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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.14728 | Vitaly Vanchurin | Vitaly Vanchurin | Geometric Learning Dynamics | 15 pages | null | null | null | cs.LG q-bio.PE quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a unified geometric framework for modeling learning dynamics in
physical, biological, and machine learning systems. The theory reveals three
fundamental regimes, each emerging from the power-law relationship $g \propto
\kappa^a$ between the metric tensor $g$ in the space of trainable variables and
the noise covariance matrix $\kappa$. The quantum regime corresponds to $a = 1$
and describes Schr\"odinger-like dynamics that emerges from a discrete shift
symmetry. The efficient learning regime corresponds to $a = \tfrac{1}{2}$ and
describes very fast machine learning algorithms. The equilibration regime
corresponds to $a = 0$ and describes classical models of biological evolution.
We argue that the emergence of the intermediate regime $a = \tfrac{1}{2}$ is a
key mechanism underlying the emergence of biological complexity.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 19:56:41 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Vanchurin",
"Vitaly",
""
]
] |
2504.14730 | Atefeh Gilani | Atefeh Gilani, Juan Felipe Gomez, Shahab Asoodeh, Flavio P. Calmon,
Oliver Kosut, Lalitha Sankar | Optimal Additive Noise Mechanisms for Differential Privacy | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | We propose a unified optimization framework for designing continuous and
discrete noise distributions that ensure differential privacy (DP) by
minimizing R\'enyi DP, a variant of DP, under a cost constraint. R\'enyi DP has
the advantage that by considering different values of the R\'enyi parameter
$\alpha$, we can tailor our optimization for any number of compositions. To
solve the optimization problem, we reduce it to a finite-dimensional convex
formulation and perform preconditioned gradient descent. The resulting noise
distributions are then compared to their Gaussian and Laplace counterparts.
Numerical results demonstrate that our optimized distributions are consistently
better, with significant improvements in $(\varepsilon, \delta)$-DP guarantees
in the moderate composition regimes, compared to Gaussian and Laplace
distributions with the same variance.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:04:41 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Gilani",
"Atefeh",
""
],
[
"Gomez",
"Juan Felipe",
""
],
[
"Asoodeh",
"Shahab",
""
],
[
"Calmon",
"Flavio P.",
""
],
[
"Kosut",
"Oliver",
""
],
[
"Sankar",
"Lalitha",
""
]
] |
2504.14731 | Kai Fu | Kai Fu | Siegel-Veech Measures of Convex Flat Cone Spheres | null | null | null | null | math.GT math.DG | http://creativecommons.org/licenses/by/4.0/ | A classical theorem of Siegel gives the average number of lattice points in
bounded subsets of $\mathbb{R}^n$. Motivated by this result, Veech introduced
an analogue for translation surfaces, known as the Siegel-Veech formula, which
describes the average number of saddle connections of bounded length on the
moduli space of translation surfaces. However, no such formula is known for
flat surfaces with cone angles that are irrational multiples of $\pi$.
A convex flat cone sphere is a Riemann sphere equipped with a conformal flat
metric with conical singularities, all of whose cone angles lie in the interval
$(0, 2\pi)$. In this paper, we extend the Siegel-Veech formula to this setting.
We define a generalized Siegel-Veech transform and prove that it belongs to
$L^\infty$ on the moduli space. This leads to the definition of a Siegel-Veech
measure on $\mathbb{R}_{>0}$, obtained by integrating the Siegel-Veech
transform over the moduli space. This measure can be viewed as a generalization
of the classical Siegel-Veech formula. We show that it is absolutely continuous
and piecewise real analytic.
Finally, we study the asymptotic behavior of this measure on small intervals
$(0,\varepsilon)$ as $\varepsilon \to 0$, providing an analogue of Siegel-Veech
constants for convex flat cone spheres.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:06:55 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Fu",
"Kai",
""
]
] |
2504.14733 | Lingyun Ding | Lingyun Ding | Long-Time Asymptotics of Passive Scalar Transport in Periodically
Modulated Channels | null | null | null | null | physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work investigates the long-time asymptotic behavior of a diffusing
passive scalar advected by fluid flow in a straight channel with a periodically
varying cross-section. The goal is to derive an asymptotic expansion for the
scalar field and estimate the timescale over which this expansion remains
valid, thereby generalizing Taylor dispersion theory to periodically modulated
channels. By formulating the eigenvalue problem for the advection-diffusion
operator on a unit cell and employing a biorthogonal eigenfunction expansion,
we generalize the classical Fourier integral used in the flat-channel case to
obtain an integral representation of the scalar field. This representation
reveals a slow manifold that governs the algebraically decaying dynamics, while
the difference between the scalar field and the slow manifold decays
exponentially in time. Building on this, we derive a long-time asymptotic
expansion of the scalar field. We show that the validity timescale of the
expansion is determined by the real part of the eigenvalues of a modified
advection-diffusion operator, which depends solely on the flow and geometry
within a single unit cell. This framework offers a rigorous and systematic
method for estimating mixing timescales in channels with complex geometries. We
show that non-flat channel boundaries tend to increase the timescale, while
transverse velocity components tend to decrease it. The approach developed here
is broadly applicable and can be extended to derive long-time asymptotics for
other systems with periodic coefficients or periodic microstructures.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:13:27 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ding",
"Lingyun",
""
]
] |
2504.14734 | Nicola Bamonti | Nicola Bamonti | Apples Falling, Buckets Rolling, and Why Inertia Keeps Trolling | null | null | null | null | physics.hist-ph gr-qc | http://creativecommons.org/licenses/by/4.0/ | Inertia is the most fundamental and yet elusive concept in physics. Modern
geniuses such as Newton and Einstein have wondered about its nature, but the
last word has yet to be said and "the question of questions" still remains:
what is inertia? In the paper, analysing the properties and criticalities of
the concept of geodesic, I propose to separate the notion of geodesic inertial
motion from that of natural motion, which I identify as motion characterised by
a generalised version of the geodesic deviation. This strategy succeeds in
shedding new light on the motion of bodies.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:33:33 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Bamonti",
"Nicola",
""
]
] |
2504.14735 | Chin-Yun Yu | Chin-Yun Yu, Marco A. Mart\'inez-Ram\'irez, Junghyun Koo, Ben Hayes,
Wei-Hsiang Liao, Gy\"orgy Fazekas, Yuki Mitsufuji | DiffVox: A Differentiable Model for Capturing and Analysing Professional
Effects Distributions | Submitted to DAFx 2025 | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study introduces a novel and interpretable model, DiffVox, for matching
vocal effects in music production. DiffVox, short for ``Differentiable Vocal
Fx", integrates parametric equalisation, dynamic range control, delay, and
reverb with efficient differentiable implementations to enable gradient-based
optimisation for parameter estimation. Vocal presets are retrieved from two
datasets, comprising 70 tracks from MedleyDB and 365 tracks from a private
collection. Analysis of parameter correlations highlights strong relationships
between effects and parameters, such as the high-pass and low-shelf filters
often behaving together to shape the low end, and the delay time correlates
with the intensity of the delayed signals. Principal component analysis reveals
connections to McAdams' timbre dimensions, where the most crucial component
modulates the perceived spaciousness while the secondary components influence
spectral brightness. Statistical testing confirms the non-Gaussian nature of
the parameter distribution, highlighting the complexity of the vocal effects
space. These initial findings on the parameter distributions set the foundation
for future research in vocal effects modelling and automatic mixing. Our source
code and datasets are accessible at https://github.com/SonyResearch/diffvox.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:52:58 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yu",
"Chin-Yun",
""
],
[
"Martínez-Ramírez",
"Marco A.",
""
],
[
"Koo",
"Junghyun",
""
],
[
"Hayes",
"Ben",
""
],
[
"Liao",
"Wei-Hsiang",
""
],
[
"Fazekas",
"György",
""
],
[
"Mitsufuji",
"Yuki",
""
]
] |
2504.14736 | Nicol\'as Gaggion Ph.D. | Nicol\'as Gaggion, Rodrigo Bonazzola, Mar\'ia Florencia Legascue,
Mar\'ia Florencia Mammarella, Florencia Sol Rodriguez, Federico Emanuel
Aballay, Florencia Bel\'en Catulo, Andana Barrios, Franco Accavallo, Santiago
Nahuel Villarreal, Martin Crespi, Martiniano Mar\'ia Ricardi, Ezequiel
Petrillo, Thomas Blein, Federico Ariel, Enzo Ferrante | ChronoRoot 2.0: An Open AI-Powered Platform for 2D Temporal Plant
Phenotyping | null | null | null | null | cs.CV q-bio.QM | http://creativecommons.org/licenses/by-sa/4.0/ | The analysis of plant developmental plasticity, including root system
architecture, is fundamental to understanding plant adaptability and
development, particularly in the context of climate change and agricultural
sustainability. While significant advances have been made in plant phenotyping
technologies, comprehensive temporal analysis of root development remains
challenging, with most existing solutions providing either limited throughput
or restricted structural analysis capabilities. Here, we present ChronoRoot
2.0, an integrated open-source platform that combines affordable hardware with
advanced artificial intelligence to enable sophisticated temporal plant
phenotyping. The system introduces several major advances, offering an integral
perspective of seedling development: (i) simultaneous multi-organ tracking of
six distinct plant structures, (ii) quality control through real-time
validation, (iii) comprehensive architectural measurements including novel
gravitropic response parameters, and (iv) dual specialized user interfaces for
both architectural analysis and high-throughput screening. We demonstrate the
system's capabilities through three use cases for Arabidopsis thaliana:
characterization of circadian growth patterns under different light conditions,
detailed analysis of gravitropic responses in transgenic plants, and
high-throughput screening of etiolation responses across multiple genotypes.
ChronoRoot 2.0 maintains its predecessor's advantages of low cost and
modularity while significantly expanding its capabilities, making sophisticated
temporal phenotyping more accessible to the broader plant science community.
The system's open-source nature, combined with extensive documentation and
containerized deployment options, ensures reproducibility and enables
community-driven development of new analytical capabilities.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:56:25 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Gaggion",
"Nicolás",
""
],
[
"Bonazzola",
"Rodrigo",
""
],
[
"Legascue",
"María Florencia",
""
],
[
"Mammarella",
"María Florencia",
""
],
[
"Rodriguez",
"Florencia Sol",
""
],
[
"Aballay",
"Federico Emanuel",
""
],
[
"Catulo",
"Florencia Belén",
""
],
[
"Barrios",
"Andana",
""
],
[
"Accavallo",
"Franco",
""
],
[
"Villarreal",
"Santiago Nahuel",
""
],
[
"Crespi",
"Martin",
""
],
[
"Ricardi",
"Martiniano María",
""
],
[
"Petrillo",
"Ezequiel",
""
],
[
"Blein",
"Thomas",
""
],
[
"Ariel",
"Federico",
""
],
[
"Ferrante",
"Enzo",
""
]
] |
2504.14737 | Shuang Zeng | Shuang Zeng and Lei Zhu and Xinliang Zhang and Hangzhou He and Yanye
Lu | SuperCL: Superpixel Guided Contrastive Learning for Medical Image
Segmentation Pre-training | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Medical image segmentation is a critical yet challenging task, primarily due
to the difficulty of obtaining extensive datasets of high-quality,
expert-annotated images. Contrastive learning presents a potential but still
problematic solution to this issue. Because most existing methods focus on
extracting instance-level or pixel-to-pixel representation, which ignores the
characteristics between intra-image similar pixel groups. Moreover, when
considering contrastive pairs generation, most SOTA methods mainly rely on
manually setting thresholds, which requires a large number of gradient
experiments and lacks efficiency and generalization. To address these issues,
we propose a novel contrastive learning approach named SuperCL for medical
image segmentation pre-training. Specifically, our SuperCL exploits the
structural prior and pixel correlation of images by introducing two novel
contrastive pairs generation strategies: Intra-image Local Contrastive Pairs
(ILCP) Generation and Inter-image Global Contrastive Pairs (IGCP) Generation.
Considering superpixel cluster aligns well with the concept of contrastive
pairs generation, we utilize the superpixel map to generate pseudo masks for
both ILCP and IGCP to guide supervised contrastive learning. Moreover, we also
propose two modules named Average SuperPixel Feature Map Generation (ASP) and
Connected Components Label Generation (CCL) to better exploit the prior
structural information for IGCP. Finally, experiments on 8 medical image
datasets indicate our SuperCL outperforms existing 12 methods. i.e. Our SuperCL
achieves a superior performance with more precise predictions from
visualization figures and 3.15%, 5.44%, 7.89% DSC higher than the previous best
results on MMWHS, CHAOS, Spleen with 10% annotations. Our code will be released
after acceptance.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 20:57:03 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zeng",
"Shuang",
""
],
[
"Zhu",
"Lei",
""
],
[
"Zhang",
"Xinliang",
""
],
[
"He",
"Hangzhou",
""
],
[
"Lu",
"Yanye",
""
]
] |
2504.14738 | Reya Vir | Reya Vir, Shreya Shankar, Harrison Chase, Will Fu-Hinthorn, Aditya
Parameswaran | PROMPTEVALS: A Dataset of Assertions and Guardrails for Custom
Production Large Language Model Pipelines | Accepted to NAACL 2025 Main Conference | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are increasingly deployed in specialized
production data processing pipelines across diverse domains -- such as finance,
marketing, and e-commerce. However, when running them in production across many
inputs, they often fail to follow instructions or meet developer expectations.
To improve reliability in these applications, creating assertions or guardrails
for LLM outputs to run alongside the pipelines is essential. Yet, determining
the right set of assertions that capture developer requirements for a task is
challenging. In this paper, we introduce PROMPTEVALS, a dataset of 2087 LLM
pipeline prompts with 12623 corresponding assertion criteria, sourced from
developers using our open-source LLM pipeline tools. This dataset is 5x larger
than previous collections. Using a hold-out test split of PROMPTEVALS as a
benchmark, we evaluated closed- and open-source models in generating relevant
assertions. Notably, our fine-tuned Mistral and Llama 3 models outperform
GPT-4o by 20.93% on average, offering both reduced latency and improved
performance. We believe our dataset can spur further research in LLM
reliability, alignment, and prompt engineering.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:04:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Vir",
"Reya",
""
],
[
"Shankar",
"Shreya",
""
],
[
"Chase",
"Harrison",
""
],
[
"Fu-Hinthorn",
"Will",
""
],
[
"Parameswaran",
"Aditya",
""
]
] |
2504.14739 | Arpit Agarwal Mr. | Arpit Agarwal (1), Mohammad Amin Mirzaee (2), Xiping Sun (2) and
Wenzhen Yuan (2) ((1) Carnegie Mellon University, USA, (2) University of
Illinois Urbana-Champaign, USA) | A Modularized Design Approach for GelSight Family of Vision-based
Tactile Sensors | The paper is accepted to International Journal of Robotics Research
with DOI 10.1177/02783649251339680 | null | 10.1177/02783649251339680 | null | cs.RO cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | GelSight family of vision-based tactile sensors has proven to be effective
for multiple robot perception and manipulation tasks. These sensors are based
on an internal optical system and an embedded camera to capture the deformation
of the soft sensor surface, inferring the high-resolution geometry of the
objects in contact. However, customizing the sensors for different robot hands
requires a tedious trial-and-error process to re-design the optical system. In
this paper, we formulate the GelSight sensor design process as a systematic and
objective-driven design problem and perform the design optimization with a
physically accurate optical simulation. The method is based on modularizing and
parameterizing the sensor's optical components and designing four generalizable
objective functions to evaluate the sensor. We implement the method with an
interactive and easy-to-use toolbox called OptiSense Studio. With the toolbox,
non-sensor experts can quickly optimize their sensor design in both forward and
inverse ways following our predefined modules and steps. We demonstrate our
system with four different GelSight sensors by quickly optimizing their initial
design in simulation and transferring it to the real sensors.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:07:41 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Agarwal",
"Arpit",
""
],
[
"Mirzaee",
"Mohammad Amin",
""
],
[
"Sun",
"Xiping",
""
],
[
"Yuan",
"Wenzhen",
""
]
] |
2504.14740 | Elena Bratkovskaya | Marcus Bleicher and Elena Bratkovskaya | The quest for the quark-gluon plasma from the perspective of dynamical
models of relativistic heavy-ion collisions | 12 pages, 9 figures; contribution to the Astronomische Nachrichten,
Special issue: IWARA2024 Reviews | null | 10.1002/asna.70006 | null | nucl-th hep-ex hep-ph nucl-ex | http://creativecommons.org/licenses/by/4.0/ | The physics of heavy-ion collisions is one of the most exciting and
challenging directions of science for the last four decades. On the theoretical
side one deals with a non-abelian field theory, while on the experimental side
today's largest accelerators are needed to enable these studies. The discovery
of a new stage of matter - called the quark-gluon plasma (QGP) - and the study
of its properties is one of the major achievements of modern physics. In this
contribution we briefly review the history of theoretical descriptions of
heavy-ion collisions based on dynamical models, focusing on the personal
experiences in this inspiring field.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:07:45 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Bleicher",
"Marcus",
""
],
[
"Bratkovskaya",
"Elena",
""
]
] |
2504.14741 | Namrata Vaswani | Namrata Vaswani | AltGDmin: Alternating GD and Minimization for Partly-Decoupled
(Federated) Optimization | To appear in Foundations and Trends in Optimization (NOW publishers) | null | null | null | cs.LG math.OC stat.ME | http://creativecommons.org/licenses/by/4.0/ | This article describes a novel optimization solution framework, called
alternating gradient descent (GD) and minimization (AltGDmin), that is useful
for many problems for which alternating minimization (AltMin) is a popular
solution. AltMin is a special case of the block coordinate descent algorithm
that is useful for problems in which minimization w.r.t one subset of variables
keeping the other fixed is closed form or otherwise reliably solved. Denote the
two blocks/subsets of the optimization variables Z by Za, Zb, i.e., Z = {Za,
Zb}. AltGDmin is often a faster solution than AltMin for any problem for which
(i) the minimization over one set of variables, Zb, is much quicker than that
over the other set, Za; and (ii) the cost function is differentiable w.r.t. Za.
Often, the reason for one minimization to be quicker is that the problem is
``decoupled" for Zb and each of the decoupled problems is quick to solve. This
decoupling is also what makes AltGDmin communication-efficient for federated
settings.
