Rabeeh Karimi, Rahul Kandu, Ran
El-Yaniv, Raviraj Joshi, Roger Waleffe, Ruoxi Zhang, Sabrina Kavanaugh, Sahil
Jain, Samuel Kriman, Sangkug Lym, Sanjeev Satheesh, Saurav Muralidharan, Sean
Narenthiran, Selvaraj Anandaraj, Seonmyeong Bak, Sergey Kashirsky, Seungju
Han, Shantanu Acharya, Shaona Ghosh, Sharath Turuvekere Sreenivas, Sharon
Clay, Shelby Thomas, Shrimai Prabhumoye, Shubham Pachori, Shubham Toshniwal,
Shyamala Prayaga, Siddhartha Jain, Sirshak Das, Slawek Kierat, Somshubra
Majumdar, Song Han, Soumye Singhal, Sriharsha Niverty, Stefania Alborghetti,
Suseella Panguluri, Swetha Bhendigeri, Syeda Nahida Akter, Szymon Migacz, Tal
Shiri, Terry Kong, Timo Roman, Tomer Ronen, Trisha Saar, Tugrul Konuk, Tuomas
Rintamaki, Tyler Poon, Ushnish De, Vahid Noroozi, Varun Singh, Vijay
Korthikanti, Vitaly Kurin, Wasi Uddin Ahmad, Wei Du, Wei Ping, Wenliang Dai,
Wonmin Byeon, Xiaowei Ren, Yao Xu, Yejin Choi, Yian Zhang, Ying Lin, Yoshi
Suhara, Zhiding Yu, Zhiqi Li, Zhiyu Li, Zhongbo Zhu, Zhuolin Yang, Zijia Chen | Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer
Models | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | As inference-time scaling becomes critical for enhanced reasoning
capabilities, it is increasingly becoming important to build models that are
efficient to infer. We introduce Nemotron-H, a family of 8B and 56B/47B hybrid
Mamba-Transformer models designed to reduce inference cost for a given accuracy
level. To achieve this goal, we replace the majority of self-attention layers
in the common Transformer model architecture with Mamba layers that perform
constant computation and require constant memory per generated token. We show
that Nemotron-H models offer either better or on-par accuracy compared to other
similarly-sized state-of-the-art open-sourced Transformer models (e.g.,
Qwen-2.5-7B/72B and Llama-3.1-8B/70B), while being up to 3$\times$ faster at
inference. To further increase inference speed and reduce the memory required
at inference time, we created Nemotron-H-47B-Base from the 56B model using a
new compression via pruning and distillation technique called MiniPuzzle.
Nemotron-H-47B-Base achieves similar accuracy to the 56B model, but is 20%
faster to infer. In addition, we introduce an FP8-based training recipe and
show that it can achieve on par results with BF16-based training. This recipe
is used to train the 56B model. We are releasing Nemotron-H base model
checkpoints with support in Hugging Face and NeMo.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 17:41:58 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 05:31:53 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 14:36:01 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"NVIDIA",
"",
""
],
[
":",
"",
""
],
[
"Blakeman",
"Aaron",
""
],
[
"Basant",
"Aarti",
""
],
[
"Khattar",
"Abhinav",
""
],
[
"Renduchintala",
"Adithya",
""
],
[
"Bercovich",
"Akhiad",
""
],
[
"Ficek",
"Aleksander",
""
],
[
"Bjorlin",
"Alexis",
""
],
[
"Taghibakhshi",
"Ali",
""
],
[
"Deshmukh",
"Amala Sanjay",
""
],
[
"Mahabaleshwarkar",
"Ameya Sunil",
""
],
[
"Tao",
"Andrew",
""
],
[
"Shors",
"Anna",
""
],
[
"Aithal",
"Ashwath",
""
],
[
"Poojary",
"Ashwin",
""
],
[
"Dattagupta",
"Ayush",
""
],
[
"Buddharaju",
"Balaram",
""
],
[
"Chen",
"Bobby",
""
],
[
"Ginsburg",
"Boris",
""
],
[
"Wang",
"Boxin",
""
],
[
"Norick",
"Brandon",
""
],
[
"Butterfield",
"Brian",
""
],
[
"Catanzaro",
"Bryan",
""
],
[
"del Mundo",
"Carlo",
""
],
[
"Dong",
"Chengyu",
""
],
[
"Harvey",
"Christine",
""
],
[
"Parisien",
"Christopher",
""
],
[
"Su",
"Dan",
""
],
[
"Korzekwa",
"Daniel",
""
],
[
"Yin",
"Danny",
""
],
[
"Gitman",
"Daria",
""
],
[
"Mosallanezhad",
"David",
""
],
[
"Narayanan",
"Deepak",
""
],
[
"Fridman",
"Denys",
""
],
[
"Rekesh",
"Dima",
""
],
[
"Ma",
"Ding",
""
],
[
"Pykhtar",
"Dmytro",
""
],
[
"Ahn",
"Dong",
""
],
[
"Riach",
"Duncan",
""
],
[
"Stosic",
"Dusan",
""
],
[
"Long",
"Eileen",
""
],
[
"Segal",
"Elad",
""
],
[
"Evans",
"Ellie",
""
],
[
"Chung",
"Eric",
""
],
[
"Galinkin",
"Erick",
""
],
[
"Bakhturina",
"Evelina",
""
],
[
"Dobrowolska",
"Ewa",
""
],
[
"Jia",
"Fei",
""
],
[
"Liu",
"Fuxiao",
""
],
[
"Prasad",
"Gargi",
""
],
[
"Shen",
"Gerald",
""
],
[
"Liu",
"Guilin",
""
],
[
"Chen",
"Guo",
""
],
[
"Qian",
"Haifeng",
""
],
[
"Ngo",
"Helen",
""
],
[
"Liu",
"Hongbin",
""
],
[
"Li",
"Hui",
""
],
[
"Gitman",
"Igor",
""
],
[
"Karmanov",
"Ilia",
""
],
[
"Moshkov",
"Ivan",
""
],
[
"Golan",
"Izik",
""
],
[
"Kautz",
"Jan",
""
],
[
"Scowcroft",
"Jane Polak",
""
],
[
"Casper",
"Jared",
""
],
[
"Seppanen",
"Jarno",
""
],
[
"Lu",
"Jason",
""
],
[
"Sewall",
"Jason",
""
],
[
"Zeng",
"Jiaqi",
""
],
[
"You",
"Jiaxuan",
""
],
[
"Zhang",
"Jimmy",
""
],
[
"Zhang",
"Jing",
""
],
[
"Huang",
"Jining",
""
],
[
"Xue",
"Jinze",
""
],
[
"Huang",
"Jocelyn",
""
],
[
"Conway",
"Joey",
""
],
[
"Kamalu",
"John",
""
],
[
"Barker",
"Jon",
""
],
[
"Cohen",
"Jonathan",
""
],
[
"Jennings",
"Joseph",
""
],
[
"Parmar",
"Jupinder",
""
],
[
"Sapra",
"Karan",
""
],
[
"Briski",
"Kari",
""
],
[
"Chumachenko",
"Kateryna",
""
],
[
"Luna",
"Katherine",
""
],
[
"Santhanam",
"Keshav",
""
],
[
"Kong",
"Kezhi",
""
],
[
"Sivamani",
"Kirthi",
""
],
[
"Pawelec",
"Krzysztof",
""
],
[
"Anik",
"Kumar",
""
],
[
"Li",
"Kunlun",
""
],
[
"McAfee",
"Lawrence",
""
],
[
"Derczynski",
"Leon",
""
],
[
"Pavao",
"Lindsey",
""
],
[
"Vega",
"Luis",
""
],
[
"Voegtle",
"Lukas",
""
],
[
"Bala",
"Maciej",
""
],
[
"de Melo",
"Maer Rodrigues",
""
],
[
"Sreedhar",
"Makesh Narsimhan",
""
],
[
"Chochowski",
"Marcin",
""
],
[
"Kliegl",
"Markus",
""
],
[
"Stepniewska-Dziubinska",
"Marta",
""
],
[
"Le",
"Matthieu",
""
],
[
"Novikov",
"Matvei",
""
],
[
"Samadi",
"Mehrzad",
""
],
[
"Andersch",
"Michael",
""
],
[
"Evans",
"Michael",
""
],
[
"Martinez",
"Miguel",
""
],
[
"Chrzanowski",
"Mike",
""
],
[
"Ranzinger",
"Mike",
""
],
[
"Blaz",
"Mikolaj",
""
],
[
"Smelyanskiy",
"Misha",
""
],
[
"Fawzy",
"Mohamed",
""
],
[
"Shoeybi",
"Mohammad",
""
],
[
"Patwary",
"Mostofa",
""
],
[
"Lee",
"Nayeon",
""
],
[
"Tajbakhsh",
"Nima",
""
],
[
"Xu",
"Ning",
""
],
[
"Rybakov",
"Oleg",
""
],
[
"Kuchaiev",
"Oleksii",
""
],
[
"Delalleau",
"Olivier",
""
],
[
"Nitski",
"Osvald",
""
],
[
"Chadha",
"Parth",
""
],
[
