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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.09033 | Snigdha Agarwal | Snigdha Agarwal, Neelam Sinha | Chest X-ray Classification using Deep Convolution Models on
Low-resolution images with Uncertain Labels | 5 pages, 5 figures | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep Convolutional Neural Networks have consistently proven to achieve
state-of-the-art results on a lot of imaging tasks over the past years'
majority of which comprise of high-quality data. However, it is important to
work on low-resolution images since it could be a cheaper alternative for
remote healthcare access where the primary need of automated pathology
identification models occurs. Medical diagnosis using low-resolution images is
challenging since critical details may not be easily identifiable. In this
paper, we report classification results by experimenting on different input
image sizes of Chest X-rays to deep CNN models and discuss the feasibility of
classification on varying image sizes. We also leverage the noisy labels in the
dataset by proposing a Randomized Flipping of labels techniques. We use an
ensemble of multi-label classification models on frontal and lateral studies.
Our models are trained on 5 out of the 14 chest pathologies of the publicly
available CheXpert dataset. We incorporate techniques such as augmentation,
regularization for model improvement and use class activation maps to visualize
the neural network's decision making. Comparison with classification results on
data from 200 subjects, obtained on the corresponding high-resolution images,
reported in the original CheXpert paper, has been presented. For pathologies
Cardiomegaly, Consolidation and Edema, we obtain 3% higher accuracy with our
model architecture.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:13:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Agarwal",
"Snigdha",
""
],
[
"Sinha",
"Neelam",
""
]
] |
2504.09034 | Gary Guth | Gary Guth and Ciprian Manolescu | Real Heegaard Floer Homology | 44 pages, 14 figures | null | null | null | math.GT math.SG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We define an invariant of three-manifolds with an involution with non-empty
fixed point set of codimension $2$; in particular, this applies to double
branched covers over knots. Our construction gives the Heegaard Floer analogue
of Li's real monopole Floer homology. It is a special case of a real version of
Lagrangian Floer homology, which may be of independent interest to symplectic
geometers. The Euler characteristic of the real Heegaard Floer homology is the
analogue of Miyazawa's invariant, and can be computed combinatorially for all
knots.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:13:25 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Guth",
"Gary",
""
],
[
"Manolescu",
"Ciprian",
""
]
] |
2504.09035 | Shubham Aggarwal | Shubham Aggarwal, Dipankar Maity, and Tamer Ba\c{s}ar | InterQ: A DQN Framework for Optimal Intermittent Control | Submitted to IEEE for possible publication | null | null | null | math.OC cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this letter, we explore the communication-control co-design of
discrete-time stochastic linear systems through reinforcement learning.
Specifically, we examine a closed-loop system involving two sequential
decision-makers: a scheduler and a controller. The scheduler continuously
monitors the system's state but transmits it to the controller intermittently
to balance the communication cost and control performance. The controller, in
turn, determines the control input based on the intermittently received
information. Given the partially nested information structure, we show that the
optimal control policy follows a certainty-equivalence form. Subsequently, we
analyze the qualitative behavior of the scheduling policy. To develop the
optimal scheduling policy, we propose InterQ, a deep reinforcement learning
algorithm which uses a deep neural network to approximate the Q-function.
Through extensive numerical evaluations, we analyze the scheduling landscape
and further compare our approach against two baseline strategies: (a) a
multi-period periodic scheduling policy, and (b) an event-triggered policy. The
results demonstrate that our proposed method outperforms both baselines. The
open source implementation can be found at https://github.com/AC-sh/InterQ.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:18:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Aggarwal",
"Shubham",
""
],
[
"Maity",
"Dipankar",
""
],
[
"Başar",
"Tamer",
""
]
] |
2504.09036 | Rahul Jain | R. Jain, E. F. Brown, H. Schatz, A. V. Afanasjev, M. Beard, L. R.
Gasques, J. Grace, A. Heger, G. W. Hitt, W. R. Hix, R. Lau, W.-J. Ong, M.
Wiescher, Y. Xu | Crust composition and the Shallow Heat Source in KS 1731-260 | null | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The presence of a shallow heat source of unknown origin in accreting neutron
star crusts has been inferred by analyzing their cooling behavior in
quiescence. To investigate a diverse bursting history for KS 1731-260 during
accretion outbursts, we use realistic crust compositions and nuclear heating
and cooling sources from detailed nuclear reaction network calculations to
interpret observed cooling curves. We find that the required strength of the
shallow heat source is reduced by more than a factor of 3 compared to previous
analysis, and obtain constraints on the most likely dominant surface burning
modes of KS 1731-260 over its history. Our analysis suggests an impure nuclear
pasta layer in the inner crust, though future observations will provide more
stringent constraints.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:20:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jain",
"R.",
""
],
[
"Brown",
"E. F.",
""
],
[
"Schatz",
"H.",
""
],
[
"Afanasjev",
"A. V.",
""
],
[
"Beard",
"M.",
""
],
[
"Gasques",
"L. R.",
""
],
[
"Grace",
"J.",
""
],
[
"Heger",
"A.",
""
],
[
"Hitt",
"G. W.",
""
],
[
"Hix",
"W. R.",
""
],
[
"Lau",
"R.",
""
],
[
"Ong",
"W. -J.",
""
],
[
"Wiescher",
"M.",
""
],
[
"Xu",
"Y.",
""
]
] |
2504.09037 | Fangkai Jiao | Zixuan Ke, Fangkai Jiao, Yifei Ming, Xuan-Phi Nguyen, Austin Xu, Do
Xuan Long, Minzhi Li, Chengwei Qin, Peifeng Wang, Silvio Savarese, Caiming
Xiong, Shafiq Joty | A Survey of Frontiers in LLM Reasoning: Inference Scaling, Learning to
Reason, and Agentic Systems | 72 pages, 6 figures | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Reasoning is a fundamental cognitive process that enables logical inference,
problem-solving, and decision-making. With the rapid advancement of large
language models (LLMs), reasoning has emerged as a key capability that
distinguishes advanced AI systems from conventional models that empower
chatbots. In this survey, we categorize existing methods along two orthogonal
dimensions: (1) Regimes, which define the stage at which reasoning is achieved
(either at inference time or through dedicated training); and (2)
Architectures, which determine the components involved in the reasoning
process, distinguishing between standalone LLMs and agentic compound systems
that incorporate external tools, and multi-agent collaborations. Within each
dimension, we analyze two key perspectives: (1) Input level, which focuses on
techniques that construct high-quality prompts that the LLM condition on; and
(2) Output level, which methods that refine multiple sampled candidates to
enhance reasoning quality. This categorization provides a systematic
understanding of the evolving landscape of LLM reasoning, highlighting emerging
trends such as the shift from inference-scaling to learning-to-reason (e.g.,
DeepSeek-R1), and the transition to agentic workflows (e.g., OpenAI Deep
Research, Manus Agent). Additionally, we cover a broad spectrum of learning
algorithms, from supervised fine-tuning to reinforcement learning such as PPO
and GRPO, and the training of reasoners and verifiers. We also examine key
designs of agentic workflows, from established patterns like
generator-evaluator and LLM debate to recent innovations. ...
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:27:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ke",
"Zixuan",
""
],
[
"Jiao",
"Fangkai",
""
],
[
"Ming",
"Yifei",
""
],
[
"Nguyen",
"Xuan-Phi",
""
],
[
"Xu",
"Austin",
""
],
[
"Long",
"Do Xuan",
""
],
[
"Li",
"Minzhi",
""
],
[
"Qin",
"Chengwei",
""
],
[
"Wang",
"Peifeng",
""
],
[
"Savarese",
"Silvio",
""
],
[
"Xiong",
"Caiming",
""
],
[
"Joty",
"Shafiq",
""
]
] |
2504.09038 | Paul Lutkus | Paul Lutkus, Michelle S. Chong, and Lars Lindemann | Nonconvex Obstacle Avoidance using Efficient Sampling-Based Distance
Functions | null | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider nonconvex obstacle avoidance where a robot described by nonlinear
dynamics and a nonconvex shape has to avoid nonconvex obstacles. Obstacle
avoidance is a fundamental problem in robotics and well studied in control.
However, existing solutions are computationally expensive (e.g., model
predictive controllers), neglect nonlinear dynamics (e.g., graph-based
planners), use diffeomorphic transformations into convex domains (e.g., for
star shapes), or are conservative due to convex overapproximations. The key
challenge here is that the computation of the distance between the shapes of
the robot and the obstacles is a nonconvex problem. We propose efficient
computation of this distance via sampling-based distance functions. We quantify
the sampling error and show that, for certain systems, such sampling-based
distance functions are valid nonsmooth control barrier functions. We also study
how to deal with disturbances on the robot dynamics in our setting. Finally, we
illustrate our method on a robot navigation task involving an omnidirectional
robot and nonconvex obstacles. We also analyze performance and computational
efficiency of our controller as a function of the number of samples.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:27:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lutkus",
"Paul",
""
],
[
"Chong",
"Michelle S.",
""
],
[
"Lindemann",
"Lars",
""
]
] |
2504.09039 | Gen Li | Gen Li, Yang Xiao, Jie Ji, Kaiyuan Deng, Bo Hui, Linke Guo, Xiaolong
Ma | Sculpting Memory: Multi-Concept Forgetting in Diffusion Models via
Dynamic Mask and Concept-Aware Optimization | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Text-to-image (T2I) diffusion models have achieved remarkable success in
generating high-quality images from textual prompts. However, their ability to
store vast amounts of knowledge raises concerns in scenarios where selective
forgetting is necessary, such as removing copyrighted content, reducing biases,
or eliminating harmful concepts. While existing unlearning methods can remove
certain concepts, they struggle with multi-concept forgetting due to
instability, residual knowledge persistence, and generation quality
degradation. To address these challenges, we propose \textbf{Dynamic Mask
coupled with Concept-Aware Loss}, a novel unlearning framework designed for
multi-concept forgetting in diffusion models. Our \textbf{Dynamic Mask}
mechanism adaptively updates gradient masks based on current optimization
states, allowing selective weight modifications that prevent interference with
unrelated knowledge. Additionally, our \textbf{Concept-Aware Loss} explicitly
guides the unlearning process by enforcing semantic consistency through
superclass alignment, while a regularization loss based on knowledge
distillation ensures that previously unlearned concepts remain forgotten during
sequential unlearning. We conduct extensive experiments to evaluate our
approach. Results demonstrate that our method outperforms existing unlearning
techniques in forgetting effectiveness, output fidelity, and semantic
coherence, particularly in multi-concept scenarios. Our work provides a
principled and flexible framework for stable and high-fidelity unlearning in
generative models. The code will be released publicly.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:38:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Gen",
""
],
[
"Xiao",
"Yang",
""
],
[
"Ji",
"Jie",
""
],
[
"Deng",
"Kaiyuan",
""
],
[
"Hui",
"Bo",
""
],
[
"Guo",
"Linke",
""
],
[
"Ma",
"Xiaolong",
""
]
] |
2504.09040 | Steven K. Lamoreaux | Steven K. Lamoreaux | The Chemistry and Physics of $^{199}$Hg Nuclear Spin Polarization
Relaxation in Quantum Magnetometry Cells | 33 pages, 6 figures | null | null | null | nucl-ex physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum spin magnetometry using optically pumped $^{199}$Hg has been
successfully used in many fundamental physics experiments. A serious problem
that has not been resolved is the instability of the $^{199}$Hg spin relaxation
rate in atomic vapor cells under irradiation with 254 nm Hg resonance light. In
this paper, previously obtained data are re-analyzed or analyzed for the first
time. The effects of impurities of H$_2$ and O$_2$ are elucidated, and possible
ways to stabilize cells are discussed. Surface states originating from the an
der Waals interaction of \hg with fused silica are analyzed and shown to be
critical to understanding relaxation mechanisms. A discussion of the possible
use of a mixture of N$_2$O and other gases is presented.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:43:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lamoreaux",
"Steven K.",
""
]
] |
2504.09041 | Kannan Premnath | William Schupbach, Kannan Premnath | Fokker-Planck Model Based Central Moment Lattice Boltzmann Method for
Effective Simulations of Thermal Convective Flows | 65 pages, 13 figures | null | null | null | physics.flu-dyn physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | The Fokker-Planck (FP) equation represents the drift-diffusive processes in
kinetic models. It can also be regarded as a model for the collision integral
of the Boltzmann-type equation to represent thermo-hydrodynamic processes in
fluids. The lattice Boltzmann method (LBM) is a drastically simplified
discretization of the Boltzmann equation. We construct two new FP-based LBMs,
one for recovering the Navier-Stokes equations and the other for simulating the
energy equation, where, in each case, the effect of collisions is represented
as relaxations of different central moments to their respective attractors.
Such attractors are obtained by matching the changes in various discrete
central moments under collision with the continuous central moments prescribed
by the FP model. As such, the resulting central moment attractors depend on the
lower order moments and the diffusion tensor parameters and significantly
differ from those based on the Maxwell distribution. The diffusion tensor
parameters for evolving higher moments in simulating fluid motions at
relatively low viscosities are chosen based on a renormalization principle. The
use of such central moment formulations in modeling the collision step offers
significant improvements in numerical stability, especially for simulations of
thermal convective flows with a wide range of variations in the transport
coefficients. We develop new FP central moment LBMs for thermo-hydrodynamics in
both two- and three-dimensions and demonstrate the ability of our approach to
accurately simulate various cases involving thermal convective buoyancy-driven
flows, especially at high Rayleigh numbers. Moreover, we show significant
improvements in numerical stability of our FP central moment LBMs when compared
to other existing central moment LBMs using the Maxwell distribution in
achieving higher Peclet numbers for mixed convection flows.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:46:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Schupbach",
"William",
""
],
[
"Premnath",
"Kannan",
""
]
] |
2504.09042 | Vine Madukpe Mr | Vine Nwabuisi Madukpe, Bright Chukwuma Ugoala, Nur Fariha Syaqina
Zulkepli | A Comprehensive Review of the Mapper Algorithm, a Topological Data
Analysis Technique, and Its Applications Across Various Fields (2007-2025) | 31 pages, 9 figures | null | null | null | math.GN | http://creativecommons.org/licenses/by/4.0/ | The Mapper algorithm, a technique within topological data analysis (TDA),
constructs a simplified graphical representation of high-dimensional data to
uncover its underlying shape and structural patterns. The algorithm has
attracted significant attention from researchers and has been applied across
various disciplines. However, to the best of the authors' knowledge, no
comprehensive review currently exists on the Mapper algorithm and its variants
as applied across different fields of study between 2007 and 2025. This review
addresses this gap and serves as a valuable resource for researchers and
practitioners aiming to apply or advance the algorithm. The reviewed literature
comprises peer-reviewed articles retrieved from major academic databases,
including Google Scholar, Web of Science, Scopus, JSTOR, PubMed, and IEEE
Xplore, using the keywords 'topological data analysis,' 'mapper algorithm,' and
'topological graph.' The study further provides an overview and a comparative
analysis of the suitability of the most commonly used filter functions and
clustering algorithms within the Mapper framework. Additionally, it examines
current trends, identifies limitations, and proposes future research directions
for the Mapper algorithm and its variants, emphasizing the need for developing
effective methodologies to streamline the analysis of high-dimensional data in
the age of big data proliferation.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:48:25 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Madukpe",
"Vine Nwabuisi",
""
],
[
"Ugoala",
"Bright Chukwuma",
""
],
[
"Zulkepli",
"Nur Fariha Syaqina",
""
]
] |
2504.09043 | G\"ok\c{c}e Dayan{\i}kl{\i} | Gokce Dayanikli and Mathieu Lauriere | Cooperation, Competition, and Common Pool Resources in Mean Field Games | 53 pages, 4 figures | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mean field games (MFGs) have been introduced to study Nash equilibria in very
large population of self-interested agents. However, when applied to common
pool resource (CPR) games, MFG equilibria lead to the so-called tragedy of the
commons (TOTC). Empirical studies have shown that in many situations, TOTC does
not materialize which hints at the fact that standard MFG models cannot explain
the behavior of agents in CPR games. In this work, we study two models which
incorporate a mix of cooperative and non-cooperative behaviors, either at the
individual level or the population level. After defining these models, we study
optimality conditions in the form of forward-backward stochastic differential
equations and we prove that the mean field models provide approximate
equilibria controls for corresponding finite-agent games. We then show an
application to a model of fish stock management, for which the solution can be
computed by solving systems of ordinary differential equations, which we prove
to have a unique solution. Numerical results illustrate the impact of the level
of cooperation at the individual and the population levels on the CPR.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:55:35 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dayanikli",
"Gokce",
""
],
[
"Lauriere",
"Mathieu",
""
]
] |
2504.09044 | Xiaofeng Dong | Xiaofeng Dong and Yanyong Hong | On quadratic Novikov algebras | null | null | null | null | math.RA math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | A quadratic Novikov algebra is a Novikov algebra $(A, \circ)$ with a
symmetric and nondegenerate bilinear form $B(\cdot,\cdot)$ satisfying $B(a\circ
b, c)=-B(b, a\circ c+c\circ a)$ for all $a$, $b$, $c\in A$. This notion
appeared in the theory of Novikov bialgebras. In this paper, we first
investigate some properties of quadratic Novikov algebras and give a
decomposition theorem of quadratic Novikov algebras. Then we present a
classification of quadratic Novikov algebras of dimensions $2$ and $3$ over
$\mathbb{C}$ up to isomorphism. Finally, a construction of quadratic Novikov
algebras called double extension is presented and we show that any quadratic
Novikov algebra containing a nonzero isotropic ideal can be obtained by double
extensions. Based on double extension, an example of quadratic Novikov algebras
of dimension 4 is given.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:56:39 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dong",
"Xiaofeng",
""
],
[
"Hong",
"Yanyong",
""
]
] |
2504.09045 | Genan Zhao | Genan Zhao | Classification of Solutions with Polynomial Energy Growth for the SU (n
+ 1) Toda System on the Punctured Complex Plane | null | null | null | null | nlin.SI math.AP math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates the classification of solutions satisfying the
polynomial energy growth condition near both the origin and infinity to the
${\mathrm SU}(n+1)$ Toda system on the punctured complex plane $\mathbb{C}^*$.