Important examples where this assumption holds include (a) low rank
column-wise compressive sensing (LRCS), low rank matrix completion (LRMC), (b)
their outlier-corrupted extensions such as robust PCA, robust LRCS and robust
LRMC; (c) phase retrieval and its sparse and low-rank model based extensions;
(d) tensor extensions of many of these problems such as tensor LRCS and tensor
completion; and (e) many partly discrete problems where GD does not apply --
such as clustering, unlabeled sensing, and mixed linear regression. LRCS finds
important applications in multi-task representation learning and few shot
learning, federated sketching, and accelerated dynamic MRI. LRMC and robust PCA
find important applications in recommender systems, computer vision and video
analytics.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:07:59 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Vaswani",
"Namrata",
""
]
] |
2504.14742 | Jeongheon Choe | Jeongheon Choe, David Lujan, Gaihua Ye, Cynthia Nnokwe, Bowen Ma,
Jiaming He, Frank Y. Gao, T. Nathan Nunley, Aritz Leonardo, Mikel
Arruabarrena, Andres Ayuela, Jianshi Zhou, Martin Rodriguez-Vega, Gregory A.
Fiete, Rui He, and Xiaoqin Li | Long-lived Zone-boundary Magnons in an Antiferromagnet | 31 pages, 4 figures | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Antiferromagnetic (AFM) insulators exhibit many desirable features for
spintronic applications such as fast dynamics in the THz range and robustness
to fluctuating external fields. However, large damping typically associated
with THz magnons presents a serious challenge for THz magnonic applications.
Here, we report long-lived short-wavelength zone boundary magnons in the
honeycomb AFM insulator CoTiO3, recently found to host topological magnons. We
find that its zone-boundary THz magnons exhibit longer lifetimes than its
zone-center magnons. This unusual momentum-dependent long magnon lifetime
originates from several factors including the antiferromagnetic order, exchange
anisotropy, a finite magnon gap, and magnon band dispersion. Our work suggests
that magnon-magnon interaction may not be detrimental to magnon lifetimes and
should be included in future searches for topological magnons.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:08:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Choe",
"Jeongheon",
""
],
[
"Lujan",
"David",
""
],
[
"Ye",
"Gaihua",
""
],
[
"Nnokwe",
"Cynthia",
""
],
[
"Ma",
"Bowen",
""
],
[
"He",
"Jiaming",
""
],
[
"Gao",
"Frank Y.",
""
],
[
"Nunley",
"T. Nathan",
""
],
[
"Leonardo",
"Aritz",
""
],
[
"Arruabarrena",
"Mikel",
""
],
[
"Ayuela",
"Andres",
""
],
[
"Zhou",
"Jianshi",
""
],
[
"Rodriguez-Vega",
"Martin",
""
],
[
"Fiete",
"Gregory A.",
""
],
[
"He",
"Rui",
""
],
[
"Li",
"Xiaoqin",
""
]
] |
2504.14743 | Elizabeth Stephenson | Herbert Edelsbrunner, Elizabeth Stephenson, Martin Hafskjold Thoresen | The Mid-sphere Cousin of the Medial Axis Transform | null | null | null | null | cs.CG math.AT | http://creativecommons.org/licenses/by/4.0/ | The medial axis of a smoothly embedded surface in $\mathbb{R}^3$ consists of
all points for which the Euclidean distance function on the surface has at
least two minima. We generalize this notion to the mid-sphere axis, which
consists of all points for which the Euclidean distance function has two
interchanging saddles that swap their partners in the pairing by persistent
homology. It offers a discrete-algebraic multi-scale approach to computing
ridge-like structures on the surface. As a proof of concept, an algorithm that
computes stair-case approximations of the mid-sphere axis is provided.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:24:41 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Edelsbrunner",
"Herbert",
""
],
[
"Stephenson",
"Elizabeth",
""
],
[
"Thoresen",
"Martin Hafskjold",
""
]
] |
2504.14744 | Mustafa Cavus | Yigitcan Yard{\i}mc{\i} and Mustafa Cavus | On the Tunability of Random Survival Forests Model for Predictive
Maintenance | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the tunability of the Random Survival Forest (RSF)
model in predictive maintenance, where accurate time-to-failure estimation is
crucial. Although RSF is widely used due to its flexibility and ability to
handle censored data, its performance is sensitive to hyperparameter
configurations. However, systematic evaluations of RSF tunability remain
limited, especially in predictive maintenance contexts. We introduce a
three-level framework to quantify tunability: (1) a model-level metric
measuring overall performance gain from tuning, (2) a hyperparameter-level
metric assessing individual contributions, and (3) identification of optimal
tuning ranges. These metrics are evaluated across multiple datasets using
survival-specific criteria: the C-index for discrimination and the Brier score
for calibration. Experiments on four CMAPSS dataset subsets, simulating
aircraft engine degradation, reveal that hyperparameter tuning consistently
improves model performance. On average, the C-index increased by 0.0547, while
the Brier score decreased by 0.0199. These gains were consistent across all
subsets. Moreover, ntree and mtry showed the highest average tunability, while
nodesize offered stable improvements within the range of 10 to 30. In contrast,
splitrule demonstrated negative tunability on average, indicating that improper
tuning may reduce model performance. Our findings emphasize the practical
importance of hyperparameter tuning in survival models and provide actionable
insights for optimizing RSF in real-world predictive maintenance applications.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:27:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yardımcı",
"Yigitcan",
""
],
[
"Cavus",
"Mustafa",
""
]
] |
2504.14745 | Rawlings Ntassah | Rawlings Ntassah, Gian Michele Dell'Aera, Fabrizio Granelli | Interference-Aware PMI selection for MIMO systems in an O-RAN scenario | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The optimization of Precoding Matrix Indicators (PMIs) is crucial for
enhancing the performance of 5G networks, particularly in dense deployments
where inter-cell interference is a significant challenge. Some approaches have
leveraged AI/ML techniques for beamforming and beam selection, however, these
methods often overlook the multi-objective nature of PMI selection, which
requires balancing spectral efficiency (SE) and interference reduction. This
paper proposes an interference-aware PMI selection method using an Advantage
Actor-Critic (A2C) reinforcement learning model, designed for deployment within
an O-RAN framework as an xApp. The proposed model prioritizes user equipment
(UE) based on a novel strategy and adjusts PMI values accordingly, with
interference management and efficient resource utilization. Experimental
results in an O-RAN environment demonstrate the approach's effectiveness in
improving network performance metrics, including SE and interference
mitigation.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:33:07 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ntassah",
"Rawlings",
""
],
[
"Dell'Aera",
"Gian Michele",
""
],
[
"Granelli",
"Fabrizio",
""
]
] |
2504.14746 | Timo Kist | Timo Kist, Joseph F. Hennawi, Frederick B. Davies | A local, topology-independent parameterization of quasar IGM damping
wings | Submitted to MNRAS | null | null | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Lyman-$\alpha$ damping wings towards quasars provide a unique probe of
reionization because their strength correlates strongly with the global
volume-averaged neutral hydrogen (HI) fraction of the intergalactic medium
(IGM). Cosmic variance in the IGM, however, is a major source of stochasticity
since the local neutral environment around a quasar varies significantly even
at fixed global neutral fraction. We show that the IGM damping wing carries
additional information about this local ionization topology, unexploited by
current analysis frameworks. We introduce a set of two new physically motivated
summary statistics encoding the local information about the HI distribution in
the IGM before it is altered by ionization radiation from the quasar,
encompassing 1) the HI column density, weighted by a Lorentzian profile
mimicking the frequency dependence of the Lyman-$\alpha$ cross section, and 2)
the distance from the quasar to the first neutral patch. This description, when
combined with the quasar's lifetime as a third parameter, reduces the IGM
transmission scatter in the damping wing region of the spectrum to $\lesssim
1\,\%$ across the full range of physical parameter space. We introduce a simple
procedure for generating synthetic HI sightlines around quasars and demonstrate
that the resulting damping wing profiles are statistically indistinguishable
from a realistic reionization topology. This opens the door for optimally
extracting the salient local information encoded in the imprint in a
model-independent fashion. In the context of a specific reionization model,
measurements of these local parameters can be translated into constraints on
the global timing of reionization, but in addition, they provide information
about the reionization topology, hitherto unused. A marginally modified version
of our framework can also be employed in the context of damping wings towards
galaxies.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:40:03 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Kist",
"Timo",
""
],
[
"Hennawi",
"Joseph F.",
""
],
[
"Davies",
"Frederick B.",
""
]
] |
2504.14747 | Zhen Tian | Qinghao Li, Zhen Tian, Xiaodan Wang, Jinming Yang, Zhihao Lin | Adaptive Field Effect Planner for Safe Interactive Autonomous Driving on
Curved Roads | The 45th IEEE International Conference on Distributed Computing
Systems Workshop (ICDCSW) has accepted this paper
(https://icdcs2025.icdcs.org/accepted-papers/ In Conjunction Events/ Page 4/
Number 175) | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Autonomous driving has garnered significant attention for its potential to
improve safety, traffic efficiency, and user convenience. However, the dynamic
and complex nature of interactive driving poses significant challenges,
including the need to navigate non-linear road geometries, handle dynamic
obstacles, and meet stringent safety and comfort requirements. Traditional
approaches, such as artificial potential fields (APF), often fall short in
addressing these complexities independently, necessitating the development of
integrated and adaptive frameworks. This paper presents a novel approach to
autonomous vehicle navigation that integrates artificial potential fields,
Frenet coordinates, and improved particle swarm optimization (IPSO). A dynamic
risk field, adapted from traditional APF, is proposed to ensure interactive
safety by quantifying risks and dynamically adjusting lane-changing intentions
based on surrounding vehicle behavior. Frenet coordinates are utilized to
simplify trajectory planning on non-straight roads, while an enhanced quintic
polynomial trajectory generator ensures smooth and comfortable path
transitions. Additionally, an IPSO algorithm optimizes trajectory selection in
real time, balancing safety and user comfort within a feasible input range. The
proposed framework is validated through extensive simulations and real-world
scenarios, demonstrating its ability to navigate complex traffic environments,
maintain safety margins, and generate smooth, dynamically feasible
trajectories.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:41:19 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Li",
"Qinghao",
""
],
[
"Tian",
"Zhen",
""
],
[
"Wang",
"Xiaodan",
""
],
[
"Yang",
"Jinming",
""
],
[
"Lin",
"Zhihao",
""
]
] |
2504.14748 | Krishna Padmanabhan | Krishna Padmanabhan, Cyrus Mehta | Statistical Design and Planning of an Adaptive Trial using Hierarchical
Composite Outcomes: A Practical example | null | null | null | null | stat.ME | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Hierarchical composite endpoints, such as those analyzed using the
Finkelstein-Schoenfeld (FS) statistic, are increasingly used in clinical trials
for their ability to incorporate clinically prioritized outcomes. However,
adaptive design methods for these endpoints remain underdeveloped. This paper
presents a practical framework for implementing sample size re-estimation (SSR)
in trials using hierarchical composites, motivated by a cardiovascular trial
with mortality, hospitalization, and a functional response as prioritized
endpoints. We use a two-stage adaptive design with a single interim analysis
for illustration. The interim analysis incorporates predictive probabilities to
determine whether the trial should stop for futility, continue as planned, or
increase the sample size to maintain power. The decision framework is based on
predefined zones for predictive probability, with corresponding adjustments to
the stage 2 sample size. Simulation studies across various treatment scenarios
demonstrate strong type I error control and increased power compared to a fixed
design, particularly for treatment effects that are clinically relevant but
lower than the alternative hypothesis. We also explore an alternative
conditional power approach for SSR, offering further sample size optimization.
Our results support the use of SSR with hierarchical composite outcomes using
an FS statistic, enhancing trial efficiency.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:41:29 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Padmanabhan",
"Krishna",
""
],
[
"Mehta",
"Cyrus",
""
]
] |
2504.14749 | Rawlings Ntassah | Rawlings Ntassah, Gian Michele Dell'Aera, Fabrizio Granelli | PPO-EPO: Energy and Performance Optimization for O-RAN Using
Reinforcement Learning | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Energy consumption in mobile communication networks has become a significant
challenge due to its direct impact on Capital Expenditure (CAPEX) and
Operational Expenditure (OPEX). The introduction of Open RAN (O-RAN) enables
telecommunication providers to leverage network intelligence to optimize energy
efficiency while maintaining Quality of Service (QoS). One promising approach
involves traffic-aware cell shutdown strategies, where underutilized cells are
selectively deactivated without compromising overall network performance.
However, achieving this balance requires precise traffic steering mechanisms
that account for throughput performance, power efficiency, and network
interference constraints.
This work proposes a reinforcement learning (RL) model based on the Proximal
Policy Optimization (PPO) algorithm to optimize traffic steering and energy
efficiency. The objective is to maximize energy efficiency and performance
gains while strategically shutting down underutilized cells. The proposed RL
model learns adaptive policies to make optimal shutdown decisions by
considering throughput degradation constraints, interference thresholds, and
PRB utilization balance. Experimental validation using TeraVM Viavi RIC tester
data demonstrates that our method significantly improves the network's energy
efficiency and downlink throughput.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:42:25 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ntassah",
"Rawlings",
""
],
[
"Dell'Aera",
"Gian Michele",
""
],
[
"Granelli",
"Fabrizio",
""
]
] |
2504.14750 | Zhen Tian | Yaoze Liu, Zhen Tian, Jinming Yang, Zhihao Lin | Data-Driven Evolutionary Game-Based Model Predictive Control for Hybrid
Renewable Energy Dispatch in Autonomous Ships | This paper has been accepted by the 2025 4th International Conference
on New Energy System and Power Engineering (NESP 2025) | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose a data-driven Evolutionary Game-Based Model
Predictive Control (EG-MPC) framework for the energy dispatch of a hybrid
renewable energy system powering an autonomous ship. The system integrates
solar photovoltaic and wind turbine generation with battery energy storage and
diesel backup power to ensure reliable operation. Given the uncertainties in
renewable generation and dynamic energy demands, an optimal dispatch strategy
is crucial to minimize operational costs while maintaining system reliability.
To address these challenges, we formulate a cost minimization problem that
considers both battery degradation costs and diesel fuel expenses, leveraging
real-world data to enhance modeling accuracy. The EG-MPC approach integrates
evolutionary game dynamics within a receding-horizon optimization framework,
enabling adaptive and near-optimal control solutions in real time. Simulation
results based on site-specific data demonstrate that the proposed method
achieves cost-effective, reliable, and adaptive energy dispatch, outperforming
conventional rule-based and standard MPC approaches, particularly under
uncertainty.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 21:51:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Liu",
"Yaoze",
""
],
[
"Tian",
"Zhen",
""
],
[
"Yang",
"Jinming",
""
],
[
"Lin",
"Zhihao",
""
]
] |
2504.14751 | Jianyu Zhang | Jianyu Zhang | AI for the Open-World: the Learning Principles | PhD thesis. This is not a compilation of published papers, but a new
one | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | During the past decades, numerous successes of AI has been made on "specific
capabilities", named closed-world, such as artificial environments or specific
real-world tasks. This well-defined narrow capability brings two nice benefits,
a clear criterion of success and the opportunity to collect a lot of examples.
The criteria not only reveal whether a machine has achieved a goal, but reveal
how the machine falls short of the goal. As a result, human designers can fix
the problems one after the other until the machine is deemed good enough for
the task. Furthermore, the large set of collected examples reduces the
difficulty of this problem-fixing process (by the central limit theorem).
Do the success in closed-world translate into broad open-world, where a
machine is required to perform any task that a human could possibly undertake
with fewer examples and less priori knowledge from human designers? No. Because
competence in a specific task provides little insight in handling other tasks,
the valuable criteria for specific tasks become helpless when handling broader
unseen tasks. Furthermore, due to the shortage of examples in unseen tasks,
central limit theorem does not stand on our side. At the end, human designers
lose the oscilloscope to "hack" an AI system for the open-world.
Achieving AI for the open-world requires unique learning principles and
innovated techniques, which are different from the ones in building AI for the
closed-world. This thesis explores necessary learning principles required to
construct AI for the open-world, including rich features (analogy a large tool
box), disentangled representation (an organized tool box), and inference-time
learning (a tool-savvy hand). Driven by the learning principles, this thesis
further proposes techniques to use the learning principles, conducts enormous
large-scale experiments to verify the learning principles.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:22:00 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhang",
"Jianyu",
""
]
] |
2504.14752 | Hadi Elzayn | Hadi Elzayn, Jacob Goldin, Cameron Guage, Daniel E. Ho, Claire Morton | Monotone Ecological Inference | null | null | null | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study monotone ecological inference, a partial identification approach to
ecological inference. The approach exploits information about one or both of
the following conditional associations: (1) outcome differences between groups
within the same neighborhood, and (2) outcomes differences within the same
group across neighborhoods with different group compositions. We show how
assumptions about the sign of these conditional associations, whether
individually or in relation to one another, can yield informative sharp bounds
in ecological inference settings. We illustrate our proposed approach using
county-level data to study differences in Covid-19 vaccination rates among
Republicans and Democrats in the United States.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:25:45 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Elzayn",
"Hadi",
""
],
[
"Goldin",
"Jacob",
""
],
[
"Guage",
"Cameron",
""
],
[
"Ho",
"Daniel E.",
""
],
[
"Morton",
"Claire",
""
]
] |
2504.14753 | Guodong Shen | Guodong Shen, Yuqi Ouyang, Junru Lu, Yixuan Yang, Victor Sanchez | Advancing Video Anomaly Detection: A Bi-Directional Hybrid Framework for
Enhanced Single- and Multi-Task Approaches | Accepted by IEEE Transactions on Image Processing (TIP) | IEEE Transactions on Image Processing, vol. 33, pp. 6865-6880,
2024 | 10.1109/TIP.2024.3512369 | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Despite the prevailing transition from single-task to multi-task approaches
in video anomaly detection, we observe that many adopt sub-optimal frameworks
for individual proxy tasks. Motivated by this, we contend that optimizing
single-task frameworks can advance both single- and multi-task approaches.
