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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.08821 | Enming Zhang | Enming Zhang, Zheng Liu, Yu Xiang, Yanwen Qu | Probabilistic QoS Metric Forecasting in Delay-Tolerant Networks Using
Conditional Diffusion Models on Latent Dynamics | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Active QoS metric prediction, commonly employed in the maintenance and
operation of DTN, could enhance network performance regarding latency,
throughput, energy consumption, and dependability. Naturally formulated as a
multivariate time series forecasting problem, it attracts substantial research
efforts. Traditional mean regression methods for time series forecasting cannot
capture the data complexity adequately, resulting in deteriorated performance
in operational tasks in DTNs such as routing. This paper formulates the
prediction of QoS metrics in DTN as a probabilistic forecasting problem on
multivariate time series, where one could quantify the uncertainty of forecasts
by characterizing the distribution of these samples. The proposed approach
hires diffusion models and incorporates the latent temporal dynamics of
non-stationary and multi-mode data into them. Extensive experiments demonstrate
the efficacy of the proposed approach by showing that it outperforms the
popular probabilistic time series forecasting methods.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 15:40:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Enming",
""
],
[
"Liu",
"Zheng",
""
],
[
"Xiang",
"Yu",
""
],
[
"Qu",
"Yanwen",
""
]
] |
2504.08822 | Jose Maria Grau | L. Bay\'on, P. Fortuny ayuso, J.M. Grau, A.M. Oller-Marc\'en and M.M
Ru\'iz | Nash Equilibria in the Showcase Showdown game with unlimited spins | 3 figures | null | 10.1016/j.amc.2025.129368 | null | math.OC math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The game of \emph{Showcase Showdown} with unlimited spins is investigated as
an $n$-players continuous game, and the Nash Equilibrium strategies for the
players are obtained. The sequential game with information on the results of
the previous players is studied, as well as three variants: no information,
possibility of draw, and different modalities of winner payoff.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 19:07:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bayón",
"L.",
""
],
[
"ayuso",
"P. Fortuny",
""
],
[
"Grau",
"J. M.",
""
],
[
"Oller-Marcén",
"A. M.",
""
],
[
"Ruíz",
"M. M",
""
]
] |
2504.08823 | Xiaobing Yu | Xiaobing Yu, Jin Yang, Xiao Wu, Peijie Qiu, Xiaofeng Liu | FM-LoRA: Factorized Low-Rank Meta-Prompting for Continual Learning | 8 Pages, 4 figures | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | How to adapt a pre-trained model continuously for sequential tasks with
different prediction class labels and domains and finally learn a generalizable
model across diverse tasks is a long-lasting challenge. Continual learning (CL)
has emerged as a promising approach to leverage pre-trained models (e.g.,
Transformers) for sequential tasks. While many existing CL methods
incrementally store additional learned structures, such as Low-Rank Adaptation
(LoRA) adapters or prompts and sometimes even preserve features from previous
samples to maintain performance. This leads to unsustainable parameter growth
and escalating storage costs as the number of tasks increases. Moreover,
current approaches often lack task similarity awareness, which further hinders
the models ability to effectively adapt to new tasks without interfering with
previously acquired knowledge. To address these challenges, we propose FM-LoRA,
a novel and efficient low-rank adaptation method that integrates both a dynamic
rank selector (DRS) and dynamic meta-prompting (DMP). This framework allocates
model capacity more effectively across tasks by leveraging a shared low-rank
subspace critical for preserving knowledge, thereby avoiding continual
parameter expansion. Extensive experiments on various CL benchmarks, including
ImageNet-R, CIFAR100, and CUB200 for class-incremental learning (CIL), and
DomainNet for domain-incremental learning (DIL), with Transformers backbone
demonstrate that FM-LoRA effectively mitigates catastrophic forgetting while
delivering robust performance across a diverse range of tasks and domains.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 19:36:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yu",
"Xiaobing",
""
],
[
"Yang",
"Jin",
""
],
[
"Wu",
"Xiao",
""
],
[
"Qiu",
"Peijie",
""
],
[
"Liu",
"Xiaofeng",
""
]
] |
2504.08825 | Rafael Hern\'andez-Jim\'enez | Juan Ignacio Musmarra, Claudia Moreno, Rafael Hern\'andez-Jim\'enez | Conformal metric perturbations and boundary term as physical source | 15 pages, 7 figures | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the context of the Relativistic Quantum Geometry formalism, where the
cosmological constant is promoted to a dynamical variable by attributing it a
geometric interpretation as a result of a flux on the boundary of a manifold
and establishing a connection between the perturbations of the Ricci tensor and
the metric tensor, we propose an approach for the perturbations of the metric
tensor. From this, imposing that Einstein's equations must hold for the tensors
defined from the perturbed quantities obtained from conformal transformations,
we derive a functional form for the cosmological parameter $\Lambda$ in terms
of the cosmological parameter $\bar{\Lambda}$ of the perturbed manifold. We
then use the obtained equations to propose a cosmological model based on the
Friedmann-Lema\^{i}tre-Robertson-Walker metric with no spatial curvature,
fitting the free parameters using observational data from Hubble and Type Ia
Supernovae. The model is statistically comparable to $\Lambda$CDM; although,
the joint analysis produces a smaller $H_{0}^{\rm Conformal}=69.80\rm\,\, Km
\,s^{-1}\,Mpc^{-1}$ in contrast to the flat $\Lambda$CDM result
$H_{0}^{\Lambda\rm CDM}=70.52\rm\,\, Km \,s^{-1}\,Mpc^{-1}$. An evident
singularity occurs when the conformal factor $\Xi^{2}=2$, yields an early
universe dominated only by matter $\rho_{m}$, which undoubtedly does not
correspond to a viable history of our cosmos. Despite these limitations, a
specific scenario remains feasible. This study aims to offer insights into the
acceleration of the universe and addresses key questions in contemporary
cosmology.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 21:15:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Musmarra",
"Juan Ignacio",
""
],
[
"Moreno",
"Claudia",
""
],
[
"Hernández-Jiménez",
"Rafael",
""
]
] |
2504.08826 | Stepan Paul | Stepan Paul | A piecewise-linear isometrically immersed flat Klein bottle in Euclidean
3-space | null | null | null | null | math.GT math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present numerical polyhedron data for the image of a piecewise-linear map
from a zero-curvature Klein bottle into Euclidean 3-space such that every point
in the domain has a neighborhood which is isometrically embedded. To the
author's knowledge, this is the first explicit piecewise-smooth isometric
immersion of a flat Klein bottle. Intuitively, the surface can be locally made
from origami and but for the self-intersections has the global topology of a
Klein bottle.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 00:37:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Paul",
"Stepan",
""
]
] |
2504.08827 | Samy Vilhes | Samy-Melwan Vilhes (LITIS), Gilles Gasso (LITIS), Mokhtar Z Alaya
(LMAC) | PatchTrAD: A Patch-Based Transformer focusing on Patch-Wise
Reconstruction Error for Time Series Anomaly Detection | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time series anomaly detection (TSAD) focuses on identifying whether
observations in streaming data deviate significantly from normal patterns. With
the prevalence of connected devices, anomaly detection on time series has
become paramount, as it enables real-time monitoring and early detection of
irregular behaviors across various application domains. In this work, we
introduce PatchTrAD, a Patch-based Transformer model for time series anomaly
detection. Our approach leverages a Transformer encoder along with the use of
patches under a reconstructionbased framework for anomaly detection. Empirical
evaluations on multiple benchmark datasets show that PatchTrAD is on par, in
terms of detection performance, with state-of-the-art deep learning models for
anomaly detection while being time efficient during inference.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 07:58:55 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Vilhes",
"Samy-Melwan",
"",
"LITIS"
],
[
"Gasso",
"Gilles",
"",
"LITIS"
],
[
"Alaya",
"Mokhtar Z",
"",
"LMAC"
]
] |
2504.08828 | Mazyar Dawoodian | Mazyar Dawoodian and Ould el Moctar | Enhanced Classical Nucleation Theory for Cavitation Inception in the
Presence of Gaseous Nuclei | null | null | null | null | cond-mat.soft physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | This paper introduces an enhanced Classical Nucleation Theory model to
predict the cavitation inception pressure and to describe the behavior of
nanoscale gaseous nuclei during cavitation. Validation is achieved through
molecular dynamics simulations. The findings highlight the significant role of
nanoscale gaseous nuclei in lowering the tensile strength required for
cavitation initiation. The results show that our enhanced CNT model predicts
lower cavitation pressures than the Blake threshold, closely matching molecular
dynamics simulations. Finally, our results illustrate that differences between
cavitation pressures using the Van der Waals and ideal gas models are greatest
for smaller nuclei and lower temperatures.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 08:00:21 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dawoodian",
"Mazyar",
""
],
[
"Moctar",
"Ould el",
""
]
] |
2504.08829 | G. Charbel KINDJI | G. Charbel N. Kindji (IRISA, MALT), Elisa Fromont (MALT, IRISA), Lina
Maria Rojas-Barahona, Tanguy Urvoy | Datum-wise Transformer for Synthetic Tabular Data Detection in the Wild | null | null | null | null | cs.LG cs.AI cs.DB cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The growing power of generative models raises major concerns about the
authenticity of published content. To address this problem, several synthetic
content detection methods have been proposed for uniformly structured media
such as image or text. However, little work has been done on the detection of
synthetic tabular data, despite its importance in industry and government. This
form of data is complex to handle due to the diversity of its structures: the
number and types of the columns may vary wildly from one table to another. We
tackle the tough problem of detecting synthetic tabular data ''in the wild'',
i.e. when the model is deployed on table structures it has never seen before.