The ${\mathrm SU}(n+1)$ Toda system is a class of nonlinear elliptic partial
differential equations of second order with significant implications in
integrable systems, quantum field theory, and differential geometry. Building
on the work of A. Eremenko (J. Math. Phys. Anal. Geom., Volume 3 p.39-46),
Jingyu Mu's thesis, and others, we obtain the classification of such solutions
by leveraging techniques from the Nevanlinna theory. In particular, we prove
that the unitary curve corresponding to a solution with polynomial energy
growth to the ${\mathrm SU}(n+1)$ Toda system on $\mathbb{C}^*$ gives a set of
fundamental solutions to a linear homogeneous ODE of $(n+1)^{th}$ order, and
each coefficient of the ODE can be written as a sum of a polynomial in $z$ and
another one in $\frac{1}{z}$.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 01:57:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhao",
"Genan",
""
]
] |
2504.09047 | Rayan Bahrami | Rayan Bahrami and Hamidreza Jafarnejadsani | Multi-Robot Coordination with Adversarial Perception | to appear at the 2025 Int'l Conference on Unmanned Aircraft Systems
(ICUAS) | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the resilience of perception-based multi-robot
coordination with wireless communication to online adversarial perception. A
systematic study of this problem is essential for many safety-critical robotic
applications that rely on the measurements from learned perception modules. We
consider a (small) team of quadrotor robots that rely only on an Inertial
Measurement Unit (IMU) and the visual data measurements obtained from a learned
multi-task perception module (e.g., object detection) for downstream tasks,
including relative localization and coordination. We focus on a class of
adversarial perception attacks that cause misclassification, mislocalization,
and latency. We propose that the effects of adversarial misclassification and
mislocalization can be modeled as sporadic (intermittent) and spurious
measurement data for the downstream tasks. To address this, we present a
framework for resilience analysis of multi-robot coordination with adversarial
measurements. The framework integrates data from Visual-Inertial Odometry (VIO)
and the learned perception model for robust relative localization and state
estimation in the presence of adversarially sporadic and spurious measurements.
The framework allows for quantifying the degradation in system observability
and stability in relation to the success rate of adversarial perception.
Finally, experimental results on a multi-robot platform demonstrate the
real-world applicability of our methodology for resource-constrained robotic
platforms.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:04:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bahrami",
"Rayan",
""
],
[
"Jafarnejadsani",
"Hamidreza",
""
]
] |
2504.09049 | Pedro Henrique Vaz Valois | Adrianna Romanowski, Pedro H. V. Valois, Kazuhiro Fukui | From Punchlines to Predictions: A Metric to Assess LLM Performance in
Identifying Humor in Stand-Up Comedy | Accepted to CMCL2025 @ NAACL | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Comedy serves as a profound reflection of the times we live in and is a
staple element of human interactions. In light of the widespread adoption of
Large Language Models (LLMs), the intersection of humor and AI has become no
laughing matter. Advancements in the naturalness of human-computer interaction
correlates with improvements in AI systems' abilities to understand humor. In
this study, we assess the ability of models in accurately identifying humorous
quotes from a stand-up comedy transcript. Stand-up comedy's unique comedic
narratives make it an ideal dataset to improve the overall naturalness of
comedic understanding. We propose a novel humor detection metric designed to
evaluate LLMs amongst various prompts on their capability to extract humorous
punchlines. The metric has a modular structure that offers three different
scoring methods - fuzzy string matching, sentence embedding, and subspace
similarity - to provide an overarching assessment of a model's performance. The
model's results are compared against those of human evaluators on the same
task. Our metric reveals that regardless of prompt engineering, leading models,
ChatGPT, Claude, and DeepSeek, achieve scores of at most 51% in humor
detection. Notably, this performance surpasses that of humans who achieve a
score of 41%. The analysis of human evaluators and LLMs reveals variability in
agreement, highlighting the subjectivity inherent in humor and the complexities
involved in extracting humorous quotes from live performance transcripts. Code
available at https://github.com/swaggirl9000/humor.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:19:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Romanowski",
"Adrianna",
""
],
[
"Valois",
"Pedro H. V.",
""
],
[
"Fukui",
"Kazuhiro",
""
]
] |
2504.09050 | Giovanni Otalora | L. K. Duchaniya, B. Mishra, G. Otalora, M. Gonzalez-Espinoza | Late-time acceleration and structure formation in interacting
$\alpha$-attractor dark energy models | 16 pages, 13 figures, 2 tables | null | null | null | gr-qc astro-ph.CO hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the cosmological dynamics of interacting dark energy within
the framework of $\alpha$-attractor models. Specifically, we analyze the
associated autonomous system, focusing on its fixed points that represent dark
energy and scaling solutions, along with their stability conditions. We employ
center manifold theory to address cases where some fixed points display
eigenvalues with zero and negative real parts. The model reveals attractors
describing dark energy, enabling a smooth transition from the
radiation-dominated era to the matter-dominated era, and ultimately into the
dark-energy-dominated phase. Additionally, we identify a scaling matter
solution capable of modifying the growth rate of matter perturbations during
the matter-dominated epoch. Consequently, we study the evolution of matter
perturbations by obtaining both analytical and numerical solutions to the
density contrast evolution equation. Based on these results, we compute
numerical solutions for the weighted growth rate $f\sigma_{8}$, indicating that
interacting $\alpha$-attractor dark energy models may provide a better fit to
structure formation data than the standard $\Lambda$CDM scenario.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:20:41 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Duchaniya",
"L. K.",
""
],
[
"Mishra",
"B.",
""
],
[
"Otalora",
"G.",
""
],
[
"Gonzalez-Espinoza",
"M.",
""
]
] |
2504.09051 | Yuanfan Zhuo | Yuanfan Zhuo, Xingliang Liang, Yanan Wu, Xianzhong Zhao | The varieties generated by 3-hypergraph semirings | null | null | null | null | math.RA math.CO | http://creativecommons.org/licenses/by/4.0/ | In this paper the 3-hypergraph semigroups and 3-hypergraph semirings from
3-hypergraphs $\mathbb{H}$ are introduced and the varieties generated by them
are studied. It is shown that all 3-hypergraph semirings $S_{\scriptscriptstyle
\mathbb{H}}$ are nonfinitely based and subdirectly irreducible. Also, it is
proved that each variety generated by 3-hypergraph semirings is equal to a
variety generated by 3-uniform hypergraph semirings. It is well known that both
variety $\mathbf{V}(S_c(abc))$ (see, J. Algebra 611: 211--245, 2022 and J.
Algebra 623: 64--85, 2023) and variety $\mathbf{V}(S_{\scriptscriptstyle
\mathbb{H}})$ play key role in the theory of variety of ai-semirings, where
3-uniform hypergraph $\mathbb{H}$ is a 3-cycle. They are shown that each
variety generated by 2-robustly strong 3-colorable 3-uniform hypergraph
semirings is equal to variety $\mathbf{V}(S_c(abc))$, and each variety
generated by so-called beam-type hypergraph semirings or fan-type hypergraph
semirings is equal to the variety $\mathbf{V}(S_{\scriptscriptstyle
\mathbb{H}})$ generated by a 3-uniform 3-cycle hypergraph semiring
$S_{\scriptscriptstyle \mathbb{H}}$. Finally, an infinite ascending chain is
provided in the lattice of subvarieties of the variety generated by all
3-uniform hypergraph semirings. This implies that the variety generated by all
3-uniform hypergraph semirings has infinitely many subvarieties.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:40:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhuo",
"Yuanfan",
""
],
[
"Liang",
"Xingliang",
""
],
[
"Wu",
"Yanan",
""
],
[
"Zhao",
"Xianzhong",
""
]
] |
2504.09052 | Zhi Ling | Zhi Ling, Shozen Dan | Bayesian shrinkage priors subject to linear constraints | null | null | null | null | stat.ME math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In Bayesian regression models with categorical predictors, constraints are
needed to ensure identifiability when using all $K$ levels of a factor. The
sum-to-zero constraint is particularly useful as it allows coefficients to
represent deviations from the population average. However, implementing such
constraints in Bayesian settings is challenging, especially when assigning
appropriate priors that respect these constraints and general principles. Here
we develop a multivariate normal prior family that satisfies arbitrary linear
constraints while preserving the local adaptivity properties of shrinkage
priors, with an efficient implementation algorithm for probabilistic
programming languages. Our approach applies broadly to various shrinkage
frameworks including Bayesian Ridge, horseshoe priors and their variants,
demonstrating excellent performance in simulation studies. The covariance
structure we derive generalizes beyond regression models to any Bayesian
analysis requiring linear constraints on parameters, providing practitioners
with a principled approach to parameter identification while maintaining proper
uncertainty quantification and interpretability.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:45:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ling",
"Zhi",
""
],
[
"Dan",
"Shozen",
""
]
] |
2504.09053 | Sukirt Thakur | Sukirt Thakur, Maziar Raissi | ELPINN: Eulerian Lagrangian Physics-Informed Neural Network | null | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Physics-Informed Neural Networks (PINNs) have gained widespread popularity
for solving inverse and forward problems across a range of scientific and
engineering domains. However, most existing PINN frameworks are limited to the
Eulerian domain, where physical quantities are described at fixed spatial
locations. In this work, we propose a novel PINN-based framework that couples
Eulerian and Lagrangian perspectives by using particle trajectory data to
reconstruct Eulerian velocity and pressure fields.
We evaluate the performance of our method across three distinct fluid flow
scenarios: two-dimensional external flow past a cylinder, two-dimensional
internal flow in a confined geometry, and three-dimensional internal flow
inside an airplane cabin. In all three cases, we successfully reconstruct the
velocity field from Lagrangian particle data. Moreover, for the 2D external and
internal flows, we recover the pressure field solely through the
physics-informed learning process, without using any direct pressure
measurements.
We also conduct a sensitivity analysis to understand the effects of temporal
resolution and particle count on the reconstruction accuracy. Our results show
that smaller time-step sizes significantly improve the predictions, while the
total number of particles has a comparatively smaller influence.
These findings establish the potential of our coupled Eulerian-Lagrangian
PINN framework as a powerful tool for enhancing experimental methods such as
Particle Tracking Velocimetry (PTV). Looking ahead, this approach may be
extended to infer hidden quantities such as pressure in three-dimensional flows
or material properties like viscosity, opening new avenues for data-driven
fluid dynamics in complex geometries.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:47:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Thakur",
"Sukirt",
""
],
[
"Raissi",
"Maziar",
""
]
] |
2504.09054 | Peng-Ju Wu | Peng-Ju Wu | Comparison of dark energy models using late-universe observations | 17 pages, 13 figures | null | null | null | astro-ph.CO gr-qc hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the framework of general relativity, dark energy was proposed to explain
the cosmic acceleration. A pivotal inquiry in cosmology is to determine whether
dark energy is the cosmological constant, and if not, the challenge lies in
constraining how it evolves with time. In this paper, we utilize the latest
observational data to constrain some typical dark energy models, and make a
comparison for them according to their capabilities of fitting the current
data. Our study is confined to late-universe observations, including the baryon
acoustic oscillation, type Ia supernova, cosmic chronometer, and strong
gravitational lensing time delay data. We employ the Akaike information
criterion (AIC) and Bayesian information criterion (BIC) to assess the worth of
models. The AIC analysis indicates that all dark energy models outperform the
$\Lambda$CDM model. However, the BIC analysis leaves room for $\Lambda$CDM due
to its heavier penalty on the model complexity. Compared to $\Lambda$CDM, most
dark energy models are robustly supported by AIC while being explicitly
disfavored by BIC. The models that are robustly favored by AIC and not
explicitly disfavored by BIC include the $w$CDM, interacting dark energy, and
Ricci dark energy models. Furthermore, we observe that an alternative modified
gravity model exhibits superior performance when compared with $\Lambda$CDM
from both the AIC and BIC perspectives.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:48:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wu",
"Peng-Ju",
""
]
] |
2504.09055 | Rinku Maji | Rinku Maji, Qaisar Shafi | Superheavy Metastable Strings in SO(10) | 14 pages, 9 captioned figures | null | null | null | hep-ph astro-ph.CO hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The spontaneous breaking of $SO(10)$ grand unified symmetry to $SU(3)_c
\times SU(2)_L \times U(1)_Y \times U(1)_\chi$ yields the GUT monopole as well
as a comparably heavy $U(1)_\chi$ monopole which also carries $U(1)_Y$ flux. A
metastable string scenario in this case requires that the $U(1)_\chi$ symmetry
is necessarily broken close to the GUT scale, thus resulting in a dimensionless
string tension $G \mu \sim 10^{-6}$. We show that the $\chi$ monopole does not
carry any unconfined flux following the electroweak symmetry breaking. Coupled
with $G \mu \sim 10^{-6}$, this metastable string network appears to provide a
good fit to the recent Pulsar Timing Array data on the stochastic gravitational
background. Gauge coupling unification, especially in the presence of low scale
supersymmetry, determines the GUT scale and, in combination with constraints
from proton decay experiments, one is able to constrain some of the key
parameters in this setup. The breaking of $SO(10)$ via $SU(5) \times U(1)_\chi$
also yields superheavy metastable strings with no unconfined flux associated
with the monopoles. Finally, we consider $SO(10)$ breaking via $SU(4)_c \times
SU(2)_L \times U(1)_R$, $SU(3)_c \times SU(2)_L \times SU(2)_R \times
U(1)_{B-L}$ and flipped $SU(5)$ that yield metastable strings where the
associated monopoles carry unconfined flux after the electroweak breaking.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:54:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Maji",
"Rinku",
""
],
[
"Shafi",
"Qaisar",
""
]
] |
2504.09056 | Daniel Larsen | Daniel Larsen | Carmichael Numbers in All Possible Arithmetic Progressions | 55 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove that every arithmetic progression either contains infinitely many
Carmichael numbers or none at all. Furthermore, there is a simple criterion for
determining which category a given arithmetic progression falls into. In
particular, if $m$ is any integer such that $(m,2\phi(m))=1$ then there exist
infinitely many Carmichael numbers divisible by $m$. As a consequence, we are
able to prove that $\liminf_{n\text{ Carmichael}}\frac{\phi(n)}{n}=0$,
resolving a question of Alford, Granville, and Pomerance.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:55:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Larsen",
"Daniel",
""
]
] |
2504.09057 | Yuyang Zhang | Yuyang Zhang, Xinhe Zhang, Jia Liu, Na Li | Sample Efficient Algorithms for Linear System Identification under Noisy
Observations | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | In this paper, we focus on learning linear dynamical systems under noisy
observations. In this setting, existing algorithms either yield biased
parameter estimates, or suffer from large sample complexities. To address these
issues, we adapt the instrumental variable method and the bias compensation
method, originally proposed for error-in-variables models, to our setting and
provide refined non-asymptotic analysis. Under mild conditions, our algorithms
achieve superior sample complexities that match the best-known sample
complexity for learning a fully observable system without observation noise.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:23:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Yuyang",
""
],
[
"Zhang",
"Xinhe",
""
],
[
"Liu",
"Jia",
""
],
[
"Li",
"Na",
""
]
] |
2504.09058 | Chengyuan Liu | Chengyuan Liu, Shihang Wang, Lizhi Qing, Kaisong Song, Junjie Cao, Jun
Lin, Ji Zhang, Ang Li, Kun Kuang, Fei Wu | Towards Stepwise Domain Knowledge-Driven Reasoning Optimization and
Reflection Improvement | Under review | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, stepwise supervision on Chain of Thoughts (CoTs) presents an
enhancement on the logical reasoning tasks such as coding and math, with the
help of Monte Carlo Tree Search (MCTS). However, its contribution to tasks
requiring domain-specific expertise and knowledge remains unexplored. Motivated
by the interest, we identify several potential challenges of vanilla MCTS
within this context, and propose the framework of Stepwise Domain
Knowledge-Driven Reasoning Optimization, employing the MCTS algorithm to
develop step-level supervision for problems that require essential
comprehension, reasoning, and specialized knowledge. Additionally, we also
introduce the Preference Optimization towards Reflection Paths, which
iteratively learns self-reflection on the reasoning thoughts from better
perspectives. We have conducted extensive experiments to evaluate the advantage
of the methodologies. Empirical results demonstrate the effectiveness on
various legal-domain problems. We also report a diverse set of valuable
findings, hoping to encourage the enthusiasm to the research of domain-specific
LLMs and MCTS.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:25:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Chengyuan",
""
],
[
"Wang",
"Shihang",
""
],
[
"Qing",
"Lizhi",
""
],
[
"Song",
"Kaisong",
""
],
[
"Cao",
"Junjie",
""
],
[
"Lin",
"Jun",
""
],
[
"Zhang",
"Ji",
""
],
[
"Li",
"Ang",
""
],
[
"Kuang",
"Kun",
""
],
[
"Wu",
"Fei",
""
]
] |
2504.09059 | Seyyedreza Madani | Seyyedreza Madani, Ahmadreza Tavasoli, Zahra Khoshtarash Astaneh,
Pierre-Olivier Pineau | Large Language Models integration in Smart Grids | null | null | null | null | cs.CY cs.ET | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) are changing the way we operate our society and
will undoubtedly impact power systems as well - but how exactly? By integrating
various data streams - including real-time grid data, market dynamics, and
consumer behaviors - LLMs have the potential to make power system operations
more adaptive, enhance proactive security measures, and deliver personalized
energy services. This paper provides a comprehensive analysis of 30 real-world
applications across eight key categories: Grid Operations and Management,
Energy Markets and Trading, Personalized Energy Management and Customer
Engagement, Grid Planning and Education, Grid Security and Compliance, Advanced
Data Analysis and Knowledge Discovery, Emerging Applications and Societal
Impact, and LLM-Enhanced Reinforcement Learning. Critical technical hurdles,
such as data privacy and model reliability, are examined, along with possible
solutions. Ultimately, this review illustrates how LLMs can significantly
contribute to building more resilient, efficient, and sustainable energy
infrastructures, underscoring the necessity of their responsible and equitable
deployment.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:29:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Madani",
"Seyyedreza",
""
],
[
"Tavasoli",
"Ahmadreza",
""
],
[
"Astaneh",
"Zahra Khoshtarash",
""
],
[
"Pineau",
"Pierre-Olivier",
""
]
] |
2504.09060 | Minghao Yang | Minghao Yang, Pengteng Li, Yan Liang, Qianyi Cai, Zhihang Zheng,
Shichen Zhang, Pengfei Zhang, Zhi-An Huang, Hui Xiong | Multimodal 3D Genome Pre-training | null | null | null | null | cs.LG cs.AI q-bio.GN | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Deep learning techniques have driven significant progress in various
analytical tasks within 3D genomics in computational biology. However, a
holistic understanding of 3D genomics knowledge remains underexplored. Here, we
propose MIX-HIC, the first multimodal foundation model of 3D genome that
integrates both 3D genome structure and epigenomic tracks, which obtains
unified and comprehensive semantics. For accurate heterogeneous semantic
fusion, we design the cross-modal interaction and mapping blocks for robust
unified representation, yielding the accurate aggregation of 3D genome
knowledge. Besides, we introduce the first large-scale dataset comprising over
1 million pairwise samples of Hi-C contact maps and epigenomic tracks for
high-quality pre-training, enabling the exploration of functional implications
in 3D genomics. Extensive experiments show that MIX-HIC can significantly
surpass existing state-of-the-art methods in diverse downstream tasks. This
work provides a valuable resource for advancing 3D genomics research.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:31:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Minghao",
""
],
[
"Li",
"Pengteng",
""
],
[
"Liang",
"Yan",
""
],
[
"Cai",
"Qianyi",
""
],
[
"Zheng",
"Zhihang",
""
],
[
"Zhang",
"Shichen",
""
],
[
"Zhang",
"Pengfei",
""
],
[
"Huang",
"Zhi-An",
""
],
[
"Xiong",
"Hui",
""
]
] |
2504.09061 | Yan Peng | Yan Peng | An upper bound on the minimum orbital period of black holes | 8 pages, 2 figures | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Previous research has focused on establishing lower bounds on the minimum
orbital period of black holes. In this work, we explore the complementary
question of whether an upper bound exists for the minimum orbital period of
black holes. We investigate the minimum orbital periods of three types of black
holes: Schwarzschild, Reissner-Nordstr\"{o}m and Kerr-Newman black holes. We
find that the minimum orbital period of these black holes is bounded by an
upper limit $T_{min} \leqslant 6\sqrt{3}\pi M$, where $M$ is the black hole
mass. Our results suggest that this upper bound on the minimum orbital period
may be a general property in black hole spacetimes.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:36:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Peng",
"Yan",
""
]
] |
2504.09062 | Zhijie Shen | Zhijie Shen, Chunyu Lin, Shujuan Huang, Lang Nie, Kang Liao, and Yao
Zhao | You Need a Transition Plane: Bridging Continuous Panoramic 3D
Reconstruction with Perspective Gaussian Splatting | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, reconstructing scenes from a single panoramic image using advanced
3D Gaussian Splatting (3DGS) techniques has attracted growing interest.