Accordingly, we leverage middle-frame prediction as the primary proxy task, and
introduce an effective hybrid framework designed to generate accurate
predictions for normal frames and flawed predictions for abnormal frames. This
hybrid framework is built upon a bi-directional structure that seamlessly
integrates both vision transformers and ConvLSTMs. Specifically, we utilize
this bi-directional structure to fully analyze the temporal dimension by
predicting frames in both forward and backward directions, significantly
boosting the detection stability. Given the transformer's capacity to model
long-range contextual dependencies, we develop a convolutional temporal
transformer that efficiently associates feature maps from all context frames to
generate attention-based predictions for target frames. Furthermore, we devise
a layer-interactive ConvLSTM bridge that facilitates the smooth flow of
low-level features across layers and time-steps, thereby strengthening
predictions with fine details. Anomalies are eventually identified by
scrutinizing the discrepancies between target frames and their corresponding
predictions. Several experiments conducted on public benchmarks affirm the
efficacy of our hybrid framework, whether used as a standalone single-task
approach or integrated as a branch in a multi-task approach. These experiments
also underscore the advantages of merging vision transformers and ConvLSTMs for
video anomaly detection.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:27:24 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Shen",
"Guodong",
""
],
[
"Ouyang",
"Yuqi",
""
],
[
"Lu",
"Junru",
""
],
[
"Yang",
"Yixuan",
""
],
[
"Sanchez",
"Victor",
""
]
] |
2504.14754 | Timothy Rupert | Tianjiao Lei, Esther Hessong, Brandon Fields, Raphael Pierre Thiraux,
Daniel S. Gianola, Timothy J. Rupert | Bulk nanocrystalline Al-Mg-Y alloys with amorphous grain boundary
complexions display high strength and compressive plasticity | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although nanocrystalline alloys regularly exhibit high strengths, their use
in structural applications often face challenges due to sample size
limitations, unstable microstructures, and the limited ability to plastically
deform. The incorporation of amorphous grain boundary complexions has been
proposed to address these issues, by simultaneously stabilizing nanocrystalline
grain structures for scale-up processing and improving alloy toughness. In the
present study, the mechanical behavior of bulk nanocrystalline Al-Mg-Y is
examined with macroscale compression testing, probing a length scale that is
relevant to real-world structural applications. Bulk samples were fabricated
via a simple powder metallurgy approach, with different pressing temperatures
and times employed for consolidation in order to investigate microstructural
and property evolution. All of the specimens contained primary face-centered
cubic Al and secondary Al4C3 and Al3Y phases, with the Al3Y particles
exhibiting two populations of small equiaxed and larger elongated particles.
Appreciable plasticity was measured along with high ultimate stresses over 800
MPa due to the presence of amorphous grain boundary complexions.
Microstructural characterization of fracture surfaces revealed that the area
fraction of dimpled regions increased with longer hot-pressing time. Most
importantly, the elongated Al3Y particles formed regular cellular patterns with
increasing hot-pressing time, delaying shear localization and significantly
enhancing plasticity. The hierarchy present in the microstructure of the
Al-Mg-Y alloy, from amorphous grain boundary complexions to secondary phases,
gives rise to excellent bulk mechanical properties, which are attractive for
structural applications.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:29:05 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Lei",
"Tianjiao",
""
],
[
"Hessong",
"Esther",
""
],
[
"Fields",
"Brandon",
""
],
[
"Thiraux",
"Raphael Pierre",
""
],
[
"Gianola",
"Daniel S.",
""
],
[
"Rupert",
"Timothy J.",
""
]
] |
2504.14755 | Andrew Sabelhaus | Akua K. Dickson, Juan C. Pacheco Garcia, Meredith L. Anderson, Ran
Jing, Sarah Alizadeh-Shabdiz, Audrey X. Wang, Charles DeLorey, Zach J.
Patterson, Andrew P. Sabelhaus | Safe Autonomous Environmental Contact for Soft Robots using Control
Barrier Functions | 10 pages, 10 figures | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robots built from soft materials will inherently apply lower environmental
forces than their rigid counterparts, and therefore may be more suitable in
sensitive settings with unintended contact. However, these robots' applied
forces result from both their design and their control system in closed-loop,
and therefore, ensuring bounds on these forces requires controller synthesis
for safety as well. This article introduces the first feedback controller for a
soft manipulator that formally meets a safety specification with respect to
environmental contact. In our proof-of-concept setting, the robot's environment
has known geometry and is deformable with a known elastic modulus. Our approach
maps a bound on applied forces to a safe set of positions of the robot's tip
via predicted deformations of the environment. Then, a quadratic program with
Control Barrier Functions in its constraints is used to supervise a nominal
feedback signal, verifiably maintaining the robot's tip within this safe set.
Hardware experiments on a multi-segment soft pneumatic robot demonstrate that
the proposed framework successfully constrains its environmental contact
forces. This framework represents a fundamental shift in perspective on control
and safety for soft robots, defining and implementing a formally verifiable
logic specification on their pose and contact forces.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:31:55 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Dickson",
"Akua K.",
""
],
[
"Garcia",
"Juan C. Pacheco",
""
],
[
"Anderson",
"Meredith L.",
""
],
[
"Jing",
"Ran",
""
],
[
"Alizadeh-Shabdiz",
"Sarah",
""
],
[
"Wang",
"Audrey X.",
""
],
[
"DeLorey",
"Charles",
""
],
[
"Patterson",
"Zach J.",
""
],
[
"Sabelhaus",
"Andrew P.",
""
]
] |
2504.14756 | Alfred Michel Grundland | {\L}ukasz Chomienia, Alfred Michel Grundland | Lie modules analysis of hydrodynamic-type systems | null | null | null | null | math-ph math.AP math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The objective of this paper is to study nonlinear superpositions of Riemann
wave solutions admitted by quasilinear hyperbolic first-order systems of
partial differential equations. Particular attention is devoted to the analysis
of non-elastic wave superpositions that cannot be decomposed into pairwise
independent interactions of waves (quasi-rectifiability). In the case of the
compressible Euler system, we describe the structure of the
infinite-dimensional Lie algebra of vector fields associated with waves. We
prove that a certain class of Lie modules associated with a hydrodynamic-type
systems can be uniquely transformed into a real Lie algebra in an
angle-preserving manner. For the Euler system, we then demonstrate the
connection between the transformed finite-dimensional Lie algebra and the
infinite-dimensional algebra associated with waves. This enables a detailed
investigation of the geometry of wave superpositions using tools from
differential geometry and Lie group theory. In particular, we study the
geometry of the manifold of wave superpositions in terms of deformations of
submanifolds corresponding to Lie subalgebras associated with waves.
Additionally, we present new methods and criteria for determining the
quasi-rectifiability of Riemann k-waves.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:33:57 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Chomienia",
"Łukasz",
""
],
[
"Grundland",
"Alfred Michel",
""
]
] |
2504.14757 | Nghi D. Q. Bui | Minh V.T. Pham, Huy N. Phan, Hoang N. Phan, Cuong Le Chi, Tien N.
Nguyen, Nghi D. Q. Bui | SWE-Synth: Synthesizing Verifiable Bug-Fix Data to Enable Large Language
Models in Resolving Real-World Bugs | Work in progress | null | null | null | cs.SE cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | Large language models (LLMs) are transforming automated program repair (APR)
through agent-based approaches that localize bugs, generate patches, and verify
fixes. However, the lack of high-quality, scalable training datasets,
especially those with verifiable outputs and intermediate reasoning
traces-limits progress, particularly for open-source models. In this work, we
present SWE-Synth, a framework for synthesizing realistic, verifiable, and
process-aware bug-fix datasets at the repository level. SWE-Synth leverages LLM
agents to simulate debugging workflows, producing not only bug-fix pairs but
also test cases and structured repair trajectories. Compared to manually
curated datasets, our method scales with minimal human effort while preserving
contextual richness and correctness. Experiments show that models trained on
SWE-Synth outperform those trained on real-world datasets by 2.3% on SWE-Bench
Lite. Our results highlight the potential of synthetic, agent-generated data to
advance the state of the art in APR and software engineering automation.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:37:43 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Pham",
"Minh V. T.",
""
],
[
"Phan",
"Huy N.",
""
],
[
"Phan",
"Hoang N.",
""
],
[
"Chi",
"Cuong Le",
""
],
[
"Nguyen",
"Tien N.",
""
],
[
"Bui",
"Nghi D. Q.",
""
]
] |
2504.14758 | Trung-Phuc Vo | Trung-Phuc Vo, Olena Tkach, Aki Pulkkinen, Didier Sebilleau, Aimo
Winkelmann, Olena Fedchenko, Yaryna Lytvynenko, Dmitry Vasilyev, Hans-Joachim
Elmers, Gerd Schonhense and Jan Minar | Unveiling Fine Structure and Energy-driven Transition of Photoelectron
Kikuchi Diffraction | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The intricate fine structure of Kikuchi diffraction plays a vital role in
probing phase transformations and strain distributions in functional materials,
particularly in electron microscopy. Beyond these applications, it also proves
essential in photoemission spectroscopy (PES) at high photon energies, aiding
in the disentanglement of complex angle-resolved PES data and enabling
emitter-site-specific studies. However, the detection and analysis of these
rich faint structures in photoelectron diffraction (PED), especially in the
hard X-ray regime, remain highly challenging, with only a limited number of
simulations successfully reproducing these patterns. The strong energy
dependence of Kikuchi patterns further complicates their interpretation,
necessitating advanced theoretical approaches. To enhance structural analysis,
we present a comprehensive theoretical study of fine diffraction patterns and
their evolution with energy by simulating core-level emissions from Ge(100) and
Si(100). Using multiple-scattering theory and the fully relativistic one-step
photoemission model, we simulate faint pattern networks for various core levels
across different kinetic energies (106 eV - 4174 eV), avoiding cluster size
convergence issues inherent in cluster-based methods. Broadening in patterns is
discussed via the inelastic scattering treatment. For the first time, circular
dichroism has been observed and successfully reproduced in the angular
distribution of Si (100) 1s, revealing detailed features and asymmetries up to
31%. Notably, we successfully replicate experimental bulk and more
"surface-sensitivity" diffraction features, further validating the robustness
of our simulations. The results show remarkable agreement with the experimental
data obtained using circularly polarized radiations, demonstrating the
potential of this methodology for advancing high-energy PES investigations.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:48:12 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Vo",
"Trung-Phuc",
""
],
[
"Tkach",
"Olena",
""
],
[
"Pulkkinen",
"Aki",
""
],
[
"Sebilleau",
"Didier",
""
],
[
"Winkelmann",
"Aimo",
""
],
[
"Fedchenko",
"Olena",
""
],
[
"Lytvynenko",
"Yaryna",
""
],
[
"Vasilyev",
"Dmitry",
""
],
[
"Elmers",
"Hans-Joachim",
""
],
[
"Schonhense",
"Gerd",
""
],
[
"Minar",
"Jan",
""
]
] |
2504.14759 | Dongryul Kim | Hyungryul Baik and Dongryul M. Kim | Normal generators for mapping class groups | 26 pages, 8 figures. This survey article was written for inclusion in
"In the Tradition of Thurston: Geometry and Topology'', Vol. 4. which is
edited by A. Papadopoulos and K. Ohshika | null | null | null | math.GT math.DS math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this expository note, we discuss normal generators for mapping class
groups of surfaces. Especially, we focus on the relation between normal
generation of a mapping class with its asymptotic translation lengths on the
Teichm\"uller space and the curve graph of the underlying surface. We also
discuss several open questions.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:56:49 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Baik",
"Hyungryul",
""
],
[
"Kim",
"Dongryul M.",
""
]
] |
2504.14760 | Surya Teja Avirneni | Surya Teja Avirneni | Establishing Workload Identity for Zero Trust CI/CD: From Secrets to
SPIFFE-Based Authentication | 13 pages, 3 figures. Originally authored in November 2024. Submitted
as part of a three-part series on Zero Trust CI/CD and Workload Identity.
Discusses SPIFFE, OIDC federation, Non-Human Identity (NHI) governance, and
CI/CD security patterns (NIST SP 800-204) | null | null | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | CI/CD systems have become privileged automation agents in modern
infrastructure, but their identity is still based on secrets or temporary
credentials passed between systems. In enterprise environments, these platforms
are centralized and shared across teams, often with broad cloud permissions and
limited isolation. These conditions introduce risk, especially in the era of
supply chain attacks, where implicit trust and static credentials leave systems
exposed. This paper describes the shift from static credentials to OpenID
Connect (OIDC) federation, and introduces SPIFFE (Secure Production Identity
Framework for Everyone) as a runtime-issued, platform-neutral identity model
for non-human actors. SPIFFE decouples identity from infrastructure, enabling
strong, portable authentication across job runners and deployed workloads. We
show how SPIFFE identities support policy alignment, workload attestation, and
mutual authentication. The paper concludes by outlining next steps in enabling
policy-based access, forming the basis of a broader Zero Trust architecture for
CI/CD.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:06:03 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Avirneni",
"Surya Teja",
""
]
] |
2504.14761 | Surya Teja Avirneni | Surya Teja Avirneni | Decoupling Identity from Access: Credential Broker Patterns for Secure
CI/CD | 10 pages, 2 figures. Originally authored in November 2024. Second
paper in a three-part series on Zero Trust CI/CD and dynamic Workload
Identity enforcement. Builds on SPIFFE-based authentication to introduce
credential broker patterns and policy-based access control | null | null | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | Credential brokers offer a way to separate identity from access in CI/CD
systems. This paper shows how verifiable identities issued at runtime, such as
those from SPIFFE, can be used with brokers to enable short-lived,
policy-driven credentials for pipelines and workloads. We walk through
practical design patterns, including brokers that issue tokens just in time,
apply access policies, and operate across trust domains. These ideas help
reduce static permissions, improve auditability, and support Zero Trust goals
in deployment workflows. This is the second paper in a three-part series on
secure CI/CD identity architecture.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:08:17 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Avirneni",
"Surya Teja",
""
]
] |
2504.14762 | Sahil Rajesh Dhayalkar | Sahil Rajesh Dhayalkar | A Combinatorial Theory of Dropout: Subnetworks, Graph Geometry, and
Generalization | 17 pages (9 pages main content and remaining pages are references,
appendix which includes 7 figures, proofs and derivations) | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | We propose a combinatorial and graph-theoretic theory of dropout by modeling
training as a random walk over a high-dimensional graph of binary subnetworks.
Each node represents a masked version of the network, and dropout induces
stochastic traversal across this space. We define a subnetwork contribution
score that quantifies generalization and show that it varies smoothly over the
graph. Using tools from spectral graph theory, PAC-Bayes analysis, and
combinatorics, we prove that generalizing subnetworks form large, connected,
low-resistance clusters, and that their number grows exponentially with network
width. This reveals dropout as a mechanism for sampling from a robust,
structured ensemble of well-generalizing subnetworks with built-in redundancy.
Extensive experiments validate every theoretical claim across diverse
architectures. Together, our results offer a unified foundation for
understanding dropout and suggest new directions for mask-guided regularization
and subnetwork optimization.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:09:20 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Dhayalkar",
"Sahil Rajesh",
""
]
] |
2504.14763 | Soobin Cho | Soobin Cho and Renming Song | Approximate factorizations for non-symmetric jump processes | 55 pages | null | null | null | math.PR math.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we first extend the approximate factorization for purely
discontinuous Markov process established in \cite{CKSV20} by getting rid of
some of the conditions imposed in \cite{CKSV20}. Then we apply the approximate
factorization to obtain two-sided heat kernel estimates for three classes of
processes: stable-like processes with critical killings in $C^{1, {\rm Dini}}$
open sets; killed stable-like processes in the setting of \cite{KW24} in $C^{1,
\varepsilon}$ open sets; and non-symmetric stable processes in what we call
$C^{1,2{\text - \rm Dini}}$ open sets. In particular, we obtain explicit sharp
two-sided heat kernel estimates of killed $\alpha$-stable processes in $C^{1,
{\rm Dini}}$ open sets for all $\alpha\in (0, 2)$ and of censored
$\alpha$-stable processes in $C^{1, {\rm Dini}}$ open sets for all $\alpha\in
(1, 2)$.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:13:48 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Cho",
"Soobin",
""
],
[
"Song",
"Renming",
""
]
] |
2504.14764 | Shreya Shankar | Shreya Shankar, Bhavya Chopra, Mawil Hasan, Stephen Lee, Bj\"orn
Hartmann, Joseph M. Hellerstein, Aditya G. Parameswaran, Eugene Wu | Steering Semantic Data Processing With DocWrangler | 18 pages; 11 figures; 3 tables | null | null | null | cs.HC cs.DB | http://creativecommons.org/licenses/by/4.0/ | Unstructured text has long been difficult to automatically analyze at scale.
Large language models (LLMs) now offer a way forward by enabling {\em semantic
data processing}, where familiar data processing operators (e.g., map, reduce,
filter) are powered by LLMs instead of code. However, building effective
semantic data processing pipelines presents a departure from traditional data
pipelines: users need to understand their data to write effective pipelines,
yet they need to construct pipelines to extract the data necessary for that
understanding -- all while navigating LLM idiosyncrasies and inconsistencies.
We present \docwrangler, a mixed-initiative integrated development environment
(IDE) for semantic data processing with three novel features to address the
gaps between the user, their data, and their pipeline: {\em (i) In-Situ User
Notes} that allows users to inspect, annotate, and track observations across
documents and LLM outputs, {\em (ii) LLM-Assisted Prompt Refinement} that
transforms user notes into improved operations, and {\em (iii) LLM-Assisted
Operation Decomposition} that identifies when operations or documents are too
complex for the LLM to correctly process and suggests decompositions. Our
evaluation combines a think-aloud study with 10 participants and a
public-facing deployment (available at
\href{https://docetl.org/playground}{docetl.org/playground}) with 1,500+
recorded sessions, revealing how users develop systematic strategies for their
semantic data processing tasks; e.g., transforming open-ended operations into
classifiers for easier validation and intentionally using vague prompts to
learn more about their data or LLM capabilities.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:35:41 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Shankar",
"Shreya",
""
],
[
"Chopra",
"Bhavya",
""
],
[
"Hasan",
"Mawil",
""
],
[
"Lee",
"Stephen",
""
],
[
"Hartmann",
"Björn",
""
],
[
"Hellerstein",
"Joseph M.",
""
],
[
"Parameswaran",
"Aditya G.",
""
],
[
"Wu",
"Eugene",
""
]
] |
2504.14765 | Alejandro Lopez-Lira | Alejandro Lopez-Lira, Yuehua Tang, Mingyin Zhu | The Memorization Problem: Can We Trust LLMs' Economic Forecasts? | null | null | null | null | q-fin.GN q-fin.ST | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Large language models (LLMs) cannot be trusted for economic forecasts during
periods covered by their training data. We provide the first systematic
evaluation of LLMs' memorization of economic and financial data, including
major economic indicators, news headlines, stock returns, and conference calls.
Our findings show that LLMs can perfectly recall the exact numerical values of
key economic variables from before their knowledge cutoff dates. This recall
appears to be randomly distributed across different dates and data types. This
selective perfect memory creates a fundamental issue -- when testing
forecasting capabilities before their knowledge cutoff dates, we cannot
distinguish whether LLMs are forecasting or simply accessing memorized data.