We introduce a novel datum-wise transformer architecture and show that it
outperforms existing models. Furthermore, we investigate the application of
domain adaptation techniques to enhance the effectiveness of our model, thereby
providing a more robust data-forgery detection solution.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 08:01:34 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kindji",
"G. Charbel N.",
"",
"IRISA, MALT"
],
[
"Fromont",
"Elisa",
"",
"MALT, IRISA"
],
[
"Rojas-Barahona",
"Lina Maria",
""
],
[
"Urvoy",
"Tanguy",
""
]
] |
2504.08830 | Chi Ho Wong | Chi Ho Wong, Zong Liang Guo, King Cheong Lam, Chun Pong Chau, Wing Yu
Chan, Chak-yin Tang, Yuen Hong Tsang, Leung Yuk Frank Lam, Xijun Hu | Tuning Charge Density Wave in the Transition from Magnetically
Frustrated Conductor to Ferrimagnetic Insulator in Carbon Nanowire within
Boron Nitride Nanotube | null | null | null | null | cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The emergence of exotic charge density wave (CDW) alongside ferrimagnetism
materials opens exciting new possibilities for quantum switching, particularly
in field-tuning CDW electronics. However, these two phenomena often compete and
rely heavily on strong electronic correlations. While carbon nanowire arrays
have been experimentally shown to exhibit ferromagnetism above 400 K, our
research shows that encapsulating a linear carbon chain (LCC) within zigzag
boron nitride nanotubes (BNT) induces a short-range CDW state under a competing
effect of ferrimagnetism and magnetic frustrations. However, for this exotic
feature to occur, the LCC needs to break the symmetry along the circular plane
of the BNT. Then we utilize a Monte Carlo model to identify the optimal length
of LCC@BNT to tackle its size effect, while also comparing the stability of
chains provided by carbon nanotubes. The shorter LCC@BNT displays a more
prominent long-range CDW pattern with a tunneling barrier of 2.3 eV on the
Fermi surface, transitioning into an unconventional insulator. Meanwhile,
magnetic frustrations disappear, and ferrimagnetism remains stable up to 280 K.
Our discovery of ferrimagnetic CDW carbyne insulators, which function without
conventional periodic lattice distortion, spin-orbit coupling, or complex d and
f hybridization represents a groundbreaking shift in thinking, which
demonstrates that such exotic properties are not exclusive to transition metal
elements. We anticipate that spin fluctuations in LCC@BNT could enable
fine-tuning of the CDW pattern, and applying an electric excitation of 2.3 eV
triggers an abrupt insulator-to-conductor transition for quantum switching
applications.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 08:21:55 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wong",
"Chi Ho",
""
],
[
"Guo",
"Zong Liang",
""
],
[
"Lam",
"King Cheong",
""
],
[
"Chau",
"Chun Pong",
""
],
[
"Chan",
"Wing Yu",
""
],
[
"Tang",
"Chak-yin",
""
],
[
"Tsang",
"Yuen Hong",
""
],
[
"Lam",
"Leung Yuk Frank",
""
],
[
"Hu",
"Xijun",
""
]
] |
2504.08831 | Mehdi Heydari Shahna | Mehdi Heydari Shahna, Pauli Mustalahti, and Jouni Mattila | Anti-Slip AI-Driven Model-Free Control with Global Exponential Stability
in Skid-Steering Robots | This paper has been submitter for the IEEE consideration | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | Undesired lateral and longitudinal wheel slippage can disrupt a mobile
robot's heading angle, traction, and, eventually, desired motion. This issue
makes the robotization and accurate modeling of heavy-duty machinery very
challenging because the application primarily involves off-road terrains, which
are susceptible to uneven motion and severe slippage. As a step toward
robotization in skid-steering heavy-duty robot (SSHDR), this paper aims to
design an innovative robust model-free control system developed by neural
networks to strongly stabilize the robot dynamics in the presence of a broad
range of potential wheel slippages. Before the control design, the dynamics of
the SSHDR are first investigated by mathematically incorporating slippage
effects, assuming that all functional modeling terms of the system are unknown
to the control system. Then, a novel tracking control framework to guarantee
global exponential stability of the SSHDR is designed as follows: 1) the
unknown modeling of wheel dynamics is approximated using radial basis function
neural networks (RBFNNs); and 2) a new adaptive law is proposed to compensate
for slippage effects and tune the weights of the RBFNNs online during
execution. Simulation and experimental results verify the proposed tracking
control performance of a 4,836 kg SSHDR operating on slippery terrain.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 09:21:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shahna",
"Mehdi Heydari",
""
],
[
"Mustalahti",
"Pauli",
""
],
[
"Mattila",
"Jouni",
""
]
] |
2504.08832 | Mahshid Sadeghpour PhD | Mahshid Sadeghpour, Arathi Arakala, Asha Rao | Generative AI in Collaborative Academic Report Writing: Advantages,
Disadvantages, and Ethical Considerations | 21 pages, 5 figures | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The availability and abundance of GenAI tools to administer tasks
traditionally managed by people have raised concerns, particularly within the
education and academic sectors, as some students may highly rely on these tools
to complete the assignments designed to enable learning. This article focuses
on informing students about the significance of investing their time during
their studies on developing essential life-long learning skills using their own
critical thinking, rather than depending on AI models that are susceptible to
misinformation, hallucination, and bias. As we transition to an AI-centric era,
it is important to educate students on how these models work, their pitfalls,
and the ethical concerns associated with feeding data to such tools.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 09:22:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sadeghpour",
"Mahshid",
""
],
[
"Arakala",
"Arathi",
""
],
[
"Rao",
"Asha",
""
]
] |
2504.08833 | Alessia Biondi | Alessia Biondi | Effective-Field Theories of Analogue Gravity | Master's thesis, University of Pisa (2023). Supervisors: Maria Luisa
Chiofalo, Massimo Mannarelli. Collaborators: Silvia Trabucco, Stefano
Liberati, Dario Grasso | null | null | null | gr-qc cond-mat.quant-gas hep-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We develop a novel method for building a gravitational analog model for a
flowing Bose-Einstein condensate. The analogue metric is obtained using
effective field theory methods, integrating out the heavy radial fluctuations.
In this way, we also obtain interaction terms up to the quartic order in the
fields. The microscopic Lagrangian describes a complex massive scalar field,
with a global U(1) symmetry, that is spontaneously broken. From the quadratic
effective Lagrangian, we obtain a dispersion law for phonons in presence of an
acoustic horizon generated by the background flow. We observe that the phonon
dispersion relation may exhibit a non-monotonic behavior determined by the
Lagrangian's parameters. In this case, a non-trivial minimum appears associated
with a characteristic length scale, indicating the breaking of translation
symmetry. Then, we determine the constraints on the Lagrangian's parameters to
make this happen. Finally, we design an original procedure to calculate the
correlation functions through field theory tools. We apply this method to the
density-density correlation function, reproducing known results. Moreover, we
study the case of a non-monotonic phonon dispersion law, finding the existence
of long range order.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 09:51:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Biondi",
"Alessia",
""
]
] |
2504.08834 | Sara Habibi | Sara Habibi, Naghmeh Ivaki, Jo\~ao Barata | A Systematic Literature Review of Unmanned Aerial Vehicles for
Healthcare and Emergency Services | 39 pages, 14 figures, 6 tables,164 references | null | null | null | cs.CY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Unmanned aerial vehicles (UAVs), initially developed for military
applications, are now used in various fields. As UAVs become more common across
multiple industries, it is crucial to understand how to adopt them effectively,
efficiently, and safely. The utilization of UAVs in healthcare and emergency
services has evolved significantly in recent years, with these aerial vehicles
potentially contributing to increased survival rates and enhanced healthcare
services.
This paper presents a two-stage systematic literature review, including a
tertiary study of 15 review papers and an in-depth assessment of 136 primary
publications focused on using UAVs in healthcare and emergency services. The
research demonstrates how civilian UAVs have been used in numerous
applications, such as healthcare emergencies, medical supply delivery, and
disaster management, for diverse use cases such as Automated External
Defibrillator (AED) delivery, blood delivery, and search and rescue.