Panoramic images offer a 360$\times$ 180 field of view (FoV), capturing the
entire scene in a single shot. However, panoramic images introduce severe
distortion, making it challenging to render 3D Gaussians into 2D distorted
equirectangular space directly. Converting equirectangular images to cubemap
projections partially alleviates this problem but introduces new challenges,
such as projection distortion and discontinuities across cube-face boundaries.
To address these limitations, we present a novel framework, named TPGS, to
bridge continuous panoramic 3D scene reconstruction with perspective Gaussian
splatting. Firstly, we introduce a Transition Plane between adjacent cube faces
to enable smoother transitions in splatting directions and mitigate
optimization ambiguity in the boundary region. Moreover, an intra-to-inter face
optimization strategy is proposed to enhance local details and restore visual
consistency across cube-face boundaries. Specifically, we optimize 3D Gaussians
within individual cube faces and then fine-tune them in the stitched panoramic
space. Additionally, we introduce a spherical sampling technique to eliminate
visible stitching seams. Extensive experiments on indoor and outdoor,
egocentric, and roaming benchmark datasets demonstrate that our approach
outperforms existing state-of-the-art methods. Code and models will be
available at https://github.com/zhijieshen-bjtu/TPGS.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:42:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shen",
"Zhijie",
""
],
[
"Lin",
"Chunyu",
""
],
[
"Huang",
"Shujuan",
""
],
[
"Nie",
"Lang",
""
],
[
"Liao",
"Kang",
""
],
[
"Zhao",
"Yao",
""
]
] |
2504.09063 | Bryan Yao Siow | Bryan Y. Siow | A Practical Approach to using Supervised Machine Learning Models to
Classify Aviation Safety Occurrences | 9 pages, 3 figures, 3 tables | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper describes a practical approach of using supervised machine
learning (ML) models to assist safety investigators to classify aviation
occurrences into either incident or serious incident categories. Our
implementation currently deployed as a ML web application is trained on a
labelled dataset derived from publicly available aviation investigation
reports. A selection of five supervised learning models (Support Vector
Machine, Logistic Regression, Random Forest Classifier, XGBoost and K-Nearest
Neighbors) were evaluated. This paper showed the best performing ML algorithm
was the Random Forest Classifier with accuracy = 0.77, F1 Score = 0.78 and MCC
= 0.51 (average of 100 sample runs). The study had also explored the effect of
applying Synthetic Minority Over-sampling Technique (SMOTE) to the imbalanced
dataset, and the overall observation ranged from no significant effect to
substantial degradation in performance for some of the models after the SMOTE
adjustment.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:46:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Siow",
"Bryan Y.",
""
]
] |
2504.09064 | Vikas Natesh | Vikas Natesh, H.T. Kung | PQS (Prune, Quantize, and Sort): Low-Bitwidth Accumulation of Dot
Products in Neural Network Computations | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present PQS, which uses three techniques together - Prune, Quantize, and
Sort - to achieve low-bitwidth accumulation of dot products in neural network
computations. In conventional quantized (e.g., 8-bit) dot products, partial
results are accumulated into wide (e.g., 32-bit) accumulators to avoid
overflows when accumulating intermediate partial sums. However, such wide
accumulators increase memory bandwidth usage and reduce energy efficiency. We
show that iterative N:M pruning in floating point followed by quantization to 8
(or fewer) bits, and accumulation of partial products in a sorted order ("small
to large") allows for accurate, compressed models with short dot product
lengths that do not require wide accumulators. We design, analyze, and
implement the PQS algorithm to eliminate accumulation overflows at inference
time for several neural networks. Our method offers a 2.5x reduction in
accumulator bitwidth while achieving model accuracy on par with floating-point
baselines for multiple image classification tasks.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:51:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Natesh",
"Vikas",
""
],
[
"Kung",
"H. T.",
""
]
] |
2504.09065 | Huiju Wang Dr | Huiju Wang, Yuanyuan Gao, Zhengkui Wang, Xiao Yue | Substitutability-Based Graph Node Pricing | 12 pages,7 figures | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the era o fdat commodification,the pricing o fgraph data presents unique
challenges that differ significantly from traditional data markets. This paper
addresses the critical issue of node pricing within graph structures, an area
that has been largely overlooked in existing literature. We introduce a novel
pricing mechanism based on the concept of substitutability, inspired by
economic principles, to better reflect the ntrinsic value of nodes in a graph.
Unlike previous studies that assumed known prices for nodes or subgraphs, our
approach emphasizes the structural significance of nodes by employing a
dominator tree, utilizing the Lengauer-Tarjan algorithm to extract dominance
relationships. This innovative framework allows us to derive a more realistic
pricing strategy that accounts for the unique connectivity and roles of nodes
within their respective networks. Our comparative experiments demonstrate that
the proposed method significantly outperforms existing pricing strategies,
yielding high-quality solutions across various datasets. This research aims to
contribute to the existing literature by addressing an important gap and
providing insights that may assist in the more effective valuation of graph
data, potentially supporting improved decision-making in data-driven
environments.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:51:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Huiju",
""
],
[
"Gao",
"Yuanyuan",
""
],
[
"Wang",
"Zhengkui",
""
],
[
"Yue",
"Xiao",
""
]
] |
2504.09066 | Yifan Yang | Yifan Yang, Lei Zou, Bing Zhou, Daoyang Li, Binbin Lin, Joynal Abedin,
Mingzheng Yang | Hyperlocal disaster damage assessment using bi-temporal street-view
imagery and pre-trained vision models | 27 pages,9 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Street-view images offer unique advantages for disaster damage estimation as
they capture impacts from a visual perspective and provide detailed,
on-the-ground insights. Despite several investigations attempting to analyze
street-view images for damage estimation, they mainly focus on post-disaster
images. The potential of time-series street-view images remains underexplored.
Pre-disaster images provide valuable benchmarks for accurate damage estimations
at building and street levels. These images could aid annotators in objectively
labeling post-disaster impacts, improving the reliability of labeled data sets
for model training, and potentially enhancing the model performance in damage
evaluation. The goal of this study is to estimate hyperlocal, on-the-ground
disaster damages using bi-temporal street-view images and advanced pre-trained
vision models. Street-view images before and after 2024 Hurricane Milton in
Horseshoe Beach, Florida, were collected for experiments. The objectives are:
(1) to assess the performance gains of incorporating pre-disaster street-view
images as a no-damage category in fine-tuning pre-trained models, including
Swin Transformer and ConvNeXt, for damage level classification; (2) to design
and evaluate a dual-channel algorithm that reads pair-wise pre- and
post-disaster street-view images for hyperlocal damage assessment. The results
indicate that incorporating pre-disaster street-view images and employing a
dual-channel processing framework can significantly enhance damage assessment
accuracy. The accuracy improves from 66.14% with the Swin Transformer baseline
to 77.11% with the dual-channel Feature-Fusion ConvNeXt model. This research
enables rapid, operational damage assessments at hyperlocal spatial
resolutions, providing valuable insights to support effective decision-making
in disaster management and resilience planning.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:52:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Yifan",
""
],
[
"Zou",
"Lei",
""
],
[
"Zhou",
"Bing",
""
],
[
"Li",
"Daoyang",
""
],
[
"Lin",
"Binbin",
""
],
[
"Abedin",
"Joynal",
""
],
[
"Yang",
"Mingzheng",
""
]
] |
2504.09067 | Decai Huang | Anghao Li, Zaizheng Wang, Haoyu Shi, Min Sun, Decai Huang | Density-driven segregation of binary granular mixtures in a vertically
vibrating drum the role of filling fraction | 23 pages, 9 figures | null | null | null | cond-mat.soft | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper investigates the influence of filling fraction on segregation
patterns of binary granular mixtures in a vertically vibrating drum through
experiments and simulations. Glass and stainless steel spherical grains, which
differ in mass density, are used to form density-driven segregation. The
results reveal four segregation patterns, including the Brazil-nut-effect (BNE)
segregation, counterclockwise two-eye-like segregation, dumpling-like
segregation, and clockwise two-eye-like segregation. The theoretical analysis
demonstrates that grains predominantly exhibit counterclockwise convection at
low filling fractions, while clockwise convection dominates at high filling
fractions. The competition between buoyancy and convection forces determines
the final stable segregation pattern. These findings provide valuable insights
into controlling segregation in granular systems, which is crucial for
optimizing industrial processes in fields such as pharmaceuticals and chemical
engineering.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:52:48 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Anghao",
""
],
[
"Wang",
"Zaizheng",
""
],
[
"Shi",
"Haoyu",
""
],
[
"Sun",
"Min",
""
],
[
"Huang",
"Decai",
""
]
] |
2504.09068 | Praveen Tharuka Wijewardana Hettige | Praveen T. W. Hettige, Benjamin W. Ong | An Incremental Non-Linear Manifold Approximation Method | 19 pages | null | null | null | stat.ML cs.GR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Analyzing high-dimensional data presents challenges due to the "curse of
dimensionality'', making computations intensive. Dimension reduction
techniques, categorized as linear or non-linear, simplify such data. Non-linear
methods are particularly essential for efficiently visualizing and processing
complex data structures in interactive and graphical applications. This
research develops an incremental non-linear dimension reduction method using
the Geometric Multi-Resolution Analysis (GMRA) framework for streaming data.
The proposed method enables real-time data analysis and visualization by
incrementally updating the cluster map, PCA basis vectors, and wavelet
coefficients. Numerical experiments show that the incremental GMRA accurately
represents non-linear manifolds even with small initial samples and aligns
closely with batch GMRA, demonstrating efficient updates and maintaining the
multiscale structure. The findings highlight the potential of Incremental GMRA
for real-time visualization and interactive graphics applications that require
adaptive high-dimensional data representations.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 03:54:05 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hettige",
"Praveen T. W.",
""
],
[
"Ong",
"Benjamin W.",
""
]
] |
2504.09069 | Zhuoran Zheng | Shuning Sun, Yu Zhang, Chen Wu, Dianjie Lu, Dianjie Lu, Guijuan Zhan,
Yang Weng, Zhuoran Zheng | UniFlowRestore: A General Video Restoration Framework via Flow Matching
and Prompt Guidance | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Video imaging is often affected by complex degradations such as blur, noise,
and compression artifacts. Traditional restoration methods follow a
"single-task single-model" paradigm, resulting in poor generalization and high
computational cost, limiting their applicability in real-world scenarios with
diverse degradation types. We propose UniFlowRestore, a general video
restoration framework that models restoration as a time-continuous evolution
under a prompt-guided and physics-informed vector field. A physics-aware
backbone PhysicsUNet encodes degradation priors as potential energy, while
PromptGenerator produces task-relevant prompts as momentum. These components
define a Hamiltonian system whose vector field integrates inertial dynamics,
decaying physical gradients, and prompt-based guidance. The system is optimized
via a fixed-step ODE solver to achieve efficient and unified restoration across
tasks. Experiments show that UniFlowRestore delivers stateof-the-art
performance with strong generalization and efficiency. Quantitative results
demonstrate that UniFlowRestore achieves state-of-the-art performance,
attaining the highest PSNR (33.89 dB) and SSIM (0.97) on the video denoising
task, while maintaining top or second-best scores across all evaluated tasks.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:06:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sun",
"Shuning",
""
],
[
"Zhang",
"Yu",
""
],
[
"Wu",
"Chen",
""
],
[
"Lu",
"Dianjie",
""
],
[
"Lu",
"Dianjie",
""
],
[
"Zhan",
"Guijuan",
""
],
[
"Weng",
"Yang",
""
],
[
"Zheng",
"Zhuoran",
""
]
] |
2504.09070 | Santabrata Das | Seshadri Majumder (IITG), Santabrata Das (IITG) and Anuj Nandi (URSC) | Long-term evolution of BH-ULX candidates: An `unusual' $L_{\rm
disc}$-$T_{\rm col}$ correlation associated with spectral states | 14 pages, 6 figures, 3 tabes, Accepted for publication in MNRAS | null | null | null | astro-ph.HE astro-ph.GA astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present the long-term spectral evolution of eight black hole
ultra-luminous X-ray sources (BH-ULXs), namely NGC1313 X-1, NGC5408 X-1,
NGC6946 X-1, IC342 X-1, NGC55 ULX1, NGC4395 ULX1, NGC5204 X-1 and NGC4190 ULX1
using {\it XMM-Newton} monitoring data spanning over a decade or more. An
in-depth spectral modeling with thermal Comptonization ({\it nthComp}) and
standard disc ({\it diskbb}) components reveals NGC5204 X-1, IC342 X-1, NGC4190
ULX1 and NGC1313 X-1 exhibiting harder spectral characteristics with dominant
effect of Comptonization ($F_{nth}>F_{disc}$, $\Gamma_{nth}\lesssim2$).
However, NGC6946 X-1 and NGC55 ULX1 remain in a disc-dominated state
($F_{disc}\sim2F_{nth}$, $\Gamma_{nth}\gtrsim2$), while NGC5408 X-1 shows
intermediate spectral characteristics. The spectral analyses indicate an
anti-correlation between disc luminosity ($L_{disc}$) and temperature
($T_{col}$) for all sources except NGC5204 X-1. These anti-correlations follow
a relation $L_{disc}\propto T_{col}^{\alpha}$ with steeper exponents of
$\alpha=-6.01\pm0.25$ (NGC55 ULX1), $-8.93\pm0.11$ (NGC6946 X-1), and
$-10.31\pm0.10$ (NGC5408 X-1) for sources with softer or intermediate spectral
characteristics. For harder sources, NGC1313 X-1 and IC342 X-1, the combined
results provide $\alpha=-3.58\pm0.04$. However, for NGC5204 X-1, a positive
correlation is found, yielding $\alpha=1.4\pm0.1$, suggesting that the emission
mechanism is associated with the transition from the `standard disc' to the
`slim disc' scenario. These findings suggest that the observed
$L_{disc}-T_{col}$ correlations, along with the overall spectro-temporal
properties of BH-ULXs, seems to be governed by disc-corona-wind driven
accretion processes at various inclinations. Finally, we report a QPO-like
feature ($\sim20$ mHz) with $rms\%\sim6.6$, Q-factor $\sim6.7$ and significant
$2.8\sigma$ in NGC55 ULX1.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:09:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Majumder",
"Seshadri",
"",
"IITG"
],
[
"Das",
"Santabrata",
"",
"IITG"
],
[
"Nandi",
"Anuj",
"",
"URSC"
]
] |
2504.09071 | Matt Grenander | Matt Grenander, Siddharth Varia, Paula Czarnowska, Yogarshi Vyas,
Kishaloy Halder, Bonan Min | Exploration of Plan-Guided Summarization for Narrative Texts: the Case
of Small Language Models | Accepted to the 7th Workshop on Narrative Understanding (WNU),
co-located with NAACL 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Plan-guided summarization attempts to reduce hallucinations in small language
models (SLMs) by grounding generated summaries to the source text, typically by
targeting fine-grained details such as dates or named entities. In this work,
we investigate whether plan-based approaches in SLMs improve summarization in
long document, narrative tasks. Narrative texts' length and complexity often
mean they are difficult to summarize faithfully. We analyze existing
plan-guided solutions targeting fine-grained details, and also propose our own
higher-level, narrative-based plan formulation. Our results show that neither
approach significantly improves on a baseline without planning in either
summary quality or faithfulness. Human evaluation reveals that while
plan-guided approaches are often well grounded to their plan, plans are equally
likely to contain hallucinations compared to summaries. As a result, the
plan-guided summaries are just as unfaithful as those from models without
planning. Our work serves as a cautionary tale to plan-guided approaches to
summarization, especially for long, complex domains such as narrative texts.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:11:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Grenander",
"Matt",
""
],
[
"Varia",
"Siddharth",
""
],
[
"Czarnowska",
"Paula",
""
],
[
"Vyas",
"Yogarshi",
""
],
[
"Halder",
"Kishaloy",
""
],
[
"Min",
"Bonan",
""
]
] |
2504.09072 | Vikas Natesh | Vikas Natesh, H.T. Kung, David Kong | MGS: Markov Greedy Sums for Accurate Low-Bitwidth Floating-Point
Accumulation | null | null | null | null | cs.AR cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We offer a novel approach, MGS (Markov Greedy Sums), to improve the accuracy
of low-bitwidth floating-point dot products in neural network computations. In
conventional 32-bit floating-point summation, adding values with different
exponents may lead to loss of precision in the mantissa of the smaller term,
which is right-shifted to align with the larger term's exponent. Such shifting
(a.k.a. 'swamping') is a significant source of numerical errors in accumulation
when implementing low-bitwidth dot products (e.g., 8-bit floating point) as the
mantissa has a small number of bits. We avoid most swamping errors by arranging
the terms in dot product summation based on their exponents and summing the
mantissas without overflowing the low-bitwidth accumulator. We design, analyze,
and implement the algorithm to minimize 8-bit floating point error at inference
time for several neural networks. In contrast to traditional sequential
summation, our method has significantly lowered numerical errors, achieving
classification accuracy on par with high-precision floating-point baselines for
multiple image classification tasks. Our dMAC hardware units can reduce power
consumption by up to 34.1\% relative to conventional MAC units.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:19:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Natesh",
"Vikas",
""
],
[
"Kung",
"H. T.",
""
],
[
"Kong",
"David",
""
]
] |
2504.09073 | Akanksha Mehndiratta | Akanksha Mehndiratta, Krishna Asawa | A Multi-view Discourse Framework for Integrating Semantic and Syntactic
Features in Dialog Agents | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Multiturn dialogue models aim to generate human-like responses by leveraging
conversational context, consisting of utterances from previous exchanges.