Explicit instructions to respect historical data boundaries fail to prevent
LLMs from achieving recall-level accuracy in forecasting tasks. Further, LLMs
seem exceptional at reconstructing masked entities from minimal contextual
clues, suggesting that masking provides inadequate protection against motivated
reasoning. Our findings raise concerns about using LLMs to forecast historical
data or backtest trading strategies, as their apparent predictive success may
merely reflect memorization rather than genuine economic insight. Any
application where future knowledge would change LLMs' outputs can be affected
by memorization. In contrast, consistent with the absence of data
contamination, LLMs cannot recall data after their knowledge cutoff date.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:36:27 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Lopez-Lira",
"Alejandro",
""
],
[
"Tang",
"Yuehua",
""
],
[
"Zhu",
"Mingyin",
""
]
] |
2504.14766 | Saniya Karwa | Saniya Karwa and Navpreet Singh | Disentangling Linguistic Features with Dimension-Wise Analysis of Vector
Embeddings | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Understanding the inner workings of neural embeddings, particularly in models
such as BERT, remains a challenge because of their high-dimensional and opaque
nature. This paper proposes a framework for uncovering the specific dimensions
of vector embeddings that encode distinct linguistic properties (LPs). We
introduce the Linguistically Distinct Sentence Pairs (LDSP-10) dataset, which
isolates ten key linguistic features such as synonymy, negation, tense, and
quantity. Using this dataset, we analyze BERT embeddings with various methods,
including the Wilcoxon signed-rank test, mutual information, and recursive
feature elimination, to identify the most influential dimensions for each LP.
We introduce a new metric, the Embedding Dimension Impact (EDI) score, which
quantifies the relevance of each embedding dimension to a LP. Our findings show
that certain properties, such as negation and polarity, are robustly encoded in
specific dimensions, while others, like synonymy, exhibit more complex
patterns. This study provides insights into the interpretability of embeddings,
which can guide the development of more transparent and optimized language
models, with implications for model bias mitigation and the responsible
deployment of AI systems.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:38:16 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Karwa",
"Saniya",
""
],
[
"Singh",
"Navpreet",
""
]
] |
2504.14767 | Mohamed El Machkouri | Rafik Aguech and Samir Ben Hariz and Mohamed El Machkouri and Youssef
Faouzi | On a class of unbalanced step-reinforced random walks | 24 pages | null | null | null | math.PR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | A step-reinforced random walk is a discrete-time stochastic process with
long-range dependence. At each step, with a fixed probability $\alpha$, the
so-called positively step-reinforced random walk repeats one of its previous
steps, chosen randomly and uniformly from its entire history. Alternatively,
with probability $1-\alpha$, it makes an independent move. For the so-called
negatively step-reinforced random walk, the process is similar, but any
repeated step is taken with its direction reversed. These random walks have
been introduced respectively by Simon (1955) and Bertoin (2024) and are
sometimes refered to the self-confident step-reinforced random walk and the
counterbalanced step-reinforced random walk respectively. In this work, we
introduce a new class of unbalanced step-reinforced random walks for which we
prove the strong law of large numbers and the central limit theorem. In
particular, our work provides a unified treatment of the elephant random walk
introduced by Schutz and Trimper (2004) and the positively and negatively
step-reinforced random walks
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:38:49 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Aguech",
"Rafik",
""
],
[
"Hariz",
"Samir Ben",
""
],
[
"Machkouri",
"Mohamed El",
""
],
[
"Faouzi",
"Youssef",
""
]
] |
2504.14768 | Takashi Goda | Takashi Goda | A note on unshifted lattice rules for high-dimensional integration in
weighted unanchored Sobolev spaces | 6 pages | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This short article studies a deterministic quasi-Monte Carlo lattice rule in
weighted unanchored Sobolev spaces of smoothness $1$. Building on the error
analysis by Kazashi and Sloan, we prove the existence of unshifted rank-1
lattice rules that achieve a worst-case error of $O(n^{-1/4}(\log n)^{1/2})$,
with the implied constant independent of the dimension, under certain
summability conditions on the weights. Although this convergence rate is
inferior to the one achievable for the shifted-averaged root mean squared
worst-case error, the result does not rely on random shifting or transformation
and holds unconditionally without any conjecture, as assumed by Kazashi and
Sloan.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:39:30 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Goda",
"Takashi",
""
]
] |
2504.14769 | Luis Morales-Navarro | Luis Morales-Navarro, Daniel J. Noh, Yasmin B. Kafai | Building babyGPTs: Youth Engaging in Data Practices and Ethical
Considerations through the Construction of Generative Language Models | null | null | 10.1145/3713043.3731525 | null | cs.HC cs.CY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | As generative language models (GLMs) have gained popularity, youth are
increasingly using them in their everyday lives. As such, most research has
centered on supporting youth as users of GLM-powered systems. However, we know
little of how to engage youth in the design of these models. Building on the
rich legacy of child-computer interaction research that positions youth as
designers of computing systems, we explore how to support young people in
designing GLMs. Through a case study of three teenagers (ages 14-15) building a
babyGPT screenplay generator, we illustrate how the team developed a model
while engaging in artificial intelligence/machine learning-relevant data
practices and addressing ethical issues. This paper contributes a case study
that demonstrates the feasibility of engaging youth in building GLMs.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:40:29 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Morales-Navarro",
"Luis",
""
],
[
"Noh",
"Daniel J.",
""
],
[
"Kafai",
"Yasmin B.",
""
]
] |
2504.14770 | Jos\'e Rom\'an Aranda Cuevas | Rom\'an Aranda, Noah Crawford, Andrew Maas, Nicole Marienau, Erica
Pearce, Renzo Sarreal, Savannah Schutte, Ransom Sterns, and Eric Woods | Region colorings of surfaces in 4-space | 20 pages, 23 figures, comments are welcome | null | null | null | math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Niebrzydowski introduced a theory of region colorings for surface links. In
this paper, we translate the coloring invariant to the context of triplane
diagrams and movies of knots. We provide inequalities between the number of
region colorings and topological quantities of $F$, such as the number of
saddles in a movie and the bridge index of a triplane diagram of $F$. As an
application, we show that Yoshikawa's 2-knots $9_1$ and $10_2$ are
non-invertible; that is $F\not=-F$.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:42:42 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Aranda",
"Román",
""
],
[
"Crawford",
"Noah",
""
],
[
"Maas",
"Andrew",
""
],
[
"Marienau",
"Nicole",
""
],
[
"Pearce",
"Erica",
""
],
[
"Sarreal",
"Renzo",
""
],
[
"Schutte",
"Savannah",
""
],
[
"Sterns",
"Ransom",
""
],
[
"Woods",
"Eric",
""
]
] |
2504.14771 | Charlotte Zimmerman | Charlotte Zimmerman, Alexis Olsho, Trevor I. Smith, Philip Eaton,
Suzanne White Brahmia | Assessing physics quantitative literacy development in algebra-based
physics | null | null | null | null | physics.ed-ph | http://creativecommons.org/licenses/by/4.0/ | Quantitative reasoning is an essential learning objective of physics
instruction. The Physics Inventory for Quantitative Literacy (PIQL) is an
assessment tool that has been developed for calculus-based physics courses to
help instructors evaluate whether their students learn to reason this way
(White Brahmia, et al., 2019). However, the PIQL is not appropriate for the
large population of students taking physics who are not enrolled in, or have
not completed, calculus. To address this need, we have developed the General
Equation-based Reasoning inventory of QuaNtity (GERQN). The GERQN is an
algebra-based version of the PIQL and is appropriate for most physics students;
the only requirement is that students have taken algebra so they are familiar
with the use of variable, negative quantities, and linear functions. In this
paper we present the development and validation of the GERQN, and a short
discussion on how the GERQN can be used by instructors to help their students
learn.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:46:54 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zimmerman",
"Charlotte",
""
],
[
"Olsho",
"Alexis",
""
],
[
"Smith",
"Trevor I.",
""
],
[
"Eaton",
"Philip",
""
],
[
"Brahmia",
"Suzanne White",
""
]
] |
2504.14772 | Tao Wang | Luyang Fang, Xiaowei Yu, Jiazhang Cai, Yongkai Chen, Shushan Wu,
Zhengliang Liu, Zhenyuan Yang, Haoran Lu, Xilin Gong, Yufang Liu, Terry Ma,
Wei Ruan, Ali Abbasi, Jing Zhang, Tao Wang, Ehsan Latif, Wei Liu, Wei Zhang,
Soheil Kolouri, Xiaoming Zhai, Dajiang Zhu, Wenxuan Zhong, Tianming Liu, and
Ping Ma | Knowledge Distillation and Dataset Distillation of Large Language
Models: Emerging Trends, Challenges, and Future Directions | null | null | null | null | cs.CL cs.LG stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The exponential growth of Large Language Models (LLMs) continues to highlight
the need for efficient strategies to meet ever-expanding computational and data
demands. This survey provides a comprehensive analysis of two complementary
paradigms: Knowledge Distillation (KD) and Dataset Distillation (DD), both
aimed at compressing LLMs while preserving their advanced reasoning
capabilities and linguistic diversity. We first examine key methodologies in
KD, such as task-specific alignment, rationale-based training, and
multi-teacher frameworks, alongside DD techniques that synthesize compact,
high-impact datasets through optimization-based gradient matching, latent space
regularization, and generative synthesis. Building on these foundations, we
explore how integrating KD and DD can produce more effective and scalable
compression strategies. Together, these approaches address persistent
challenges in model scalability, architectural heterogeneity, and the
preservation of emergent LLM abilities. We further highlight applications
across domains such as healthcare and education, where distillation enables
efficient deployment without sacrificing performance. Despite substantial
progress, open challenges remain in preserving emergent reasoning and
linguistic diversity, enabling efficient adaptation to continually evolving
teacher models and datasets, and establishing comprehensive evaluation
protocols. By synthesizing methodological innovations, theoretical foundations,
and practical insights, our survey charts a path toward sustainable,
resource-efficient LLMs through the tighter integration of KD and DD
principles.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 23:50:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Fang",
"Luyang",
""
],
[
"Yu",
"Xiaowei",
""
],
[
"Cai",
"Jiazhang",
""
],
[
"Chen",
"Yongkai",
""
],
[
"Wu",
"Shushan",
""
],
[
"Liu",
"Zhengliang",
""
],
[
"Yang",
"Zhenyuan",
""
],
[
"Lu",
"Haoran",
""
],
[
"Gong",
"Xilin",
""
],
[
"Liu",
"Yufang",
""
],
[
"Ma",
"Terry",
""
],
[
"Ruan",
"Wei",
""
],
[
"Abbasi",
"Ali",
""
],
[
"Zhang",
"Jing",
""
],
[
"Wang",
"Tao",
""
],
[
"Latif",
"Ehsan",
""
],
[
"Liu",
"Wei",
""
],
[
"Zhang",
"Wei",
""
],
[
"Kolouri",
"Soheil",
""
],
[
"Zhai",
"Xiaoming",
""
],
[
"Zhu",
"Dajiang",
""
],
[
"Zhong",
"Wenxuan",
""
],
[
"Liu",
"Tianming",
""
],
[
"Ma",
"Ping",
""
]
] |
2504.14773 | Fei Liu | Haoming Li and Zhaoliang Chen and Jonathan Zhang and Fei Liu | PLANET: A Collection of Benchmarks for Evaluating LLMs' Planning
Capabilities | 10 pages | null | null | null | cs.AI cs.CL cs.LG cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Planning is central to agents and agentic AI. The ability to plan, e.g.,
creating travel itineraries within a budget, holds immense potential in both
scientific and commercial contexts. Moreover, optimal plans tend to require
fewer resources compared to ad-hoc methods. To date, a comprehensive
understanding of existing planning benchmarks appears to be lacking. Without
it, comparing planning algorithms' performance across domains or selecting
suitable algorithms for new scenarios remains challenging. In this paper, we
examine a range of planning benchmarks to identify commonly used testbeds for
algorithm development and highlight potential gaps. These benchmarks are
categorized into embodied environments, web navigation, scheduling, games and
puzzles, and everyday task automation. Our study recommends the most
appropriate benchmarks for various algorithms and offers insights to guide
future benchmark development.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:02:50 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Li",
"Haoming",
""
],
[
"Chen",
"Zhaoliang",
""
],
[
"Zhang",
"Jonathan",
""
],
[
"Liu",
"Fei",
""
]
] |
2504.14774 | Mario Martinez-Saito | Mario Martinez-Saito | Defining and finding lifelike entities with a lazy filter | null | null | null | null | q-bio.OT | http://creativecommons.org/licenses/by/4.0/ | Our binary intuitive understanding of life and lifelikeness is good enough
for daily life, but not for research in the natural sciences. Here we propose
an operational definition of lifeness of a particular entity as a scalar,
product of its defining information (algorithmic complexity) integrated over
its lifetime. We provide a parameter-free dynamical filtering algorithm that
can efficiently and on-the-fly fit configurations of entities constituted by
moving particles to a tree structure that can model a wide range of
hierarchical system entities, while extracting and calculating the complexity,
lifespan, and lifeness of the entity and all of its constituting subtrees or
subentities. We simulated 41 interacting particle worlds and found preliminary
evidence suggesting that the lifeness of entities is associated with the
distance to criticality, as roughly measured by the range of the pairwise
interaction forces of the elemental particles of the worlds they inhabit. This
study is a proof of concept that defining, measuring, and quantifying lifeness
is (1) feasible, useful for (2) simplifying theoretical discussions, for (3)
hierarchically assessing biotic properties such as number of hierarchical
levels and predicted longevity and for (4) classifying both artificial and
biological entities, respectively via computer simulations and a combination of
evolutionary and molecular biology approaches.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:06:32 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Martinez-Saito",
"Mario",
""
]
] |
2504.14775 | Tianyu Guo | Tianyu Guo, Xianwei Zhang, Jiangsu Du, Zhiguang Chen, Nong Xiao,
Yutong Lu | gLLM: Global Balanced Pipeline Parallelism System for Distributed LLM
Serving with Token Throttling | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pipeline parallelism has emerged as a predominant approach for deploying
large language models (LLMs) across distributed nodes, owing to its lower
communication overhead compared to tensor parallelism. While demonstrating high
throughput in request serving, pipeline parallelism often suffers from
performance limitations caused by pipeline bubbles, which are primarily
resulted from imbalanced computation delays across batches. Existing methods
like Sarathi-Serve attempt to address this through hybrid scheduling of chunked
prefill and decode tokens using a fixed token budget. However, such methods may
experience significant fluctuations due to either insufficient prefill tokens
or uneven distribution of decode tokens, ultimately leading to computational
imbalance. To overcome these inefficiencies, we present gLLM, a globally
balanced pipeline parallelism system incorporating Token Throttling to
effectively mitigate the pipeline bubbles. Our Token Throttling mechanism is a
fine-grained scheduling policy that independently regulates the quantities of
prefill and decode tokens, thus enabling balanced computation by leveraging
global information from the inference system. Specifically, for decode tokens,
gLLM maintains near-consistent token count across processing batches. For
prefill tokens, it dynamically adjusts batch sizes based on both total pending
tokens and the memory utilization rates of key-value cache (KV cache).
Furthermore, gLLM runtime adopts an asynchronous execution and message passing
architecture specifically optimized for pipeline parallelism characteristics.
Experimental evaluations with representative LLMs show that gLLM achieves
significant performance improvements, delivering 11% to 398% higher maximum
throughput compared to state-of-the-art pipeline or tensor parallelism systems,
while simultaneously maintaining lower latency.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:07:49 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Guo",
"Tianyu",
""
],
[
"Zhang",
"Xianwei",
""
],
[
"Du",
"Jiangsu",
""
],
[
"Chen",
"Zhiguang",
""
],
[
"Xiao",
"Nong",
""
],
[
"Lu",
"Yutong",
""
]
] |
2504.14776 | Zhecheng Wang | Zhecheng Wang, Jiaju Ma, Eitan Grinspun, Bryan Wang, Tovi Grossman | Script2Screen: Supporting Dialogue Scriptwriting with Interactive
Audiovisual Generation | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Scriptwriting has traditionally been text-centric, a modality that only
partially conveys the produced audiovisual experience. A formative study with
professional writers informed us that connecting textual and audiovisual
modalities can aid ideation and iteration, especially for writing dialogues. In
this work, we present Script2Screen, an AI-assisted tool that integrates
scriptwriting with audiovisual scene creation in a unified, synchronized
workflow. Focusing on dialogues in scripts, Script2Screen generates expressive
scenes with emotional speeches and animated characters through a novel
text-to-audiovisual-scene pipeline. The user interface provides fine-grained
controls, allowing writers to fine-tune audiovisual elements such as character
gestures, speech emotions, and camera angles. A user study with both novice and
professional writers from various domains demonstrated that Script2Screen's
interactive audiovisual generation enhances the scriptwriting process,
facilitating iterative refinement while complementing, rather than replacing,
their creative efforts.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:11:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Wang",
"Zhecheng",
""
],
[
"Ma",
"Jiaju",
""
],
[
"Grinspun",
"Eitan",
""
],
[
"Wang",
"Bryan",
""
],
[
"Grossman",
"Tovi",
""
]
] |
2504.14777 | Surya Teja Avirneni | Surya Teja Avirneni | Intent-Aware Authorization for Zero Trust CI/CD | 13 pages. Originally authored in Dec 2024. Third paper in a
three-part series on Zero Trust CI/CD and Workload Identity. Expands on
SPIFFE-based authentication and credential brokers with policy-based,
intent-aware authorization patterns using OPA and Cedar | null | null | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | This paper introduces intent-aware authorization for Zero Trust CI/CD
systems. Identity establishes who is making the request, but additional signals
are required to decide whether access should be granted. We describe a control
loop architecture where policy engines such as OPA and Cedar evaluate runtime
context, justification, and human approvals before issuing access credentials.