"Shamis",
"Pasha",
""
],
[
"Micikevicius",
"Paulius",
""
],
[
"Molchanov",
"Pavlo",
""
],
[
"Dykas",
"Peter",
""
],
[
"Fischer",
"Philipp",
""
],
[
"Aquilanti",
"Pierre-Yves",
""
],
[
"Bialecki",
"Piotr",
""
],
[
"Varshney",
"Prasoon",
""
],
[
"Gundecha",
"Pritam",
""
],
[
"Tredak",
"Przemek",
""
],
[
"Karimi",
"Rabeeh",
""
],
[
"Kandu",
"Rahul",
""
],
[
"El-Yaniv",
"Ran",
""
],
[
"Joshi",
"Raviraj",
""
],
[
"Waleffe",
"Roger",
""
],
[
"Zhang",
"Ruoxi",
""
],
[
"Kavanaugh",
"Sabrina",
""
],
[
"Jain",
"Sahil",
""
],
[
"Kriman",
"Samuel",
""
],
[
"Lym",
"Sangkug",
""
],
[
"Satheesh",
"Sanjeev",
""
],
[
"Muralidharan",
"Saurav",
""
],
[
"Narenthiran",
"Sean",
""
],
[
"Anandaraj",
"Selvaraj",
""
],
[
"Bak",
"Seonmyeong",
""
],
[
"Kashirsky",
"Sergey",
""
],
[
"Han",
"Seungju",
""
],
[
"Acharya",
"Shantanu",
""
],
[
"Ghosh",
"Shaona",
""
],
[
"Sreenivas",
"Sharath Turuvekere",
""
],
[
"Clay",
"Sharon",
""
],
[
"Thomas",
"Shelby",
""
],
[
"Prabhumoye",
"Shrimai",
""
],
[
"Pachori",
"Shubham",
""
],
[
"Toshniwal",
"Shubham",
""
],
[
"Prayaga",
"Shyamala",
""
],
[
"Jain",
"Siddhartha",
""
],
[
"Das",
"Sirshak",
""
],
[
"Kierat",
"Slawek",
""
],
[
"Majumdar",
"Somshubra",
""
],
[
"Han",
"Song",
""
],
[
"Singhal",
"Soumye",
""
],
[
"Niverty",
"Sriharsha",
""
],
[
"Alborghetti",
"Stefania",
""
],
[
"Panguluri",
"Suseella",
""
],
[
"Bhendigeri",
"Swetha",
""
],
[
"Akter",
"Syeda Nahida",
""
],
[
"Migacz",
"Szymon",
""
],
[
"Shiri",
"Tal",
""
],
[
"Kong",
"Terry",
""
],
[
"Roman",
"Timo",
""
],
[
"Ronen",
"Tomer",
""
],
[
"Saar",
"Trisha",
""
],
[
"Konuk",
"Tugrul",
""
],
[
"Rintamaki",
"Tuomas",
""
],
[
"Poon",
"Tyler",
""
],
[
"De",
"Ushnish",
""
],
[
"Noroozi",
"Vahid",
""
],
[
"Singh",
"Varun",
""
],
[
"Korthikanti",
"Vijay",
""
],
[
"Kurin",
"Vitaly",
""
],
[
"Ahmad",
"Wasi Uddin",
""
],
[
"Du",
"Wei",
""
],
[
"Ping",
"Wei",
""
],
[
"Dai",
"Wenliang",
""
],
[
"Byeon",
"Wonmin",
""
],
[
"Ren",
"Xiaowei",
""
],
[
"Xu",
"Yao",
""
],
[
"Choi",
"Yejin",
""
],
[
"Zhang",
"Yian",
""
],
[
"Lin",
"Ying",
""
],
[
"Suhara",
"Yoshi",
""
],
[
"Yu",
"Zhiding",
""
],
[
"Li",
"Zhiqi",
""
],
[
"Li",
"Zhiyu",
""
],
[
"Zhu",
"Zhongbo",
""
],
[
"Yang",
"Zhuolin",
""
],
[
"Chen",
"Zijia",
""
]
] |
2504.03699 | Chi-Sheng Chen | Ying-Jung Chen and Ahmad Albarqawi and Chi-Sheng Chen | Reinforcing Clinical Decision Support through Multi-Agent Systems and
Ethical AI Governance | null | null | null | null | cs.AI cs.CY cs.LG cs.MA q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in the data-driven medicine approach, which integrates
ethically managed and explainable artificial intelligence into clinical
decision support systems (CDSS), are critical to ensure reliable and effective
patient care. This paper focuses on comparing novel agent system designs that
use modular agents to analyze laboratory results, vital signs, and clinical
context, and to predict and validate results. We implement our agent system
with the eICU database, including running lab analysis, vitals-only
interpreters, and contextual reasoners agents first, then sharing the memory
into the integration agent, prediction agent, transparency agent, and a
validation agent. Our results suggest that the multi-agent system (MAS)
performed better than the single-agent system (SAS) with mortality prediction
accuracy (59%, 56%) and the mean error for length of stay (LOS)(4.37 days, 5.82
days), respectively. However, the transparency score for the SAS (86.21) is
slightly better than the transparency score for MAS (85.5). Finally, this study
suggests that our agent-based framework not only improves process transparency
and prediction accuracy but also strengthens trustworthy AI-assisted decision
support in an intensive care setting.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 05:32:43 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 15:38:04 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 05:26:26 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Chen",
"Ying-Jung",
""
],
[
"Albarqawi",
"Ahmad",
""
],
[
"Chen",
"Chi-Sheng",
""
]
] |
2504.03784 | Kai Ye | Kai Ye, Hongyi Zhou, Jin Zhu, Francesco Quinzan, Chengchung Shi | Robust Reinforcement Learning from Human Feedback for Large Language
Models Fine-Tuning | null | null | null | null | stat.ML cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reinforcement learning from human feedback (RLHF) has emerged as a key
technique for aligning the output of large language models (LLMs) with human
preferences. To learn the reward function, most existing RLHF algorithms use
the Bradley-Terry model, which relies on assumptions about human preferences
that may not reflect the complexity and variability of real-world judgments. In
this paper, we propose a robust algorithm to enhance the performance of
existing approaches under such reward model misspecifications. Theoretically,
our algorithm reduces the variance of reward and policy estimators, leading to
improved regret bounds. Empirical evaluations on LLM benchmark datasets
demonstrate that the proposed algorithm consistently outperforms existing
methods, with 77-81% of responses being favored over baselines on the Anthropic
Helpful and Harmless dataset.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 16:16:35 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 03:41:09 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 09:29:06 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ye",
"Kai",
""
],
[
"Zhou",
"Hongyi",
""
],
[
"Zhu",
"Jin",
""
],
[
"Quinzan",
"Francesco",
""
],
[
"Shi",
"Chengchung",
""
]
] |
2504.03786 | Sifan Li | Sifan Li, Yujun Cai, Bryan Hooi, Nanyun Peng, Yiwei Wang | Do "New Snow Tablets" Contain Snow? Large Language Models Over-Rely on
Names to Identify Ingredients of Chinese Drugs | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional Chinese Medicine (TCM) has seen increasing adoption in
healthcare, with specialized Large Language Models (LLMs) emerging to support
clinical applications. A fundamental requirement for these models is accurate
identification of TCM drug ingredients. In this paper, we evaluate how general
and TCM-specialized LLMs perform when identifying ingredients of Chinese drugs.