Existing methods often neglect the interactions between these utterances or
treat all of them as equally significant. This paper introduces a
discourse-aware framework for response selection in retrieval-based dialogue
systems. The proposed model first encodes each utterance and response with
contextual, positional, and syntactic features using Multi-view Canonical
Correlation Analysis (MCCA). It then learns discourse tokens that capture
relationships between an utterance and its surrounding turns in a shared
subspace via Canonical Correlation Analysis (CCA). This two-step approach
effectively integrates semantic and syntactic features to build discourse-level
understanding. Experiments on the Ubuntu Dialogue Corpus demonstrate that our
model achieves significant improvements in automatic evaluation metrics,
highlighting its effectiveness in response selection.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:22:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mehndiratta",
"Akanksha",
""
],
[
"Asawa",
"Krishna",
""
]
] |
2504.09074 | Dhruv Kulkarni | Dhruv Kulkarni, Bharat Bhammar, Henil Thaker, Pranav Dhobi, R.P.
Gohil, Sai Manoj Pudukotai Dinkarrao | A Case for Kolmogorov-Arnold Networks in Prefetching: Towards
Low-Latency, Generalizable ML-Based Prefetchers | null | null | null | null | cs.AR | http://creativecommons.org/licenses/by-sa/4.0/ | The memory wall problem arises due to the disparity between fast processors
and slower memory, causing significant delays in data access, even more so on
edge devices. Data prefetching is a key strategy to address this, with
traditional methods evolving to incorporate Machine Learning (ML) for improved
accuracy. Modern prefetchers must balance high accuracy with low latency to
further practicality. We explore the applicability of utilizing
Kolmogorov-Arnold Networks (KAN) with learnable activation functions,a
prefetcher we implemented called KANBoost, to further this aim. KANs are a
novel, state-of-the-art model that work on breaking down continuous, bounded
multi-variate functions into functions of their constituent variables, and use
these constitutent functions as activations on each individual neuron. KANBoost
predicts the next memory access by modeling deltas between consecutive
addresses, offering a balance of accuracy and efficiency to mitigate the memory
wall problem with minimal overhead, instead of relying on address-correlation
prefetching. Initial results indicate that KAN-based prefetching reduces
inference latency (18X lower than state-of-the-art ML prefetchers) while
achieving moderate IPC improvements (2.5\% over no-prefetching). While KANs
still face challenges in capturing long-term dependencies, we propose that
future research should explore hybrid models that combine KAN efficiency with
stronger sequence modeling techniques, paving the way for practical ML-based
prefetching in edge devices and beyond.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:29:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kulkarni",
"Dhruv",
""
],
[
"Bhammar",
"Bharat",
""
],
[
"Thaker",
"Henil",
""
],
[
"Dhobi",
"Pranav",
""
],
[
"Gohil",
"R. P.",
""
],
[
"Dinkarrao",
"Sai Manoj Pudukotai",
""
]
] |
2504.09075 | Yinzhi Wang | Sasmita Mohapatra, Weiming Yang, Zhengtang Yang, Chenxiao Wang, Jinxin
Ma, Gary L. Pavlis, Yinzhi Wang | Parallel Seismic Data Processing Performance with Cloud-based Storage | null | null | null | null | physics.geo-ph cs.DC | http://creativecommons.org/licenses/by/4.0/ | This article introduces a general processing framework to effectively utilize
waveform data stored on modern cloud platforms. The focus is hybrid processing
schemes where a local system drives processing. We show that downloading files
and doing all processing locally is problematic even when the local system is a
high-performance compute cluster. Benchmark tests with parallel processing show
that approach always creates a bottleneck as the volume of data being handled
increases with more processes pulling data. We find a hybrid model where
processing to reduce the volume of data transferred from the cloud servers to
the local system can dramatically improve processing time. Tests implemented
with Massively Parallel Analysis System for Seismology (MsPASS) utilizing
Amazon Web Service's Lamba service yield throughput comparable to processing
day files on a local HPC file system. Given the ongoing migration of seismology
data to cloud storage, our results show doing some or all processing on the
cloud will be essential for any processing involving large volumes of data.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:30:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mohapatra",
"Sasmita",
""
],
[
"Yang",
"Weiming",
""
],
[
"Yang",
"Zhengtang",
""
],
[
"Wang",
"Chenxiao",
""
],
[
"Ma",
"Jinxin",
""
],
[
"Pavlis",
"Gary L.",
""
],
[
"Wang",
"Yinzhi",
""
]
] |
2504.09076 | Mk Bashar | Mk Bashar, Ocean Monjur, Samia Islam, Mohammad Galib Shams, Niamul
Quader | Exploring Synergistic Ensemble Learning: Uniting CNNs, MLP-Mixers, and
Vision Transformers to Enhance Image Classification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, Convolutional Neural Networks (CNNs), MLP-mixers, and Vision
Transformers have risen to prominence as leading neural architectures in image
classification. Prior research has underscored the distinct advantages of each
architecture, and there is growing evidence that combining modules from
different architectures can boost performance. In this study, we build upon and
improve previous work exploring the complementarity between different
architectures. Instead of heuristically merging modules from various
architectures through trial and error, we preserve the integrity of each
architecture and combine them using ensemble techniques. By maintaining the
distinctiveness of each architecture, we aim to explore their inherent
complementarity more deeply and with implicit isolation. This approach provides
a more systematic understanding of their individual strengths.
In addition to uncovering insights into architectural complementarity, we
showcase the effectiveness of even basic ensemble methods that combine models
from diverse architectures. These methods outperform ensembles comprised of
similar architectures. Our straightforward ensemble framework serves as a
foundational strategy for blending complementary architectures, offering a
solid starting point for further investigations into the unique strengths and
synergies among different architectures and their ensembles in image
classification. A direct outcome of this work is the creation of an ensemble of
classification networks that surpasses the accuracy of the previous
state-of-the-art single classification network on ImageNet, setting a new
benchmark, all while requiring less overall latency.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:32:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bashar",
"Mk",
""
],
[
"Monjur",
"Ocean",
""
],
[
"Islam",
"Samia",
""
],
[
"Shams",
"Mohammad Galib",
""
],
[
"Quader",
"Niamul",
""
]
] |
2504.09077 | BingYu Nan | Bingyu Nan, Feng Liu, Xuezhong Qian and Wei Song | A Visual Self-attention Mechanism Facial Expression Recognition Network
beyond Convnext | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Facial expression recognition is an important research direction in the field
of artificial intelligence. Although new breakthroughs have been made in recent
years, the uneven distribution of datasets and the similarity between different
categories of facial expressions, as well as the differences within the same
category among different subjects, remain challenges. This paper proposes a
visual facial expression signal feature processing network based on truncated
ConvNeXt approach(Conv-cut), to improve the accuracy of FER under challenging
conditions. The network uses a truncated ConvNeXt-Base as the feature
extractor, and then we designed a Detail Extraction Block to extract detailed
features, and introduced a Self-Attention mechanism to enable the network to
learn the extracted features more effectively. To evaluate the proposed
Conv-cut approach, we conducted experiments on the RAF-DB and FERPlus datasets,
and the results show that our model has achieved state-of-the-art performance.
Our code could be accessed at Github.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:35:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Nan",
"Bingyu",
""
],
[
"Liu",
"Feng",
""
],
[
"Qian",
"Xuezhong",
""
],
[
"Song",
"Wei",
""
]
] |
2504.09078 | D. Bhanu Prakash | D Bhanu Prakash, D K K Vamsi | Role of Intra-specific Competition and Additional Food on Prey-Predator
Systems exhibiting Holling Type-IV Functional Response | null | null | null | null | math.DS math.OC | http://creativecommons.org/licenses/by/4.0/ | In recent years, the study on the impact of competition on additional food
provided prey-predator systems have gained significant attention from
researchers in the field of mathematical biology. In this study, we consider an
additional food provided prey-predator model exhibiting Holling type-IV
functional response and the intra-specific competition among predators. We
prove the existence and uniqueness of global positive solutions for the
proposed model. We study the existence and stability of equilibrium points and
further explore the codimension-$1$ and $2$ bifurcations with respect to the
additional food and competition. We further study the global dynamics of the
system and discuss the consequences of providing additional food. Later, we do
the time-optimal control studies with respect to the quality and quantity of
additional food as control variables by transforming the independent variable
in the control system. Making use of the Pontraygin maximum principle, we
characterize the optimal quality of additional food and optimal quantity of
additional food. We show that the findings of these dynamics and control
studies have the potential to be applied to a variety of problems in pest
management.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:36:54 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Prakash",
"D Bhanu",
""
],
[
"Vamsi",
"D K K",
""
]
] |
2504.09079 | Saail Narvekar | Saail Narvekar, Soofiyan Atar, Vishal Gupta, Lohit Penubaku, Kavi Arya | agriFrame: Agricultural framework to remotely control a rover inside a
greenhouse environment | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | The growing demand for innovation in agriculture is essential for food
security worldwide and more implicit in developing countries. With growing
demand comes a reduction in rapid development time. Data collection and
analysis are essential in agriculture. However, considering a given crop, its
cycle comes once a year, and researchers must wait a few months before
collecting more data for the given crop. To overcome this hurdle, researchers
are venturing into digital twins for agriculture. Toward this effort, we
present an agricultural framework(agriFrame). Here, we introduce a simulated
greenhouse environment for testing and controlling a robot and remotely
controlling/implementing the algorithms in the real-world greenhouse setup.
This work showcases the importance/interdependence of network setup, remotely
controllable rover, and messaging protocol. The sophisticated yet simple-to-use
agriFrame has been optimized for the simulator on minimal laptop/desktop
specifications.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:44:25 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Narvekar",
"Saail",
""
],
[
"Atar",
"Soofiyan",
""
],
[
"Gupta",
"Vishal",
""
],
[
"Penubaku",
"Lohit",
""
],
[
"Arya",
"Kavi",
""
]
] |
2504.09080 | Tomoki Kurikawa | Tomoki Kurikawa, Kunihiko Kaneko | Stability Control of Metastable States as a Unified Mechanism for
Flexible Temporal Modulation in Cognitive Processing | null | null | null | null | q-bio.NC cond-mat.dis-nn nlin.AO physics.bio-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Flexible modulation of temporal dynamics in neural sequences underlies many
cognitive processes. For instance, we can adaptively change the speed of motor
sequences and speech. While such flexibility is influenced by various factors
such as attention and context, the common neural mechanisms responsible for
this modulation remain poorly understood. We developed a biologically plausible
neural network model that incorporates neurons with multiple timescales and
Hebbian learning rules. This model is capable of generating simple sequential
patterns as well as performing delayed match-to-sample (DMS) tasks that require
the retention of stimulus identity. Fast neural dynamics establish metastable
states, while slow neural dynamics maintain task-relevant information and
modulate the stability of these states to enable temporal processing. We
systematically analyzed how factors such as neuronal gain, external input
strength (contextual cues), and task difficulty influence the temporal
properties of neural activity sequences - specifically, dwell time within
patterns and transition times between successive patterns. We found that these
factors flexibly modulate the stability of metastable states. Our findings
provide a unified mechanism for understanding various forms of temporal
modulation and suggest a novel computational role for neural timescale
diversity in dynamically adapting cognitive performance to changing
environmental demands.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:44:34 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kurikawa",
"Tomoki",
""
],
[
"Kaneko",
"Kunihiko",
""
]
] |
2504.09082 | Roberto da Silva | Roberto da Silva, Eliseu Venites Filho, Henrique Almeida Fernandes,
Paulo F. Gomes | Revisiting the Contact Model with Diffusion Beyond the Conventional
Methods | 10 pages, 5 figures | null | null | null | cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The contact process is a non-equilibrium Hamiltonian model that, even in one
dimension, lacks an exact solution and has been extensively studied via Monte
Carlo simulations, both in steady-state and time-dependent scenarios. Although
the effects of particle mobility/diffusion on criticality have been
preliminarily investigated, they remain incompletely understood. In this work,
we examine how the critical rate of the model varies with the probability of
particle mobility. By analyzing different stochastic evolutions of the system,
we employ two modern approaches: 1) Random Matrix Theory (RMT): By building on
the success of RMT, particularly Wishart-like matrices, in studying statistical
physics of systems with up-down symmetry via magnetization dynamics [R. da
Silva, IJMPC 2022], we demonstrate its applicability to models with an
absorbing state. 2) Optimized Temporal Power Laws: By using short-time
dynamics, we optimize power laws derived from ensemble-averaged evolutions of
the system. Both methods consistently reveal that the critical rate decays with
mobility according to a simple Belehradek function. Additionally, a
straightforward mean-field analysis supports the decay of the critical
parameter with mobility, although it predicts a simpler linear dependence.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:46:39 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"da Silva",
"Roberto",
""
],
[
"Filho",
"Eliseu Venites",
""
],
[
"Fernandes",
"Henrique Almeida",
""
],
[
"Gomes",
"Paulo F.",
""
]
] |
2504.09083 | Qipei Mei | Muhammad Adil, Gaang Lee, Vicente A. Gonzalez, Qipei Mei | Using Vision Language Models for Safety Hazard Identification in
Construction | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Safety hazard identification and prevention are the key elements of proactive
safety management. Previous research has extensively explored the applications
of computer vision to automatically identify hazards from image clips collected
from construction sites. However, these methods struggle to identify
context-specific hazards, as they focus on detecting predefined individual
entities without understanding their spatial relationships and interactions.
Furthermore, their limited adaptability to varying construction site guidelines
and conditions hinders their generalization across different projects. These
limitations reduce their ability to assess hazards in complex construction
environments and adaptability to unseen risks, leading to potential safety
gaps. To address these challenges, we proposed and experimentally validated a
Vision Language Model (VLM)-based framework for the identification of
construction hazards. The framework incorporates a prompt engineering module
that structures safety guidelines into contextual queries, allowing VLM to
process visual information and generate hazard assessments aligned with the
regulation guide. Within this framework, we evaluated state-of-the-art VLMs,
including GPT-4o, Gemini, Llama 3.2, and InternVL2, using a custom dataset of
1100 construction site images. Experimental results show that GPT-4o and Gemini
1.5 Pro outperformed alternatives and displayed promising BERTScore of 0.906
and 0.888 respectively, highlighting their ability to identify both general and
context-specific hazards. However, processing times remain a significant
challenge, impacting real-time feasibility. These findings offer insights into
the practical deployment of VLMs for construction site hazard detection,
thereby contributing to the enhancement of proactive safety management.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:11:23 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Adil",
"Muhammad",
""
],
[
"Lee",
"Gaang",
""
],
[
"Gonzalez",
"Vicente A.",
""
],
[
"Mei",
"Qipei",
""
]
] |
2504.09084 | Chengyang Wu | Danijela Damjanovi\'c, Amie Wilkinson, Chengyang Wu, Disheng Xu | The symmetries of affine $K$-systems and a program for centralizer
rigidity | null | null | null | null | math.DS | http://creativecommons.org/licenses/by/4.0/ | Let Aff(X) be the group of affine diffeomorphisms of a closed homogeneous
manifold X=G/B admitting a G-invariant Lebesgue-Haar probability measure $\mu$.
For $f_0\in$ Aff(X), let $Z^\infty(f_0)$ be the group of $C^\infty$
diffeomorphisms of X commuting with $f_0$. This paper addresses the question:
for which $f_0\in$ Aff(X) is $Z^\infty(f_0)$ a Lie subgroup of
$Diff^\infty(X)$? Among our main results are the following.
(1) If $f_0\in$ Aff(X) is weakly mixing with respect to $\mu$, then
$Z^\infty(f_0)<$ Aff(X), and hence is a Lie group.
(2) If $f_0\in$ Aff(X) is ergodic with respect to $\mu$, then $Z^\infty(f_0)$
is a (necessarily $C^0$ closed) Lie subgroup of $Diff^\infty(X)$ (although not
necessarily a subgroup of Aff(X)).
(3) If $f_0\in$ Aff(X) fails to be a K-system with respect to $\mu$, then
there exists $f\in$ Aff(X) arbitrarily close to $f_0$ such that $Z^\infty(f)$
is not a Lie group, containing as a continuously embedded subgroup either the
abelian group $C^\infty_c((0,1))$ (under addition) or the simple group
$Diff^\infty_c((0,1))$ (under composition).
(4) Considering perturbations of $f_0$ by left translations, we conclude that
$f_0$ is stably ergodic if and only if the condition $Z^\infty<$ Aff(X) holds
in a neighborhood of $f_0$ in Aff(X). (Note that by BS97, Dani77, $f_0\in$
Aff(X) is stably ergodic in Aff(X) if and only if $f_0$ is a K-system.)