The studies indicate that UAVs significantly improve response times in
emergency situations, enhance survival rates by ensuring the timely delivery of
critical medical supplies such as AEDs, and prove to be cost-effective
alternatives to traditional delivery methods, especially in remote or
inaccessible areas. The studies also highlight the need for ongoing research
and development to address existing challenges, such as regulatory frameworks,
security, privacy and safety concerns, infrastructure development, and ethical
and social issues. Effectively understanding and tackling these challenges is
essential for maximizing the benefits of UAV technology in healthcare and
emergency services, ultimately leading to safer, more resilient, and responsive
systems that can better serve public health needs.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 10:02:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Habibi",
"Sara",
""
],
[
"Ivaki",
"Naghmeh",
""
],
[
"Barata",
"João",
""
]
] |
2504.08835 | Delfim F. M. Torres | M\'arcia Lemos-Silva, Delfim F. M. Torres | The Omega Calculus of Variations | This is a preprint of a paper accepted 27-March-2025 to Applied
Mathematics E-Notes (ISSN 1607-2510). Available free at mirror sites of
[http://www.math.nthu.edu.tw/~amen/] | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove a necessary optimality condition of Euler--Lagrange type for the
calculus of variations with Omega derivatives, which turns out to be sufficient
under jointly convexity of the Lagrangian.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 14:40:41 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lemos-Silva",
"Márcia",
""
],
[
"Torres",
"Delfim F. M.",
""
]
] |
2504.08836 | Chris Hays | Chris Hays, Manish Raghavan | Double Machine Learning for Causal Inference under Shared-State
Interference | 48 pages, 6 figures | null | null | null | stat.ML cs.LG econ.EM | http://creativecommons.org/licenses/by/4.0/ | Researchers and practitioners often wish to measure treatment effects in
settings where units interact via markets and recommendation systems. In these
settings, units are affected by certain shared states, like prices, algorithmic
recommendations or social signals. We formalize this structure, calling it
shared-state interference, and argue that our formulation captures many
relevant applied settings. Our key modeling assumption is that individuals'
potential outcomes are independent conditional on the shared state. We then
prove an extension of a double machine learning (DML) theorem providing
conditions for achieving efficient inference under shared-state interference.
We also instantiate our general theorem in several models of interest where it
is possible to efficiently estimate the average direct effect (ADE) or global
average treatment effect (GATE).
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 16:45:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hays",
"Chris",
""
],
[
"Raghavan",
"Manish",
""
]
] |
2504.08837 | Haozhe Wang | Haozhe Wang, Chao Qu, Zuming Huang, Wei Chu, Fangzhen Lin, Wenhu Chen | VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models
with Reinforcement Learning | submitted to NeurIPS | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recently, slow-thinking systems like GPT-o1 and DeepSeek-R1 have demonstrated
great potential in solving challenging problems through explicit reflection.
They significantly outperform the best fast-thinking models, such as GPT-4o, on
various math and science benchmarks. However, their multimodal reasoning
capabilities remain on par with fast-thinking models. For instance, GPT-o1's
performance on benchmarks like MathVista, MathVerse, and MathVision is similar
to fast-thinking models. In this paper, we aim to enhance the slow-thinking
capabilities of vision-language models using reinforcement learning (without
relying on distillation) to advance the state of the art. First, we adapt the
GRPO algorithm with a novel technique called Selective Sample Replay (SSR) to
address the vanishing advantages problem. While this approach yields strong
performance, the resulting RL-trained models exhibit limited self-reflection or
self-verification. To further encourage slow-thinking, we introduce Forced
Rethinking, which appends a textual rethinking trigger to the end of initial
rollouts in RL training, explicitly enforcing a self-reflection reasoning step.
By combining these two techniques, our model, VL-Rethinker, advances
state-of-the-art scores on MathVista, MathVerse, and MathVision to achieve
80.3%, 61.8%, and 43.9% respectively. VL-Rethinker also achieves open-source
SoTA on multi-disciplinary benchmarks such as MMMU-Pro, EMMA, and MEGA-Bench,
narrowing the gap with GPT-o1.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 17:41:56 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Haozhe",
""
],
[
"Qu",
"Chao",
""
],
[
"Huang",
"Zuming",
""
],
[
"Chu",
"Wei",
""
],
[
"Lin",
"Fangzhen",
""
],
[
"Chen",
"Wenhu",
""
]
] |
2504.08838 | Mike Lasby | Mike Lasby, Nish Sinnadurai, Valavan Manohararajah, Sean Lie,
Vithursan Thangarasa | SD$^2$: Self-Distilled Sparse Drafters | 21 pages | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Speculative decoding is a powerful technique for reducing the latency of
Large Language Models (LLMs), offering a fault-tolerant framework that enables
the use of highly compressed draft models. In this work, we introduce
Self-Distilled Sparse Drafters (SD$^2$), a novel methodology that leverages
self-data distillation and fine-grained weight sparsity to produce highly
efficient and well-aligned draft models. SD$^2$ systematically enhances draft
token acceptance rates while significantly reducing Multiply-Accumulate
operations (MACs), even in the Universal Assisted Generation (UAG) setting,
where draft and target models originate from different model families. On a
Llama-3.1-70B target model, SD$^2$ provides a $\times$1.59 higher Mean Accepted
Length (MAL) compared to layer-pruned draft models and reduces MACs by over
43.87% with a 8.36% reduction in MAL compared to a dense draft models. Our
results highlight the potential of sparsity-aware fine-tuning and compression
strategies to improve LLM inference efficiency while maintaining alignment with
target models.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 18:21:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lasby",
"Mike",
""
],
[
"Sinnadurai",
"Nish",
""
],
[
"Manohararajah",
"Valavan",
""
],
[
"Lie",
"Sean",
""
],
[
"Thangarasa",
"Vithursan",
""
]
] |
2504.08839 | Scott Field | Scott E. Field, Vijay Varma, Jonathan Blackman, Bhooshan Gadre, Chad
R. Galley, Tousif Islam, Keefe Mitman, Michael P\"urrer, Adhrit Ravichandran,
Mark A. Scheel, Leo C. Stein, Jooheon Yoo | GWSurrogate: A Python package for gravitational wave surrogate models | 5 pages and 1 figure. Published in the Journal of Open Source
Software (JOSS) | null | 10.21105/joss.07073 | null | astro-ph.IM gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fast and accurate waveform models are fundamentally important to modern
gravitational wave astrophysics, enabling the study of merging compact objects
like black holes and neutron stars. However, generating high-fidelity
gravitational waveforms through numerical relativity simulations is
computationally intensive, often requiring days to months of computation time
on supercomputers. Surrogate models provide a practical solution to
dramatically accelerate waveform evaluations (typically tens of milliseconds
per evaluation) while retaining the accuracy of computationally expensive
simulations. The GWSurrogate Python package provides easy access to these
gravitational wave surrogate models through a user-friendly interface.
Currently, the package supports 16 surrogate models, each varying in duration,
included physical effects (e.g., nonlinear memory, tidal forces, harmonic
modes, eccentricity, mass ratio range, precession effects), and underlying
solution methods (e.g., Effective One Body, numerical relativity, black hole
perturbation theory). GWSurrogate models follow the waveform model conventions
used by the LIGO-Virgo-Kagra collaboration, making the package immediately
suitable for both theoretical studies and practical gravitational wave data
analysis. By enabling rapid and precise waveform generation, GWSurrogate serves
as a production-level tool for diverse applications, including parameter
estimation, template bank generation, and tests of general relativity.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 18:59:29 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Field",
"Scott E.",
""
],
[
"Varma",
"Vijay",
""
],
[
"Blackman",
"Jonathan",
""
],
[
"Gadre",
"Bhooshan",
""
],
[
"Galley",
"Chad R.",
""
],
[
"Islam",
"Tousif",
""
],
[
"Mitman",
"Keefe",
""
],
[
"Pürrer",
"Michael",
""
],
[
"Ravichandran",
"Adhrit",
""
],
[
"Scheel",
"Mark A.",
""
],
[
"Stein",
"Leo C.",
""
],
[
"Yoo",
"Jooheon",
""
]
] |
2504.08840 | Vasiliki Tassopoulou | Vasiliki Tassopoulou, Haochang Shou, Christos Davatzikos | Adaptive Shrinkage Estimation For Personalized Deep Kernel Regression In
Modeling Brain Trajectories | null | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Longitudinal biomedical studies monitor individuals over time to capture
dynamics in brain development, disease progression, and treatment effects.
However, estimating trajectories of brain biomarkers is challenging due to
biological variability, inconsistencies in measurement protocols (e.g.,
differences in MRI scanners), scarcity, and irregularity in longitudinal
measurements. Herein, we introduce a novel personalized deep kernel regression
framework for forecasting brain biomarkers, with application to regional
volumetric measurements. Our approach integrates two key components: a
population model that captures brain trajectories from a large and diverse
cohort, and a subject-specific model that captures individual trajectories. To
optimally combine these, we propose Adaptive Shrinkage Estimation, which
effectively balances population and subject-specific models. We assess our
model's performance through predictive accuracy metrics, uncertainty
quantification, and validation against external clinical studies. Benchmarking
against state-of-the-art statistical and machine learning models -- including
linear mixed effects models, generalized additive models, and deep learning
methods -- demonstrates the superior predictive performance of our approach.
Additionally, we apply our method to predict trajectories of composite
neuroimaging biomarkers, which highlights the versatility of our approach in
modeling the progression of longitudinal neuroimaging biomarkers. Furthermore,
validation on three external neuroimaging studies confirms the robustness of
our method across different clinical contexts. We make the code available at
https://github.com/vatass/AdaptiveShrinkageDKGP.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 19:13:44 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Tassopoulou",
"Vasiliki",
""
],
[
"Shou",
"Haochang",
""
],
[
"Davatzikos",
"Christos",
""
]
] |
2504.08841 | Luis F Recalde | Luis F. Recalde, Mrunal Sarvaiya, Giuseppe Loianno and Guanrui Li | ES-HPC-MPC: Exponentially Stable Hybrid Perception Constrained MPC for
Quadrotor with Suspended Payloads | The first two listed authors contributed equally | null | null | null | eess.SY cs.RO cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Aerial transportation using quadrotors with cable-suspended payloads holds
great potential for applications in disaster response, logistics, and
infrastructure maintenance. However, their hybrid and underactuated dynamics
pose significant control and perception challenges. Traditional approaches
often assume a taut cable condition, limiting their effectiveness in real-world
applications where slack-to-taut transitions occur due to disturbances. We
introduce ES-HPC-MPC, a model predictive control framework that enforces
exponential stability and perception-constrained control under hybrid dynamics.