Our systematic analysis reveals consistent failure patterns: models often
interpret drug names literally, overuse common herbs regardless of relevance,
and exhibit erratic behaviors when faced with unfamiliar formulations. LLMs
also fail to understand the verification task. These findings demonstrate that
current LLMs rely primarily on drug names rather than possessing systematic
pharmacological knowledge. To address these limitations, we propose a Retrieval
Augmented Generation (RAG) approach focused on ingredient names. Experiments
across 220 TCM formulations show our method significantly improves accuracy
from approximately 50% to 82% in ingredient verification tasks. Our work
highlights critical weaknesses in current TCM-specific LLMs and offers a
practical solution for enhancing their clinical reliability.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 17:43:45 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 12:33:54 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 08:51:02 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Li",
"Sifan",
""
],
[
"Cai",
"Yujun",
""
],
[
"Hooi",
"Bryan",
""
],
[
"Peng",
"Nanyun",
""
],
[
"Wang",
"Yiwei",
""
]
] |
2504.04118 | Darshan Kumar | Darshan Kumar (HNAS), Praveen Kumar Dhankar (SITN), Saibal Ray (GLA)
and Fengge Zhang (HNAS) | Joint Analysis of Constraints on f(R) Parametrization from Recent
Cosmological Observations | 15 pages, 13 figures, and 6 tables; v2: references added | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | In this study, we present constraints on the parameters of three well-known
$f(R)$ gravity models, viz. (i) Hu-Sawicki, (ii) Starobinsky, and (iii) ArcTanh
by using a joint analysis of recent cosmological observations. We perform
analytical approximations for the Hubble parameter, $H(z)$, and cosmological
distances in terms of the Hubble constant $(H_0)$, matter density
$(\Omega_{m0})$, and a deviation parameter $b$ for each model. Our analysis
uses data from four cosmological observations: (a) Hubble parameter
measurements (Cosmic Chronometers), (b) Type Ia Supernovae (Union 3.0), (c)
Baryon Acoustic Oscillations (DESI-2024), and (d) Gamma-Ray Bursts (GRBs). We
first optimize the models using each dataset independently, and subsequently,
we perform a comprehensive joint analysis combining all four datasets. Our
results show that the Hu-Sawicki and ArcTanh models do not deviate
significantly from the $\Lambda$CDM model at 68\% confidence level for
individual datasets and remain consistent at 99\% confidence level in the joint
analysis. In contrast, the Starobinsky model shows a strong deviation and
appears as a viable alternative to $\Lambda$CDM. We also constrain the
transition redshift parameter ($z_t$), and check that the obtained value agrees
with the values inferred from both early-time measurement (Planck) and
late-time data from Type Ia Supernovae. These results support the potential
support of $f(R)$ gravity to explain the late-time cosmic acceleration
effectively.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 09:19:22 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 12:01:11 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Kumar",
"Darshan",
"",
"HNAS"
],
[
"Dhankar",
"Praveen Kumar",
"",
"SITN"
],
[
"Ray",
"Saibal",
"",
"GLA"
],
[
"Zhang",
"Fengge",
"",
"HNAS"
]
] |
2504.04222 | Tianyu Cui | Tianyu Cui, Xinjie Lin, Sijia Li, Miao Chen, Qilei Yin, Qi Li, Ke Xu | TrafficLLM: Enhancing Large Language Models for Network Traffic Analysis
with Generic Traffic Representation | null | null | null | null | cs.LG cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | Machine learning (ML) powered network traffic analysis has been widely used
for the purpose of threat detection. Unfortunately, their generalization across
different tasks and unseen data is very limited. Large language models (LLMs),
known for their strong generalization capabilities, have shown promising
performance in various domains. However, their application to the traffic
analysis domain is limited due to significantly different characteristics of
network traffic. To address the issue, in this paper, we propose TrafficLLM,
which introduces a dual-stage fine-tuning framework to learn generic traffic
representation from heterogeneous raw traffic data. The framework uses
traffic-domain tokenization, dual-stage tuning pipeline, and extensible
adaptation to help LLM release generalization ability on dynamic traffic
analysis tasks, such that it enables traffic detection and traffic generation
across a wide range of downstream tasks. We evaluate TrafficLLM across 10
distinct scenarios and 229 types of traffic. TrafficLLM achieves F1-scores of
0.9875 and 0.9483, with up to 80.12% and 33.92% better performance than
existing detection and generation methods. It also shows strong generalization
on unseen traffic with an 18.6% performance improvement. We further evaluate
TrafficLLM in real-world scenarios. The results confirm that TrafficLLM is easy
to scale and achieves accurate detection performance on enterprise traffic.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 16:18:33 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:30:04 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Cui",
"Tianyu",
""
],
[
"Lin",
"Xinjie",
""
],
[
"Li",
"Sijia",
""
],
[
"Chen",
"Miao",
""
],
[
"Yin",
"Qilei",
""
],
[
"Li",
"Qi",
""
],
[
"Xu",
"Ke",
""
]
] |
2504.04383 | Ximing Lu | Ximing Lu, Seungju Han, David Acuna, Hyunwoo Kim, Jaehun Jung, Shrimai
Prabhumoye, Niklas Muennighoff, Mostofa Patwary, Mohammad Shoeybi, Bryan
Catanzaro, Yejin Choi | Retro-Search: Exploring Untaken Paths for Deeper and Efficient Reasoning | Code and data will be publicly released upon internal approval | null | null | null | cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large reasoning models exhibit remarkable reasoning capabilities via long,
elaborate reasoning trajectories. Supervised fine-tuning on such reasoning