The affine K-systems are precisely those that are partially hyperbolic and
essentially accessible, belonging to a class of diffeomorphisms whose dynamics
have been extensively studied. In addition, the properties of partial
hyperbolicity and accessibility are stable under $C^1$-small perturbation, and
in some contexts, essential accessibility has been shown to be stable under
smooth perturbation. Considering the smooth perturbations of affine K-systems,
we outline a full program for (local) centralizer rigidity.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:11:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Damjanović",
"Danijela",
""
],
[
"Wilkinson",
"Amie",
""
],
[
"Wu",
"Chengyang",
""
],
[
"Xu",
"Disheng",
""
]
] |
2504.09085 | Marek Herde | Marek Herde, Lukas L\"uhrs, Denis Huseljic, Bernhard Sick | crowd-hpo: Realistic Hyperparameter Optimization and Benchmarking for
Learning from Crowds with Noisy Labels | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Crowdworking is a cost-efficient solution to acquire class labels. Since
these labels are subject to noise, various approaches to learning from crowds
have been proposed. Typically, these approaches are evaluated with default
hyperparameters, resulting in suboptimal performance, or with hyperparameters
tuned using a validation set with ground truth class labels, representing an
often unrealistic scenario. Moreover, both experimental setups can produce
different rankings of approaches, complicating comparisons between studies.
Therefore, we introduce crowd-hpo as a realistic benchmark and experimentation
protocol including hyperparameter optimization under noisy crowd-labeled data.
At its core, crowd-hpo investigates model selection criteria to identify
well-performing hyperparameter configurations only with access to noisy
crowd-labeled validation data. Extensive experimental evaluations with neural
networks show that these criteria are effective for optimizing hyperparameters
in learning from crowds approaches. Accordingly, incorporating such criteria
into experimentation protocols is essential for enabling more realistic and
fair benchmarking.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:36:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Herde",
"Marek",
""
],
[
"Lührs",
"Lukas",
""
],
[
"Huseljic",
"Denis",
""
],
[
"Sick",
"Bernhard",
""
]
] |
2504.09086 | Yunfei Long | Yunfei Long, Abhinav Kumar, Xiaoming Liu, Daniel Morris | RICCARDO: Radar Hit Prediction and Convolution for Camera-Radar 3D
Object Detection | CVPR 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radar hits reflect from points on both the boundary and internal to object
outlines. This results in a complex distribution of radar hits that depends on
factors including object category, size, and orientation. Current radar-camera
fusion methods implicitly account for this with a black-box neural network. In
this paper, we explicitly utilize a radar hit distribution model to assist
fusion. First, we build a model to predict radar hit distributions conditioned
on object properties obtained from a monocular detector. Second, we use the
predicted distribution as a kernel to match actual measured radar points in the
neighborhood of the monocular detections, generating matching scores at nearby
positions. Finally, a fusion stage combines context with the kernel detector to
refine the matching scores. Our method achieves the state-of-the-art
radar-camera detection performance on nuScenes. Our source code is available at
https://github.com/longyunf/riccardo.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:37:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Long",
"Yunfei",
""
],
[
"Kumar",
"Abhinav",
""
],
[
"Liu",
"Xiaoming",
""
],
[
"Morris",
"Daniel",
""
]
] |
2504.09087 | Dane Spaeth | Dane Spaeth, Sabine Reffert, Trifon Trifonov, Adrian Kaminski, Simon
Albrecht, Frank Grundahl, Mads Fredslund Andersen, Andreas Quirrenbach, Pere
L. Pall\'e | Precise radial velocities of giant stars -- XVII. Distinguishing planets
from intrinsically induced radial velocity signals in evolved stars | 23 pages, 11 figures, 4 appendices, Accepted for publication in
Astronomy & Astrophysics, Abstract abridged for arXiv submission | null | 10.1051/0004-6361/202453500 | null | astro-ph.EP astro-ph.IM astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | From a long-term Doppler monitoring campaign of 373 giant stars, we have
identified ten giants with periodic radial velocity variations that are
challenging to associate with planets. Similar cases in the literature are
attributed to poorly understood intrinsic processes. Our goal is to confirm or
refute the presence of planets around these ten evolved stars. Additionally, we
evaluate the reliability and sensitivity of planet-confirmation metrics when
applied to giant stars and present cases of intrinsically induced radial
velocity variations, aiming to enhance the physical understanding of the
phenomenon. We combined 25 years of radial velocity data from the
Hamilton/Lick, SONG, and CARMENES spectrographs. To assess consistency with
Keplerian models, we examined the residuals and tracked changes in statistical
significance as new data were incorporated. Additionally, we compared radial
velocity amplitudes across optical and infrared wavelengths, searched for
periodic variations of activity indicators, and examined their correlations
with radial velocities. Seven of the ten giants exhibit intrinsically induced
radial velocity variations. The strongest arguments against planets orbiting
the giants are guided by long-term radial velocity monitoring that detects
changing periodicity on long timescales or detects systematics close to the
original period in the radial velocity residuals. While activity indicators
offer some support, their signals are generally weak. Comparing optical and
infrared radial velocity amplitudes also proves insufficient for confirming or
refuting planets. We find HIP 64823 remains a promising candidate for hosting a
giant exoplanet with orbital period $P\sim$ 7.75 yr. For two stars, the
evidence remains inconclusive. Long-term radial velocity monitoring is
essential for distinguishing planetary companions from intrinsic variations in
evolved stars.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:48:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Spaeth",
"Dane",
""
],
[
"Reffert",
"Sabine",
""
],
[
"Trifonov",
"Trifon",
""
],
[
"Kaminski",
"Adrian",
""
],
[
"Albrecht",
"Simon",
""
],
[
"Grundahl",
"Frank",
""
],
[
"Andersen",
"Mads Fredslund",
""
],
[
"Quirrenbach",
"Andreas",
""
],
[
"Pallé",
"Pere L.",
""
]
] |
2504.09088 | Yonghao Huang | Yonghao Huang, Leiting Chen, Chuan Zhou | Multi-Modal Brain Tumor Segmentation via 3D Multi-Scale Self-attention
and Cross-attention | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the success of CNN-based and Transformer-based models in various
computer vision tasks, recent works study the applicability of CNN-Transformer
hybrid architecture models in 3D multi-modality medical segmentation tasks.
Introducing Transformer brings long-range dependent information modeling
ability in 3D medical images to hybrid models via the self-attention mechanism.
However, these models usually employ fixed receptive fields of 3D volumetric
features within each self-attention layer, ignoring the multi-scale volumetric
lesion features. To address this issue, we propose a CNN-Transformer hybrid 3D
medical image segmentation model, named TMA-TransBTS, based on an
encoder-decoder structure. TMA-TransBTS realizes simultaneous extraction of
multi-scale 3D features and modeling of long-distance dependencies by
multi-scale division and aggregation of 3D tokens in a self-attention layer.
Furthermore, TMA-TransBTS proposes a 3D multi-scale cross-attention module to
establish a link between the encoder and the decoder for extracting rich volume
representations by exploiting the mutual attention mechanism of cross-attention
and multi-scale aggregation of 3D tokens. Extensive experimental results on
three public 3D medical segmentation datasets show that TMA-TransBTS achieves
higher averaged segmentation results than previous state-of-the-art CNN-based
3D methods and CNN-Transform hybrid 3D methods for the segmentation of 3D
multi-modality brain tumors.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 05:53:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Huang",
"Yonghao",
""
],
[
"Chen",
"Leiting",
""
],
[
"Zhou",
"Chuan",
""
]
] |
2504.09090 | Peixuan Lei | Yilin Wang, Peixuan Lei, Xuyang Wang, Liangliang Jiang, Liming Xuan,
Wei Cheng, Honghua Zhao, Yuanxiang Li | Leveraging Large Self-Supervised Time-Series Models for Transferable
Diagnosis in Cross-Aircraft Type Bleed Air System | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bleed Air System (BAS) is critical for maintaining flight safety and
operational efficiency, supporting functions such as cabin pressurization, air
conditioning, and engine anti-icing. However, BAS malfunctions, including
overpressure, low pressure, and overheating, pose significant risks such as
cabin depressurization, equipment failure, or engine damage. Current diagnostic
approaches face notable limitations when applied across different aircraft
types, particularly for newer models that lack sufficient operational data. To
address these challenges, this paper presents a self-supervised learning-based
foundation model that enables the transfer of diagnostic knowledge from mature
aircraft (e.g., A320, A330) to newer ones (e.g., C919). Leveraging
self-supervised pretraining, the model learns universal feature representations
from flight signals without requiring labeled data, making it effective in
data-scarce scenarios. This model enhances both anomaly detection and baseline
signal prediction, thereby improving system reliability. The paper introduces a
cross-model dataset, a self-supervised learning framework for BAS diagnostics,
and a novel Joint Baseline and Anomaly Detection Loss Function tailored to
real-world flight data. These innovations facilitate efficient transfer of
diagnostic knowledge across aircraft types, ensuring robust support for early
operational stages of new models. Additionally, the paper explores the
relationship between model capacity and transferability, providing a foundation
for future research on large-scale flight signal models.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:06:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Yilin",
""
],
[
"Lei",
"Peixuan",
""
],
[
"Wang",
"Xuyang",
""
],
[
"Jiang",
"Liangliang",
""
],
[
"Xuan",
"Liming",
""
],
[
"Cheng",
"Wei",
""
],
[
"Zhao",
"Honghua",
""
],
[
"Li",
"Yuanxiang",
""
]
] |
2504.09091 | Ryan Lau | R. Y. Lau, D. J. Strozzi, M. Sherlock, M. Weis, A. S. Joglekar, W. A.
Farmer, Y. Shi, and J. R. Cary | Nonlocal effects on Thermal Transport in MagLIF-Relevant Gaspipes on NIF | 12 pages, 10 figures | null | null | null | physics.plasm-ph | http://creativecommons.org/licenses/by/4.0/ | We present simulations of heat flow relevant to gaspipe experiments on the
National Ignition Facility (NIF) to investigate kinetic effects on transport
phenomena. D2 and neopentane (C5H12) filled targets are used to study the laser
preheat stage of a MagLIF scheme where anaxial magnetic field is sometimes
applied to the target. Simulations were done with the radiation-MHD code HYDRA
with a collision-dominated fluid model and the Schurtz nonlocal electron
thermal conduction model. Using the Schurtz model to evolve the electron
temperature increased the heat front propagation of neopentane gas targets
compared to a local model by limiting radial heat flow. This increases electron
temperature near the axis, which decreases laser absorption. We find the effect
of heat flow models on temperature profiles and laser propagation is modest.
Beyond the Schurtz model, we utilize HYDRA to initialize plasma conditions for
the Vlasov Fokker-Planck K2 code. We run K2 until a quasi-steady state is
reached and examine the impact of kinetic effects on heat transport. Although
axial heat flow is well predicted by fluid models, the fluid model consistently
over predicts radial heat flow up to 150% in regions with the largest
temperature gradient of D2 filled gaspipes. On the other hand, the Schurtz
nonlocal electron conduction model is found to be adequate for capturing
kinetic heat flow in gaspipes.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:10:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lau",
"R. Y.",
""
],
[
"Strozzi",
"D. J.",
""
],
[
"Sherlock",
"M.",
""
],
[
"Weis",
"M.",
""
],
[
"Joglekar",
"A. S.",
""
],
[
"Farmer",
"W. A.",
""
],
[
"Shi",
"Y.",
""
],
[
"Cary",
"J. R.",
""
]
] |
2504.09092 | Zipeng Wang | Zipeng Wang | Fractional integrals associated with Zygmund dilations | null | null | null | null | math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a family of fractional integral operators defined in $\mathbb{R}^3$
whose kernels are distributions associated with Zygmund dilations: $(x_1, x_2,
x_3) \rightarrow (\delta_1 x_1, \delta_2 x_2, \delta_1\delta_2 x_3)$ for
$\delta_1,\delta_2>0$ having singularity on every coordinate subspace. As a
result, we obtain a Hardy-Littlewood-Sobolev type inequality.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:13:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Zipeng",
""
]
] |
2504.09093 | Shigeru Yamagami | Shigeru Yamagami | Boundary measures of holomorphic functions on the imaginary domain | 24 pages, 3 figures | null | null | null | math.CV math.FA | http://creativecommons.org/licenses/by/4.0/ | In connection with the Herglotz-Nevanlinna integral representation of
so-called Pick functions,
we introduce the notion of boundary measure of holomorphic functions on the
imaginary domain and elucidate some of basic properties.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:16:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yamagami",
"Shigeru",
""
]
] |
2504.09094 | Akanksha Mehndiratta | Akanksha Mehndiratta, Krishna Asawa | Enhancing Dialogue Systems with Discourse-Level Understanding Using Deep
Canonical Correlation Analysis | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The evolution of conversational agents has been driven by the need for more
contextually aware systems that can effectively manage dialogue over extended
interactions. To address the limitations of existing models in capturing and
utilizing long-term conversational history, we propose a novel framework that
integrates Deep Canonical Correlation Analysis (DCCA) for discourse-level
understanding. This framework learns discourse tokens to capture relationships
between utterances and their surrounding context, enabling a better
understanding of long-term dependencies. Experiments on the Ubuntu Dialogue
Corpus demonstrate significant enhancement in response selection, based on the
improved automatic evaluation metric scores. The results highlight the
potential of DCCA in improving dialogue systems by allowing them to filter out
irrelevant context and retain critical discourse information for more accurate
response retrieval.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:19:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mehndiratta",
"Akanksha",
""
],
[
"Asawa",
"Krishna",
""
]
] |
2504.09096 | Binghui Peng | Binghui Peng | High dimensional online calibration in polynomial time | null | null | null | null | cs.LG cs.DS cs.GT stat.ML | http://creativecommons.org/licenses/by/4.0/ | In online (sequential) calibration, a forecaster predicts probability
distributions over a finite outcome space $[d]$ over a sequence of $T$ days,
with the goal of being calibrated. While asymptotically calibrated strategies
are known to exist, they suffer from the curse of dimensionality: the best
known algorithms require $\exp(d)$ days to achieve non-trivial calibration.
In this work, we present the first asymptotically calibrated strategy that
guarantees non-trivial calibration after a polynomial number of rounds.
Specifically, for any desired accuracy $\epsilon > 0$, our forecaster becomes
$\epsilon$-calibrated after $T = d^{O(1/\epsilon^2)}$ days. We complement this
result with a lower bound, proving that at least $T =
d^{\Omega(\log(1/\epsilon))}$ rounds are necessary to achieve
$\epsilon$-calibration. Our results resolve the open questions posed by
[Abernethy-Mannor'11, Hazan-Kakade'12].
Our algorithm is inspired by recent breakthroughs in swap regret minimization
[Peng-Rubinstein'24, Dagan et al.'24]. Despite its strong theoretical
guarantees, the approach is remarkably simple and intuitive: it randomly
selects among a set of sub-forecasters, each of which predicts the empirical
outcome frequency over recent time windows.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:28:05 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Peng",
"Binghui",
""
]
] |
2504.09097 | Jeongwan On | Jeongwan On, Kyeonghwan Gwak, Gunyoung Kang, Junuk Cha, Soohyun Hwang,
Hyein Hwang, Seungryul Baek | BIGS: Bimanual Category-agnostic Interaction Reconstruction from
Monocular Videos via 3D Gaussian Splatting | Accepted to CVPR 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reconstructing 3Ds of hand-object interaction (HOI) is a fundamental problem
that can find numerous applications. Despite recent advances, there is no
comprehensive pipeline yet for bimanual class-agnostic interaction
reconstruction from a monocular RGB video, where two hands and an unknown
object are interacting with each other. Previous works tackled the limited
hand-object interaction case, where object templates are pre-known or only one
hand is involved in the interaction. The bimanual interaction reconstruction
exhibits severe occlusions introduced by complex interactions between two hands
and an object. To solve this, we first introduce BIGS (Bimanual Interaction 3D
Gaussian Splatting), a method that reconstructs 3D Gaussians of hands and an
unknown object from a monocular video. To robustly obtain object Gaussians
avoiding severe occlusions, we leverage prior knowledge of pre-trained
diffusion model with score distillation sampling (SDS) loss, to reconstruct
unseen object parts. For hand Gaussians, we exploit the 3D priors of hand model
(i.e., MANO) and share a single Gaussian for two hands to effectively
accumulate hand 3D information, given limited views. To further consider the 3D
alignment between hands and objects, we include the interacting-subjects
optimization step during Gaussian optimization. Our method achieves the
state-of-the-art accuracy on two challenging datasets, in terms of 3D hand pose
estimation (MPJPE), 3D object reconstruction (CDh, CDo, F10), and rendering
quality (PSNR, SSIM, LPIPS), respectively.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:30:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"On",
"Jeongwan",
""
],
[
"Gwak",
"Kyeonghwan",
""
],
[
"Kang",
"Gunyoung",
""
],
[
"Cha",
"Junuk",
""
],
[
"Hwang",
"Soohyun",
""
],
[
"Hwang",
"Hyein",
""
],
[
"Baek",
"Seungryul",
""
]
] |
2504.09098 | Daniel Herden | Daniel Bossaller, Daniel Herden, Indalecio Ruiz-Bola\~nos | The trace dual of nonlinear skew cyclic codes | 16 pages | null | null | null | cs.IT math.IT math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Codes which have a finite field $\mathbb{F}_{q^m}$ as their alphabet but
which are only linear over a subfield $\mathbb{F}_q$ are a topic of much recent
interest due to their utility in constructing quantum error correcting codes.