Our method leverages Exponentially Stabilizing Control Lyapunov Functions
(ES-CLFs) to enforce stability during the tasks and Control Barrier Functions
(CBFs) to maintain the payload within the onboard camera's field of view (FoV).
We validate our method through both simulation and real-world experiments,
demonstrating stable trajectory tracking and reliable payload perception. We
validate that our method maintains stability and satisfies perception
constraints while tracking dynamically infeasible trajectories and when the
system is subjected to hybrid mode transitions caused by unexpected
disturbances.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 19:43:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Recalde",
"Luis F.",
""
],
[
"Sarvaiya",
"Mrunal",
""
],
[
"Loianno",
"Giuseppe",
""
],
[
"Li",
"Guanrui",
""
]
] |
2504.08842 | Micah Adler Ph.D. | Micah Adler and Dan Alistarh and Nir Shavit | Towards Combinatorial Interpretability of Neural Computation | 47 Pages | null | null | null | cs.LG cs.NE | http://creativecommons.org/licenses/by/4.0/ | We introduce combinatorial interpretability, a methodology for understanding
neural computation by analyzing the combinatorial structures in the sign-based
categorization of a network's weights and biases. We demonstrate its power
through feature channel coding, a theory that explains how neural networks
compute Boolean expressions and potentially underlies other categories of
neural network computation. According to this theory, features are computed via
feature channels: unique cross-neuron encodings shared among the inputs the
feature operates on. Because different feature channels share neurons, the
neurons are polysemantic and the channels interfere with one another, making
the computation appear inscrutable.
We show how to decipher these computations by analyzing a network's feature
channel coding, offering complete mechanistic interpretations of several small
neural networks that were trained with gradient descent. Crucially, this is
achieved via static combinatorial analysis of the weight matrices, without
examining activations or training new autoencoding networks. Feature channel
coding reframes the superposition hypothesis, shifting the focus from neuron
activation directionality in high-dimensional space to the combinatorial
structure of codes. It also allows us for the first time to exactly quantify
and explain the relationship between a network's parameter size and its
computational capacity (i.e. the set of features it can compute with low
error), a relationship that is implicitly at the core of many modern scaling
laws.
Though our initial studies of feature channel coding are restricted to
Boolean functions, we believe they provide a rich, controlled, and informative
research space, and that the path we propose for combinatorial interpretation
of neural computation can provide a basis for understanding both artificial and
biological neural circuits.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 21:28:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Adler",
"Micah",
""
],
[
"Alistarh",
"Dan",
""
],
[
"Shavit",
"Nir",
""
]
] |
2504.08843 | Sai Nandan Morapakula | Sai Nandan Morapakula, Sangram Deshpande, Rakesh Yata, Rushikesh
Ubale, Uday Wad, Kazuki Ikeda | End-to-End Portfolio Optimization with Quantum Annealing | 9 pages, 4 figures, 2 tables | null | null | null | quant-ph econ.GN math.OC q-fin.EC q-fin.PM q-fin.RM | http://creativecommons.org/publicdomain/zero/1.0/ | With rapid technological progress reshaping the financial industry, quantum
technology plays a critical role in advancing risk management, asset
allocation, and financial strategies. Realizing its full potential requires
overcoming challenges like quantum hardware limits, algorithmic stability, and
implementation barriers. This research explores integrating quantum annealing
with portfolio optimization, highlighting quantum methods' ability to enhance
investment strategy efficiency and speed. Using hybrid quantum-classical
models, the study shows combined approaches effectively handle complex
optimization better than classical methods. Empirical results demonstrate a
portfolio increase of 200,000 Indian Rupees over the benchmark. Additionally,
using rebalancing leads to a portfolio that also surpasses the benchmark value.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 21:31:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Morapakula",
"Sai Nandan",
""
],
[
"Deshpande",
"Sangram",
""
],
[
"Yata",
"Rakesh",
""
],
[
"Ubale",
"Rushikesh",
""
],
[
"Wad",
"Uday",
""
],
[
"Ikeda",
"Kazuki",
""
]
] |
2504.08844 | Di Xu | Di Xu | Artificial Intelligence Augmented Medical Imaging Reconstruction in
Radiation Therapy | PhD thesis | null | null | null | eess.IV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Efficiently acquired and precisely reconstructed imaging are crucial to the
success of modern radiation therapy (RT). Computed tomography (CT) and magnetic
resonance imaging (MRI) are two common modalities for providing RT treatment
planning and delivery guidance/monitoring. In recent decades, artificial
intelligence (AI) has emerged as a powerful and widely adopted technique across
various fields, valued for its efficiency and convenience enabled by implicit
function definition and data-driven feature representation learning. Here, we
present a series of AI-driven medical imaging reconstruction frameworks for
enhanced radiotherapy, designed to improve CT image reconstruction quality and
speed, refine dual-energy CT (DECT) multi-material decomposition (MMD), and
significantly accelerate 4D MRI acquisition.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 23:02:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Di",
""
]
] |
2504.08845 | Zheng-Chuan Wang | Zheng-Chuan Wang | The temperature dependent thermal vector potential in spinor Boltzmann
equation | 6 figures | null | null | null | cond-mat.stat-mech quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The thermal scalar and vector potential were introduced to investigate the
thermal transport under a temperature gradient in terms of linear response
theory[1,2]. However, the microscopic origin of these phenomenological thermal
potentials had not been addressed clearly. In this manuscript, we try to derive
a temperature dependent damping force based on the spinor Boltzmann equation
(SBE), and relate it with the thermal gauge potential, which is exactly the
temperature dependent thermal scalar and vector potential. It is shown that the
thermal potential originates from the scattering of conduction electrons and
impurity or other scattering mechanisms. We also derive a temperature dependent
inverse relaxation time, which depends on momentum, it is different from the
usual constant relaxation time. We evaluate the temperature dependent damping
force by an approximated analytical solution of SBE. The other physical
observable such as charge current and spin current are also explored.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 23:24:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Zheng-Chuan",
""
]
] |
2504.08846 | Benjamin Jasperson | Mostafa Faghih Shojaei, Rahul Gulati, Benjamin A. Jasperson,
Shangshang Wang, Simone Cimolato, Dangli Cao, Willie Neiswanger, Krishna
Garikipati | AI-University: An LLM-based platform for instructional alignment to
scientific classrooms | 10 pages, 3 figures | null | null | null | cs.CY cs.AI cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce AI University (AI-U), a flexible framework for AI-driven course
content delivery that adapts to instructors' teaching styles. At its core, AI-U
fine-tunes a large language model (LLM) with retrieval-augmented generation
(RAG) to generate instructor-aligned responses from lecture videos, notes, and
textbooks. Using a graduate-level finite-element-method (FEM) course as a case
study, we present a scalable pipeline to systematically construct training
data, fine-tune an open-source LLM with Low-Rank Adaptation (LoRA), and
optimize its responses through RAG-based synthesis. Our evaluation - combining
cosine similarity, LLM-based assessment, and expert review - demonstrates
strong alignment with course materials. We also have developed a prototype web
application, available at https://my-ai-university.com, that enhances
traceability by linking AI-generated responses to specific sections of the
relevant course material and time-stamped instances of the open-access video
lectures. Our expert model is found to have greater cosine similarity with a
reference on 86% of test cases. An LLM judge also found our expert model to
outperform the base Llama 3.2 model approximately four times out of five. AI-U
offers a scalable approach to AI-assisted education, paving the way for broader
adoption in higher education. Here, our framework has been presented in the
setting of a class on FEM - a subject that is central to training PhD and
Master students in engineering science. However, this setting is a particular
instance of a broader context: fine-tuning LLMs to research content in science.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 01:26:34 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shojaei",
"Mostafa Faghih",
""
],
[
"Gulati",
"Rahul",
""
],
[
"Jasperson",
"Benjamin A.",
""
],
[
"Wang",
"Shangshang",
""
],
[
"Cimolato",
"Simone",
""
],
[
"Cao",
"Dangli",
""
],
[
"Neiswanger",
"Willie",
""
],
[
"Garikipati",
"Krishna",
""
]
] |
2504.08847 | Qiang Zou | Guoyue Luo, Qiang Zou | Soap Film-inspired Subdivisional Lattice Structure Construction | null | null | null | null | cs.CG cs.GR | http://creativecommons.org/licenses/by/4.0/ | Lattice structures, distinguished by their customizable geometries at the
microscale and outstanding mechanical performance, have found widespread
application across various industries. One fundamental process in their design
and manufacturing is constructing boundary representation (B-rep) models, which
are essential for running advanced applications like simulation, optimization,
and process planning. However, this construction process presents significant
challenges due to the high complexity of lattice structures, particularly in
generating nodal shapes where robustness and smoothness issues can arise from
the complex intersections between struts. To address these challenges, this
paper proposes a novel approach for lattice structure construction by cutting
struts and filling void regions with subdivisional nodal shapes. Inspired by
soap films, the method generates smooth, shape-preserving control meshes using
Laplacian fairing and subdivides them through the point-normal Loop (PN-Loop)
subdivision scheme to obtain subdivisional nodal shapes. The proposed method
ensures robust model construction with reduced shape deviations, enhanced
surface fairness, and smooth transitions between subdivisional nodal shapes and