The system builds on SPIFFE-based workload identity and credential brokers, and
enables fine-grained, auditable authorization. This is the third paper in a
series on Zero Trust CI/CD design patterns.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:25:35 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Avirneni",
"Surya Teja",
""
]
] |
2504.14778 | Chentao Yue | Yonghui Li, Chentao Yue, Branka Vucetic | Optimal Linear MAP Decoding of Convolutional Codes | Accepted by ISIT 2025 | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a linear representation of BCJR maximum a
posteriori probability (MAP) decoding of a rate 1/2 convolutional code (CC),
referred to as the linear MAP decoding (LMAP). We discover that the MAP forward
and backward decoding can be implemented by the corresponding dual soft input
and soft output (SISO) encoders using shift registers. The bidrectional MAP
decoding output can be obtained by combining the contents of respective forward
and backward dual encoders. Represented using simple shift-registers, LMAP
decoder maps naturally to hardware registers and thus can be easily
implemented. Simulation results demonstrate that the LMAP decoding achieves the
same performance as the BCJR MAP decoding, but has a significantly reduced
decoding delay. For the block length 64, the CC of the memory length 14 with
LMAP decoding surpasses the random coding union (RCU) bound by approximately
0.5 dB at a BLER of $10^{-3}$, and closely approaches both the normal
approximation (NA) and meta-converse (MC) bounds.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:31:22 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Li",
"Yonghui",
""
],
[
"Yue",
"Chentao",
""
],
[
"Vucetic",
"Branka",
""
]
] |
2504.14779 | Janet Johnson | Janet G. Johnson, Macarena Peralta, Mansanjam Kaur, Ruijie Sophia
Huang, Sheng Zhao, Ruijia Guan, Shwetha Rajaram, Michael Nebeling | Exploring Collaborative GenAI Agents in Synchronous Group Settings:
Eliciting Team Perceptions and Design Considerations for the Future of Work | To be published in ACM Conference on Computer-Supported Cooperative
Work and Social Computing (CSCW 2025). 33 pages, 11 figures, 1 table | null | null | null | cs.HC cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | While generative artificial intelligence (GenAI) is finding increased
adoption in workplaces, current tools are primarily designed for individual
use. Prior work established the potential for these tools to enhance personal
creativity and productivity towards shared goals; however, we don't know yet
how to best take into account the nuances of group work and team dynamics when
deploying GenAI in work settings. In this paper, we investigate the potential
of collaborative GenAI agents to augment teamwork in synchronous group settings
through an exploratory study that engaged 25 professionals across 6 teams in
speculative design workshops and individual follow-up interviews. Our workshops
included a mixed reality provotype to simulate embodied collaborative GenAI
agents capable of actively participating in group discussions. Our findings
suggest that, if designed well, collaborative GenAI agents offer valuable
opportunities to enhance team problem-solving by challenging groupthink,
bridging communication gaps, and reducing social friction. However, teams'
willingness to integrate GenAI agents depended on its perceived fit across a
number of individual, team, and organizational factors. We outline the key
design tensions around agent representation, social prominence, and engagement
and highlight the opportunities spatial and immersive technologies could offer
to modulate GenAI influence on team outcomes and strike a balance between
augmentation and agency.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:38:02 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Johnson",
"Janet G.",
""
],
[
"Peralta",
"Macarena",
""
],
[
"Kaur",
"Mansanjam",
""
],
[
"Huang",
"Ruijie Sophia",
""
],
[
"Zhao",
"Sheng",
""
],
[
"Guan",
"Ruijia",
""
],
[
"Rajaram",
"Shwetha",
""
],
[
"Nebeling",
"Michael",
""
]
] |
2504.14780 | Jianxiu Li | Jianxiu Li and Urbashi Mitra | Delay-Angle Information Spoofing for Channel State Information-Free
Location-Privacy Enhancement | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a delay-angle information spoofing (DAIS) strategy is proposed
to enhance the location privacy at the physical layer. More precisely, the
location-relevant delays and angles are artificially shifted without the aid of
channel state information (CSI) at the transmitter, such that the location
perceived by the eavesdropper is incorrect and distinct from the true one. By
leveraging the intrinsic structure of the wireless channel, a precoder is
designed to achieve DAIS while the legitimate localizer can remove the
obfuscation via securely receiving a modest amount of information, i.e., the
delay-angle shifts. A lower bound on eavesdropper's localization error is
derived, revealing that location privacy is enhanced not only due to estimation
error, but also by the geometric mismatch introduced by DAIS. Furthermore, the
lower bound is explicitly expressed as a function of the delay-angle shifts,
characterizing performance trends and providing the appropriate design of these
shift parameters. The statistical hardness of maliciously inferring the
delay-angle shifts by a single-antenna eavesdropper as well as the challenges
for a multi-antenna eavesdropper are investigated to assess the robustness of
the proposed DAIS strategy. Numerical results show that the proposed DAIS
strategy results in more than 15 dB performance degradation for the
eavesdropper as compared with that for the legitimate localizer at high
signal-to-noise ratios, and provides more effective location-privacy
enhancement than the prior art.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:40:45 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Li",
"Jianxiu",
""
],
[
"Mitra",
"Urbashi",
""
]
] |
2504.14781 | Puqing Jiang | Mingzhen Zhang, Tao Chen, Shangzhi Song, Yunjia Bao, Ruiqiang Guo,
Weidong Zheng, Puqing Jiang, Ronggui Yang | Extending the Low-Frequency Limit of Time-Domain Thermoreflectance via
Periodic Waveform Analysis | 19 pages, 4 figures | null | null | null | physics.app-ph cond-mat.other | http://creativecommons.org/licenses/by/4.0/ | Time-domain thermoreflectance (TDTR) is a powerful technique for
characterizing the thermal properties of layered materials. However, its
effectiveness at modulation frequencies below 0.1 MHz is hindered by pulse
accumulation effects, limiting its ability to accurately measure in-plane
thermal conductivities below 6 W/(m K). Here, we present a periodic waveform
analysis-based TDTR (PWA-TDTR) method that extends the measurable frequency
range down to 50 Hz with minimal modifications to the conventional setup. This
advancement greatly enhances measurement sensitivity, enabling accurate
measurements of in-plane thermal conductivities as low as 0.2 W/(m K). We
validate the technique by measuring polymethyl methacrylate (PMMA) and fused
silica, using PWA-TDTR to obtain in-plane thermal diffusivity and conventional
TDTR to measure cross-plane thermal effusivity. Together, these allow the
extraction of both thermal conductivity and volumetric heat capacity, with
results in excellent agreement with literature values. We further demonstrate
the versatility of PWA-TDTR through (1) thermal conductivity and heat capacity
measurements of thin liquid films and (2) depth-resolved thermal conductivity
profiling in lithium niobate crystals, revealing point defect-induced
inhomogeneities at depths up to 100 um. By overcoming frequency and sensitivity
constraints, PWA-TDTR significantly expands the applicability of TDTR, enabling
detailed investigations of thermal transport in materials and conditions that
were previously challenging to study.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:42:14 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhang",
"Mingzhen",
""
],
[
"Chen",
"Tao",
""
],
[
"Song",
"Shangzhi",
""
],
[
"Bao",
"Yunjia",
""
],
[
"Guo",
"Ruiqiang",
""
],
[
"Zheng",
"Weidong",
""
],
[
"Jiang",
"Puqing",
""
],
[
"Yang",
"Ronggui",
""
]
] |
2504.14782 | Ahmed Sobhi Saleh Ahmed Dr. | Ahmed Sobhi Saleh, Kristof Croes, Hajdin Ceric, Ingrid De Wolf, Houman
Zahedmanesh | Novel Concept-Oriented Synthetic Data approach for Training Generative
AI-Driven Crystal Grain Analysis Using Diffusion Model | 19 Pages, 5 Figures | Computational Materials Science, Vol 251 (2025) | 10.1016/j.commatsci.2025.113723 | null | cs.LG cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | The traditional techniques for extracting polycrystalline grain structures
from microscopy images, such as transmission electron microscopy (TEM) and
scanning electron microscopy (SEM), are labour-intensive, subjective, and
time-consuming, limiting their scalability for high-throughput analysis. In
this study, we present an automated methodology integrating edge detection with
generative diffusion models to effectively identify grains, eliminate noise,
and connect broken segments in alignment with predicted grain boundaries. Due
to the limited availability of adequate images preventing the training of deep
machine learning models, a new seven-stage methodology is employed to generate
synthetic TEM images for training. This concept-oriented synthetic data
approach can be extended to any field of interest where the scarcity of data is
a challenge. The presented model was applied to various metals with average
grain sizes down to the nanoscale, producing grain morphologies from
low-resolution TEM images that are comparable to those obtained from advanced
and demanding experimental techniques with an average accuracy of 97.23%.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:46:28 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Saleh",
"Ahmed Sobhi",
""
],
[
"Croes",
"Kristof",
""
],
[
"Ceric",
"Hajdin",
""
],
[
"De Wolf",
"Ingrid",
""
],
[
"Zahedmanesh",
"Houman",
""
]
] |
2504.14783 | Wenhui Zhu | Wenhui Zhu, Peijie Qiu, Xiwen Chen, Zhangsihao Yang, Aristeidis
Sotiras, Abolfazl Razi, and Yalin Wang | How Effective Can Dropout Be in Multiple Instance Learning ? | null | null | null | null | cs.CV cs.AI eess.IV stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Multiple Instance Learning (MIL) is a popular weakly-supervised method for
various applications, with a particular interest in histological whole slide
image (WSI) classification. Due to the gigapixel resolution of WSI,
applications of MIL in WSI typically necessitate a two-stage training scheme:
first, extract features from the pre-trained backbone and then perform MIL
aggregation. However, it is well-known that this suboptimal training scheme
suffers from "noisy" feature embeddings from the backbone and inherent weak
supervision, hindering MIL from learning rich and generalizable features.
However, the most commonly used technique (i.e., dropout) for mitigating this
issue has yet to be explored in MIL. In this paper, we empirically explore how
effective the dropout can be in MIL. Interestingly, we observe that dropping
the top-k most important instances within a bag leads to better performance and
generalization even under noise attack. Based on this key observation, we
propose a novel MIL-specific dropout method, termed MIL-Dropout, which
systematically determines which instances to drop. Experiments on five MIL
benchmark datasets and two WSI datasets demonstrate that MIL-Dropout boosts the
performance of current MIL methods with a negligible computational cost. The
code is available at https://github.com/ChongQingNoSubway/MILDropout.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:46:31 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhu",
"Wenhui",
""
],
[
"Qiu",
"Peijie",
""
],
[
"Chen",
"Xiwen",
""
],
[
"Yang",
"Zhangsihao",
""
],
[
"Sotiras",
"Aristeidis",
""
],
[
"Razi",
"Abolfazl",
""
],
[
"Wang",
"Yalin",
""
]
] |
2504.14784 | Piyush A | Piyush Akimitsu | Effects of Price Regulations on Service Utilization and Public Insurance
Costs: Evidence from Telehealth Parity Laws | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | This study explores Telehealth Parity Laws (TPLs) and their heterogeneous
treatment effects by policy type on outpatient utilization and Medicare costs,
considering broadband and licensure infrastructure. State-specific legislative
framings create varied Price (physician reimbursement) and Cost (consumer
expense) control combinations within a quasi-experimental design. Partial
equilibrium causal estimates reveal negligible impacts on Medicare enrollment,
indicating that Medicare cost shifts stem purely from outpatient utilization
changes. Broadband access correlates with increased preventable hospital stays
but lower Medicare costs and enrollment. Additionally, the Interstate Licensure
Compact increases enrollment among the aged and disabled, possibly addressing
previously unmet demand for healthcare services.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:48:57 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Akimitsu",
"Piyush",
""
]
] |
2504.14785 | Zhenyu Yu | Zhenyu Yu, Mohd Yamani Idna Idris, Pei Wang | When Cloud Removal Meets Diffusion Model in Remote Sensing | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cloud occlusion significantly hinders remote sensing applications by
obstructing surface information and complicating analysis. To address this, we
propose DC4CR (Diffusion Control for Cloud Removal), a novel multimodal
diffusion-based framework for cloud removal in remote sensing imagery. Our
method introduces prompt-driven control, allowing selective removal of thin and
thick clouds without relying on pre-generated cloud masks, thereby enhancing
preprocessing efficiency and model adaptability. Additionally, we integrate
low-rank adaptation for computational efficiency, subject-driven generation for
improved generalization, and grouped learning to enhance performance on small
datasets. Designed as a plug-and-play module, DC4CR seamlessly integrates into
existing cloud removal models, providing a scalable and robust solution.
Extensive experiments on the RICE and CUHK-CR datasets demonstrate
state-of-the-art performance, achieving superior cloud removal across diverse
conditions. This work presents a practical and efficient approach for remote
sensing image processing with broad real-world applications.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 00:56:57 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yu",
"Zhenyu",
""
],
[
"Idris",
"Mohd Yamani Idna",
""
],
[
"Wang",
"Pei",
""
]
] |
2504.14786 | Mayira Sharif | Mayira Sharif, Guangzeng Han, Weisi Liu and Xiaolei Huang | Cultivating Multidisciplinary Research and Education on GPU
Infrastructure for Mid-South Institutions at the University of Memphis:
Practice and Challenge | PEARC'25: Workforce Development, Training, Education and Community -
Short Papers | null | null | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | To support rapid scientific advancement and promote access to large-scale
computing resources for under-resourced institutions at the Mid-South region,
the University of Memphis (UofM) established the first regional mid-scale GPU
cluster, iTiger, a valuable high-performance computing (HPC) infrastructure. In
this study, we present our continuous efforts to manage the critical
cyberinfrastructure and provide essential computing supports for educators,
students, and researchers in AI, data sciences, and related scientific fields
in the Mid-South region, such as precision agriculture, smart transportation,
and health informatics. We outline our initiatives to broaden CI adoptions
across regional computing-related scientific and engineering fields, such as
seed grant, workshop trainings, course integration, and other outreach
activities. While we've observed promising outcomes of regional CI adoptions,
we will discuss insights and challenges of Mid-South CI users, which can
inspire other institutions to implement similar programs.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:01:18 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Sharif",
"Mayira",
""
],
[
"Han",
"Guangzeng",
""
],
[
"Liu",
"Weisi",
""
],
[
"Huang",
"Xiaolei",
""
]
] |
2504.14787 | Sirui Zeng | Sirui Zeng, Xifeng Yan | ADL: A Declarative Language for Agent-Based Chatbots | null | null | null | null | cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There are numerous agent frameworks capable of creating and orchestrating
agents to address complex tasks. However, these frameworks are often too
complicated for customer service professionals, who may not have much
programming experience but still need an easy way to create chatbots with rich
business logic. In this work, we introduce ADL, a Declarative Language for
Agent-Based Chatbots. ADL simplifies chatbot development by using natural
language programming at its core, making it easier for a broad audience to
customize and build task-oriented chatbots. It includes four types of agents
and supports integration with custom functions, tool use, and third-party
agents. ADL abstracts away implementation details, offering a declarative way
to define agents and their interactions, which could ease prompt engineering,
testing and debugging. MICA, a multi-agent system designed to interpret and
execute ADL programs, has been developed and is now available as an open-source
project at https://github.com/Mica-labs/MICA. Its user documentation can be
found at https://mica-labs.github.io/.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:02:08 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zeng",
"Sirui",
""
],
[
"Yan",
"Xifeng",
""
]
] |
2504.14788 | Junchen Fu | Junchen Fu, Xuri Ge, Xin Xin, Haitao Yu, Yue Feng, Alexandros
Karatzoglou, Ioannis Arapakis, Joemon M. Jose | The 1st EReL@MIR Workshop on Efficient Representation Learning for
Multimodal Information Retrieval | WWW2025 Workshop Summary | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal representation learning has garnered significant attention in the
AI community, largely due to the success of large pre-trained multimodal
foundation models like LLaMA, GPT, Mistral, and CLIP. These models have
achieved remarkable performance across various tasks of multimodal information
retrieval (MIR), including web search, cross-modal retrieval, and recommender
systems, etc. However, due to their enormous parameter sizes, significant
efficiency challenges emerge across training, deployment, and inference stages
when adapting these models' representation for IR tasks. These challenges
present substantial obstacles to the practical adaptation of foundation models
for representation learning in information retrieval tasks.
To address these pressing issues, we propose organizing the first EReL@MIR
workshop at the Web Conference 2025, inviting participants to explore novel
solutions, emerging problems, challenges, efficiency evaluation metrics and
benchmarks. This workshop aims to provide a platform for both academic and
industry researchers to engage in discussions, share insights, and foster
collaboration toward achieving efficient and effective representation learning
for multimodal information retrieval in the era of large foundation models.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:10:59 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Fu",
"Junchen",
""
],
[
"Ge",
"Xuri",
""
],
[
"Xin",
"Xin",
""
],
[
"Yu",
"Haitao",
""
],
[
"Feng",
"Yue",
""
],
[
"Karatzoglou",
"Alexandros",
""
],
[
"Arapakis",
"Ioannis",
""
],
[
"Jose",
"Joemon M.",
""
]
] |
2504.14789 | Dong Yang | Zongzhen Xie, Dong Yang and Houjun Zhang | Classification of silted algebras for two quivers of Dynkin type
$\mathbb{A}_{n}$ | null | null | null | null | math.RT math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we give a complete classification of silted algebras for the
quiver $\overrightarrow{\mathbb{A}}_{n}$ of type $\mathbb{A}_{n}$ with linear
orientation and for the quiver obtained from $\overrightarrow{\mathbb{A}}_{n}$
by reversing the arrow at the unique source. Based on the classification, we
also compute the number of silted algebras for these two quivers.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:22:01 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xie",
"Zongzhen",
""
],
[
"Yang",
"Dong",
""
],
[
"Zhang",
"Houjun",
""
]
] |
2504.14790 | Jun Yang | Jun Yang, Shintaro Yamasaki | Enhanced Data-driven Topology Design Methodology with Multi-level Mesh
and Correlation-based Mutation for Stress-related Multi-objective
Optimization | 23 pages, 22 figures | null | null | null | cs.LG cs.NA math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Topology optimization (TO) serves as a widely applied structural design
approach to tackle various engineering problems. Nevertheless,
sensitivity-based TO methods usually struggle with solving strongly nonlinear
optimization problems. By leveraging high capacity of deep generative model,
which is an influential machine learning technique, the sensitivity-free
data-driven topology design (DDTD) methodology is regarded as an effective
means of overcoming these issues. The DDTD methodology depends on initial
dataset with a certain regularity, making its results highly sensitive to
initial dataset quality. This limits its effectiveness and generalizability,
especially for optimization problems without priori information. In this
research, we proposed a multi-level mesh DDTD-based method with
correlation-based mutation module to escape from the limitation of the quality
of the initial dataset on the results and enhance computational efficiency. The
core is to employ a correlation-based mutation module to assign new geometric
features with physical meaning to the generated data, while utilizing a
multi-level mesh strategy to progressively enhance the refinement of the
structural representation, thus avoiding the maintenance of a high
degree-of-freedom (DOF) representation throughout the iterative process. The
proposed multi-level mesh DDTD-based method can be driven by a low quality
initial dataset without the need for time-consuming construction of a specific
dataset, thus significantly increasing generality and reducing application
difficulty, while further lowering computational cost of DDTD methodology.
Various comparison experiments with the traditional sensitivity-based TO
methods on stress-related strongly nonlinear problems demonstrate the
generality and effectiveness of the proposed method.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:33:56 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yang",
"Jun",
""
],
[
"Yamasaki",
"Shintaro",
""
]
] |
2504.14791 | Xing Wang | Xing M. Wang | Quantum Measurement Without Collapse or Many Worlds: The Branched
Hilbert Subspace Interpretation | 10 pages | null | null | null | quant-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The interpretation of quantum measurements presents a fundamental challenge
in quantum mechanics, with concepts such as the Copenhagen Interpretation (CI),
Many-Worlds Interpretation (MWI), and Bohmian Mechanics (BM) offering distinct
perspectives. We propose the Branched Hilbert Subspace Interpretation (BHSI),
which describes measurement as branching the local Hilbert space of a system
into parallel subspaces. We formalize the mathematical framework of BHSI using
branching and the engaging and disengaging unitary operators to relationally
and causally update the states of observers. Unlike the MWI, BHSI avoids the
ontological proliferation of worlds and copies of observers, realizing the Born
rule based on branch weights. Unlike the CI, BHSI retains the essential
features of the MWI: unitary evolution and no wavefunction collapse. Unlike the
BM, BHSI does not depend on a nonlocal structure, which may conflict with
relativity. We apply BHSI to examples such as the double-slit experiment, the
Bell test, Wagner and his friend, and the black hole information paradox. In
addition, we explore whether recohering branches can be achieved in BHSI.