In this article, we find generators for trace dual spaces of different families
of $\mathbb{F}_q$-linear codes over $\mathbb{F}_{q^2}$. In particular, given
the field extension $\mathbb{F}_q\leq \mathbb{F}_{q^2}$ with $q$ an odd prime
power, we determine the trace Euclidean and trace Hermitian dual codes for the
general $\mathbb{F}_q$-linear cyclic $\mathbb{F}_{q^2}$-code. In addition, we
also determine the trace Euclidean and trace Hermitian duals for general
$\mathbb{F}_q$-linear skew cyclic $\mathbb{F}_{q^2}$-codes, which are defined
to be left $\mathbb{F}_q[X]$-submodules of
$\mathbb{F}_{q^2}[X;\sigma]/(X^n-1)$, where $\sigma$ denotes the Frobenius
automorphism and $\mathbb{F}_{q^2}[X;\sigma]$ the induced skew polynomial ring.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:35:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bossaller",
"Daniel",
""
],
[
"Herden",
"Daniel",
""
],
[
"Ruiz-Bolaños",
"Indalecio",
""
]
] |
2504.09099 | Florian Bemmann | Florian Bemmann, Doruntina Murtezaj | Rethinking News and Media System Design Towards Positive Societal
Implications | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Since this century, the speed, availability, and plethora of online
informational content have made it increasingly difficult for humans to keep an
overview of real-world situations, build a personal opinion, and sometimes even
decide on the truth. Thereby, personal opinion-making and public discourse
became harder - two essential building blocks that keep a democratic society
alive. HCI thus needs to rethink news, information, and social media systems to
mitigate such negative effects. Instead of polarising through emotional and
extremely framed messages, informational content online should make people
think about other opinions and discuss constructively. Instead, through
polarization and filter bubble effects, people lose openness and tolerance for
the existence of opposing opinions. In this workshop, we will discuss how we
can redesign our information technology for a better societal impact. We will
present key takeaways from the social sciences and discuss how we can implement
them using recent HCI findings and digital technologies.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:39:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bemmann",
"Florian",
""
],
[
"Murtezaj",
"Doruntina",
""
]
] |
2504.09100 | Chengyu Wang | Chengyu Wang, Taolin Zhang, Richang Hong, Jun Huang | A Short Survey on Small Reasoning Models: Training, Inference,
Applications and Research Directions | null | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, the reasoning capabilities of large reasoning models (LRMs), such
as DeepSeek-R1, have seen significant advancements through the slow thinking
process. Despite these achievements, the substantial computational demands of
LRMs present considerable challenges. In contrast, small reasoning models
(SRMs), often distilled from larger ones, offer greater efficiency and can
exhibit distinct capabilities and cognitive trajectories compared to LRMs. This
work surveys around 170 recently published papers on SRMs for tackling various
complex reasoning tasks. We review the current landscape of SRMs and analyze
diverse training and inference techniques related to SRMs. Furthermore, we
provide a comprehensive review of SRMs for domain-specific applications and
discuss possible future research directions. This survey serves as an essential
reference for researchers to leverage or develop SRMs for advanced reasoning
functionalities with high efficiency.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:45:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Chengyu",
""
],
[
"Zhang",
"Taolin",
""
],
[
"Hong",
"Richang",
""
],
[
"Huang",
"Jun",
""
]
] |
2504.09101 | Abdulmajid Murad | Abdulmajid Murad, Massimiliano Ruocco | Synthetic Aircraft Trajectory Generation Using Time-Based VQ-VAE | This paper was presented at the 25th Integrated Communications,
Navigation and Surveillance Conference (ICNS 2025), April 8--10, 2025,
Brussels, Belgium | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | In modern air traffic management, generating synthetic flight trajectories
has emerged as a promising solution for addressing data scarcity, protecting
sensitive information, and supporting large-scale analyses. In this paper, we
propose a novel method for trajectory synthesis by adapting the Time-Based
Vector Quantized Variational Autoencoder (TimeVQVAE). Our approach leverages
time-frequency domain processing, vector quantization, and transformer-based
priors to capture both global and local dynamics in flight data. By
discretizing the latent space and integrating transformer priors, the model
learns long-range spatiotemporal dependencies and preserves coherence across
entire flight paths. We evaluate the adapted TimeVQVAE using an extensive suite
of quality, statistical, and distributional metrics, as well as a flyability
assessment conducted in an open-source air traffic simulator. Results indicate
that TimeVQVAE outperforms a temporal convolutional VAE baseline, generating
synthetic trajectories that mirror real flight data in terms of spatial
accuracy, temporal consistency, and statistical properties. Furthermore, the
simulator-based assessment shows that most generated trajectories maintain
operational feasibility, although occasional outliers underscore the potential
need for additional domain-specific constraints. Overall, our findings
underscore the importance of multi-scale representation learning for capturing
complex flight behaviors and demonstrate the promise of TimeVQVAE in producing
representative synthetic trajectories for downstream tasks such as model
training, airspace design, and air traffic forecasting.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:46:51 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Murad",
"Abdulmajid",
""
],
[
"Ruocco",
"Massimiliano",
""
]
] |
2504.09102 | Foad Parsaei | Foad Parsaei, and Sara Rastgoo | Wormholes in $f(T,\mathcal{T})$ gravity | 14 pages, 15 figures | null | null | null | gr-qc | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study aims to investigate the physical properties of wormhole geometry
within the context of $f(T,\mathcal{T})$ gravity, which serves as a
teleparallel formulation of general relativity. We study a linear model,
$f(T,\mathcal{T})=\alpha T+\beta\mathcal{T}$, to explore traversable wormholes.
A linear equation of state is utilized for radial pressure, leading to a
power-law shaped function. It was found that the violation of energy
conditions, depends on the $\alpha$ and $\beta$ parameters. A diverse array of
intriguing wormhole solutions was identified, contingent upon the specific
model parameters employed. It is demonstrated that isotropic wormhole solutions
cannot be attained within this framework. Additionally, solutions characterized
by a variable equation of state parameter are introduced. A comparative
analysis of wormhole solutions in the context of $f(T,\mathcal{T})$ and
curvature-based gravity, specifically $f(R,T)$, is also provided.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:48:54 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Parsaei",
"Foad",
""
],
[
"Rastgoo",
"Sara",
""
]
] |
2504.09103 | Chengran Yuan | Jiawei Sun, Xibin Yue, Jiahui Li, Tianle Shen, Chengran Yuan, Shuo
Sun, Sheng Guo, Quanyun Zhou and Marcelo H Ang Jr | IMPACT: Behavioral Intention-aware Multimodal Trajectory Prediction with
Adaptive Context Trimming | under review | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While most prior research has focused on improving the precision of
multimodal trajectory predictions, the explicit modeling of multimodal
behavioral intentions (e.g., yielding, overtaking) remains relatively
underexplored. This paper proposes a unified framework that jointly predicts
both behavioral intentions and trajectories to enhance prediction accuracy,
interpretability, and efficiency. Specifically, we employ a shared context
encoder for both intention and trajectory predictions, thereby reducing
structural redundancy and information loss. Moreover, we address the lack of
ground-truth behavioral intention labels in mainstream datasets (Waymo,
Argoverse) by auto-labeling these datasets, thus advancing the community's
efforts in this direction. We further introduce a vectorized occupancy
prediction module that infers the probability of each map polyline being
occupied by the target vehicle's future trajectory. By leveraging these
intention and occupancy prediction priors, our method conducts dynamic,
modality-dependent pruning of irrelevant agents and map polylines in the
decoding stage, effectively reducing computational overhead and mitigating
noise from non-critical elements. Our approach ranks first among LiDAR-free
methods on the Waymo Motion Dataset and achieves first place on the Waymo
Interactive Prediction Dataset. Remarkably, even without model ensembling, our
single-model framework improves the soft mean average precision (softmAP) by 10
percent compared to the second-best method in the Waymo Interactive Prediction
Leaderboard. Furthermore, the proposed framework has been successfully deployed
on real vehicles, demonstrating its practical effectiveness in real-world
applications.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:56:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sun",
"Jiawei",
""
],
[
"Yue",
"Xibin",
""
],
[
"Li",
"Jiahui",
""
],
[
"Shen",
"Tianle",
""
],
[
"Yuan",
"Chengran",
""
],
[
"Sun",
"Shuo",
""
],
[
"Guo",
"Sheng",
""
],
[
"Zhou",
"Quanyun",
""
],
[
"Ang",
"Marcelo H",
"Jr"
]
] |
2504.09104 | Lucio Davide Spano | Alessandro Carcangiu, Marco Manca, Jacopo Mereu, Carmen Santoro,
Ludovica Simeoli, Lucio Davide Spano | Tell-XR: Conversational End-User Development of XR Automations | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The availability of extended reality (XR) devices has widened their adoption,
yet authoring interactive experiences remains complex for non-programmers. We
introduce Tell-XR, an intelligent agent leveraging large language models (LLMs)
to guide end-users in defining the interaction in XR settings using automations
described as Event-Condition-Action (ECA) rules. Through a formative study, we
identified the key conversation stages to define and refine automations, which
informed the design of the system architecture. The evaluation study in two
scenarios (a VR museum and an AR smart home) demonstrates the effectiveness of
Tell-XR across different XR interaction settings.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:00:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Carcangiu",
"Alessandro",
""
],
[
"Manca",
"Marco",
""
],
[
"Mereu",
"Jacopo",
""
],
[
"Santoro",
"Carmen",
""
],
[
"Simeoli",
"Ludovica",
""
],
[
"Spano",
"Lucio Davide",
""
]
] |
2504.09105 | Jose Angel Pelaez | Carmen Cascante, Joan F\'abrega, Daniel Pascuas and Jos\'e \'Angel
Pel\'aez | Words of analytic paraproducts on Bergman spaces | null | null | null | null | math.CV math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For a fixed analytic function g on the unit disc, we consider the analytic
paraproducts induced by g, which are formally defined by
$T_gf(z)=\int_0^zf(\zeta)g'(\zeta)d\zeta$,
$S_gf(z)=\int_0^zf'(\zeta)g(\zeta)d\zeta$, and $M_gf(z)=g(z)f(z)$. An
$N$-letter $g$-word is an operator of the form $L=L_1\cdots L_N$, where each
$L_j$ is either $M_g$, $S_g$ or $T_g$. It has been recently proved, in a recent
paper by A. Aleman and the authors of this paper, that understanding the
boundedness of a $g$-word on classical Hardy and Bergman spaces is a
challenging problem due to the potential cancellations involved. Our main
result provides a complete quantitative characterization of the boundedness of
an arbitrary $g$-word on a weighted Bergman space $A^p_{\omega^{p/2}}$, where
$\omega=e^{-2\varphi}$ is a smooth rapidly decreasing weight. In particular, it
states that any $N$-letter $g$-word such that $\#\{j:L_j=T_g\}=n\ge 1$ is
bounded on $A^p_{\omega^{p/2}}$ if and only if $g$ satisfies the "fractional"
Bloch-type condition \[ \|g\|_{\mathcal{B}^s_\varphi}^s=
\sup_{z\in\mathbb{D}}\frac{s|g(z)|^{s-1}|g'(z)|}{1+\varphi'(|z|)} <\infty, \]
where $s=\frac{N}{n}$, and $\|L\|_{A^p_{\omega^{p/2}}}\simeq
\|g\|_{\mathcal{B}^s_\varphi}^N$.
The class of smooth rapidly decreasing weights contains
the radial weights \begin{equation*}
\omega_n(z)=e^{-2\exp_{n}(g_{\alpha,c}(|z|))}, \quad\mbox{where}\quad
g_{\alpha,c}(r)=\tfrac{c}{(1-r^2)^{\alpha}}, \quad\mbox{for $c,\alpha>0$,}
\end{equation*} $\exp_0(x)=x$ and $\exp_n(x)=e^{\exp_{n-1}(x)}$, for
$n\in\mathbb{N}$. Therefore it contains
weights which decrease arbitrarily rapidly to zero as $|z|\to 1^-$.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:00:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Cascante",
"Carmen",
""
],
[
"Fábrega",
"Joan",
""
],
[
"Pascuas",
"Daniel",
""
],
[
"Peláez",
"José Ángel",
""
]
] |
2504.09106 | Yonghao Huang | Yonghao Huang, Leiting Chen, Chuan Zhou | Multi-modal and Multi-view Fundus Image Fusion for Retinopathy Diagnosis
via Multi-scale Cross-attention and Shifted Window Self-attention | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The joint interpretation of multi-modal and multi-view fundus images is
critical for retinopathy prevention, as different views can show the complete
3D eyeball field and different modalities can provide complementary lesion
areas. Compared with single images, the sequence relationships in multi-modal
and multi-view fundus images contain long-range dependencies in lesion
features. By modeling the long-range dependencies in these sequences, lesion
areas can be more comprehensively mined, and modality-specific lesions can be
detected. To learn the long-range dependency relationship and fuse
complementary multi-scale lesion features between different fundus modalities,
we design a multi-modal fundus image fusion method based on multi-scale
cross-attention, which solves the static receptive field problem in previous
multi-modal medical fusion methods based on attention. To capture multi-view
relative positional relationships between different views and fuse
comprehensive lesion features between different views, we design a multi-view
fundus image fusion method based on shifted window self-attention, which also
solves the computational complexity of the multi-view fundus fusion method
based on self-attention is quadratic to the size and number of multi-view
fundus images. Finally, we design a multi-task retinopathy diagnosis framework
to help ophthalmologists reduce workload and improve diagnostic accuracy by
combining the proposed two fusion methods. The experimental results of
retinopathy classification and report generation tasks indicate our method's
potential to improve the efficiency and reliability of retinopathy diagnosis in
clinical practice, achieving a classification accuracy of 82.53\% and a report
generation BlEU-1 of 0.543.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:06:15 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Huang",
"Yonghao",
""
],
[
"Chen",
"Leiting",
""
],
[
"Zhou",
"Chuan",
""
]
] |
2504.09107 | Miao Cheng | Miao Cheng and Feiyan Zhou and Hongwei Zou and Limin Wang | Shrinkage Initialization for Smooth Learning of Neural Networks | 6 pages, 4 figures | null | 10.1145/3695220.3695226 | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The successes of intelligent systems have quite relied on the artificial
learning of information, which lead to the broad applications of neural
learning solutions. As a common sense, the training of neural networks can be
largely improved by specifically defined initialization, neuron layers as well
as the activation functions. Though there are sequential layer based
initialization available, the generalized solution to initial stages is still
desired. In this work, an improved approach to initialization of neural
learning is presented, which adopts the shrinkage approach to initialize the
transformation of each layer of networks. It can be universally adapted for the
structures of any networks with random layers, while stable performance can be
attained. Furthermore, the smooth learning of networks is adopted in this work,
due to the diverse influence on neural learning. Experimental results on
several artificial data sets demonstrate that, the proposed method is able to
present robust results with the shrinkage initialization, and competent for
smooth learning of neural networks.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:11:35 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Cheng",
"Miao",
""
],
[
"Zhou",
"Feiyan",
""
],
[
"Zou",
"Hongwei",
""
],
[
"Wang",
"Limin",
""
]
] |
2504.09108 | Telem Ibungochouba Singh | Yenshembam Priyobarta Singh, Irengbam Roshila Devi, Telem Ibungochouba
Singh | Quasinormal modes of spherically symmetric black hole with cosmological
constant and global monopole in bumblebee gravity | 28 pages and 39 figures | null | null | null | gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | This paper studies the scalar, electromagnetic field and Dirac field
perturbations of spherically symmetric black hole within the framework of
Einstein-bumblebee gravity model with global monopole and cosmological
constant. We investigate the effective potentials, greybody factors and
quasinormal modes (QNMs) by applying the Klein-Gordon equation, electromagnetic
field equation and Dirac equation expressed in Newman-Penrose (NP) formalism.
Using the general method of rigorous bound, the greybody factors of scalar,
electromagnetic and Dirac field are derived. Applying the sixth order WKB
approximation and Pad\'{e} approximation, the QNM frequencies are derived. We
also discuss the impact of global monopole $\eta$, cosmological constant
$\Lambda$ and Lorentz violation parameter $L$ to the effective potential,
greybody factor and QNMs. Increasing the parameter $\eta$ prevents the rise of
effective potential for both Schwarzschild-de Sitter (SdS)-like and
Schwarzschild-Anti de Sitter (SAdS)-like black holes with global monopole and
consequently increases the greybody factors. However, decreasing the parameter
$L$ reduces the rise of effective potential for the SAdS-like black hole with
global monopole and it increases the greybody factor but increasing the
parameter $L$ has an opposite effect for SdS-like black hole with global
monopole. It is also shown that the shadow radius increases with increasing the
parameter $\eta$ for both dS and AdS cases. Increasing the value of $L$ tends
to increase the shadow radius for dS black hole but it has an opposite effect
in AdS case. A careful studying is being carried out to investigate how the
absorption cross-section gets affected when the parameter $L$ appears into the
picture.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:11:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Singh",
"Yenshembam Priyobarta",
""
],
[
"Devi",
"Irengbam Roshila",
""
],
[
"Singh",
"Telem Ibungochouba",
""
]
] |
2504.09110 | Haruma Furukawa | Ryosuke Yoneda, Haruma Furukawa, Daigo Fujiwara, Toshio Aoyagi | Gaussian process regression with additive periodic kernels for two-body
interaction analysis in coupled phase oscillators | null | null | null | null | nlin.AO | http://creativecommons.org/licenses/by/4.0/ | We propose a Gaussian process regression framework with additive periodic
kernels for the analysis of two-body interactions in coupled oscillator
systems. While finite-order Fourier expansions determined by Bayesian methods
can still yield artifacts such as a high-amplitude, high-frequency vibration,
our additive periodic kernel approach has been demonstrated to effectively
circumvent these issues. Furthermore, by exploiting the additive and periodic
nature of the coupling functions, we significantly reduce the effective
dimensionality of the inference problem. We first validate our method on simple
coupled phase oscillators and demonstrate its robustness to more complex
systems, including Van der Pol and FitzHugh-Nagumo oscillators, under
conditions of biased or limited data. We next apply our approach to spiking
neural networks modeled by Hodgkin-Huxley equations, in which we successfully
recover the underlying interaction functions. These results highlight the
flexibility and stability of Gaussian process regression in capturing
nonlinear, periodic interactions in oscillator networks. Our framework provides
a practical alternative to conventional methods, enabling data-driven studies
of synchronized rhythmic systems across physics, biology, and engineering.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:42:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yoneda",
"Ryosuke",
""
],
[
"Furukawa",
"Haruma",
""
],
[
"Fujiwara",
"Daigo",
""
],
[
"Aoyagi",
"Toshio",
""
]
] |
2504.09111 | Stefanie Schwaar | Stefanie Schwaar, Franziska Diez, Michael Trebing, Nils Witznick | Study on Text Classification for Public Administration | null | null | null | null | stat.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In German public administration, there are 45 different offices to which
incoming messages need to be distributed. Since these messages are often
unstructured, the system has to be based at least partly on message content.