retained struts. The effectiveness of the method has been demonstrated by a
series of examples and comparisons. The code will be open-sourced upon
publication.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 01:33:23 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Luo",
"Guoyue",
""
],
[
"Zou",
"Qiang",
""
]
] |
2504.08848 | Bibek Upadhayay | Bibek Upadhayay, Vahid Behzadan, Ph.D | X-Guard: Multilingual Guard Agent for Content Moderation | 34 pages, 15 figures | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have rapidly become integral to numerous
applications in critical domains where reliability is paramount. Despite
significant advances in safety frameworks and guardrails, current protective
measures exhibit crucial vulnerabilities, particularly in multilingual
contexts. Existing safety systems remain susceptible to adversarial attacks in
low-resource languages and through code-switching techniques, primarily due to
their English-centric design. Furthermore, the development of effective
multilingual guardrails is constrained by the scarcity of diverse cross-lingual
training data. Even recent solutions like Llama Guard-3, while offering
multilingual support, lack transparency in their decision-making processes. We
address these challenges by introducing X-Guard agent, a transparent
multilingual safety agent designed to provide content moderation across diverse
linguistic contexts. X-Guard effectively defends against both conventional
low-resource language attacks and sophisticated code-switching attacks. Our
approach includes: curating and enhancing multiple open-source safety datasets
with explicit evaluation rationales; employing a jury of judges methodology to
mitigate individual judge LLM provider biases; creating a comprehensive
multilingual safety dataset spanning 132 languages with 5 million data points;
and developing a two-stage architecture combining a custom-finetuned mBART-50
translation module with an evaluation X-Guard 3B model trained through
supervised finetuning and GRPO training. Our empirical evaluations demonstrate
X-Guard's effectiveness in detecting unsafe content across multiple languages
while maintaining transparency throughout the safety evaluation process. Our
work represents a significant advancement in creating robust, transparent, and
linguistically inclusive safety systems for LLMs and its integrated systems.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 01:58:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Upadhayay",
"Bibek",
""
],
[
"Behzadan",
"Vahid",
""
],
[
"D",
"Ph.",
""
]
] |
2504.08849 | Mallika Mainali | Mallika Mainali, Rosina O. Weber | Exploring Cognitive Attributes in Financial Decision-Making | 7 pages, 2 figures. Presented in SIAM International Conference on
Data Mining (SDM25) METACOG-25: 2nd Workshop on Metacognitive Prediction of
AI Behavior | null | null | null | cs.CY cs.AI cs.CE | http://creativecommons.org/licenses/by-sa/4.0/ | Cognitive attributes are fundamental to metacognition, shaping how
individuals process information, evaluate choices, and make decisions. To
develop metacognitive artificial intelligence (AI) models that reflect human
reasoning, it is essential to account for the attributes that influence
reasoning patterns and decision-maker behavior, often leading to different or
even conflicting choices. This makes it crucial to incorporate cognitive
attributes in designing AI models that align with human decision-making
processes, especially in high-stakes domains such as finance, where decisions
have significant real-world consequences. However, existing AI alignment
research has primarily focused on value alignment, often overlooking the role
of individual cognitive attributes that distinguish decision-makers. To address
this issue, this paper (1) analyzes the literature on cognitive attributes, (2)
establishes five criteria for defining them, and (3) categorizes 19
domain-specific cognitive attributes relevant to financial decision-making.
These three components provide a strong basis for developing AI systems that
accurately reflect and align with human decision-making processes in financial
contexts.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 02:11:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mainali",
"Mallika",
""
],
[
"Weber",
"Rosina O.",
""
]
] |
2504.08850 | Jiaming Xu | Jiaming Xu, Jiayi Pan, Yongkang Zhou, Siming Chen, Jinhao Li, Yaoxiu
Lian, Junyi Wu, Guohao Dai | SpecEE: Accelerating Large Language Model Inference with Speculative
Early Exiting | Accepted by ISCA 2025 | null | null | null | cs.DC cs.AI | http://creativecommons.org/licenses/by/4.0/ | Early exiting has recently emerged as a promising technique for accelerating
large language models (LLMs) by effectively reducing the hardware computation
and memory access. In this paper, we present SpecEE, a fast LLM inference
engine with speculative early exiting. (1) At the algorithm level, we propose
the speculation-based lightweight predictor design by exploiting the
probabilistic correlation between the speculative tokens and the correct
results and high parallelism of GPUs. (2) At the system level, we point out
that not all layers need a predictor and design the two-level heuristic
predictor scheduling engine based on skewed distribution and contextual
similarity. (3) At the mapping level, we point out that different decoding
methods share the same essential characteristics, and propose the context-aware
merged mapping for predictor with efficient GPU implementations to support
speculative decoding, and form a framework for various existing orthogonal
acceleration techniques (e.g., quantization and sparse activation) on cloud and
personal computer (PC) scenarios, successfully pushing the Pareto frontier of
accuracy and speedup. It is worth noting that SpecEE can be applied to any LLM
by negligible training overhead in advance without affecting the model original
parameters. Extensive experiments show that SpecEE achieves 2.25x and 2.43x
speedup with Llama2-7B on cloud and PC scenarios respectively.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 02:38:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Jiaming",
""
],
[
"Pan",
"Jiayi",
""
],
[
"Zhou",
"Yongkang",
""
],
[
"Chen",
"Siming",
""
],
[
"Li",
"Jinhao",
""
],
[
"Lian",
"Yaoxiu",
""
],
[
"Wu",
"Junyi",
""
],
[
"Dai",
"Guohao",
""
]
] |
2504.08851 | Yuchu Jiang | Yuchu Jiang, Jiale Fu, Chenduo Hao, Xinting Hu, Yingzhe Peng, Xin
Geng, Xu Yang | Mimic In-Context Learning for Multimodal Tasks | 14 pages, 7 figures,CVPR 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recently, In-context Learning (ICL) has become a significant inference
paradigm in Large Multimodal Models (LMMs), utilizing a few in-context
demonstrations (ICDs) to prompt LMMs for new tasks. However, the synergistic
effects in multimodal data increase the sensitivity of ICL performance to the
configurations of ICDs, stimulating the need for a more stable and general
mapping function. Mathematically, in Transformer-based models, ICDs act as
``shift vectors'' added to the hidden states of query tokens. Inspired by this,
we introduce Mimic In-Context Learning (MimIC) to learn stable and
generalizable shift effects from ICDs. Specifically, compared with some
previous shift vector-based methods, MimIC more strictly approximates the shift
effects by integrating lightweight learnable modules into LMMs with four key
enhancements: 1) inserting shift vectors after attention layers, 2) assigning a
shift vector to each attention head, 3) making shift magnitude query-dependent,
and 4) employing a layer-wise alignment loss. Extensive experiments on two LMMs
(Idefics-9b and Idefics2-8b-base) across three multimodal tasks (VQAv2, OK-VQA,
Captioning) demonstrate that MimIC outperforms existing shift vector-based
methods. The code is available at https://github.com/Kamichanw/MimIC.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 03:37:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jiang",
"Yuchu",
""
],
[
"Fu",
"Jiale",
""
],
[
"Hao",
"Chenduo",
""
],
[
"Hu",
"Xinting",
""
],
[
"Peng",
"Yingzhe",
""
],
[
"Geng",
"Xin",
""
],
[
"Yang",
"Xu",
""
]
] |
2504.08852 | Lizhong Chen | Raymond Baartmans, Andrew Ensinger, Victor Agostinelli, Lizhong Chen | ML For Hardware Design Interpretability: Challenges and Opportunities | null | null | null | null | cs.LG cs.AI cs.AR | http://creativecommons.org/licenses/by/4.0/ | The increasing size and complexity of machine learning (ML) models have
driven the growing need for custom hardware accelerators capable of efficiently
supporting ML workloads. However, the design of such accelerators remains a
time-consuming process, heavily relying on engineers to manually ensure design
interpretability through clear documentation and effective communication.
Recent advances in large language models (LLMs) offer a promising opportunity
to automate these design interpretability tasks, particularly the generation of
natural language descriptions for register-transfer level (RTL) code, what we
refer to as "RTL-to-NL tasks." In this paper, we examine how design
interpretability, particularly in RTL-to-NL tasks, influences the efficiency of
the hardware design process. We review existing work adapting LLMs for these
tasks, highlight key challenges that remain unaddressed, including those
related to data, computation, and model development, and identify opportunities
to address them. By doing so, we aim to guide future research in leveraging ML
to automate RTL-to-NL tasks and improve hardware design interpretability,
thereby accelerating the hardware design process and meeting the increasing
demand for custom hardware accelerators in machine learning and beyond.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 03:47:51 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Baartmans",
"Raymond",
""
],
[
"Ensinger",
"Andrew",
""
],
[
"Agostinelli",
"Victor",
""
],
[
"Chen",
"Lizhong",
""
]
] |
2504.08853 | Thomas Montefiore Dr | Paul Formosa, In\^es Hip\'olito, Thomas Montefiore | Artificial Intelligence (AI) and the Relationship between Agency,
Autonomy, and Moral Patiency | null | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The proliferation of Artificial Intelligence (AI) systems exhibiting complex
and seemingly agentive behaviours necessitates a critical philosophical
examination of their agency, autonomy, and moral status. In this paper we
undertake a systematic analysis of the differences between basic, autonomous,
and moral agency in artificial systems. We argue that while current AI systems
are highly sophisticated, they lack genuine agency and autonomy because: they
operate within rigid boundaries of pre-programmed objectives rather than
exhibiting true goal-directed behaviour within their environment; they cannot
authentically shape their engagement with the world; and they lack the critical
self-reflection and autonomy competencies required for full autonomy.