traces, also known as distillation, can be a cost-effective way to boost
reasoning capabilities of student models. However, empirical observations
reveal that these reasoning trajectories are often suboptimal, switching
excessively between different lines of thought, resulting in under-thinking,
over-thinking, and even degenerate responses. We introduce Retro-Search, an
MCTS-inspired search algorithm, for distilling higher quality reasoning paths
from large reasoning models. Retro-Search retrospectively revises reasoning
paths to discover better, yet shorter traces, which can then lead to student
models with enhanced reasoning capabilities with shorter, thus faster
inference. Our approach can enable two use cases: self-improvement, where
models are fine-tuned on their own Retro-Search-ed thought traces, and
weak-to-strong improvement, where a weaker model revises stronger model's
thought traces via Retro-Search. For self-improving, R1-distill-7B, fine-tuned
on its own Retro-Search-ed traces, reduces the average reasoning length by
31.2% while improving performance by 7.7% across seven math benchmarks. For
weak-to-strong improvement, we retrospectively revise R1-671B's traces from the
OpenThoughts dataset using R1-distill-32B as the Retro-Search-er, a model 20x
smaller. Qwen2.5-32B, fine-tuned on this refined data, achieves performance
comparable to R1-distill-32B, yielding an 11.3% reduction in reasoning length
and a 2.4% performance improvement compared to fine-tuning on the original
OpenThoughts data. Our work counters recently emergent viewpoints that question
the relevance of search algorithms in the era of large reasoning models, by
demonstrating that there are still opportunities for algorithmic advancements,
even for frontier models.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 06:23:27 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:07:31 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Lu",
"Ximing",
""
],
[
"Han",
"Seungju",
""
],
[
"Acuna",
"David",
""
],
[
"Kim",
"Hyunwoo",
""
],
[
"Jung",
"Jaehun",
""
],
[
"Prabhumoye",
"Shrimai",
""
],
[
"Muennighoff",
"Niklas",
""
],
[
"Patwary",
"Mostofa",
""
],
[
"Shoeybi",
"Mohammad",
""
],
[
"Catanzaro",
"Bryan",
""
],
[
"Choi",
"Yejin",
""
]
] |
2504.04558 | F. Javier Garc\'ia de Abajo | F. Javier Garc\'ia de Abajo, D. N. Basov, Frank H. L. Koppens, Lorenzo
Orsini, Matteo Ceccanti, Sebasti\'an Castilla, Lorenzo Cavicchi, Marco
Polini, P. A. D. Gon\c{c}alves, A. T. Costa, N. M. R. Peres, N. Asger
Mortensen, Sathwik Bharadwaj, Zubin Jacob, P. J. Schuck, A. N. Pasupathy,
Milan Delor, M. K. Liu, Aitor Mugarza, Pablo Merino, Marc G. Cuxart, Emigdio
Ch\'avez-Angel, Martin Svec, Luiz H. G. Tizei, Florian Dirnberger, Hui Deng,
Christian Schneider, Vinod Menon, Thorsten Deilmann, Alexey Chernikov,
Kristian S. Thygesen, Yohannes Abate, Mauricio Terrones, Vinod K. Sangwan,
Mark C. Hersam, Leo Yu, Xueqi Chen, Tony F. Heinz, Puneet Murthy, Martin
Kroner, Tomasz Smolenski, Deepankur Thureja, Thibault Chervy, Armando Genco,
Chiara Trovatello, Giulio Cerullo, Stefano Dal Conte, Daniel Timmer,
Antonietta De Sio, Christoph Lienau, Nianze Shang, Hao Hong, Kaihui Liu,
Zhipei Sun, Lee A. Rozema, Philip Walther, Andrea Al\`u, Michele Cotrufo,
Raquel Queiroz, X.-Y. Zhu, Joel D. Cox, Eduardo J. C. Dias, \'Alvaro
Rodr\'iguez Echarri, Fadil Iyikanat, Andrea Marini, Paul Herrmann, Nele
Tornow, Sebastian Klimmer, Jan Wilhelm, Giancarlo Soavi, Zeyuan Sun, Shiwei
Wu, Ying Xiong, Oles Matsyshyn, Roshan Krishna Kumar, Justin C. W. Song,
Tomer Bucher, Alexey Gorlach, Shai Tsesses, Ido Kaminer, Julian Schwab,
Florian Mangold, Harald Giessen, M. S\'anchez S\'anchez, D. K. Efetov, T.
Low, G. G\'omez-Santos, T. Stauber, Gonzalo \'Alvarez-P\'erez, Jiahua Duan,
Luis Mart\'in-Moreno, Alexander Paarmann, Joshua D. Caldwell, Alexey Y.
Nikitin, Pablo Alonso-Gonz\'alez, Niclas S. Mueller, Valentyn Volkov, Deep
Jariwala, Timur Shegai, Jorik van de Groep, Alexandra Boltasseva, Igor V.
Bondarev, Vladimir M. Shalaev, Jeffrey Simon, Colton Fruhling, Guangzhen
Shen, Dino Novko, Shijing Tan, Bing Wang, Hrvoje Petek, Vahagn Mkhitaryan,
Renwen Yu, Alejandro Manjavacas, J. Enrique Ortega, Xu Cheng, Ruijuan Tian,
Dong Mao, Dries Van Thourhout, Xuetao Gan, Qing Dai, Aaron Sternbach, You
Zhou, Mohammad Hafezi, Dmitrii Litvinov, Magdalena Grzeszczyk, Kostya S.
Novoselov, Maciej Koperski, Sotirios Papadopoulos, Lukas Novotny, Leonardo
Viti, Miriam Serena Vitiello, Nathan D. Cottam, Benjamin T. Dewes, Oleg
Makarovsky, Amalia Patan\`e, Yihao Song, Mingyang Cai, Jiazhen Chen, Doron
Naveh, Houk Jang, Suji Park, Fengnian Xia, Philipp K. Jenke, Josip Bajo,
Benjamin Braun, Kenneth S. Burch, Liuyan Zhao, and Xiaodong Xu | Roadmap for Photonics with 2D Materials | 199 pages, 42 figures, 1154 references | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Triggered by the development of exfoliation and the identification of a wide
range of extraordinary physical properties in self-standing films consisting of
one or few atomic layers, two-dimensional (2D) materials such as graphene,
transition metal dichalcogenides (TMDs), and other van der Waals (vdW) crystals
currently constitute a wide research field protruding in multiple directions in
combination with layer stacking and twisting, nanofabrication, surface-science
methods, and integration into nanostructured environments. Photonics
encompasses a multidisciplinary collection of those directions, where 2D
materials contribute with polaritons of unique characteristics such as strong
spatial confinement, large optical-field enhancement, long lifetimes, high
sensitivity to external stimuli (e.g., electric and magnetic fields, heating,
and strain), a broad spectral range from the far infrared to the ultraviolet,
and hybridization with spin and momentum textures of electronic band
structures. The explosion of photonics with 2D materials as a vibrant research
area is producing breakthroughs, including the discovery and design of new