For public service no data are given so far and no pretrained model is
available. The data we used are conducted by Governikus KG and are of highly
different length. To handle those data with standard methods different
approaches are known, like normalization or segmentation. However, text
classification is highly dependent on the data structure, a study for public
administration data is missing at the moment. We conducted such a study
analyzing different techniques of classification based on segments,
normalization and feature selection. Thereby, we used different methods, this
means neural nets, random forest, logistic regression, SVM classifier and SVAE.
The comparison shows for the given public service data a classification
accuracy of above 80\% can be reached based on cross validation. We further
show that normalization is preferable, while the difference to the segmentation
approach depends mainly on the choice of algorithm.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:44:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Schwaar",
"Stefanie",
""
],
[
"Diez",
"Franziska",
""
],
[
"Trebing",
"Michael",
""
],
[
"Witznick",
"Nils",
""
]
] |
2504.09112 | Imam Makhfudz | Imam Makhfudz | Controllable and Non-Dissipative Inertial Dynamics of Skyrmion in a
Bosonic Platform | 5.5 pages main text + 3.5pages supplementary materials. Comments are
welcome | null | null | null | cond-mat.mes-hall cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | It has been understood in the past a decade or two that the dynamics of spin
or magnetization in ultrafast regime necessarily involves inertial term that
reflects the reluctance to follow abrupt or sudden change in the spin or
magnetization orientation. The role of inertial spin dynamics in governing the
motion of Skyrmion, a topological spin texture, is elucidated. Using
nonequilibrium Green's function Keldysh formalism, an equation of motion is
derived in terms of collective coordinates for a Skyrmion coupled via a
''minimal coupling'' to a bath of harmonic oscillators of frequency $\omega$. A
deterministic and non-dissipative dynamics equation of motion is obtained with
an explicit mass term for the Skyrmion emerging due the coupling, even within
rigid Skyrmion picture. This results in a cyclotronic motion of Skyrmion, with
a frequency that can go ultrafast, depending on that of the oscillator.
Controlling the oscillator frequency can therefore guide the Skyrmion dynamics.
Our theory bridges inertial dynamics and topology in magnetism and opens a
pathway to ultrafast control of topological spin textures.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:52:39 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Makhfudz",
"Imam",
""
]
] |
2504.09113 | Zeyan Li | Zeyan Li, Jie Song, Tieying Zhang, Tao Yang, Xiongjun Ou, Yingjie Ye,
Pengfei Duan, Muchen Lin, Jianjun Chen | Adaptive and Efficient Log Parsing as a Cloud Service | Accepted by SIGMOD'25 Industry track | null | 10.1145/3722212.3724427 | null | cs.SE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Logs are a critical data source for cloud systems, enabling advanced features
like monitoring, alerting, and root cause analysis. However, the massive scale
and diverse formats of unstructured logs pose challenges for adaptable,
efficient, and accurate parsing methods. This paper introduces
ByteBrain-LogParser, an innovative log parsing framework designed specifically
for cloud environments. ByteBrain-LogParser employs a hierarchical clustering
algorithm to allow real-time precision adjustments, coupled with optimizations
such as positional similarity distance, deduplication, and hash encoding to
enhance performance. Experiments on large-scale datasets show that it processes
229,000 logs per second on average, achieving an 840% speedup over the fastest
baseline while maintaining accuracy comparable to state-of-the-art methods.
Real-world evaluations further validate its efficiency and adaptability,
demonstrating its potential as a robust cloud-based log parsing solution.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:53:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Zeyan",
""
],
[
"Song",
"Jie",
""
],
[
"Zhang",
"Tieying",
""
],
[
"Yang",
"Tao",
""
],
[
"Ou",
"Xiongjun",
""
],
[
"Ye",
"Yingjie",
""
],
[
"Duan",
"Pengfei",
""
],
[
"Lin",
"Muchen",
""
],
[
"Chen",
"Jianjun",
""
]
] |
2504.09114 | Xianke Qiang | Xianke Qiang, Hongda Liu, Xinran Zhang, Zheng Chang, Ying-Chang Liang | Deploying Large AI Models on Resource-Limited Devices with Split
Federated Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Artificial Intelligence Models (LAMs) powered by massive datasets,
extensive parameter scales, and extensive computational resources, leading to
significant transformations across various industries. Yet, their practical
deployment on resource-limited mobile edge devices is hindered by critical
challenges such as data privacy, constrained resources, and high overhead
costs. Addressing this gap, this paper proposes a novel framework, named
Quantized Split Federated Fine-Tuning Large AI Model (SFLAM). By partitioning
the training load between edge devices and servers using a split learning
paradigm, SFLAM can facilitate the operation of large models on devices and
significantly lowers the memory requirements on edge devices. Additionally,
SFLAM incorporates quantization management, power control, and bandwidth
allocation strategies to enhance training efficiency while concurrently
reducing energy consumption and communication latency. A theoretical analysis
exploring the latency-energy trade-off is presented, and the framework's
efficacy is validated via comprehensive simulations. The findings indicate that
SFLAM achieves superior performance in terms of learning efficiency and
scalability compared to conventional methods, thereby providing a valuable
approach for enabling advanced AI services in resource-constrained scenarios.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:55:11 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Qiang",
"Xianke",
""
],
[
"Liu",
"Hongda",
""
],
[
"Zhang",
"Xinran",
""
],
[
"Chang",
"Zheng",
""
],
[
"Liang",
"Ying-Chang",
""
]
] |
2504.09116 | Lingyou Zhou | Lingyou Zhou, Xin Dong, Kehai Qiu, Gang Yu, Jie Zhang, Jiliang Zhang | Ray-Based Characterization of the AMPLE Model from 0.85 to 5 GHz | This work has been submitted to IEEE for possible publication | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we characterize the adaptive multiple path loss exponent
(AMPLE) radio propagation model under urban macrocell (UMa) and urban microcell
(UMi) scenarios from 0.85-5 GHz using Ranplan Professional. We first enhance
the original AMPLE model by introducing an additional frequency coefficient to
support path loss prediction across multiple carrier frequencies. By using
measurement-validated Ranplan Professional simulator, we simulate four cities
and validate the simulations for further path loss model characterization.
Specifically, we extract the close-in (CI) model parameters from the
simulations and compare them with parameters extracted from measurements in
other works. Under the ray-based model characterization, we compare the AMPLE
model with the 3rd Generation Partnership Project (3GPP) path loss model, the
CI model, the alpha-beta-gamma (ABG) model, and those with simulation
calibrations. In addition to standard performance metrics, we introduce the
prediction-measurement difference error (PMDE) to assess overall prediction
alignment with measurement, and mean simulation time per data point to evaluate
model complexity. The results show that the AMPLE model outperforms existing
models while maintaining similar model complexity.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:05:28 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhou",
"Lingyou",
""
],
[
"Dong",
"Xin",
""
],
[
"Qiu",
"Kehai",
""
],
[
"Yu",
"Gang",
""
],
[
"Zhang",
"Jie",
""
],
[
"Zhang",
"Jiliang",
""
]
] |
2504.09117 | Evagoras Makridis Mr | Neofytos Charalampous, Evagoras Makridis, Apostolos I. Rikos, and
Themistoklis Charalambous | HARQ-based Quantized Average Consensus over Unreliable Directed Network
Topologies | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose a distributed algorithm (herein called HARQ-QAC)
that enables nodes to calculate the average of their initial states by
exchanging quantized messages over a directed communication network. In our
setting, we assume that our communication network consists of unreliable
communication links (i.e., links suffering from packet drops). For countering
link unreliability our algorithm leverages narrowband error-free feedback
channels for acknowledging whether a packet transmission between nodes was
successful. Additionally, we show that the feedback channels play a crucial
role in enabling our algorithm to exhibit finite-time convergence. We analyze
our algorithm and demonstrate its operation via an example, where we illustrate
its operational advantages. Finally, simulations corroborate that our proposed
algorithm converges to the average of the initial quantized values in a finite
number of steps, despite the packet losses. This is the first quantized
consensus algorithm in the literature that can handle packet losses and
converge to the average. Additionally, the use of the retransmission mechanism
allows for accelerating the convergence.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:06:44 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Charalampous",
"Neofytos",
""
],
[
"Makridis",
"Evagoras",
""
],
[
"Rikos",
"Apostolos I.",
""
],
[
"Charalambous",
"Themistoklis",
""
]
] |
2504.09118 | Yifei He | Yifei He and M{\aa}ns I. Andersson and Stefano Markidis | Optimizing FDTD Solvers for Electromagnetics: A Compiler-Guided Approach
with High-Level Tensor Abstractions | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The Finite Difference Time Domain (FDTD) method is a widely used numerical
technique for solving Maxwell's equations, particularly in computational
electromagnetics and photonics. It enables accurate modeling of wave
propagation in complex media and structures but comes with significant
computational challenges. Traditional FDTD implementations rely on handwritten,
platform-specific code that optimizes certain kernels while underperforming in
others. The lack of portability increases development overhead and creates
performance bottlenecks, limiting scalability across modern hardware
architectures. To address these challenges, we introduce an end-to-end
domain-specific compiler based on the MLIR/LLVM infrastructure for FDTD
simulations. Our approach generates efficient and portable code optimized for
diverse hardware platforms.We implement the three-dimensional FDTD kernel as
operations on a 3D tensor abstraction with explicit computational semantics.
High-level optimizations such as loop tiling, fusion, and vectorization are
automatically applied by the compiler. We evaluate our customized code
generation pipeline on Intel, AMD, and ARM platforms, achieving up to
$10\times$ speedup over baseline Python implementation using NumPy.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:08:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Yifei",
""
],
[
"Andersson",
"Måns I.",
""
],
[
"Markidis",
"Stefano",
""
]
] |
2504.09119 | Dimitris Stavropoulos PhD | KM3NeT Collaboration | Measurement of the atmospheric $\nu_{\mu}$ flux with six detection units
of KM3NeT/ORCA | null | null | null | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | A measurement of the atmospheric $\nu_{\mu}+\bar{\nu}_{\mu}$ flux with
energies between 1--100 GeV is presented. The measurement has been performed
using data taken with the first six detection units of the KM3NeT/ORCA
detector, referred to as ORCA6. The data were collected between January 2020
and November 2021 and correspond to 510 days of livetime, with a total exposure
of 433 kton$\cdot$years. Using machine learning classification, 3894 neutrino
candidate events have been selected with an atmospheric muon contamination of
less than 1$\%$. The atmospheric $\nu_{\mu}+\bar{\nu}_{\mu}$ energy spectrum is
derived using an unfolding procedure and the impact of systematic uncertainties
is estimated. The atmospheric $\nu_{\mu}+\bar{\nu}_{\mu}$ flux measured using
the ORCA6 configuration is in agreement with the values measured by other
experiments.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:08:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"KM3NeT Collaboration",
"",
""
]
] |
2504.09120 | Shoichi Sasaki | Toshizo Arikawa, Keita Sakai and Shoichi Sasaki | Glueball mass spectrum at finite temperature revisited: Constant
contribution in glueball correlators in the deconfinement phase | 23 pages, 39 figures | null | null | null | hep-lat hep-ph | http://creativecommons.org/licenses/by/4.0/ | We study the glueball properties at finite temperature from the temporal
correlation in $SU(3)$ Yang-Mills theory using anisotropic lattice QCD.
Although the existence of a constant contribution to the meson correlation
function appearing in the deconfinement phase are known, its effect on the
glueball correlation function at finite temperature has not been considered in
previous studies. The present study reveals that the constant contribution to
the glueball correlation function actually occurs in the three lowest-lying
glueball states, corresponding to the $0^{++}$, $2^{++}$ and $0^{-+}$
glueballs, from near the critical temperature of the deconfinement phase
transition $T_C$ to the high temperature side. If the existence of constant
terms is taken into account in the standard pole-mass analysis, it is observed
that the pole-mass of the glueball ground state remains unchanged below $T_c$
and then increases with temperature above $T_C$. The result indicates that the
true temperature dependence of the glueball mass above $T_C$ is opposite to the
results of the previous studies.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:09:56 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Arikawa",
"Toshizo",
""
],
[
"Sakai",
"Keita",
""
],
[
"Sasaki",
"Shoichi",
""
]
] |
2504.09121 | Rafael S\'anchez | Jos\'e Balduque and Rafael S\'anchez | Quantum thermocouples: nonlocal conversion and control of heat in
semiconductor nanostructures | Short review. 16 pages + references, 6 figures. Comments are welcome! | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nanoscale conductors are interesting for thermoelectrics because of their
particular spectral features connecting separated heat and particle currents.
Multiterminal devices in the quantum regime benefit from phase-coherent
phenomena, which turns the thermoelectric effect nonlocal, and from tunable
single-particle interactions. This way one can define quantum thermocouples
which convert an injected heat current into useful power in an isothermal
conductor, or work as refrigerators. Additionally, efficient heat management
devices can be defined. We review recent theoretical and experimental progress
in the research of multiterminal thermal and thermoelectric quantum transport
leading to proposals of autonomous quantum heat engines and thermal devices.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:10:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Balduque",
"José",
""
],
[
"Sánchez",
"Rafael",
""
]
] |
2504.09122 | Krzysztof Urbanowski | Krzysztof Urbanowski | Closer look at sum uncertainty relations and related relations | 15 pages, draft | null | null | null | quant-ph physics.app-ph physics.data-an | http://creativecommons.org/licenses/by/4.0/ | We analyze the weak and critical points of various uncertainty relations that
follow from the inequalities for the norms of vectors in the Hilbert space of
states of a quantum system. There are studied uncertainty relations for sums of
standard deviations, for sums of variances, and other relations between
standard deviations or variances. The obtained results are compared with the
conclusions obtained in similar cases using the standard Heisenberg-Robertson
uncertainty relation. We also show that there exists an upper bound on the
product of standard deviations that appears in the Heisenberg-Robertson
uncertainty relation.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:11:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Urbanowski",
"Krzysztof",
""
]
] |
2504.09123 | JiSun Huh | JiSun Huh, Byung-Hak Hwang, Donghyun Kim, Jang Soo Kim, and Jaeseong
Oh | Refinement of Hikita's $e$-positivity theorem via Abreu--Nigro's
$g$-functions and restricted modular law | 33pages | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the symmetric functions \( g_{\mm,k}(x;q) \), introduced by
Abreu and Nigro for a Hessenberg function \( \mm \) and a positive
integer \( k \), which refine the chromatic symmetric function.
Building on Hikita's recent breakthrough on the Stanley--Stembridge
conjecture, we prove the \( e \)-positivity of \( g_{\mm,k}(x;1) \),
refining Hikita's result. We also provide a Schur expansion of the
sum \( \sum_{k=1}^n e_k(x) g_{\mm,n-k}(x;q) \) in terms of
\( P \)-tableaux with 1 in the upper-left corner. We introduce a
restricted version of the modular law as our main tool. Then, we
show that any function satisfying the restricted modular law is
determined by its values on disjoint unions of path graphs.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:17:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Huh",
"JiSun",
""
],
[
"Hwang",
"Byung-Hak",
""
],
[
"Kim",
"Donghyun",
""
],
[
"Kim",
"Jang Soo",
""
],
[
"Oh",
"Jaeseong",
""
]
] |
2504.09124 | Fabrice Baudoin Dr | Fabrice Baudoin, Nizar Demni, Teije Kuijper, Jing Wang | Brownian motion and stochastic areas on complex full flag manifolds | null | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We show that the Brownian motion on the complex full flag manifold can be
represented by a matrix-valued diffusion obtained from the unitary Brownian
motion. This representation actually leads to an explicit formula for the
characteristic function of the joint distribution of the stochastic areas on
the full flag manifold. The limit law for those stochastic areas is shown to be
a multivariate Cauchy distribution with independent and identically distributed
entries. Using a deep connection between area functionals on the flag manifold
and winding functionals on complex spheres, we establish new results about
simultaneous Brownian windings on the complex sphere and their asymptotics. As
a byproduct, our work also unveils a new probabilistic interpretation of the
Jacobi operators and polynomials on simplices.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:19:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Baudoin",
"Fabrice",
""
],
[
"Demni",
"Nizar",
""
],
[
"Kuijper",
"Teije",
""
],
[
"Wang",
"Jing",
""
]
] |
2504.09125 | Yash Agrawal | Yash Agrawal, Saurabh Singh, Girish B. S., Somashekar R., Srivani K.
S., Raghunathan A., Vishakha S. Pandharpure, Udaya Shankar N., Keerthipriya
S. and Mayuri Sathyanarayana Rao | Spectral Intruders: A Multi-Site Study of Radio Environment for
Cosmology Experiments | 33 pages, 18 figures, (Submitted to Experimental Astronomy) | null | null | null | astro-ph.IM astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | Radio Frequency Interference (RFI) presents a significant challenge for
carrying out precision measurements in radio astronomy. In particular, RFI can
be a show stopper when looking for faint cosmological signals such as the
red-shifted 21-cm line from cosmic dawn (CD) and epoch of reionization (EoR).