Nonetheless, we do not rule out the possibility of future systems that could
achieve a limited form of artificial moral agency without consciousness through
hybrid approaches to ethical decision-making. This leads us to suggest, by
appealing to the necessity of consciousness for moral patiency, that such
non-conscious AMAs might represent a case that challenges traditional
assumptions about the necessary connection between moral agency and moral
patiency.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 03:48:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Formosa",
"Paul",
""
],
[
"Hipólito",
"Inês",
""
],
[
"Montefiore",
"Thomas",
""
]
] |
2504.08854 | Sujan Ghimire | Sujan Ghimire, Muhtasim Alam Chowdhury, Banafsheh Saber Latibari,
Muntasir Mamun, Jaeden Wolf Carpenter, Benjamin Tan, Hammond Pearce, Pratik
Satam, and Soheil Salehi | Hardware Design and Security Needs Attention: From Survey to Path
Forward | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Recent advances in attention-based artificial intelligence (AI) models have
unlocked vast potential to automate digital hardware design while enhancing and
strengthening security measures against various threats. This rapidly emerging
field leverages Large Language Models (LLMs) to generate HDL code, identify
vulnerabilities, and sometimes mitigate them. The state of the art in this
design automation space utilizes optimized LLMs with HDL datasets, creating
automated systems for register-transfer level (RTL) generation, verification,
and debugging, and establishing LLM-driven design environments for streamlined
logic designs. Additionally, attention-based models like graph attention have
shown promise in chip design applications, including floorplanning. This survey
investigates the integration of these models into hardware-related domains,
emphasizing logic design and hardware security, with or without the use of IP
libraries. This study explores the commercial and academic landscape,
highlighting technical hurdles and future prospects for automating hardware
design and security. Moreover, it provides new insights into the study of
LLM-driven design systems, advances in hardware security mechanisms, and the
impact of influential works on industry practices. Through the examination of
30 representative approaches and illustrative case studies, this paper
underscores the transformative potential of attention-based models in
revolutionizing hardware design while addressing the challenges that lie ahead
in this interdisciplinary domain.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 03:48:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ghimire",
"Sujan",
""
],
[
"Chowdhury",
"Muhtasim Alam",
""
],
[
"Latibari",
"Banafsheh Saber",
""
],
[
"Mamun",
"Muntasir",
""
],
[
"Carpenter",
"Jaeden Wolf",
""
],
[
"Tan",
"Benjamin",
""
],
[
"Pearce",
"Hammond",
""
],
[
"Satam",
"Pratik",
""
],
[
"Salehi",
"Soheil",
""
]
] |
2504.08855 | Rhea Pritham Marpu | Kevin J McNamara, Rhea Pritham Marpu | Exponential Shift: Humans Adapt to AI Economies | null | null | null | null | cs.CY cs.AI cs.CE cs.ET | http://creativecommons.org/licenses/by/4.0/ | This paper explores how artificial intelligence (AI) and robotics are
transforming the global labor market. Human workers, limited to a 33% duty
cycle due to rest and holidays, cost $14 to $55 per hour. In contrast, digital
labor operates nearly 24/7 at just $0.10 to $0.50 per hour. We examine sectors
like healthcare, education, manufacturing, and retail, finding that 40-70% of
tasks could be automated. Yet, human skills like emotional intelligence and
adaptability remain essential. Humans process 5,000-20,000 tokens (units of
information) per hour, while AI far exceeds this, though its energy use-3.5 to
7 times higher than humans-could offset 20-40% of cost savings. Using
real-world examples, such as AI in journalism and law, we illustrate these
dynamics and propose six strategies-like a 4-day workweek and retraining-to
ensure a fair transition to an AI-driven economy.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 04:43:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"McNamara",
"Kevin J",
""
],
[
"Marpu",
"Rhea Pritham",
""
]
] |
2504.08856 | Boxuan Ma Dr. | Tianyuan Yang, Ren Baofeng, Chenghao Gu, Tianjia He, Boxuan Ma,
Shinichi Konomi | Examining GPT's Capability to Generate and Map Course Concepts and Their
Relationship | null | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | Extracting key concepts and their relationships from course information and
materials facilitates the provision of visualizations and recommendations for
learners who need to select the right courses to take from a large number of
courses. However, identifying and extracting themes manually is labor-intensive
and time-consuming. Previous machine learning-based methods to extract relevant
concepts from courses heavily rely on detailed course materials, which
necessitates labor-intensive preparation of course materials. This paper
investigates the potential of LLMs such as GPT in automatically generating
course concepts and their relations. Specifically, we design a suite of prompts
and provide GPT with the course information with different levels of detail,
thereby generating high-quality course concepts and identifying their
relations. Furthermore, we comprehensively evaluate the quality of the
generated concepts and relationships through extensive experiments. Our results
demonstrate the viability of LLMs as a tool for supporting educational content
selection and delivery.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 05:03:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Tianyuan",
""
],
[
"Baofeng",
"Ren",
""
],
[
"Gu",
"Chenghao",
""
],
[
"He",
"Tianjia",
""
],
[
"Ma",
"Boxuan",
""
],
[
"Konomi",
"Shinichi",
""
]
] |
2504.08857 | Hanyu Zhu | Han-Yu Zhu, Yin-Ting Zhang, Wen-Jie Xie, Wei-Xing Zhou | Structural robustness of the international food supply network under
external shocks and its determinants | null | null | null | null | econ.GN q-fin.EC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The stability of the global food supply network is critical for ensuring food
security. This study constructs an aggregated international food supply network
based on the trade data of four staple crops and evaluates its structural
robustness through network integrity under accumulating external shocks.
Network integrity is typically quantified in network science by the relative
size of the largest connected component, and we propose a new robustness metric
that incorporates both the broadness p and severity q of external shocks. Our
findings reveal that the robustness of the network has gradually increased over
the past decades, punctuated by temporary declines that can be explained by
major historical events. While the aggregated network remains robust under
moderate disruptions, extreme shocks targeting key suppliers such as the United
States and India can trigger systemic collapse. When the shock broadness p is
less than about 0.3 and the shock severity q is close to 1, the structural
robustness curves S(p,q) decrease linearly with respect to the shock broadness
p, suggesting that the most critical economies have relatively even influence
on network integrity. Comparing the robustness curves of the four individual
staple foods, we find that the soybean supply network is the least robust.
Furthermore, regression and machine learning analyses show that increaseing
food (particularly rice and soybean) production enhances network robustness,
while rising food prices significantly weaken it.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 05:45:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhu",
"Han-Yu",
""
],
[
"Zhang",
"Yin-Ting",
""
],
[
"Xie",
"Wen-Jie",
""
],
[
"Zhou",
"Wei-Xing",
""
]
] |
2504.08858 | Zinnat Hassan | Zinnat Hassan | Exploring Wormholes in Modified Theories of Gravity | Ph.D Thesis. Some chapters are published in the following journals:
Fortschritte der Physik 69, 2100023 (2021); General Relativity and
Gravitation 55, 90 (2023); Classical Quantum Gravity 41, 235001 (2024);
Annalen der Physik 536, 2400114 (2024) | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | This thesis investigates traversable wormhole spacetimes in modified theories
of gravity, where the matter at the wormhole throat is described by an
anisotropic energy-momentum tensor. Wormhole solutions are constructed in the
context of $f(Q)$ gravity under various equations of state, and their physical
viability and stability are thoroughly examined. The influence of the
Generalized Uncertainty Principle (GUP) on Casimir wormholes is also analyzed,
emphasizing the role of the GUP parameter in shaping wormhole geometry in
$f(Q)$ gravity. Additionally, the study investigates the observational aspects
of wormholes, such as their shadow profiles and deflection angles in the
presence of dark matter. The work is further extended to four-dimensional
Einstein-Gauss-Bonnet (EGB) gravity, where the effects of dark matter models
and the Gauss-Bonnet coupling parameter on the energy conditions are evaluated.