Compared to the CI and MWI, BHSI provides a minimalist, unitarity-preserving,
collapse-free, and probabilistically inherent alternative interpretation of
quantum measurements.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:34:57 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Wang",
"Xing M.",
""
]
] |
2504.14792 | Shidong Liu | Shidong Liu, Qi Wu, Gang Li | Dipionic transitions of $Y(4500)$ to $J/\psi$ | 15 pages, 7 figures, 2 tables, comments welcome | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We, using an effective Lagrangian approach, investigated the dipionic
transition of the $Y(4500)$ newly well-established in the process $e^+e^-\to
K^+K^- J/\psi$. In this study, the $Y(4500)$ was considered as a mixture of the
$\psi(5S)$ and $\psi(4\,{^3D_1})$, and its dipionic decay to the $J/\psi$ was
assumed to occur via charmed meson loops including box and two kinds of
triangle loops. The calculated invariant mass spectra due to the different loop
mechanisms exhibit distinct differences. These spectrum patterns are found to
be nearly independent of the cutoff parameter, especially for the box and
$\mathcal{R}$ loops. By means of comparing our calculated results with the
future experimental measurements of the $\pi^+\pi^-$ and $\pi^\pm J/\psi$
invariant mass distributions, we could judge which loop mechanism is of more
importance in this dipionic transition. The interference among the different
kinds of loops is also exhibited. Despite the unknown phase angles between
different kinds of loop diagrams, the partial decay width for the process
$Y(4500)\to\pi^+\pi^- J/\psi$ is estimated to be 250-700 keV, aligning with the
estimation obtained by combining BESIII Collaboration measurements and
theoretical predictions for $Y(4500)\to K^+K^- J/\psi$. We hope that the
present calculations would be tested by the future BESIII or Belle II
experiments.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:37:07 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Liu",
"Shidong",
""
],
[
"Wu",
"Qi",
""
],
[
"Li",
"Gang",
""
]
] |
2504.14793 | Ariana Tang | Ophir Friedler, Hu Fu, Anna Karlin, Ariana Tang | Price Stability and Improved Buyer Utility with Presentation Design: A
Theoretical Study of the Amazon Buy Box | Accepted at the Web Conference (WWW) 2025 | null | null | null | cs.GT econ.TH | http://creativecommons.org/licenses/by/4.0/ | Platforms design the form of presentation by which sellers are shown to the
buyers. This design not only shapes the buyers' experience but also leads to
different market equilibria or dynamics. One component in this design is
through the platform's mediation of the search frictions experienced by the
buyers for different sellers. We take a model of monopolistic competition and
show that, on one hand, when all sellers have the same inspection costs, the
market sees no stable price since the sellers always have incentives to
undercut each other, and, on the other hand, the platform may stabilize the
price by giving prominence to one seller chosen by a carefully designed
mechanism. This calls to mind Amazon's Buy Box. We study natural mechanisms for
choosing the prominent seller, characterize the range of equilibrium prices
implementable by them, and find that in certain scenarios the buyers' surplus
improves as the search friction increases.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:45:12 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Friedler",
"Ophir",
""
],
[
"Fu",
"Hu",
""
],
[
"Karlin",
"Anna",
""
],
[
"Tang",
"Ariana",
""
]
] |
2504.14794 | Tian Xia | Tian Xia, Quanzhou Long, Wanlong Zhang, Zhenwei Xie and Xiaocong Yuan | High-Dimensional Light Field Modulation via Conjugate Phase Encoding in
Liquid Crystal Devices | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by-nc-nd/4.0/ | High-dimensional light field modulation demands precise control over multiple
optical parameters, a capability critical for next-generation photonic systems.
While liquid crystals offer inherent advantages in dynamic birefringence
tuning, existing approaches face fundamental limitations in decoupling
interdependent phase responses across polarization states. Here, we demonstrate
a conjugate phase encoding paradigm enabling simultaneous manipulation of
wavelength-dependent wavefronts, orbital angular momentum (OAM), and
polarization states in photoaligned liquid crystal devices.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:49:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xia",
"Tian",
""
],
[
"Long",
"Quanzhou",
""
],
[
"Zhang",
"Wanlong",
""
],
[
"Xie",
"Zhenwei",
""
],
[
"Yuan",
"Xiaocong",
""
]
] |
2504.14795 | Shin-Ichi Maeda | Ryu Tadokoro, Tsukasa Takagi, Shin-ichi Maeda | Segmentation with Noisy Labels via Spatially Correlated Distributions | null | null | null | null | eess.IV cs.CV cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In semantic segmentation, the accuracy of models heavily depends on the
high-quality annotations. However, in many practical scenarios such as medical
imaging and remote sensing, obtaining true annotations is not straightforward
and usually requires significant human labor. Relying on human labor often
introduces annotation errors, including mislabeling, omissions, and
inconsistency between annotators. In the case of remote sensing, differences in
procurement time can lead to misaligned ground truth annotations. These label
errors are not independently distributed, and instead usually appear in
spatially connected regions where adjacent pixels are more likely to share the
same errors. To address these issues, we propose an approximate Bayesian
estimation based on a probabilistic model that assumes training data includes
label errors, incorporating the tendency for these errors to occur with spatial
correlations between adjacent pixels. Bayesian inference requires computing the
posterior distribution of label errors, which becomes intractable when spatial
correlations are present. We represent the correlation of label errors between
adjacent pixels through a Gaussian distribution whose covariance is structured
by a Kac-Murdock-Szeg\"{o} (KMS) matrix, solving the computational challenges.
Through experiments on multiple segmentation tasks, we confirm that leveraging
the spatial correlation of label errors significantly improves performance.
Notably, in specific tasks such as lung segmentation, the proposed method
achieves performance comparable to training with clean labels under moderate
noise levels. Code is available at
https://github.com/pfnet-research/Bayesian_SpatialCorr.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:50:10 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Tadokoro",
"Ryu",
""
],
[
"Takagi",
"Tsukasa",
""
],
[
"Maeda",
"Shin-ichi",
""
]
] |
2504.14796 | David Yang | David Yang, Mostafa Abdelmegeed, John Modl, Minjeong Kim | Edge-boosted graph learning for functional brain connectivity analysis | Accepted at IEEE International Symposium on Biomedical Imaging (ISBI)
2025, 4 pages | null | null | null | cs.LG eess.IV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Predicting disease states from functional brain connectivity is critical for
the early diagnosis of severe neurodegenerative diseases such as Alzheimer's
Disease and Parkinson's Disease. Existing studies commonly employ Graph Neural
Networks (GNNs) to infer clinical diagnoses from node-based brain connectivity
matrices generated through node-to-node similarities of regionally averaged
fMRI signals. However, recent neuroscience studies found that such node-based
connectivity does not accurately capture ``functional connections" within the
brain. This paper proposes a novel approach to brain network analysis that
emphasizes edge functional connectivity (eFC), shifting the focus to inter-edge
relationships. Additionally, we introduce a co-embedding technique to integrate
edge functional connections effectively. Experimental results on the ADNI and
PPMI datasets demonstrate that our method significantly outperforms
state-of-the-art GNN methods in classifying functional brain networks.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:53:55 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yang",
"David",
""
],
[
"Abdelmegeed",
"Mostafa",
""
],
[
"Modl",
"John",
""
],
[
"Kim",
"Minjeong",
""
]
] |
2504.14797 | Armin Moin | Clare E. Laney, Andrew Barovic and Armin Moin | Automated Duplicate Bug Report Detection in Large Open Bug Repositories | IEEE COMPSAC 2025 | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Many users and contributors of large open-source projects report software
defects or enhancement requests (known as bug reports) to the issue-tracking
systems. However, they sometimes report issues that have already been reported.
First, they may not have time to do sufficient research on existing bug
reports. Second, they may not possess the right expertise in that specific area
to realize that an existing bug report is essentially elaborating on the same
matter, perhaps with a different wording. In this paper, we propose a novel
approach based on machine learning methods that can automatically detect
duplicate bug reports in an open bug repository based on the textual data in
the reports. We present six alternative methods: Topic modeling, Gaussian Naive
Bayes, deep learning, time-based organization, clustering, and summarization
using a generative pre-trained transformer large language model. Additionally,
we introduce a novel threshold-based approach for duplicate identification, in
contrast to the conventional top-k selection method that has been widely used
in the literature. Our approach demonstrates promising results across all the
proposed methods, achieving accuracy rates ranging from the high 70%'s to the
low 90%'s. We evaluated our methods on a public dataset of issues belonging to
an Eclipse open-source project.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:55:54 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Laney",
"Clare E.",
""
],
[
"Barovic",
"Andrew",
""
],
[
"Moin",
"Armin",
""
]
] |
2504.14798 | Hao Xuan | Hao Xuan, Xingyu Li | Verifying Robust Unlearning: Probing Residual Knowledge in Unlearned
Models | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Machine Unlearning (MUL) is crucial for privacy protection and content
regulation, yet recent studies reveal that traces of forgotten information
persist in unlearned models, enabling adversaries to resurface removed
knowledge. Existing verification methods only confirm whether unlearning was
executed, failing to detect such residual information leaks. To address this,
we introduce the concept of Robust Unlearning, ensuring models are
indistinguishable from retraining and resistant to adversarial recovery. To
empirically evaluate whether unlearning techniques meet this security standard,
we propose the Unlearning Mapping Attack (UMA), a post-unlearning verification
framework that actively probes models for forgotten traces using adversarial
queries. Extensive experiments on discriminative and generative tasks show that
existing unlearning techniques remain vulnerable, even when passing existing
verification metrics. By establishing UMA as a practical verification tool,
this study sets a new standard for assessing and enhancing machine unlearning
security.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:56:15 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xuan",
"Hao",
""
],
[
"Li",
"Xingyu",
""
]
] |
2504.14800 | Shuxian Zhao | Shuxian Zhao, Jie Gui, Minjing Dong, Baosheng Yu, Zhipeng Gui, Lu
Dong, Yuan Yan Tang, and James Tin-Yau Kwok | A Survey on Small Sample Imbalance Problem: Metrics, Feature Analysis,
and Solutions | null | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The small sample imbalance (S&I) problem is a major challenge in machine
learning and data analysis. It is characterized by a small number of samples
and an imbalanced class distribution, which leads to poor model performance. In
addition, indistinct inter-class feature distributions further complicate
classification tasks. Existing methods often rely on algorithmic heuristics
without sufficiently analyzing the underlying data characteristics. We argue
that a detailed analysis from the data perspective is essential before
developing an appropriate solution. Therefore, this paper proposes a systematic
analytical framework for the S\&I problem. We first summarize imbalance metrics
and complexity analysis methods, highlighting the need for interpretable
benchmarks to characterize S&I problems. Second, we review recent solutions for
conventional, complexity-based, and extreme S&I problems, revealing
methodological differences in handling various data distributions. Our summary
finds that resampling remains a widely adopted solution. However, we conduct
experiments on binary and multiclass datasets, revealing that classifier
performance differences significantly exceed the improvements achieved through
resampling. Finally, this paper highlights open questions and discusses future
trends.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:58:29 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhao",
"Shuxian",
""
],
[
"Gui",
"Jie",
""
],
[
"Dong",
"Minjing",
""
],
[
"Yu",
"Baosheng",
""
],
[
"Gui",
"Zhipeng",
""
],
[
"Dong",
"Lu",
""
],
[
"Tang",
"Yuan Yan",
""
],
[
"Kwok",
"James Tin-Yau",
""
]
] |
2504.14801 | Gyujin Oh | Gyujin Oh, Koji Shimizu | Moduli stacks of crystals and isocrystals | 111 pages | null | null | null | math.NT math.AG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Given a liftable smooth proper variety over $\mathbb{F}_p$, we construct the
moduli stacks of crystals and isocrystals on it. We show that the former is a
formal algebraic stack over $\mathbb{Z}_p$ and the latter is an adic stack --
Artin stack in rigid geometry -- over $\mathbb{Q}_p$. Both stacks come equipped
with the Verschiebung endomorphism $V$ corresponding to the Frobenius pullback
of (iso)crystals. We study the geometry of the $V$-fixed points over the open
substack of irreducible isocrystals, which we use to geometrically count the
rank one $F$-isocrystals. Along the way, we carefully develop the theory of
adic stacks.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:01:55 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Oh",
"Gyujin",
""
],
[
"Shimizu",
"Koji",
""
]
] |
2504.14802 | Ji-Yong Shin | Kezhi Xiong, Soonwon Moon, Joshua Kang, Bryant Curto, Jieung Kim,
Ji-Yong Shin | ReCraft: Self-Contained Split, Merge, and Membership Change of Raft
Protocol | null | The 55th Annual IEEE/IFIP International Conference on Dependable
Systems and Networks (2025) | null | null | cs.DC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Designing reconfiguration schemes for consensus protocols is challenging
because subtle corner cases during reconfiguration could invalidate the
correctness of the protocol. Thus, most systems that embed consensus protocols
conservatively implement the reconfiguration and refrain from developing an
efficient scheme. Existing implementations often stop the entire system during
reconfiguration and rely on a centralized coordinator, which can become a
single point of failure. We present ReCraft, a novel reconfiguration protocol
for Raft, which supports multi- and single-cluster-level reconfigurations.
ReCraft does not rely on external coordinators and blocks minimally. ReCraft
enables the sharding of Raft clusters with split and merge reconfigurations and
adds a membership change scheme that improves Raft. We prove the safety and
liveness of ReCraft and demonstrate its efficiency through implementations in
etcd.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:05:06 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xiong",
"Kezhi",
""
],
[
"Moon",
"Soonwon",
""
],
[
"Kang",
"Joshua",
""
],
[
"Curto",
"Bryant",
""
],
[
"Kim",
"Jieung",
""
],
[
"Shin",
"Ji-Yong",
""
]
] |
2504.14803 | Jingru Zhang | Haitao Xu and Jingru Zhang | The k-Center Problem of Uncertain Points on Graphs | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study the $k$-center problem of uncertain points on a
graph. Given are an undirected graph $G = (V, E)$ and a set $\mathcal{P}$ of
$n$ uncertain points where each uncertain point with a non-negative weight has
$m$ possible locations on $G$ each associated with a probability. The problem
aims to find $k$ centers (points) on $G$ so as to minimize the maximum weighted
expected distance of uncertain points to their expected closest centers. No
previous work exist for the $k$-center problem of uncertain points on
undirected graphs. We propose exact algorithms that solve respectively the case
of $k=2$ in $O(|E|^2m^2n\log |E|mn\log mn )$ time and the problem with $k\geq
3$ in $O(\min\{|E|^km^kn^{k+1}k\log |E|mn\log m,
|E|^kn^\frac{k}{2}m^\frac{k^2}{2}\log |E|mn\})$ time, provided with the
distance matrix of $G$. In addition, an $O(|E|mn\log mn)$-time algorithmic
approach is given for the one-center case.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:05:53 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xu",
"Haitao",
""
],
[
"Zhang",
"Jingru",
""
]
] |
2504.14804 | Jiaxin Guo | Jiaxin GUO, Xiaoyu Chen, Zhiqiang Rao, Jinlong Yang, Zongyao Li,
Hengchao Shang, Daimeng Wei and Hao Yang | Automatic Evaluation Metrics for Document-level Translation: Overview,
Challenges and Trends | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | With the rapid development of deep learning technologies, the field of
machine translation has witnessed significant progress, especially with the
advent of large language models (LLMs) that have greatly propelled the
advancement of document-level translation. However, accurately evaluating the
quality of document-level translation remains an urgent issue. This paper first
introduces the development status of document-level translation and the
importance of evaluation, highlighting the crucial role of automatic evaluation
metrics in reflecting translation quality and guiding the improvement of
translation systems. It then provides a detailed analysis of the current state
of automatic evaluation schemes and metrics, including evaluation methods with
and without reference texts, as well as traditional metrics, Model-based
metrics and LLM-based metrics. Subsequently, the paper explores the challenges
faced by current evaluation methods, such as the lack of reference diversity,
dependence on sentence-level alignment information, and the bias, inaccuracy,
and lack of interpretability of the LLM-as-a-judge method. Finally, the paper
looks ahead to the future trends in evaluation methods, including the
development of more user-friendly document-level evaluation methods and more
robust LLM-as-a-judge methods, and proposes possible research directions, such
as reducing the dependency on sentence-level information, introducing
multi-level and multi-granular evaluation approaches, and training models
specifically for machine translation evaluation. This study aims to provide a
comprehensive analysis of automatic evaluation for document-level translation
and offer insights into future developments.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:08:42 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"GUO",
"Jiaxin",
""
],
[
"Chen",
"Xiaoyu",
""
],
[
"Rao",
"Zhiqiang",
""
],
[
"Yang",
"Jinlong",
""
],
[
"Li",
"Zongyao",
""
],
[
"Shang",
"Hengchao",
""
],
[
"Wei",
"Daimeng",
""
],
[
"Yang",
"Hao",
""
]
] |
2504.14805 | Jinwoo Choi | Jinwoo Choi, Seung-Woo Seo | Dynamic Contrastive Skill Learning with State-Transition Based Skill
Clustering and Dynamic Length Adjustment | ICLR 2025; 23 pages, 12 figures | null | null | null | cs.LG cs.AI cs.RO | http://creativecommons.org/licenses/by/4.0/ | Reinforcement learning (RL) has made significant progress in various domains,
but scaling it to long-horizon tasks with complex decision-making remains
challenging. Skill learning attempts to address this by abstracting actions
into higher-level behaviors. However, current approaches often fail to
recognize semantically similar behaviors as the same skill and use fixed skill
lengths, limiting flexibility and generalization. To address this, we propose
Dynamic Contrastive Skill Learning (DCSL), a novel framework that redefines
skill representation and learning. DCSL introduces three key ideas:
state-transition based skill representation, skill similarity function
learning, and dynamic skill length adjustment. By focusing on state transitions
and leveraging contrastive learning, DCSL effectively captures the semantic
context of behaviors and adapts skill lengths to match the appropriate temporal
extent of behaviors. Our approach enables more flexible and adaptive skill
extraction, particularly in complex or noisy datasets, and demonstrates
competitive performance compared to existing methods in task completion and
efficiency.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:11:39 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Choi",
"Jinwoo",
""
],
[
"Seo",
"Seung-Woo",
""
]
] |
2504.14806 | Xiongwei Zhao | Xiongwei Zhao, Yang Wang, Qihao Sun, Haojie Bai and Xingxiang Xie | An Iterative Task-Driven Framework for Resilient LiDAR Place Recognition
in Adverse Weather | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | LiDAR place recognition (LPR) plays a vital role in autonomous navigation.