materials and metasurfaces with unprecedented properties as well as
applications in integrated photonics, light emission, optical sensing, and
exciting prospects for applications in quantum information, and nanoscale
thermal transport. This Roadmap summarizes the state of the art in the field,
identifies challenges and opportunities, and discusses future goals and how to
meet them through a wide collection of topical sections prepared by leading
practitioners.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 17:15:06 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 21:11:11 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Apr 2025 17:13:42 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"de Abajo",
"F. Javier García",
""
],
[
"Basov",
"D. N.",
""
],
[
"Koppens",
"Frank H. L.",
""
],
[
"Orsini",
"Lorenzo",
""
],
[
"Ceccanti",
"Matteo",
""
],
[
"Castilla",
"Sebastián",
""
],
[
"Cavicchi",
"Lorenzo",
""
],
[
"Polini",
"Marco",
""
],
[
"Gonçalves",
"P. A. D.",
""
],
[
"Costa",
"A. T.",
""
],
[
"Peres",
"N. M. R.",
""
],
[
"Mortensen",
"N. Asger",
""
],
[
"Bharadwaj",
"Sathwik",
""
],
[
"Jacob",
"Zubin",
""
],
[
"Schuck",
"P. J.",
""
],
[
"Pasupathy",
"A. N.",
""
],
[
"Delor",
"Milan",
""
],
[
"Liu",
"M. K.",
""
],
[
"Mugarza",
"Aitor",
""
],
[
"Merino",
"Pablo",
""
],
[
"Cuxart",
"Marc G.",
""
],
[
"Chávez-Angel",
"Emigdio",
""
],
[
"Svec",
"Martin",
""
],
[
"Tizei",
"Luiz H. G.",
""
],
[
"Dirnberger",
"Florian",
""
],
[
"Deng",
"Hui",
""
],
[
"Schneider",
"Christian",
""
],
[
"Menon",
"Vinod",
""
],
[
"Deilmann",
"Thorsten",
""
],
[
"Chernikov",
"Alexey",
""
],
[
"Thygesen",
"Kristian S.",
""
],
[
"Abate",
"Yohannes",
""
],
[
"Terrones",
"Mauricio",
""
],
[
"Sangwan",
"Vinod K.",
""
],
[
"Hersam",
"Mark C.",
""
],
[
"Yu",
"Leo",
""
],
[
"Chen",
"Xueqi",
""
],
[
"Heinz",
"Tony F.",
""
],
[
"Murthy",
"Puneet",
""
],
[
"Kroner",
"Martin",
""
],
[
"Smolenski",
"Tomasz",
""
],
[
"Thureja",
"Deepankur",
""
],
[
"Chervy",
"Thibault",
""
],
[
"Genco",
"Armando",
""
],
[
"Trovatello",
"Chiara",
""
],
[
"Cerullo",
"Giulio",
""
],
[
"Conte",
"Stefano Dal",
""
],
[
"Timmer",
"Daniel",
""
],
[
"De Sio",
"Antonietta",
""
],
[
"Lienau",
"Christoph",
""
],
[
"Shang",
"Nianze",
""
],
[
"Hong",
"Hao",
""
],
[
"Liu",
"Kaihui",
""
],
[
"Sun",
"Zhipei",
""
],
[
"Rozema",
"Lee A.",
""
],
[
"Walther",
"Philip",
""
],
[
"Alù",
"Andrea",
""
],
[
"Cotrufo",
"Michele",
""
],
[
"Queiroz",
"Raquel",
""
],
[
"Zhu",
"X. -Y.",
""
],
[
"Cox",
"Joel D.",
""
],
[
"Dias",
"Eduardo J. C.",
""
],
[
"Echarri",
"Álvaro Rodríguez",
""
],
[
"Iyikanat",
"Fadil",
""
],
[
"Marini",
"Andrea",
""
],
[
"Herrmann",
"Paul",
""
],
[
"Tornow",
"Nele",
""
],
[
"Klimmer",
"Sebastian",
""
],
[
"Wilhelm",
"Jan",
""
],
[
"Soavi",
"Giancarlo",
""
],
[
"Sun",
"Zeyuan",
""
],
[
"Wu",
"Shiwei",
""
],
[
"Xiong",
"Ying",
""
],
[
"Matsyshyn",
"Oles",
""
],
[
"Kumar",
"Roshan Krishna",
""
],
[
"Song",
"Justin C. W.",
""
],
[
"Bucher",
"Tomer",
""
],
[
"Gorlach",
"Alexey",
""
],
[
"Tsesses",
"Shai",
""
],
[
"Kaminer",
"Ido",
""
],
[
"Schwab",
"Julian",
""
],
[
"Mangold",
"Florian",
""
],
[
"Giessen",
"Harald",
""
],
[
"Sánchez",
"M. Sánchez",
""
],
[
"Efetov",
"D. K.",
""
],
[
"Low",
"T.",
""
],
[
"Gómez-Santos",
"G.",
""
],
[
"Stauber",
"T.",
""
],
[
"Álvarez-Pérez",
"Gonzalo",
""
],
[
"Duan",
"Jiahua",
""
],
[
"Martín-Moreno",
"Luis",
""
],
[
"Paarmann",
"Alexander",
""
],
[
"Caldwell",
"Joshua D.",
""
],
[
"Nikitin",
"Alexey Y.",
""
],
[
"Alonso-González",
"Pablo",
""
],
[
"Mueller",
"Niclas S.",
""
],
[
"Volkov",
"Valentyn",
""
],
[
"Jariwala",
"Deep",
""
],
[
"Shegai",
"Timur",
""
],
[
"van de Groep",
"Jorik",
""
],
[
"Boltasseva",
"Alexandra",
""
],
[
"Bondarev",
"Igor V.",
""
],
[
"Shalaev",
"Vladimir M.",
""
],
[
"Simon",
"Jeffrey",
""
],
[
"Fruhling",
"Colton",
""
],
[
"Shen",
"Guangzhen",
""
],
[
"Novko",
"Dino",
""
],
[
"Tan",
"Shijing",
""
],
[
"Wang",
"Bing",
""
],
[
"Petek",
"Hrvoje",
""
],
[
"Mkhitaryan",
"Vahagn",
""
],
[
"Yu",
"Renwen",
""
],
[
"Manjavacas",
"Alejandro",
""
],
[
"Ortega",
"J. Enrique",
""
],
[
"Cheng",
"Xu",
""
],
[
"Tian",
"Ruijuan",
""
],
[
"Mao",
"Dong",
""
],
[
"Van Thourhout",
"Dries",
""
],
[
"Gan",
"Xuetao",
""
],
[
"Dai",
"Qing",
""
],
[
"Sternbach",
"Aaron",
""
],
[
"Zhou",
"You",
""
],
[
"Hafezi",
"Mohammad",
""
],
[
"Litvinov",
"Dmitrii",
""
],
[
"Grzeszczyk",
"Magdalena",
""
],
[
"Novoselov",
"Kostya S.",
""
],
[
"Koperski",
"Maciej",
""
],
[
"Papadopoulos",
"Sotirios",
""
],
[
"Novotny",
"Lukas",
""
],
[
"Viti",
"Leonardo",
""
],
[
"Vitiello",
"Miriam Serena",
""
],
[
"Cottam",
"Nathan D.",
""
],
[
"Dewes",
"Benjamin T.",
""
],
[
"Makarovsky",
"Oleg",
""
],
[
"Patanè",
"Amalia",
""
],
[
"Song",
"Yihao",
""
],
[
"Cai",
"Mingyang",
""
],
[
"Chen",
"Jiazhen",
""
],
[
"Naveh",
"Doron",
""
],
[
"Jang",
"Houk",
""
],
[
"Park",
"Suji",
""
],
[
"Xia",
"Fengnian",
""
],
[
"Jenke",
"Philipp K.",
""
],
[
"Bajo",
"Josip",
""
],
[
"Braun",
"Benjamin",
""
],
[
"Burch",
"Kenneth S.",
""
],
[
"Zhao",
"Liuyan",
""
],
[
"Xu",
"Xiaodong",
""
]
] |
2504.04808 | Tengjun Jin | Tengjun Jin, Yuxuan Zhu, Daniel Kang | ELT-Bench: An End-to-End Benchmark for Evaluating AI Agents on ELT
Pipelines | 14 pages, 18 figures | null | null | null | cs.DB cs.AI | http://creativecommons.org/licenses/by/4.0/ | Practitioners are increasingly turning to Extract-Load-Transform (ELT)
pipelines with the widespread adoption of cloud data warehouses. However,
designing these pipelines often involves significant manual work to ensure
correctness. Recent advances in AI-based methods, which have shown strong
capabilities in data tasks, such as text-to-SQL, present an opportunity to
alleviate manual efforts in developing ELT pipelines. Unfortunately, current
benchmarks in data engineering only evaluate isolated tasks, such as using data
tools and writing data transformation queries, leaving a significant gap in
evaluating AI agents for generating end-to-end ELT pipelines.