Various physical properties of the wormholes, including the complexity factor,
active gravitational mass, and total gravitational energy, are also discussed.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 06:58:48 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hassan",
"Zinnat",
""
]
] |
2504.08859 | Fanmeng Wang | Fanmeng Wang, Wentao Guo, Qi Ou, Hongshuai Wang, Haitao Lin, Hongteng
Xu, Zhifeng Gao | PolyConf: Unlocking Polymer Conformation Generation through Hierarchical
Generative Models | null | null | null | null | cond-mat.soft cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Polymer conformation generation is a critical task that enables atomic-level
studies of diverse polymer materials. While significant advances have been made
in designing various conformation generation methods for small molecules and
proteins, these methods struggle to generate polymer conformations due to
polymers' unique structural characteristics. The scarcity of polymer
conformation datasets further limits progress, making this promising area
largely unexplored. In this work, we propose PolyConf, a pioneering tailored
polymer conformation generation method that leverages hierarchical generative
models to unlock new possibilities for this task. Specifically, we decompose
the polymer conformation into a series of local conformations (i.e., the
conformations of its repeating units), generating these local conformations
through an autoregressive model. We then generate corresponding orientation
transformations via a diffusion model to assemble these local conformations
into the complete polymer conformation. Moreover, we develop the first
benchmark with a high-quality polymer conformation dataset derived from
molecular dynamics simulations to boost related research in this area. The
comprehensive evaluation demonstrates that PolyConf consistently generates
high-quality polymer conformations, facilitating advancements in polymer
modeling and simulation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 07:12:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Fanmeng",
""
],
[
"Guo",
"Wentao",
""
],
[
"Ou",
"Qi",
""
],
[
"Wang",
"Hongshuai",
""
],
[
"Lin",
"Haitao",
""
],
[
"Xu",
"Hongteng",
""
],
[
"Gao",
"Zhifeng",
""
]
] |
2504.08860 | Chen Yan | Chen Yan, Boyu Diao, Hangda Liu, Zhulin An and Yongjun Xu | A Nonlinear Hash-based Optimization Method for SpMV on GPUs | This article has been indexed by CCGrid2025 | null | null | null | cs.DC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sparse matrix-vector multiplication (SpMV) is a fundamental operation with a
wide range of applications in scientific computing and artificial intelligence.
However, the large scale and sparsity of sparse matrix often make it a
performance bottleneck. In this paper, we highlight the effectiveness of
hash-based techniques in optimizing sparse matrix reordering, introducing the
Hash-based Partition (HBP) format, a lightweight SpMV approach. HBP retains the
performance benefits of the 2D-partitioning method while leveraging the hash
transformation's ability to group similar elements, thereby accelerating the
pre-processing phase of sparse matrix reordering. Additionally, we achieve
parallel load balancing across matrix blocks through a competitive method. Our
experiments, conducted on both Nvidia Jetson AGX Orin and Nvidia RTX 4090, show
that in the pre-processing step, our method offers an average speedup of 3.53
times compared to the sorting approach and 3.67 times compared to the dynamic
programming method employed in Regu2D. Furthermore, in SpMV, our method
achieves a maximum speedup of 3.32 times on Orin and 3.01 times on RTX4090
against the CSR format in sparse matrices from the University of Florida Sparse
Matrix Collection.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 08:31:44 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yan",
"Chen",
""
],
[
"Diao",
"Boyu",
""
],
[
"Liu",
"Hangda",
""
],
[
"An",
"Zhulin",
""
],
[
"Xu",
"Yongjun",
""
]
] |
2504.08861 | Joshua Hatherley | Joshua Hatherley, Robert Sparrow | Diachronic and synchronic variation in the performance of adaptive
machine learning systems: The ethical challenges | null | 2023. Journal of the American Medical Informatics Association
30(2): 361-366 | 10.1093/jamia/ocac218 | null | cs.HC cs.AI cs.CY cs.LG | http://creativecommons.org/licenses/by/4.0/ | Objectives: Machine learning (ML) has the potential to facilitate "continual
learning" in medicine, in which an ML system continues to evolve in response to
exposure to new data over time, even after being deployed in a clinical
setting. In this paper, we provide a tutorial on the range of ethical issues
raised by the use of such "adaptive" ML systems in medicine that have, thus
far, been neglected in the literature.
Target audience: The target audiences for this tutorial are the developers of
machine learning AI systems, healthcare regulators, the broader medical
informatics community, and practicing clinicians.
Scope: Discussions of adaptive ML systems to date have overlooked the
distinction between two sorts of variance that such systems may exhibit --
diachronic evolution (change over time) and synchronic variation (difference
between cotemporaneous instantiations of the algorithm at different sites) --
and under-estimated the significance of the latter. We highlight the challenges
that diachronic evolution and synchronic variation present for the quality of
patient care, informed consent, and equity, and discuss the complex ethical
trade-offs involved in the design of such systems.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 09:01:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hatherley",
"Joshua",
""
],
[
"Sparrow",
"Robert",
""
]
] |
2504.08862 | Peiyang Wu | Peiyang Wu, Nan Guo, Junliang Lv, Xiao Xiao, Xiaochun Ye | RTLRepoCoder: Repository-Level RTL Code Completion through the
Combination of Fine-Tuning and Retrieval Augmentation | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by/4.0/ | As an essential part of modern hardware design, manually writing Register
Transfer Level (RTL) code such as Verilog is often labor-intensive. Following
the tremendous success of large language models (LLMs), researchers have begun
to explore utilizing LLMs for generating RTL code. However, current studies
primarily focus on generating simple single modules, which can not meet the
demands in real world. In fact, due to challenges in managing long-context RTL
code and complex cross-file dependencies, existing solutions cannot handle
large-scale Verilog repositories in practical hardware development. As the
first endeavor to exclusively adapt LLMs for large-scale RTL development, we
propose RTLRepoCoder, a groundbreaking solution that incorporates specific
fine-tuning and Retrieval-Augmented Generation (RAG) for repository-level
Verilog code completion. Open-source Verilog repositories from the real world,
along with an extended context size, are used for domain-specific fine-tuning.
The optimized RAG system improves the information density of the input context
by retrieving relevant code snippets. Tailored optimizations for RAG are
carried out, including the embedding model, the cross-file context splitting
strategy, and the chunk size. Our solution achieves state-of-the-art
performance on public benchmark, significantly surpassing GPT-4 and advanced
domain-specific LLMs on Edit Similarity and Exact Match rate. Comprehensive
experiments demonstrate the remarkable effectiveness of our approach and offer
insights for future work.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 09:04:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wu",
"Peiyang",
""
],
[
"Guo",
"Nan",
""
],
[
"Lv",
"Junliang",
""
],
[
"Xiao",
"Xiao",
""
],
[
"Ye",
"Xiaochun",
""
]
] |
2504.08863 | Nicholas Sukiennik | Nicholas Sukiennik, Chen Gao, Fengli Xu, Yong Li | An Evaluation of Cultural Value Alignment in LLM | Submitted to COLM 2025 | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | LLMs as intelligent agents are being increasingly applied in scenarios where
human interactions are involved, leading to a critical concern about whether
LLMs are faithful to the variations in culture across regions. Several works
have investigated this question in various ways, finding that there are biases
present in the cultural representations of LLM outputs. To gain a more
comprehensive view, in this work, we conduct the first large-scale evaluation
of LLM culture assessing 20 countries' cultures and languages across ten LLMs.
With a renowned cultural values questionnaire and by carefully analyzing LLM
output with human ground truth scores, we thoroughly study LLMs' cultural
alignment across countries and among individual models. Our findings show that
the output over all models represents a moderate cultural middle ground. Given
the overall skew, we propose an alignment metric, revealing that the United
States is the best-aligned country and GLM-4 has the best ability to align to
cultural values. Deeper investigation sheds light on the influence of model
origin, prompt language, and value dimensions on cultural output. Specifically,
models, regardless of where they originate, align better with the US than they
do with China. The conclusions provide insight to how LLMs can be better
aligned to various cultures as well as provoke further discussion of the
potential for LLMs to propagate cultural bias and the need for more culturally
adaptable models.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 09:13:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sukiennik",
"Nicholas",
""
],
[
"Gao",
"Chen",
""
],
[
"Xu",
"Fengli",
""
],
[
"Li",
"Yong",
""
]
] |
2504.08864 | Kacha Dzhaparidze | Kacha Dzhaparidze | Orthogonal series for si- and related processes, Karhunen-Lo\`eve
decompositions | null | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | This paper reproduces results from Chapter 11 of the forthcoming book
\cite{dzh25}. It discusses series expansions of processes with stationary
increments (si-processes) and certain associated processes. Making use of de
Branges theory of Hilbert spaces of entire functions, it sheds new light on the
existing literature and makes available some new results. In particular, it
provides some new decompositions of the Karhunen-Lo\`eve type.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:38:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dzhaparidze",
"Kacha",
""
]
] |
2504.08865 | Yan Haoran | Haoran Yan, Yinfang Chen, Minghua Ma, Ming Wen, Shan Lu, Shenglin
Zhang, Tianyin Xu, Rujia Wang, Chetan Bansal, Saravan Rajmohan, Chaoyun
Zhang, Dongmei Zhang | An Empirical Study of Production Incidents in Generative AI Cloud
Services | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ever-increasing demand for generative artificial intelligence (GenAI) has
motivated cloud-based GenAI services such as Azure OpenAI Service and Amazon
Bedrock. Like any large-scale cloud service, failures are inevitable in
cloud-based GenAI services, resulting in user dissatisfaction and significant
monetary losses. However, GenAI cloud services, featured by their massive
parameter scales, hardware demands, and usage patterns, present unique
challenges, including generated content quality issues and privacy concerns,
compared to traditional cloud services. To understand the production
reliability of GenAI cloud services, we analyzed production incidents from a
leading GenAI cloud service provider spanning in the past four years. Our study
(1) presents the general characteristics of GenAI cloud service incidents at
different stages of the incident life cycle; (2) identifies the symptoms and
impacts of these incidents on GenAI cloud service quality and availability; (3)
uncovers why these incidents occurred and how they were resolved; (4) discusses
open research challenges in terms of incident detection, triage, and
mitigation, and sheds light on potential solutions.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:39:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yan",
"Haoran",
""
],
[
"Chen",
"Yinfang",
""
],
[
"Ma",
"Minghua",
""
],
[
"Wen",
"Ming",
""
],
[
"Lu",
"Shan",
""
],
[
"Zhang",
"Shenglin",
""
],
[
"Xu",
"Tianyin",
""
],
[
"Wang",
"Rujia",
""
],
[
"Bansal",
"Chetan",
""
],
[
"Rajmohan",
"Saravan",
""
],
[
"Zhang",
"Chaoyun",
""
],
[
"Zhang",
"Dongmei",
""
]
] |
2504.08866 | Mohammad Jalwana | Mohammad A.A.K. Jalwana, Naveed Akhtar, Ajmal Mian, Nazanin Rahnavard,
Mubarak Shah | On Transfer-based Universal Attacks in Pure Black-box Setting | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Despite their impressive performance, deep visual models are susceptible to
transferable black-box adversarial attacks. Principally, these attacks craft
perturbations in a target model-agnostic manner. However, surprisingly, we find
that existing methods in this domain inadvertently take help from various
priors that violate the black-box assumption such as the availability of the
dataset used to train the target model, and the knowledge of the number of
classes in the target model. Consequently, the literature fails to articulate
the true potency of transferable black-box attacks. We provide an empirical
study of these biases and propose a framework that aids in a prior-free
transparent study of this paradigm. Using our framework, we analyze the role of
prior knowledge of the target model data and number of classes in attack
performance. We also provide several interesting insights based on our
analysis, and demonstrate that priors cause overestimation in transferability
scores. Finally, we extend our framework to query-based attacks. This extension
inspires a novel image-blending technique to prepare data for effective
surrogate model training.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:41:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jalwana",
"Mohammad A. A. K.",
""
],
[
"Akhtar",
"Naveed",
""
],
[
"Mian",
"Ajmal",
""
],
[
"Rahnavard",
"Nazanin",
""
],
[
"Shah",
"Mubarak",
""
]
] |
2504.08867 | Felix Benning | Felix Benning, Steffen Dereich | In almost all shallow analytic neural network optimization landscapes,
efficient minimizers have strongly convex neighborhoods | null | null | null | null | cs.LG math.PR stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Whether or not a local minimum of a cost function has a strongly convex
neighborhood greatly influences the asymptotic convergence rate of optimizers.