However, existing LPR methods struggle to maintain robustness under adverse
weather conditions such as rain, snow, and fog, where weather-induced noise and
point cloud degradation impair LiDAR reliability and perception accuracy. To
tackle these challenges, we propose an Iterative Task-Driven Framework
(ITDNet), which integrates a LiDAR Data Restoration (LDR) module and a LiDAR
Place Recognition (LPR) module through an iterative learning strategy. These
modules are jointly trained end-to-end, with alternating optimization to
enhance performance. The core rationale of ITDNet is to leverage the LDR module
to recover the corrupted point clouds while preserving structural consistency
with clean data, thereby improving LPR accuracy in adverse weather.
Simultaneously, the LPR task provides feature pseudo-labels to guide the LDR
module's training, aligning it more effectively with the LPR task. To achieve
this, we first design a task-driven LPR loss and a reconstruction loss to
jointly supervise the optimization of the LDR module. Furthermore, for the LDR
module, we propose a Dual-Domain Mixer (DDM) block for frequency-spatial
feature fusion and a Semantic-Aware Generator (SAG) block for semantic-guided
restoration. In addition, for the LPR module, we introduce a Multi-Frequency
Transformer (MFT) block and a Wavelet Pyramid NetVLAD (WPN) block to aggregate
multi-scale, robust global descriptors. Finally, extensive experiments on the
Weather-KITTI, Boreas, and our proposed Weather-Apollo datasets demonstrate
that, demonstrate that ITDNet outperforms existing LPR methods, achieving
state-of-the-art performance in adverse weather. The datasets and code will be
made publicly available at https://github.com/Grandzxw/ITDNet.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:15:27 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhao",
"Xiongwei",
""
],
[
"Wang",
"Yang",
""
],
[
"Sun",
"Qihao",
""
],
[
"Bai",
"Haojie",
""
],
[
"Xie",
"Xingxiang",
""
]
] |
2504.14807 | Deepak Ghimire | Deepak Ghimire, Sunghwan Jeong, Sunhong Yoon, Sanghyun Park, Juhwan
Choi | Real-Time Sleepiness Detection for Driver State Monitoring System | 8 pages, published in GST 2015 | Advanced Science and Technology Letters, 120, 1-8 (2015) | null | null | cs.CV cs.HC cs.LG | http://creativecommons.org/licenses/by/4.0/ | A driver face monitoring system can detect driver fatigue, which is a
significant factor in many accidents, using computer vision techniques. In this
paper, we present a real-time technique for driver eye state detection. First,
the face is detected, and the eyes are located within the face region for
tracking. A normalized cross-correlation-based online dynamic template matching
technique, combined with Kalman filter tracking, is proposed to track the
detected eye positions in subsequent image frames. A support vector machine
with histogram of oriented gradients (HOG) features is used to classify the
state of the eyes as open or closed. If the eyes remain closed for a specified
period, the driver is considered to be asleep, and an alarm is triggered.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:15:37 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ghimire",
"Deepak",
""
],
[
"Jeong",
"Sunghwan",
""
],
[
"Yoon",
"Sunhong",
""
],
[
"Park",
"Sanghyun",
""
],
[
"Choi",
"Juhwan",
""
]
] |
2504.14808 | Mario M. Kubek | Mario M. Kubek, Shiraj Pokharel, Thomas B\"ohme, Emma L. McDaniel,
Herwig Unger, Armin R. Mikler | On Self-improving Token Embeddings | 18 pages, 4 figures, 3 tables, accepted at the 2025 25th
International Conference on Innovations for Community Services (I4CS), June
11 - 13, Munich, Germany, 2025 | null | null | null | cs.CL cs.AI cs.IR cs.LG | http://creativecommons.org/licenses/by/4.0/ | This article introduces a novel and fast method for refining pre-trained
static word or, more generally, token embeddings. By incorporating the
embeddings of neighboring tokens in text corpora, it continuously updates the
representation of each token, including those without pre-assigned embeddings.
This approach effectively addresses the out-of-vocabulary problem, too.
Operating independently of large language models and shallow neural networks,
it enables versatile applications such as corpus exploration, conceptual
search, and word sense disambiguation. The method is designed to enhance token
representations within topically homogeneous corpora, where the vocabulary is
restricted to a specific domain, resulting in more meaningful embeddings
compared to general-purpose pre-trained vectors. As an example, the methodology
is applied to explore storm events and their impacts on infrastructure and
communities using narratives from a subset of the NOAA Storm Events database.
The article also demonstrates how the approach improves the representation of
storm-related terms over time, providing valuable insights into the evolving
nature of disaster narratives.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:17:19 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Kubek",
"Mario M.",
""
],
[
"Pokharel",
"Shiraj",
""
],
[
"Böhme",
"Thomas",
""
],
[
"McDaniel",
"Emma L.",
""
],
[
"Unger",
"Herwig",
""
],
[
"Mikler",
"Armin R.",
""
]
] |
2504.14810 | Jucheng Hu | Jucheng Hu, Surong Yang, Dongzhan Zhou, Lijun Wu | DONOD: Robust and Generalizable Instruction Fine-Tuning for LLMs via
Model-Intrinsic Dataset Pruning | null | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Ad-hoc instruction fine-tuning of large language models (LLMs) is widely
adopted for domain-specific adaptation. While domain-specific supervised
fine-tuning (SFT) is effective and efficient, it often weakens cross-domain
generalization and struggles with noisy training data. To address these
challenges, we propose DONOD, a lightweight model-intrinsic data pruning
method. Our approach evaluates data using two model-parameter-based metrics:
Delta of Norm (DON), which captures the cumulative influence on model weights,
and Norm of Delta (NOD), which quantifies weight instability. Moreover, by
employing the Technique for Order of Preference by Similarity to Ideal Solution
(TOPSIS) algorithm, we effectively filter noisy, unlearnable, and
generalization-harming samples without relying on auxiliary models during the
SFT process. Experiments on mathematical tasks demonstrate that data selected
by DONOD achieve superior fine-tuning efficiency and improved robustness
against noisy data. By filtering out 70% of the full dataset, we improve
target-domain accuracy by 14.90% and cross-domain accuracy by 5.67%. Meanwhile,
our selected data present superior cross-architecture generalization. Data
pruned by smaller models (e.g., Llama 3.1-8B) generalize effectively on larger
models (e.g., Llama 2-13B). Compared to existing related methodologies, DONOD
demonstrates comparable or superior performance while remaining
dataset-agnostic, enabling broader applicability.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:25:03 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Hu",
"Jucheng",
""
],
[
"Yang",
"Surong",
""
],
[
"Zhou",
"Dongzhan",
""
],
[
"Wu",
"Lijun",
""
]
] |
2504.14811 | Bowen Yang | Bowen Yang | Categorifying Clifford QCA | 31 pages | null | null | null | math-ph math.MP quant-ph | http://creativecommons.org/licenses/by/4.0/ | We provide a complete classification of Clifford quantum cellular automata
(QCAs) on arbitrary metric spaces and any qudits (of prime or composite
dimensions) in terms of algebraic L-theory. Building on the delooping formalism
of Pedersen and Weibel, we reinterpret Clifford QCAs as symmetric formations in
a filtered additive category constructed from the geometry of the underlying
space. This perspective allows us to identify the group of stabilized Clifford
QCAs, modulo circuits and separated automorphisms, with the Witt group of the
corresponding Pedersen--Weibel category. For Euclidean lattices, the
classification reproduces and expands upon known results, while for more
general spaces -- including open cones over finite simplicial complexes -- we
relate nontrivial QCAs to generalized homology theories with coefficients in
the L-theory spectrum. We also outline extensions to QCAs with symmetry and
mixed qudit dimensions, and discuss how these fit naturally into the
L-theoretic framework.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:27:52 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yang",
"Bowen",
""
]
] |
2504.14812 | Yangyang Gu | Yangyang Gu, Xianglong Li, Haolin Wu, Jing Chen, Kun He, Ruiying Du,
Cong Wu | CSI2Dig: Recovering Digit Content from Smartphone Loudspeakers Using
Channel State Information | 14 pages, 14 figures | null | null | null | cs.CR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Eavesdropping on sounds emitted by mobile device loudspeakers can capture
sensitive digital information, such as SMS verification codes, credit card
numbers, and withdrawal passwords, which poses significant security risks.
Existing schemes either require expensive specialized equipment, rely on
spyware, or are limited to close-range signal acquisition. In this paper, we
propose a scheme, CSI2Dig, for recovering digit content from Channel State
Information (CSI) when digits are played through a smartphone loudspeaker. We
observe that the electromagnetic interference caused by the audio signals from
the loudspeaker affects the WiFi signals emitted by the phone's WiFi antenna.
Building upon contrastive learning and denoising autoencoders, we develop a
two-branch autoencoder network designed to amplify the impact of this
electromagnetic interference on CSI. For feature extraction, we introduce the
TS-Net, a model that captures relevant features from both the temporal and
spatial dimensions of the CSI data. We evaluate our scheme across various
devices, distances, volumes, and other settings. Experimental results
demonstrate that our scheme can achieve an accuracy of 72.97%.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:31:19 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Gu",
"Yangyang",
""
],
[
"Li",
"Xianglong",
""
],
[
"Wu",
"Haolin",
""
],
[
"Chen",
"Jing",
""
],
[
"He",
"Kun",
""
],
[
"Du",
"Ruiying",
""
],
[
"Wu",
"Cong",
""
]
] |
2504.14813 | Hosam Mahmoud | Srinivasan Balaji, Hosam Mahmoud | Urn-driven random walks | null | null | null | null | math.PR | http://creativecommons.org/publicdomain/zero/1.0/ | The symmetric random walk is known to be recurrent in one and two dimensions,
and becomes transient in three or higher dimensions. We compare the symmetric
random walk to walks driven by certain \polya\ urns. We show that, in contrast,
if the probabilities of the random walk are instead driven by a
\polya-Eggenberger urn, the states are recurrent only in one dimension. Further
consideration of exchangeability reveals that the walk is null recurrent. As
soon as the underlying Markov chain of \polya\ walk gets in two dimensions or
higher, there is a positive probability that the walker gets lost in the space,
and the probability of her recurrence is less than 1.
On the other hand, a walk driven by Friedman urn behaves like the symmetric
random walk, being recurrent in one and two dimensions and transient in higher
dimensions. As Friedman urn scheme is not exchangeable, it is considerably
harder to determine the nature of the recurrence in one and two dimensions.
Empirical evidence through simulation suggests that in one dimension Friedman
walk is positive recurrent.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:32:51 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Balaji",
"Srinivasan",
""
],
[
"Mahmoud",
"Hosam",
""
]
] |
2504.14814 | Kazuhisa Fujita Dr. | Kazuhisa Fujita | A Basic Evaluation of Neural Networks Trained with the Error Diffusion
Learning Algorithm | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Artificial neural networks are powerful tools capable of addressing various
tasks. Although the backpropagation algorithm has become a standard training
method for these neural networks, its lack of biological plausibility has
inspired the development of alternative learning approaches. One such
alternative is Kaneko's Error Diffusion Learning Algorithm (EDLA), a
biologically motivated approach wherein a single global error signal diffuses
throughout a network composed of paired excitatory-inhibitory sublayers,
thereby eliminating the necessity for layer-wise backpropagation. This study
presents a contemporary formulation of the EDLA framework and evaluates its
effectiveness through parity check, regression, and image classification tasks.
Our experimental results indicate that EDLA networks can consistently achieve
high accuracy across these benchmarks, with performance efficiency and
convergence speed notably influenced by the choice of learning rate, neuron
count, and network depth. Further investigation of the internal representations
formed by EDLA networks reveals their capacity for meaningful feature
extraction, similar to traditional neural networks. These results suggest that
EDLA is a biologically motivated alternative for training feedforward networks
and will motivate future work on extending this method to biologically inspired
neural networks.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:41:17 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Fujita",
"Kazuhisa",
""
]
] |
2504.14815 | Xiaoyong Yuan | Xiaoyong Yuan, Xiaolong Ma, Linke Guo, Lan Zhang | What Lurks Within? Concept Auditing for Shared Diffusion Models at Scale | 17 pages, 15 figures | null | null | null | cs.LG cs.AI cs.CR cs.CV | http://creativecommons.org/licenses/by/4.0/ | Diffusion models (DMs) have revolutionized text-to-image generation, enabling
the creation of highly realistic and customized images from text prompts. With
the rise of parameter-efficient fine-tuning (PEFT) techniques like LoRA, users
can now customize powerful pre-trained models using minimal computational
resources. However, the widespread sharing of fine-tuned DMs on open platforms
raises growing ethical and legal concerns, as these models may inadvertently or
deliberately generate sensitive or unauthorized content, such as copyrighted
material, private individuals, or harmful content. Despite the increasing
regulatory attention on generative AI, there are currently no practical tools
for systematically auditing these models before deployment. In this paper, we
address the problem of concept auditing: determining whether a fine-tuned DM
has learned to generate a specific target concept. Existing approaches
typically rely on prompt-based input crafting and output-based image
classification but suffer from critical limitations, including prompt
uncertainty, concept drift, and poor scalability. To overcome these challenges,
we introduce Prompt-Agnostic Image-Free Auditing (PAIA), a novel, model-centric
concept auditing framework. By treating the DM as the object of inspection,
PAIA enables direct analysis of internal model behavior, bypassing the need for
optimized prompts or generated images. We evaluate PAIA on 320 controlled model
and 690 real-world community models sourced from a public DM sharing platform.
PAIA achieves over 90% detection accuracy while reducing auditing time by
18-40x compared to existing baselines. To our knowledge, PAIA is the first
scalable and practical solution for pre-deployment concept auditing of
diffusion models, providing a practical foundation for safer and more
transparent diffusion model sharing.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:44:59 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yuan",
"Xiaoyong",
""
],
[
"Ma",
"Xiaolong",
""
],
[
"Guo",
"Linke",
""
],
[
"Zhang",
"Lan",
""
]
] |
2504.14816 | Fan Wang | Fan Wang | Wavelet Characterization of Inhomogeneous Lipschitz Spaces on Spaces of
Homogeneous Type and Its Applications | null | null | null | null | math.CA math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article, the author establishes a wavelet characterization of
inhomogeneous Lipschitz space $\mathrm{lip}_{\theta}(\mathcal{X})$ via Carlson
sequence, where $\mathcal{X}$ is a space of homogeneous type introduced by R.