To fill this gap, we introduce ELT-Bench, an end-to-end benchmark designed to
assess the capabilities of AI agents to build ELT pipelines. ELT-Bench consists
of 100 pipelines, including 835 source tables and 203 data models across
various domains. By simulating realistic scenarios involving the integration of
diverse data sources and the use of popular data tools, ELT-Bench evaluates AI
agents' abilities in handling complex data engineering workflows. AI agents
must interact with databases and data tools, write code and SQL queries, and
orchestrate every pipeline stage. We evaluate two representative code agent
frameworks, Spider-Agent and SWE-Agent, using six popular Large Language Models
(LLMs) on ELT-Bench. The highest-performing agent, Spider-Agent
Claude-3.7-Sonnet with extended thinking, correctly generates only 3.9% of data
models, with an average cost of $4.30 and 89.3 steps per pipeline. Our
experimental results demonstrate the challenges of ELT-Bench and highlight the
need for a more advanced AI agent to reduce manual effort in ELT workflows. Our
code and data are available at https://github.com/uiuc-kang-lab/ELT-Bench.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 08:03:36 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 19:46:56 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Jin",
"Tengjun",
""
],
[
"Zhu",
"Yuxuan",
""
],
[
"Kang",
"Daniel",
""
]
] |
2504.04894 | Paula Tru\"ol | Maciej Borodzik and Paula Tru\"ol | Non-complex cobordisms between quasipositive knots | 9 pages, 1 table, comments welcome! v2: minor changes, extended
introduction. Terminology changed to `complex cobordisms' to avoid a
terminology clash (see Remark 1.7) | null | null | MPIM-Bonn-2025 | math.GT | http://creativecommons.org/licenses/by/4.0/ | We show that for every genus $g \geq 0$, there exist quasipositive knots
$K_0^g$ and $K_1^g$ such that there is a cobordism of genus
$g=|g_4(K_1^g)-g_4(K_0^g)|$ between $K_0^g$ and $K_1^g$, but there is no ribbon
cobordism of genus $g$ in either direction and thus no complex cobordism
between these two knots. This gives a negative answer to a question posed by
Feller in 2016.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 10:06:14 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 11:54:04 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Borodzik",
"Maciej",
""
],
[
"Truöl",
"Paula",
""
]
] |
2504.04982 | Zhiwei Cao | Zhiwei Cao, Minghao Li, Feng Lin, Jimin Jia, Yonggang Wen, Jianxiong
Yin, Simon See | Transforming Future Data Center Operations and Management via Physical
AI | 9 pages, 5 figures | null | null | null | cs.AI cs.DC | http://creativecommons.org/licenses/by/4.0/ | Data centers (DCs) as mission-critical infrastructures are pivotal in
powering the growth of artificial intelligence (AI) and the digital economy.
The evolution from Internet DC to AI DC has introduced new challenges in
operating and managing data centers for improved business resilience and
reduced total cost of ownership. As a result, new paradigms, beyond the
traditional approaches based on best practices, must be in order for future
data centers. In this research, we propose and develop a novel Physical AI
(PhyAI) framework for advancing DC operations and management. Our system
leverages the emerging capabilities of state-of-the-art industrial products and
our in-house research and development. Specifically, it presents three core
modules, namely: 1) an industry-grade in-house simulation engine to simulate DC
operations in a highly accurate manner, 2) an AI engine built upon NVIDIA
PhysicsNemo for the training and evaluation of physics-informed machine
learning (PIML) models, and 3) a digital twin platform built upon NVIDIA
Omniverse for our proposed 5-tier digital twin framework. This system presents
a scalable and adaptable solution to digitalize, optimize, and automate future
data center operations and management, by enabling real-time digital twins for
future data centers. To illustrate its effectiveness, we present a compelling
case study on building a surrogate model for predicting the thermal and airflow
profiles of a large-scale DC in a real-time manner. Our results demonstrate its
superior performance over traditional time-consuming Computational Fluid
Dynamics/Heat Transfer (CFD/HT) simulation, with a median absolute temperature
prediction error of 0.18 {\deg}C. This emerging approach would open doors to
several potential research directions for advancing Physical AI in future DC
operations.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 12:09:22 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:06:12 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Cao",
"Zhiwei",
""
],
[
"Li",
"Minghao",
""
],
[
"Lin",
"Feng",
""
],
[
"Jia",
"Jimin",
""
],
[
"Wen",
"Yonggang",
""
],
[
"Yin",
"Jianxiong",
""
],
[
"See",
"Simon",
""
]
] |
2504.05062 | Chenfeng Xu | Chenfeng Xu | LDGNet: A Lightweight Difference Guiding Network for Remote Sensing
Change Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid advancement of deep learning, the field of change detection
(CD) in remote sensing imagery has achieved remarkable progress. Existing
change detection methods primarily focus on achieving higher accuracy with
increased computational costs and parameter sizes, leaving development of
lightweight methods for rapid real-world processing an underexplored challenge.
To address this challenge, we propose a Lightweight Difference Guiding Network
(LDGNet), leveraging absolute difference image to guide optical remote sensing
change detection. First, to enhance the feature representation capability of
the lightweight backbone network, we propose the Difference Guiding Module
(DGM), which leverages multi-scale features extracted from the absolute
difference image to progressively influence the original image encoder at each
layer, thereby reinforcing feature extraction. Second, we propose the
Difference-Aware Dynamic Fusion (DADF) module with Visual State Space Model
(VSSM) for lightweight long-range dependency modeling. The module first uses
feature absolute differences to guide VSSM's global contextual modeling of
change regions, then employs difference attention to dynamically fuse these
long-range features with feature differences, enhancing change semantics while
suppressing noise and background. Extensive experiments on multiple datasets
demonstrate that our method achieves comparable or superior performance to
current state-of-the-art (SOTA) methods requiring several times more
computation, while maintaining only 3.43M parameters and 1.12G FLOPs.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 13:33:54 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:15:27 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Xu",
"Chenfeng",
""
]
] |
2504.05154 | Geyang Guo | Geyang Guo, Tarek Naous, Hiromi Wakaki, Yukiko Nishimura, Yuki
Mitsufuji, Alan Ritter, Wei Xu | CARE: Aligning Language Models for Regional Cultural Awareness | 24 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing language models (LMs) often exhibit a Western-centric bias and
struggle to represent diverse cultural knowledge. Previous attempts to address
this rely on synthetic data and express cultural knowledge only in English. In
this work, we study whether a small amount of human-written, multilingual
cultural preference data can improve LMs across various model families and
sizes. We first introduce CARE, a multilingual resource of 24.1k responses with
human preferences on 2,580 questions about Chinese and Arab cultures, all
carefully annotated by native speakers and offering more balanced coverage.
Using CARE, we demonstrate that cultural alignment improves existing LMs beyond
generic resources without compromising general capabilities. Moreover, we
evaluate the cultural awareness of LMs, native speakers, and retrieved web
content when queried in different languages. Our experiment reveals regional
disparities among LMs, which may also be reflected in the documentation gap:
native speakers often take everyday cultural commonsense and social norms for
granted, while non-natives are more likely to actively seek out and document
them. CARE is publicly available at https://github.com/Guochry/CARE (we plan to
add Japanese data in the near future).