As wireless communications, satellite transmissions, and other RF technologies
proliferate globally, understanding the RFI landscape has become essential for
site selection and data integrity. We present findings from RFI surveys
conducted at four distinct locations: three locations in India, the
Gauribidanur Radio Observatory in Karnataka, Twin Lakes in Ladakh, Kalpong Dam
in the Andaman Islands, and the Gruvebadet Atmosphere Laboratory in Ny-Alesund,
Svalbard, Norway. These sites, selected based on their geographical diversity
and varying levels of human activity, were studied to assess RFI presence in
30-300 MHz bands-critical for low-frequency observations and experiments
targeting the 21-cm CD/EoR signal. Using an automated RFI detection approach
via the Hampel filter and singular value decomposition, the surveys identified
both persistent and transient interference, which varies with location and
time. The results provide a comprehensive view of the RFI environment at each
site, informing the feasibility of long-term cosmological observations and
aiding in the mitigation of RFI in radio astronomical data. The methods
developed to characterize RFI can be easily generalized to any location and
experiment.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:24:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Agrawal",
"Yash",
""
],
[
"Singh",
"Saurabh",
""
],
[
"S.",
"Girish B.",
""
],
[
"R.",
"Somashekar",
""
],
[
"S.",
"Srivani K.",
""
],
[
"A.",
"Raghunathan",
""
],
[
"Pandharpure",
"Vishakha S.",
""
],
[
"N.",
"Udaya Shankar",
""
],
[
"S.",
"Keerthipriya",
""
],
[
"Rao",
"Mayuri Sathyanarayana",
""
]
] |
2504.09126 | Kanat Abdukhalikov | Kanat Abdukhalikov, Duy Ho, San Ling, Gyanendra K. Verma | Linear complementary dual quasi-cyclic codes of index 2 | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | We provide a polynomial approach to investigate linear complementary dual
(LCD) quasi-cyclic codes over finite fields. We establish necessary and
sufficient conditions for LCD quasi-cyclic codes of index 2 with respect to the
Euclidean, Hermitian, and symplectic inner products. As a consequence of these
characterizations, we derive necessary and sufficient conditions for LCD
one-generator quasi-cyclic codes.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:27:11 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Abdukhalikov",
"Kanat",
""
],
[
"Ho",
"Duy",
""
],
[
"Ling",
"San",
""
],
[
"Verma",
"Gyanendra K.",
""
]
] |
2504.09127 | Andres Contreras Hip | Andres A. Contreras Hip | Channels of Energy for the Linearized Energy Critical Wave Equation in
Even Dimensions $N\geq 8$ | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We prove an exterior energy estimate for the linearized energy critical wave
equation around a multisoliton for even dimensions $N\geq 8.$ This extends
previous work of Collot-Duyckaerts-Kenig-Merle to higher dimensions. During the
proof we encounter various additional important technical difficulties compared
to lower dimensions. In particular, we need to deal with a number of
generalized eigenfunctions of the static operator which increases linearly in
$N.$ This makes the analysis of projections onto these eigenfunctions a higher
dimensional problem, which requires linear systems to control. This is a
crucial ingredient in our upcoming work where we give an alternative proof of
the soliton resolution for the wave maps equation based on the method of
channels of energy developed by Duyckaerts-Kenig-Merle.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:28:15 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hip",
"Andres A. Contreras",
""
]
] |
2504.09128 | Gianluca Paolini | J.B. Nation and Gianluca Paolini | Elementary properties of free lattices II: decidability of the universal
theory | null | null | null | null | math.LO | http://creativecommons.org/publicdomain/zero/1.0/ | We continue our work on the model theory of free lattices, solving two of the
main open problems from our first paper on the subject. Our main result is that
the universal (existential) theory of infinite free lattices is decidable. Our
second main result is a proof that finitely generated free lattices are
positively distinguishable, as for each $n \geq 1$ there is a positive $\exists
\forall$-sentence true in $\mathbf{F}_n$ and false in $\mathbf{F}_{n+1}$.
Finally, we show that free lattices are first-order rigid in the class of
finitely generated projective lattices, and that a projective lattice has the
same existential (universal) theory of an infinite free lattice if and only if
it has breadth $> 4$ (i.e., a single existential sentence is sufficient).
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:30:41 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Nation",
"J. B.",
""
],
[
"Paolini",
"Gianluca",
""
]
] |
2504.09129 | Tze Ho Elden Tse | Jizong Peng, Tze Ho Elden Tse, Kai Xu, Wenchao Gao, Angela Yao | A Constrained Optimization Approach for Gaussian Splatting from
Coarsely-posed Images and Noisy Lidar Point Clouds | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | 3D Gaussian Splatting (3DGS) is a powerful reconstruction technique, but it
needs to be initialized from accurate camera poses and high-fidelity point
clouds. Typically, the initialization is taken from Structure-from-Motion (SfM)
algorithms; however, SfM is time-consuming and restricts the application of
3DGS in real-world scenarios and large-scale scene reconstruction. We introduce
a constrained optimization method for simultaneous camera pose estimation and
3D reconstruction that does not require SfM support. Core to our approach is
decomposing a camera pose into a sequence of camera-to-(device-)center and
(device-)center-to-world optimizations. To facilitate, we propose two
optimization constraints conditioned to the sensitivity of each parameter group
and restricts each parameter's search space. In addition, as we learn the scene
geometry directly from the noisy point clouds, we propose geometric constraints
to improve the reconstruction quality. Experiments demonstrate that the
proposed method significantly outperforms the existing (multi-modal) 3DGS
baseline and methods supplemented by COLMAP on both our collected dataset and
two public benchmarks.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:34:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Peng",
"Jizong",
""
],
[
"Tse",
"Tze Ho Elden",
""
],
[
"Xu",
"Kai",
""
],
[
"Gao",
"Wenchao",
""
],
[
"Yao",
"Angela",
""
]
] |
2504.09130 | Yikun Wang | Yikun Wang, Siyin Wang, Qinyuan Cheng, Zhaoye Fei, Liang Ding, Qipeng
Guo, Dacheng Tao, Xipeng Qiu | VisuoThink: Empowering LVLM Reasoning with Multimodal Tree Search | 12 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in Large Vision-Language Models have showcased remarkable
capabilities. However, they often falter when confronted with complex reasoning
tasks that humans typically address through visual aids and deliberate,
step-by-step thinking. While existing methods have explored text-based slow
thinking or rudimentary visual assistance, they fall short of capturing the
intricate, interleaved nature of human visual-verbal reasoning processes. To
overcome these limitations and inspired by the mechanisms of slow thinking in
human cognition, we introduce VisuoThink, a novel framework that seamlessly
integrates visuospatial and linguistic domains. VisuoThink facilitates
multimodal slow thinking by enabling progressive visual-textual reasoning and
incorporates test-time scaling through look-ahead tree search. Extensive
experiments demonstrate that VisuoThink significantly enhances reasoning
capabilities via inference-time scaling, even without fine-tuning, achieving
state-of-the-art performance in tasks involving geometry and spatial reasoning.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:37:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Yikun",
""
],
[
"Wang",
"Siyin",
""
],
[
"Cheng",
"Qinyuan",
""
],
[
"Fei",
"Zhaoye",
""
],
[
"Ding",
"Liang",
""
],
[
"Guo",
"Qipeng",
""
],
[
"Tao",
"Dacheng",
""
],
[
"Qiu",
"Xipeng",
""
]
] |
2504.09132 | Matthew Webster | Matthew B. Webster, Dongheon Lee, Joonnyong Lee | Self-Supervised Autoencoder Network for Robust Heart Rate Extraction
from Noisy Photoplethysmogram: Applying Blind Source Separation to Biosignal
Analysis | 12 pages, 5 figures, preprint | null | null | null | cs.LG eess.SP | http://creativecommons.org/licenses/by/4.0/ | Biosignals can be viewed as mixtures measuring particular physiological
events, and blind source separation (BSS) aims to extract underlying source
signals from mixtures. This paper proposes a self-supervised multi-encoder
autoencoder (MEAE) to separate heartbeat-related source signals from
photoplethysmogram (PPG), enhancing heart rate (HR) detection in noisy PPG
data. The MEAE is trained on PPG signals from a large open polysomnography
database without any pre-processing or data selection. The trained network is
then applied to a noisy PPG dataset collected during the daily activities of
nine subjects. The extracted heartbeat-related source signal significantly
improves HR detection as compared to the original PPG. The absence of
pre-processing and the self-supervised nature of the proposed method, combined
with its strong performance, highlight the potential of BSS in biosignal
analysis.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:47:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Webster",
"Matthew B.",
""
],
[
"Lee",
"Dongheon",
""
],
[
"Lee",
"Joonnyong",
""
]
] |
2504.09133 | Franz Georg Fuchs | Franz G. Fuchs, Ruben P. Bassa | Compact Circuits for Constrained Quantum Evolutions of Sparse Operators | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We introduce a general framework for constructing compact quantum circuits
that implement the real-time evolution of Hamiltonians of the form $H = \sigma
P_B$, where $\sigma$ is a Pauli string commuting with a projection operator
$P_B$ onto a subspace of the computational basis. Such Hamiltonians frequently
arise in quantum algorithms, including constrained mixers in QAOA, fermionic
and excitation operators in VQE, and lattice gauge theory applications. Our
method emphasizes the minimization of non-transversal gates, particularly
T-gates, critical for fault-tolerant quantum computing. We construct circuits
requiring $\mathcal{O}(n|B|)$ CX gates and $\mathcal{O}{n |B| + \log(|B|) \log
(1/\epsilon)}$ T-gates, where $n$ is the number of qubits, $|B|$ the dimension
of the projected subspace, and $\epsilon$ the desired approximation precision.
For group-generated subspaces, we further reduce complexity to $\mathcal{O}(n
\log |B|)$ CX gates and $\mathcal{O}{n+\log(\frac{1}{\epsilon})}$ T gates. Our
constructive proofs yield explicit algorithms and include several applications,
such as improved transposition circuits, efficient implementations of fermionic
excitations, and oracle operators for combinatorial optimization.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:47:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fuchs",
"Franz G.",
""
],
[
"Bassa",
"Ruben P.",
""
]
] |
2504.09134 | Feilong Jing | Feilong Jing, Yang Deng, Boyi Wang, Xudong Zheng, Yifan Sun, Zhang
Chen and Bin Liang | Steady-State Drifting Equilibrium Analysis of Single-Track Two-Wheeled
Robots for Controller Design | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Drifting is an advanced driving technique where the wheeled robot's
tire-ground interaction breaks the common non-holonomic pure rolling
constraint. This allows high-maneuverability tasks like quick cornering, and
steady-state drifting control enhances motion stability under lateral slip
conditions. While drifting has been successfully achieved in four-wheeled robot
systems, its application to single-track two-wheeled (STTW) robots, such as
unmanned motorcycles or bicycles, has not been thoroughly studied. To bridge
this gap, this paper extends the drifting equilibrium theory to STTW robots and
reveals the mechanism behind the steady-state drifting maneuver. Notably, the
counter-steering drifting technique used by skilled motorcyclists is explained
through this theory. In addition, an analytical algorithm based on intrinsic
geometry and kinematics relationships is proposed, reducing the computation
time by four orders of magnitude while maintaining less than 6% error compared
to numerical methods. Based on equilibrium analysis, a model predictive
controller (MPC) is designed to achieve steady-state drifting and equilibrium
points transition, with its effectiveness and robustness validated through
simulations.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:48:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jing",
"Feilong",
""
],
[
"Deng",
"Yang",
""
],
[
"Wang",
"Boyi",
""
],
[
"Zheng",
"Xudong",
""
],
[
"Sun",
"Yifan",
""
],
[
"Chen",
"Zhang",
""
],
[
"Liang",
"Bin",
""
]
] |
2504.09135 | Haowei Lin | Haotian Ye, Himanshu Jain, Chong You, Ananda Theertha Suresh, Haowei
Lin, James Zou, Felix Yu | Efficient and Asymptotically Unbiased Constrained Decoding for Large
Language Models | null | AISTATS 2025 | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In real-world applications of large language models, outputs are often
required to be confined: selecting items from predefined product or document
sets, generating phrases that comply with safety standards, or conforming to
specialized formatting styles. To control the generation, constrained decoding
has been widely adopted. However, existing prefix-tree-based constrained
decoding is inefficient under GPU-based model inference paradigms, and it
introduces unintended biases into the output distribution. This paper
introduces Dynamic Importance Sampling for Constrained Decoding (DISC) with
GPU-based Parallel Prefix-Verification (PPV), a novel algorithm that leverages
dynamic importance sampling to achieve theoretically guaranteed asymptotic
unbiasedness and overcomes the inefficiency of prefix-tree. Extensive
experiments demonstrate the superiority of our method over existing methods in
both efficiency and output quality. These results highlight the potential of
our methods to improve constrained generation in applications where adherence
to specific constraints is essential.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:49:21 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ye",
"Haotian",
""
],
[
"Jain",
"Himanshu",
""
],
[
"You",
"Chong",
""
],
[
"Suresh",
"Ananda Theertha",
""
],
[
"Lin",
"Haowei",
""
],
[
"Zou",
"James",
""
],
[
"Yu",
"Felix",
""
]
] |
2504.09136 | Yo Toda | Yo Toda and Osamu Seto | Constraints on the simultaneous variation of the fine structure constant
and electron mass in light of DESI BAO data | 16 pages, 4 figures | null | null | EPHOU-25-005 | astro-ph.CO gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | We study the cosmological constraints of the time variation of the electron
mass $m_e$ and the fine-structure constant $\alpha$, using data of cosmic
microwave background, supernovae light curve and baryon acoustic oscillation
(BAO) data including the recent DESI BAO DR2 measurements. The results are
slightly depending on the BAO data set included in the analysis. The latest
DESI BAO DR2 data strongly indicates that $m_e$ or $\alpha$ is slightly larger
than the previous data from 6DF+SDSS and DESI BAO DR1. We also compare the
varying $m_e$ model, the varying $\alpha$ model, and the simultaneous variation
of $m_e$ and $\alpha$. When considering the Hubble tension, a larger electron
mass is the most promising option and the variation of the fine-structure
constants does not alleviate the tension.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:49:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Toda",
"Yo",
""
],
[
"Seto",
"Osamu",
""
]
] |
2504.09138 | Bokai Xu | Jiayao Yang, Jiayi Zhang, Bokai Xu, Jiakang Zheng, Zhilong Liu, Ziheng
Liu, Dusit Niyato, M\'erouane Debbah, Zhu Han, and Bo Ai | White-Box AI Model: Next Frontier of Wireless Communications | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | White-box AI (WAI), or explainable AI (XAI) model, a novel tool to achieve
the reasoning behind decisions and predictions made by the AI algorithms, makes
it more understandable and transparent. It offers a new approach to address key
challenges of interpretability and mathematical validation in traditional
black-box models. In this paper, WAI-aided wireless communication systems are
proposed and investigated thoroughly to utilize the promising capabilities.
First, we introduce the fundamental principles of WAI. Then, a detailed
comparison between WAI and traditional black-box model is conducted in terms of
optimization objectives and architecture design, with a focus on deep neural
networks (DNNs) and transformer networks. Furthermore, in contrast to the
traditional black-box methods, WAI leverages theory-driven causal modeling and
verifiable optimization paths, thereby demonstrating potential advantages in
areas such as signal processing and resource allocation. Finally, we outline
future research directions for the integration of WAI in wireless communication
systems.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:57:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Jiayao",
""
],
[
"Zhang",
"Jiayi",
""
],
[
"Xu",
"Bokai",
""
],
[
"Zheng",
"Jiakang",
""
],
[
"Liu",
"Zhilong",
""
],
[
"Liu",
"Ziheng",
""
],
[
"Niyato",
"Dusit",
""
],
[
"Debbah",
"Mérouane",
""
],
[
"Han",
"Zhu",
""
],
[
"Ai",
"Bo",
""
]
] |
2504.09139 | Konstantin Osipenko | K. Yu. Osipenko | Exact inequalities and optimal recovery by inaccurate information | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | The paper considers a multidimensional problem of optimal recovery of an
operator whose action is represented by multiplying the original function by a
weight function of a special type, based on inaccurately specified information
about the values of operators of a similar type. An exact inequality for the
norms of such operators is obtained. The problem under consideration is a
generalization of the problem of optimal recovery of a derivative based on
other inaccurately specified derivatives in the space $\mathbb R^d$ and the
problem of an exact inequality, which is an analogue of the
Hardy-Littlewood-Polya inequality.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 09:03:29 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Osipenko",
"K. Yu.",
""
]
] |
2504.09140 | Peiyi He | Peiyi He, Jiade Li, Cong Li, Ning Li, Bo Han, Ruochen Shi, Ruishi Qi,
Jinlong Du, Pu Yu, Peng Gao | Strongly confined Mid-infrared to Terahertz Phonon Polaritons in
Ultra-thin SrTiO3 | null | null | null | null | physics.optics cond-mat.mtrl-sci physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Surface phonon polaritons (SPhPs) have emerged as a promising platform for
subwavelength optical manipulation, offering distinct advantages for
applications in infrared sensing, imaging, and optoelectronic devices. However,
the narrow Reststrahlen bands of conventional polar materials impose
significant limitations on their applications across the mid-infrared (MIR) to
terahertz (THz) range. Addressing this challenge requires the development of
materials capable of supporting SPhPs with broad spectral range, strong field
confinement, slow group velocity, and high quality factor. Here, using
monochromatic electron energy-loss spectroscopy in a scanning transmission
electron microscope, we demonstrate that ultra-thin SrTiO3 membranes encompass
the exceptional properties mentioned above that have long been sought after.
Systematic measurements across varying membrane thicknesses reveal two distinct
SPhP branches characterized by wide spectral dispersion, high field
confinement, and anomalously slow group velocities spanning from the MIR (68 ~
99 meV) to THz (12 ~ 59 meV) range. Notably, in membranes approaching ~ 3 nm
thickness (~ 8 unit cells), these polaritons exhibit unprecedented confinement
factors exceeding 500 and group velocities as low as ~ 7 * 10-5 c, rivaling the
best-performing van der Waals materials. These findings establish perovskite
oxide such as SrTiO3 as a versatile platform for tailoring light-matter
interactions at the nanoscale, providing critical insights for the design of
next-generation photonic devices requiring broadband operation and enhanced
optical confinement.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 09:04:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Peiyi",
""
],
[
"Li",
"Jiade",
""
],
[
"Li",
"Cong",
""
],
[
"Li",
"Ning",
""
],
[
"Han",
"Bo",
""
],
[
"Shi",
"Ruochen",
""
],
[
"Qi",
"Ruishi",
""
],
[
"Du",
"Jinlong",
""
],
[
"Yu",
"Pu",
""
],
[
"Gao",
"Peng",
""
]
] |
2504.09141 | Andres Contreras Hip | Andres A. Contreras Hip, Zijie Zhuang | Bounds on the distance exponent for higher-dimensional Liouville first
passage percolation | null | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | For $\xi \geq 0$ and $d \geq 3$, the higher-dimensional Liouville first
passage percolation (LFPP) is a random metric on $\epsilon \mathbb{Z}^d$
obtained by reweighting each vertex by $e^{\xi h_\epsilon(x)}$, where
$h_\epsilon(x)$ is a continuous mollification of the whole-space log-correlated
Gaussian field. This metric generalizes the two-dimensional LFPP, which is
related to Liouville quantum gravity. We derive several estimates for the
set-to-set distance exponent of this metric, including upper and lower bounds
and bounds on its derivative with respect to $\xi$. In the subcritical region
for $\xi$, we derive estimates for the fractal dimension and show that it is
continuous and strictly increasing with respect to $\xi$. In particular, our
result is an important step towards proving a technical assumption made in
previous work by the first author and Gwynne. These are also the first bounds
on the distance exponent for LFPP in higher dimensions.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 09:07:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hip",
"Andres A. Contreras",
""
],
[
"Zhuang",
"Zijie",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.