R. Coifman and G. Weiss. As applications, characterizations of several
geometric conditions on $\mathcal{X}$, involving the upper bound, the lower
bound, and the Ahlfors regular condition, are obtained.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:50:17 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Wang",
"Fan",
""
]
] |
2504.14817 | Byeong-Yun Ko | Byeong-Yun Ko, Deokki Min, Hyeonuk Nam, and Yong-Hwa Park | DNN based HRIRs Identification with a Continuously Rotating Speaker
Array | null | null | null | null | eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conventional static measurement of head-related impulse responses (HRIRs) is
time-consuming due to the need for repositioning a speaker array for each
azimuth angle. Dynamic approaches using analytical models with a continuously
rotating speaker array have been proposed, but their accuracy is significantly
reduced at high rotational speeds. To address this limitation, we propose a
DNN-based HRIRs identification using sequence-to-sequence learning. The
proposed DNN model incorporates fully connected (FC) networks to effectively
capture HRIR transitions and includes reset and update gates to identify HRIRs
over a whole sequence. The model updates the HRIRs vector coefficients based on
the gradient of the instantaneous square error (ISE). Additionally, we
introduce a learnable normalization process based on the speaker excitation
signals to stabilize the gradient scale of ISE across time. A training scheme,
referred to as whole-sequence updating and optimization scheme, is also
introduced to prevent overfitting. We evaluated the proposed method through
simulations and experiments. Simulation results using the FABIAN database show
that the proposed method outperforms previous analytic models, achieving over 7
dB improvement in normalized misalignment (NM) and maintaining log spectral
distortion (LSD) below 2 dB at a rotational speed of 45{\deg}/s. Experimental
results with a custom-built speaker array confirm that the proposed method
successfully preserved accurate sound localization cues, consistent with those
from static measurement. Source code is available at
https://github.com/byko0810/DNN-based-HRIRs-identification
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:50:34 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ko",
"Byeong-Yun",
""
],
[
"Min",
"Deokki",
""
],
[
"Nam",
"Hyeonuk",
""
],
[
"Park",
"Yong-Hwa",
""
]
] |
2504.14818 | Xinghua Zhang | Jing Zhang, Zhixin Liu, Shengda Zhao, Yangjun Yan, Rongxin Yue, Jiaxin
Yu, Xinjie Li and Xinghua Zhang | Partial-Wetting Phenomena in Active Matter | Main text: 6 pages, 3 figures. Supplementary Material included: 16
pages, 6 figures | null | null | null | cond-mat.soft | http://creativecommons.org/licenses/by/4.0/ | Abundant interfacial phenomena in nature, such as water droplets on lotus
leaves and water transport in plant vessels, originate from partial-wetting
phenomena, which can be well described by Young's equation. It remains an
intriguing question whether similar behaviors exist in active matter. In this
letter, we present a clear demonstration of the partial-wetting phenomenon in a
ternary laning system, which is a typical active system. A phase diagram is
constructed in which the relative drift velocities of different components
govern the transitions among drying, partial wetting, and complete wetting
states. The mechanical balance on the contact lines of the partial-wetting
phase described by Young's equation is verified. A theoretical picture is
proposed to explain the analogy of partial wetting in the laning system to that
in the equilibrium system.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:51:30 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhang",
"Jing",
""
],
[
"Liu",
"Zhixin",
""
],
[
"Zhao",
"Shengda",
""
],
[
"Yan",
"Yangjun",
""
],
[
"Yue",
"Rongxin",
""
],
[
"Yu",
"Jiaxin",
""
],
[
"Li",
"Xinjie",
""
],
[
"Zhang",
"Xinghua",
""
]
] |
2504.14819 | Xiaojuan Deng | Ao Cai and Xiaojuan Deng | Abstract Continuity Theorem for the Lyapunov Exponents of linear
cocycles | 26 pages | null | null | null | math.DS | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We prove the H\"older continuity of Lyapunov exponents for general linear
cocycles when the base measures vary in Wasserstein distance, under the
assumption of uniform large deviations type (LDT) estimates. This is a measure
version of the abstract continuity theorem (ACT) established by Duarte-Klein
[Duarte, P. and Klein, S. (2016). Lyapunov exponents of linear cocycles:
Continuity via large deviations. Atlantis Studies in Dynamical Systems, 3]. The
main obstacle here lies in the fact that the magnitude of the exceptional sets
in LDT estimates is constantly changing when the base measures deviate. We
overcome this via a combination of a Urysohn-type lemma and properties of
Wasserstein distance in every iteration step. Our measure version of ACT,
combined with the original work of Duarte-Klein, provides a complete scheme for
proving joint H\"older continuity of Lyapunov exponents with respect to both
measure and fiber which resolves all parameter dependence. This continuity
theorem is general and applicable to a wide range of mathematical models,
including product of random matrices and cocycles essentially generated by
shifts. In particular, it applies to associated Schr\"odinger operators which
are central objects in the study of mathematical physics.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:52:31 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Cai",
"Ao",
""
],
[
"Deng",
"Xiaojuan",
""
]
] |
2504.14820 | Zichun Xu | Zichun Xu, Zhaomin Wang, Yuntao Li, Lei Zhuang, Zhiyuan Zhao, Guocai
Yang, and Jingdong Zhao | Accelerating Visual Reinforcement Learning with Separate Primitive
Policy for Peg-in-Hole Tasks | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For peg-in-hole tasks, humans rely on binocular visual perception to locate
the peg above the hole surface and then proceed with insertion. This paper
draws insights from this behavior to enable agents to learn efficient assembly
strategies through visual reinforcement learning. Hence, we propose a Separate
Primitive Policy (S2P) to simultaneously learn how to derive location and
insertion actions. S2P is compatible with model-free reinforcement learning
algorithms. Ten insertion tasks featuring different polygons are developed as
benchmarks for evaluations. Simulation experiments show that S2P can boost the
sample efficiency and success rate even with force constraints. Real-world
experiments are also performed to verify the feasibility of S2P. Ablations are
finally given to discuss the generalizability of S2P and some factors that
affect its performance.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:53:17 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Xu",
"Zichun",
""
],
[
"Wang",
"Zhaomin",
""
],
[
"Li",
"Yuntao",
""
],
[
"Zhuang",
"Lei",
""
],
[
"Zhao",
"Zhiyuan",
""
],
[
"Yang",
"Guocai",
""
],
[
"Zhao",
"Jingdong",
""
]
] |
2504.14821 | Dan Wang | Dan Wang, Haowei Chen, Yu Pang, Xiaolong Zou and Wenhui Duan | Band-spin-valley coupled exciton physics in antiferromagnetic MnPS$_3$ | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The introduction of intrinsic magnetic order in two-dimensional (2D)
semiconductors offers great opportunities for investigating correlated
excitonic phenomena. Here, we employ full-spinor GW plus Bethe-Salpeter
equation methodology to reveal rich exciton physics in a prototypical 2D
N\'{e}el-type antiferromagnetic semiconductor MnPS$_3$, enabled by the
interplay among inverted dispersion of the second valence band, spin-valley
coupling and magnetic order. The negative hole mass increases the reduced mass
of the lowest-energy bright exciton, leading to exchange splitting enhancement
of the bright exciton relative to band-edge dark exciton. Notably, such
splitting couples with spontaneous valley polarization to generate distinct
excitonic fine structure between $K$ and $-K$ valleys, which dictate distinct
relaxation behaviors. Crucially, magnetic order transition from N\'{e}el
antiferromagnetic to ferromagnetic state induces significant quasiparticle band
structure reconstruction and excitonic transitions modification, with
low-energy optical excitations being exclusively contributed by majority-spin
channel. These findings establish 2D antiferromagnetic semiconductors as an
intriguing platform to study band-spin-valley coupled exciton physics.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:56:07 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Wang",
"Dan",
""
],
[
"Chen",
"Haowei",
""
],
[
"Pang",
"Yu",
""
],
[
"Zou",
"Xiaolong",
""
],
[
"Duan",
"Wenhui",
""
]
] |
2504.14822 | Rui Qiu | Rui Qiu, Shijie Chen, Yu Su, Po-Yin Yen, Han-Wei Shen | Completing A Systematic Review in Hours instead of Months with
Interactive AI Agents | null | null | null | null | cs.HC cs.CL | http://creativecommons.org/licenses/by/4.0/ | Systematic reviews (SRs) are vital for evidence-based practice in high stakes
disciplines, such as healthcare, but are often impeded by intensive labors and
lengthy processes that can take months to complete. Due to the high demand for
domain expertise, existing automatic summarization methods fail to accurately
identify relevant studies and generate high-quality summaries. To that end, we
introduce InsightAgent, a human-centered interactive AI agent powered by large
language models that revolutionize this workflow. InsightAgent partitions a
large literature corpus based on semantics and employs a multi-agent design for
more focused processing of literature, leading to significant improvement in
the quality of generated SRs. InsightAgent also provides intuitive
visualizations of the corpus and agent trajectories, allowing users to
effortlessly monitor the actions of the agent and provide real-time feedback
based on their expertise. Our user studies with 9 medical professionals
demonstrate that the visualization and interaction mechanisms can effectively
improve the quality of synthesized SRs by 27.2%, reaching 79.7% of
human-written quality. At the same time, user satisfaction is improved by
34.4%. With InsightAgent, it only takes a clinician about 1.5 hours, rather
than months, to complete a high-quality systematic review.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:57:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Qiu",
"Rui",
""
],
[
"Chen",
"Shijie",
""
],
[
"Su",
"Yu",
""
],
[
"Yen",
"Po-Yin",
""
],
[
"Shen",
"Han-Wei",
""
]
] |
2504.14823 | Zhengyan Deng | Zhengyan Deng, Yusen Zheng, Chenliang Sheng, Shaowen Qin | Optimal Repurchasing Contract Design for Efficient Utilization of
Computing Resources | null | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid advancement of AI and other emerging technologies has triggered
exponential growth in computing resources demand. Faced with prohibitive
infrastructure costs for large-scale computing clusters, users are increasingly
resorting to leased computing resources from third-party providers. However,
prevalent overestimation of operational requirements frequently leads to
substantial underutilization of the computing resources. To mitigate such
inefficiency, we propose a contract-based incentive framework for computing
resources repurchasing. Comparing to auction mechanisms, our design enables
providers to reclaim and reallocate surplus computing resources through
market-driven incentives. Our framework operates in a multi-parameter
environment where both clients' idle resource capacities and their unit
valuations of retained resources are private information, posing a significant
challenge to contract design. Two scenarios are considered based on whether all
clients possess the same amount of idle resource capacity. By transforming the
contract design problem into solving a mathematical program, we obtain the
optimal contracts for each scenario, which can maximize the utility of
computing resources providers while ensuring the requirements of incentive
compatibility (IC) and individual rationality (IR). This innovative design not
only provides an effective approach to reduce the inefficient utilization of
computing resources, but also establishes a market-oriented paradigm for
sustainable computing ecosystems.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 02:59:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Deng",
"Zhengyan",
""
],
[
"Zheng",
"Yusen",
""
],
[
"Sheng",
"Chenliang",
""
],
[
"Qin",
"Shaowen",
""
]
] |
2504.14824 | Wei Wang | Wei Wang, Nan Cheng, Conghao Zhou, Haixia Peng, Haibo Zhou, Zhou Su
and Xuemin (Sherman) Shen | An Enhanced Dual-Currency VCG Auction Mechanism for Resource Allocation
in IoV: A Value of Information Perspective | null | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Internet of Vehicles (IoV) is undergoing a transformative evolution,
enabled by advancements in future 6G network technologies, to support
intelligent, highly reliable, and low-latency vehicular services. However, the
enhanced capabilities of loV have heightened the demands for efficient network
resource allocation while simultaneously giving rise to diverse vehicular
service requirements. For network service providers (NSPs), meeting the
customized resource-slicing requirements of vehicle service providers (VSPs)
while maximizing social welfare has become a significant challenge. This paper
proposes an innovative solution by integrating a mean-field multi-agent
reinforcement learning (MFMARL) framework with an enhanced
Vickrey-Clarke-Groves (VCG) auction mechanism to address the problem of social
welfare maximization under the condition of unknown VSP utility functions. The
core of this solution is introducing the ``value of information" as a novel
monetary metric to estimate the expected benefits of VSPs, thereby ensuring the
effective execution of the VCG auction mechanism. MFMARL is employed to
optimize resource allocation for social welfare maximization while adapting to
the intelligent and dynamic requirements of IoV. The proposed enhanced VCG
auction mechanism not only protects the privacy of VSPs but also reduces the
likelihood of collusion among VSPs, and it is theoretically proven to be
dominant-strategy incentive compatible (DSIC). The simulation results
demonstrate that, compared to the VCG mechanism implemented using quantization
methods, the proposed mechanism exhibits significant advantages in convergence
speed, social welfare maximization, and resistance to collusion, providing new
insights into resource allocation in intelligent 6G networks.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:00:04 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Wang",
"Wei",
"",
"Sherman"
],
[
"Cheng",
"Nan",
"",
"Sherman"
],
[
"Zhou",
"Conghao",
"",
"Sherman"
],
[
"Peng",
"Haixia",
"",
"Sherman"
],
[
"Zhou",
"Haibo",
"",
"Sherman"
],
[
"Su",
"Zhou",
"",
"Sherman"
],
[
"Xuemin",
"",
"",
"Sherman"
],
[
"Shen",
"",
""
]
] |
2504.14825 | Zhoujie Qian | Zhoujie Qian | ECViT: Efficient Convolutional Vision Transformer with Local-Attention
and Multi-scale Stages | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision Transformers (ViTs) have revolutionized computer vision by leveraging
self-attention to model long-range dependencies. However, ViTs face challenges
such as high computational costs due to the quadratic scaling of self-attention
and the requirement of a large amount of training data. To address these
limitations, we propose the Efficient Convolutional Vision Transformer (ECViT),
a hybrid architecture that effectively combines the strengths of CNNs and
Transformers. ECViT introduces inductive biases such as locality and
translation invariance, inherent to Convolutional Neural Networks (CNNs) into
the Transformer framework by extracting patches from low-level features and
enhancing the encoder with convolutional operations. Additionally, it
incorporates local-attention and a pyramid structure to enable efficient
multi-scale feature extraction and representation. Experimental results
demonstrate that ECViT achieves an optimal balance between performance and
efficiency, outperforming state-of-the-art models on various image
classification tasks while maintaining low computational and storage
requirements. ECViT offers an ideal solution for applications that prioritize
high efficiency without compromising performance.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:00:17 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Qian",
"Zhoujie",
""
]
] |
2504.14826 | Zhuoran Zheng | Zhuoran Zheng, Xin Su, Chen Wu, Xiuyi Jia | Distribution-aware Dataset Distillation for Efficient Image Restoration | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | With the exponential increase in image data, training an image restoration
model is laborious. Dataset distillation is a potential solution to this
problem, yet current distillation techniques are a blank canvas in the field of
image restoration. To fill this gap, we propose the Distribution-aware Dataset
Distillation method (TripleD), a new framework that extends the principles of
dataset distillation to image restoration. Specifically, TripleD uses a
pre-trained vision Transformer to extract features from images for complexity
evaluation, and the subset (the number of samples is much smaller than the
original training set) is selected based on complexity. The selected subset is
then fed through a lightweight CNN that fine-tunes the image distribution to
align with the distribution of the original dataset at the feature level. To
efficiently condense knowledge, the training is divided into two stages. Early
stages focus on simpler, low-complexity samples to build foundational
knowledge, while later stages select more complex and uncertain samples as the
model matures. Our method achieves promising performance on multiple image
restoration tasks, including multi-task image restoration, all-in-one image
restoration, and ultra-high-definition image restoration tasks. Note that we
can train a state-of-the-art image restoration model on an
ultra-high-definition (4K resolution) dataset using only one consumer-grade GPU
in less than 8 hours (500 savings in computing resources and immeasurable
training time).
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:00:18 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zheng",
"Zhuoran",
""
],
[
"Su",
"Xin",
""
],
[
"Wu",
"Chen",
""
],
[
"Jia",
"Xiuyi",
""
]
] |
2504.14827 | Yenkai Huang | YenKai Huang, Zheng Ning, Ming Cheng | LACE: Exploring Turn-Taking and Parallel Interaction Modes in Human-AI
Co-Creation for Iterative Image Generation | Extended version of the short paper accepted at the GenAICHI Workshop
at CHI 2025. Includes additional results, analysis, qualitative feedback, and
discussion | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | This paper introduces LACE, a co-creative system enabling professional
artists to leverage generative AI through controlled prompting and iterative
refinement within Photoshop. Addressing challenges in precision, iterative
coherence, and workflow compatibility, LACE allows flexible control via
layer-based editing and dual-mode collaboration (turn-taking and parallel). A
pilot study (N=21) demonstrates significant improvements in user satisfaction,
ownership, usability, and artistic perception compared to standard AI
workflows. We offer comprehensive findings, system details, nuanced user
feedback, and implications for integrating generative AI in professional art
practices.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:01:27 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Huang",
"YenKai",
""
],
[
"Ning",
"Zheng",
""
],
[
"Cheng",
"Ming",
""
]
] |
2504.14828 | Ming Li | Xiaoqi Liu, Yuedi Qu, Ming Li and Shu-qian Shen | Variational quantum algorithm for the Poisson equation based on the
banded Toeplitz systems | 20 pages, 5 figures | Commun. Theor. Phys. 77 (2025) 045101 | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For solving the Poisson equation it is usually possible to discretize it into
solving the corresponding linear system $Ax=b$.Variational quantum algorithms
(VQAs) for the discreted Poisson equation have been studied before. We give a
VQA based on the banded Toeplitz systems for solving the Poisson equation with
respect to the structural features of matrix $A$. In detail, we decompose the
matrix $A$ and $A^2$ into a linear combination of the corresponding banded
Toeplitz matrix and sparse matrices with only a few non-zero elements. For the
one-dimensional Poisson equation with different boundary conditions and the
$d$-dimensional Poisson equation with Dirichlet boundary conditions, the number
of decomposition terms is less than the work in [Phys. Rev. A 108, 032418
(2023)]. Based on the decomposition of the matrix, we design quantum circuits
that evaluate efficiently the cost function.Additionally, numerical simulation
verifies the feasibility of the proposed algorithm. In the end, the VQAs for
linear systems of equations and matrix-vector multiplications with $K$-banded
Teoplitz matrix $T_n^K$ are given, where $T_n^K\in R^{n\times n}$ and $K\in
O({\rm ploy}\log n)$.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:07:49 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Liu",
"Xiaoqi",
""
],
[
"Qu",
"Yuedi",
""
],
[
"Li",
"Ming",
""
],
[
"Shen",
"Shu-qian",
""
]
] |
2504.14830 | Xin Zhang | Xin Zhang, Kaiwen Xia | Solving All Seismic Tomographic Problems using Deep Learning | null | null | null | null | physics.geo-ph physics.data-an | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In a variety of geoscientific applications scientists often need to image
properties of the Earth's interior in order to understand the heterogeneity and
processes taking place within the Earth. Seismic tomography is one such method
which has been used widely to study properties of the subsurface. In order to
solve tomographic problems efficiently, neural network-based methods have been
introduced to geophysics. However, these methods can only be applied to certain
types of problems with fixed acquisition geometry at a specific site. In this
study we extend neural network-based methods to problems with various scales
and acquisition geometries by using graph mixture density networks (MDNs). We
train a graph MDN for 2D tomographic problems using simulated velocity models
and travel time data, and apply the trained network to both synthetic and real
data problems that have various scales and station distributions at different
sites. The results demonstrate that graph MDNs can provide comparable solutions
to those obtained using traditional Bayesian methods in seconds, and therefore
provide the possibility to use graph MDNs to produce rapid solutions for all
kinds of seismic tomographic problems over the world.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:14:57 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhang",
"Xin",
""
],
[
"Xia",
"Kaiwen",
""
]
] |
2504.14831 | Mansoor Ur Rehman | Mansoor Ur Rehman and Qaisar Shafi | Supersymmetric Hybrid Inflation in light of Atacama Cosmology Telescope
Data Release 6, Planck 2018 and LB-BK18 | 4 pages, 4 figures | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | Supersymmetry-based hybrid inflation models (referred to as `spontaneously
broken supersymmetry' by the Planck collaboration) are attractive for several
reasons, including the appealing feature that inflation is associated with
local gauge symmetry breaking in the early universe. Following the Planck
collaboration's notation, the inflationary potential with sub-Planckian
inflaton field values is given by: $V = \Lambda^4 [1 + \alpha_h \log(\phi /
M_{Pl})] - m_{3/2} \Lambda^2 \phi + \Lambda^4 O((\phi / M_{Pl})^4)$. Here,
$\Lambda = \sqrt{\kappa} M$ denotes the energy scale of inflation, $M$ is the
gauge symmetry breaking scale, $\kappa$ is a dimensionless parameter that sets
the inflaton mass ($\sqrt{2} \kappa M$), and $\alpha_h$ is determined from
quantum corrections in terms of $\kappa$ and the underlying gauge group. A soft
supersymmetry-breaking term proportional to the gravitino mass $m_{3/2}$ (~10
TeV) and linear in the inflaton field $\phi$ is also present during inflation.
The final term in $V$ represents the leading-order supergravity correction.
(Note that the last two terms were not taken into account in the Planck
analysis.) We provide estimates for the parameters $\kappa$ (and $\alpha_h$)
that yield a scalar spectral index $n_s$ in the range 0.96 to 0.98, which is
fully consistent with recent P-ACT-LB measurements presented by the Atacama
Cosmology Telescope, as well as earlier measurements by Planck. We recall that
in the absence of the soft SUSY-breaking term proportional to $m_{3/2}$ in $V$,
the spectral index $n_s = 1 - 1/N = 0.98$, where $N = 50$ denotes the number of
e-foldings. The tensor-to-scalar ratio $r$ in this minimal model is tiny, but
it can reach potentially observable values ($r \lesssim 0.01$) in non-minimal
models.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:18:43 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Rehman",
"Mansoor Ur",
""
],
[
"Shafi",
"Qaisar",
""
]
] |
2504.14832 | Weizhi Liu | Yue Li, Weizhi Liu, Dongdong Lin | Protecting Your Voice: Temporal-aware Robust Watermarking | null | null | null | null | cs.CR cs.AI cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid advancement of generative models has led to the synthesis of
real-fake ambiguous voices. To erase the ambiguity, embedding watermarks into
the frequency-domain features of synthesized voices has become a common
routine. However, the robustness achieved by choosing the frequency domain
often comes at the expense of fine-grained voice features, leading to a loss of
fidelity. Maximizing the comprehensive learning of time-domain features to
enhance fidelity while maintaining robustness, we pioneer a
\textbf{\underline{t}}emporal-aware
\textbf{\underline{r}}ob\textbf{\underline{u}}st
wat\textbf{\underline{e}}rmarking (\emph{True}) method for protecting the
speech and singing voice.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:23:10 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Li",
"Yue",
""
],
[
"Liu",
"Weizhi",
""
],
[
"Lin",
"Dongdong",
""
]
] |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.