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 14:57:06 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:34:48 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Guo",
"Geyang",
""
],
[
"Naous",
"Tarek",
""
],
[
"Wakaki",
"Hiromi",
""
],
[
"Nishimura",
"Yukiko",
""
],
[
"Mitsufuji",
"Yuki",
""
],
[
"Ritter",
"Alan",
""
],
[
"Xu",
"Wei",
""
]
] |
2504.05382 | Henrik R. Larsson Dr. | Henrik R. Larsson | Benchmarking vibrational spectra: 5000 accurate eigenstates of
acetonitrile using tree tensor network states | null | J. Phys. Chem. Lett. 2025, 16, 3991-3997 | 10.1021/acs.jpclett.5c00782 | null | physics.chem-ph physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate vibrational spectra are essential for understanding how molecules
behave, yet their computation remains challenging and benchmark data to
reliably compare different methods are sparse. Here, we present high-accuracy
eigenstate computations for the six-atom, 12-dimensional acetonitrile molecule,
a prototypical, strongly coupled, anharmonic system. Using a density matrix
renormalization group (DMRG) algorithm with a tree-tensor-network-state (TTNS)
ansatz, a refinement using TTNSs as basis set, and reliable procedures to
estimate energy errors, we compute up to 5,000 vibrational states with error
estimates below 0.0007 $\mathrm{cm}^{-1}$. Our analysis reveals that previous
works underestimated the energy error by up to two orders of magnitude. Our
data serve as a benchmark for future vibrational spectroscopy methods and our
new method offers a path toward similarly precise computations of large,
complex molecular systems.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 18:00:02 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 18:00:00 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Larsson",
"Henrik R.",
""
]
] |
2504.05408 | Yujin Potter | Wenbo Guo, Yujin Potter, Tianneng Shi, Zhun Wang, Andy Zhang, Dawn
Song | Frontier AI's Impact on the Cybersecurity Landscape | null | null | null | null | cs.CR cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | As frontier AI advances rapidly, understanding its impact on cybersecurity
and inherent risks is essential to ensuring safe AI evolution (e.g., guiding
risk mitigation and informing policymakers). While some studies review AI
applications in cybersecurity, none of them comprehensively discuss AI's future
impacts or provide concrete recommendations for navigating its safe and secure
usage. This paper presents an in-depth analysis of frontier AI's impact on
cybersecurity and establishes a systematic framework for risk assessment and
mitigation. To this end, we first define and categorize the marginal risks of
frontier AI in cybersecurity and then systemically analyze the current and
future impacts of frontier AI in cybersecurity, qualitatively and
quantitatively. We also discuss why frontier AI likely benefits attackers more
than defenders in the short term from equivalence classes, asymmetry, and
economic impact. Next, we explore frontier AI's impact on future software
system development, including enabling complex hybrid systems while introducing
new risks. Based on our findings, we provide security recommendations,
including constructing fine-grained benchmarks for risk assessment, designing
AI agents for defenses, building security mechanisms and provable defenses for
hybrid systems, enhancing pre-deployment security testing and transparency, and
strengthening defenses for users. Finally, we present long-term research
questions essential for understanding AI's future impacts and unleashing its
defensive capabilities.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 18:25:18 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:35:14 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Guo",
"Wenbo",
""
],
[
"Potter",
"Yujin",
""
],
[
"Shi",
"Tianneng",
""
],
[
"Wang",
"Zhun",
""
],
[
"Zhang",
"Andy",
""
],
[
"Song",
"Dawn",
""
]
] |
2504.05481 | Ryan O'Loughlin | Ryan O'Loughlin | Describing the Numerical Range With Specht's Theorem | null | null | null | null | math.FA | http://creativecommons.org/licenses/by/4.0/ | We use Specht's Theorem, which classifies when matrices are unitarily
equivalent, to describe the numerical range of a matrix. In the case of
$2$-by-$2$ matrices we recover the Elliptical Range Theorem, which shows that
the numerical range of a $2$-by-$2$ matrix is an ellipse. We then show a
specified point lies in the numerical range of a $3$-by-$3$ matrix if and only
if a system of matrix trace equations is satisfied.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 20:23:55 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 21:50:09 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"O'Loughlin",
"Ryan",
""
]
] |
2504.05542 | Andre Severo Pereira Gomes | Wilken Aldair Misael, Lucia Amidani, Juliane M\"arz, Elena F.
Bazarkina, Kristina O. Kvashnina, Val\'erie Vallet, Andr\'e Severo Pereira
Gomes | Core-Excited States of Linear and Bent Uranyl Complexes: Insights from
High-Energy Resolution X-ray Spectroscopy and Relativistic Quantum Chemistry | 27 pages, 9 figures, 3 tables | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by-sa/4.0/ | Advanced X-ray spectroscopic techniques are widely recognized as
state-of-the-art tools for probing the electronic structure, bonding, and
chemical environments of the heaviest elements in the periodic table. In this
study, we employ X-ray absorption near-edge structure measurements in
high-energy resolution fluorescence detection (HERFD-XANES) mode to investigate
the core states arising from excitations out of the U 3d${_{3/2}}$ (M$_4$ edge)
levels for molecular complexes in which the uranyl moiety deviates from
linearity to varying degrees, and in particular systems containing the
UO$_2$Cl$_2$ group such as UO$_2$Cl$_2$.n(H$_2$O) and UO$_2$Cl$_2$(phen)$_2$,
which in the latter case exhibits a pronounced O-U-O bending angle. These U
M$_4$ edge HERFD-XANES spectra are compared to those of other linear
(Cs$_2$UO$_2$Cl$_4$) or pseudo-linear ([UO$_2$(NO$_3$)$_2$.n(H$_2$O)]) uranyl
complexes. This evaluation is complemented by ab initio relativistic quantum
chemistry simulations using 2-component Time-Dependent Density Functional
Theory (TD-DFT) with the CAM-B3LYP functional, employing the Tamm-Dancoff
approximation (2c-TDA). Our 2c-TDA simulations show modest deviations from the
HERFD-XANES data, with peak splittings differing by less than 1 eV from
experimental values. These core-excited states were further characterized by
Natural Transition Orbital (NTO) analysis. Overall, our results highlight the
influence of equatorial ligands on the spectroscopic signatures, particularly
pronounced in UO$_2$Cl$_2$(phen)$_2$, where the U 3d${_{3/2}} \rightarrow$ $5f$
$\sigma{_u}^{*}$ satellite transition appears at lower energies compared to the
other systems studied.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 22:40:52 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 11:43:15 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Misael",
"Wilken Aldair",
""
],
[
"Amidani",
"Lucia",
""
],
[
"März",
"Juliane",
""
],
[
"Bazarkina",
"Elena F.",
""
],
[
"Kvashnina",
"Kristina O.",
""
],
[
"Vallet",
"Valérie",
""
],
[
"Gomes",
"André Severo Pereira",
""
]
] |
2504.05732 | Haoyu Wang | Haoyu Wang, Yujia Fu, Zhu Zhang, Shuo Wang, Zirui Ren, Xiaorong Wang,
Zhili Li, Chaoqun He, Bo An, Zhiyuan Liu, Maosong Sun | LLM$\times$MapReduce-V2: Entropy-Driven Convolutional Test-Time Scaling
for Generating Long-Form Articles from Extremely Long Resources | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Long-form generation is crucial for a wide range of practical applications,
typically categorized into short-to-long and long-to-long generation. While
short-to-long generations have received considerable attention, generating long
texts from extremely long resources remains relatively underexplored. The
primary challenge in long-to-long generation lies in effectively integrating
and analyzing relevant information from extensive inputs, which remains
difficult for current large language models (LLMs). In this paper, we propose
LLM$\times$MapReduce-V2, a novel test-time scaling strategy designed to enhance
the ability of LLMs to process extremely long inputs. Drawing inspiration from
convolutional neural networks, which iteratively integrate local features into
higher-level global representations, LLM$\times$MapReduce-V2 utilizes stacked
convolutional scaling layers to progressively expand the understanding of input
materials. Both quantitative and qualitative experimental results demonstrate
that our approach substantially enhances the ability of LLMs to process long
inputs and generate coherent, informative long-form articles, outperforming
several representative baselines. Both LLM$\times$MapReduce-V2 and SurveyEval
are publicly available at https://github.com/thunlp/LLMxMapReduce .
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 07:03:48 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 03:28:58 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Wang",
"Haoyu",
""
],
[
"Fu",
"Yujia",
""
],
[
"Zhang",
"Zhu",
""
],
[
"Wang",
"Shuo",
""
],
[
"Ren",
"Zirui",
""
],
[
"Wang",
"Xiaorong",
""
],
[
"Li",
"Zhili",
""
],
[
"He",
"Chaoqun",
""
],
[
"An",
"Bo",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Sun",
"Maosong",
""
]
] |
2504.05747 | Peerat Limkonchotiwat | Raymond Ng, Thanh Ngan Nguyen, Yuli Huang, Ngee Chia Tai, Wai Yi
Leong, Wei Qi Leong, Xianbin Yong, Jian Gang Ngui, Yosephine Susanto,
Nicholas Cheng, Hamsawardhini Rengarajan, Peerat Limkonchotiwat, Adithya
Venkatadri Hulagadri, Kok Wai Teng, Yeo Yeow Tong, Bryan Siow, Wei Yi Teo,
Wayne Lau, Choon Meng Tan, Brandon Ong, Zhi Hao Ong, Jann Railey Montalan,
Adwin Chan, Sajeban Antonyrex, Ren Lee, Esther Choa, David Ong Tat-Wee, Bing
Jie Darius Liu, William Chandra Tjhi, Erik Cambria, Leslie Teo | SEA-LION: Southeast Asian Languages in One Network | We released our model at
https://huggingface.co/collections/aisingapore/sea-lionv3-672589a39cdadd6a5b199581 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recently, Large Language Models (LLMs) have dominated much of the artificial
intelligence scene with their ability to process and generate natural
languages. However, the majority of LLM research and development remains
English-centric, leaving low-resource languages such as those in the Southeast
Asian (SEA) region under-represented. To address this representation gap, we
introduce Llama-SEA-LION-v3-8B-IT and Gemma-SEA-LION-v3-9B-IT, two cutting-edge
multilingual LLMs designed for SEA languages. The SEA-LION family of LLMs
supports 11 SEA languages, namely English, Chinese, Indonesian, Vietnamese,
Malay, Thai, Burmese, Lao, Filipino, Tamil, and Khmer. Our work leverages
large-scale multilingual continued pre-training with a comprehensive
post-training regime involving multiple stages of instruction fine-tuning,
alignment, and model merging. Evaluation results on multilingual benchmarks
indicate that our models achieve state-of-the-art performance across LLMs
supporting SEA languages. We open-source the models to benefit the wider SEA
community.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 07:24:51 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:51:05 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ng",
"Raymond",
""
],
[
"Nguyen",
"Thanh Ngan",
""
],
[
"Huang",
"Yuli",
""
],
[
"Tai",
"Ngee Chia",
""
],
[
"Leong",
"Wai Yi",
""
],
[
"Leong",
"Wei Qi",
""
],
[
"Yong",
"Xianbin",
""
],
[
"Ngui",
"Jian Gang",
""
],
[
"Susanto",
"Yosephine",
""
],
[
"Cheng",
"Nicholas",
""
],
[
"Rengarajan",
"Hamsawardhini",
""
],
[
"Limkonchotiwat",
"Peerat",
""
],
[
"Hulagadri",
"Adithya Venkatadri",
""
],
[
"Teng",
"Kok Wai",
""
],
[
"Tong",
"Yeo Yeow",
""
],
[
"Siow",
"Bryan",
""
],
[
"Teo",
"Wei Yi",
""
],
[
"Lau",
"Wayne",
""
],
[
"Tan",
"Choon Meng",
""
],
[
"Ong",
"Brandon",
""
],
[
"Ong",
"Zhi Hao",
""
],
[
"Montalan",
"Jann Railey",
""
],
[
"Chan",
"Adwin",
""
],
[
"Antonyrex",
"Sajeban",
""
],
[
"Lee",
"Ren",
""
],
[
"Choa",
"Esther",
""
],
[
"Tat-Wee",
"David Ong",
""
],
[
"Liu",
"Bing Jie Darius",
""
],
[
"Tjhi",
"William Chandra",
""
],
[
"Cambria",
"Erik",
""
],
[
"Teo",
"Leslie",
""
]
] |
2504.05810 | Xinpeng Ding | Xinpeng Ding, Kui Zhang, Jianhua Han, Lanqing Hong, Hang Xu, Xiaomeng
Li | PaMi-VDPO: Mitigating Video Hallucinations by Prompt-Aware
Multi-Instance Video Preference Learning | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Direct Preference Optimization (DPO) helps reduce hallucinations in Video
Multimodal Large Language Models (VLLMs), but its reliance on offline
preference data limits adaptability and fails to capture true video-response
misalignment. We propose Video Direct Preference Optimization (VDPO), an online
preference learning framework that eliminates the need for preference
annotation by leveraging video augmentations to generate rejected samples while
keeping responses fixed. However, selecting effective augmentations is
non-trivial, as some clips may be semantically identical to the original under
specific prompts, leading to false rejections and disrupting alignment. To
address this, we introduce Prompt-aware Multi-instance Learning VDPO
(PaMi-VDPO), which selects augmentations based on prompt context. Instead of a
single rejection, we construct a candidate set of augmented clips and apply a
close-to-far selection strategy, initially ensuring all clips are semantically
relevant while then prioritizing the most prompt-aware distinct clip. This
allows the model to better capture meaningful visual differences, mitigating
hallucinations, while avoiding false rejections, and improving alignment.
PaMi-VDPOseamlessly integrates into existing VLLMs without additional
parameters, GPT-4/human supervision. With only 10k SFT data, it improves the
base model by 5.3% on VideoHallucer, surpassing GPT-4o, while maintaining
stable performance on general video benchmarks.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 08:41:41 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 07:20:46 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Ding",
"Xinpeng",
""
],
[
"Zhang",
"Kui",
""
],
[
"Han",
"Jianhua",
""
],
[
"Hong",
"Lanqing",
""
],
[
"Xu",
"Hang",
""
],
[
"Li",
"Xiaomeng",
""
]
] |
2504.05848 | Tommaso Favalli | Tommaso Favalli | Relativistic limits on the discretization and temporal resolution of a
quantum clock | 7 pages, no figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We provide a brief discussion regarding relativistic limits on the
discretization and temporal resolution of time values in a quantum clock. Our
clock is characterized by a time observable chosen to be the complement of a
bounded and discrete Hamiltonian which can have an equally-spaced or a generic
spectrum. In the first case the time observable can be described by an
Hermitian operator and we find a limit in the discretization for the time
eigenvalues. Nevertheless, in both cases, the time observable can be described
by a POVM and, by increasing the number of time states, we can arbitrarily
reduce the bound on the minimum time quantum, demonstrating that we can safely
take the time values as continuous when the number of time states tends to
infinity. Finally, we find a limit for temporal resolution of our time
observable when the clock is used (together with light signals) in a
relativistic framework for measuring spacetime distances.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 09:27:41 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 01:53:02 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Favalli",
"Tommaso",
""
]
] |
2504.05853 | Oliver Schnetz | Oliver Schnetz | Notes on color reductions and $\gamma$ traces | 11 pages | null | null | null | hep-ph hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We present efficient algorithms to calculate the color factors for the
$SU(N)$ gauge group and to evaluate $\gamma$ traces. The aim of these notes is
to give a self-contained, proved account of the basic results with particular
emphasis on color reductions. We fine tune existing algorithms to make
calculations at high loop orders possible.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 09:29:23 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:54:14 GMT"
}
] | 2025-04-16T00:00:00 | [
[
"Schnetz",
"Oliver",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.