In this article, we rigorously analyze the prevalence of this property for the
mean squared error induced by shallow, 1-hidden layer neural networks with
analytic activation functions when applied to regression problems. The
parameter space is divided into two domains: the 'efficient domain' (all
parameters for which the respective realization function cannot be generated by
a network having a smaller number of neurons) and the 'redundant domain' (the
remaining parameters). In almost all regression problems on the efficient
domain the optimization landscape only features local minima that are strongly
convex. Formally, we will show that for certain randomly picked regression
problems the optimization landscape is almost surely a Morse function on the
efficient domain. The redundant domain has significantly smaller dimension than
the efficient domain and on this domain, potential local minima are never
isolated.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:43:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Benning",
"Felix",
""
],
[
"Dereich",
"Steffen",
""
]
] |
2504.08868 | Maxim Krasnov | Maxim Krasnov, Ufuk Aydemir, Maxim Khlopov | Spontaneous baryogenesis with large misalignment | Prepared for submission to Phys. Lett. B | null | null | null | hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | We investigate the particle production by the Nambu-Goldstone boson in the
spontaneous baryogenesis scenario for large misalignment angles. Studying
numerically the case of an arbitrary initial phase, we reproduce the cubic
dependence on the initial value of the phase of baryon asymmetry in the case of
small oscillations and present our results for an initial phase close to {\pi}.
Our calculations indicate that there is a saturation of particle production as
the initial phase approaches {\pi}. We also discuss the probability
distribution of the baryon asymmetry.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:45:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Krasnov",
"Maxim",
""
],
[
"Aydemir",
"Ufuk",
""
],
[
"Khlopov",
"Maxim",
""
]
] |
2504.08869 | Markus Verlemann | Markus Verlemann | Gluing charged black holes into de Sitter space | 29 pages, 0 figures, MSc thesis | null | null | null | gr-qc math-ph math.AP math.DG math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend Hintz's cosmological black hole gluing result to the
Einstein-Maxwell system with positive cosmological constant by gluing multiple
Reissner-Nordstr\"om or Kerr--Newman--de Sitter black holes into neighbourhoods
of points in the conformal boundary of de Sitter space. We determine necessary
and sufficient conditions on the black hole parameters -- related to
Friedrich's conformal constraint equations -- for this gluing to be possible.
We also improve the original gluing method slightly by showing that the
construction of a solution in Taylor series may be accomplished using an
exactness argument, eliminating the need for an early gauge-fixing.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 11:24:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Verlemann",
"Markus",
""
]
] |
2504.08870 | Moritz Schlaich | Moritz Schlaich, Paul Fischer, Paul Florian Giesel, Clara Klink,
Alexandre Obertelli, Lutz Schweikhard, Frank Wienholtz | The PUMA offline ion source beamline | null | null | null | null | physics.acc-ph nucl-ex | http://creativecommons.org/licenses/by/4.0/ | The antiProton Unstable Matter Annihilation experiment (PUMA) at CERN aims to
study the nucleonic composition in the matter density tail of stable and
radioactive nuclei using low-energy antiprotons. Since there is no facility in
which both low-energy antiprotons and radioactive nuclei can be produced, the
experimental realization with exotic nuclei requires the transportation of the
antiprotons from the Extra Low ENergy Antiproton (ELENA) facility to the nearby
located Isotope mass Separator On-Line DEvice (ISOLDE). For tests and first
applications of the proposed experimental technique to stable isotopes at
ELENA, a dedicated offline ion source beamline was developed that will provide
isotopically pure, cooled and bunched ion beams with intensities of more than
$10^4$ ions per bunch while maintaining a vacuum of better than $5\times
10^{-10}$ mbar at the handover point. This offline ion source beamline is
characterized and its capabilities are demonstrated using the example of stable
krypton isotopes.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 11:28:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Schlaich",
"Moritz",
""
],
[
"Fischer",
"Paul",
""
],
[
"Giesel",
"Paul Florian",
""
],
[
"Klink",
"Clara",
""
],
[
"Obertelli",
"Alexandre",
""
],
[
"Schweikhard",
"Lutz",
""
],
[
"Wienholtz",
"Frank",
""
]
] |
2504.08871 | Mayank Raikwar | Danilo Gligoroski, Mayank Raikwar, Sonu Kumar Jha | An LLM Framework For Cryptography Over Chat Channels | 27 Pages | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Recent advancements in Large Language Models (LLMs) have transformed
communication, yet their role in secure messaging remains underexplored,
especially in surveillance-heavy environments. At the same time, many
governments all over the world are proposing legislation to detect, backdoor,
or even ban encrypted communication. That emphasizes the need for alternative
ways to communicate securely and covertly over open channels. We propose a
novel cryptographic embedding framework that enables covert Public Key or
Symmetric Key encrypted communication over public chat channels with humanlike
produced texts. Some unique properties of our framework are: 1. It is LLM
agnostic, i.e., it allows participants to use different local LLM models
independently; 2. It is pre- or post-quantum agnostic; 3. It ensures
indistinguishability from human-like chat-produced texts. Thus, it offers a
viable alternative where traditional encryption is detectable and restricted.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 11:34:14 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gligoroski",
"Danilo",
""
],
[
"Raikwar",
"Mayank",
""
],
[
"Jha",
"Sonu Kumar",
""
]
] |
2504.08872 | Shuaijun Chen | Seunghyun Lee, Omid Tavallaie, Shuaijun Chen, Kanchana Thilakarathna,
Suranga Seneviratne, Adel Nadjaran Toosi, Albert Y. Zomaya | Personalizing Federated Learning for Hierarchical Edge Networks with
Non-IID Data | null | null | null | null | cs.LG cs.AI cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accommodating edge networks between IoT devices and the cloud server in
Hierarchical Federated Learning (HFL) enhances communication efficiency without
compromising data privacy. However, devices connected to the same edge often
share geographic or contextual similarities, leading to varying edge-level data
heterogeneity with different subsets of labels per edge, on top of device-level
heterogeneity. This hierarchical non-Independent and Identically Distributed
(non-IID) nature, which implies that each edge has its own optimization goal,
has been overlooked in HFL research. Therefore, existing edge-accommodated HFL
demonstrates inconsistent performance across edges in various hierarchical
non-IID scenarios. To ensure robust performance with diverse edge-level non-IID
data, we propose a Personalized Hierarchical Edge-enabled Federated Learning
(PHE-FL), which personalizes each edge model to perform well on the unique
class distributions specific to each edge. We evaluated PHE-FL across 4
scenarios with varying levels of edge-level non-IIDness, with extreme IoT
device level non-IIDness. To accurately assess the effectiveness of our
personalization approach, we deployed test sets on each edge server instead of
the cloud server, and used both balanced and imbalanced test sets. Extensive
experiments show that PHE-FL achieves up to 83 percent higher accuracy compared
to existing federated learning approaches that incorporate edge networks, given
the same number of training rounds. Moreover, PHE-FL exhibits improved
stability, as evidenced by reduced accuracy fluctuations relative to the
state-of-the-art FedAvg with two-level (edge and cloud) aggregation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 11:42:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lee",
"Seunghyun",
""
],
[
"Tavallaie",
"Omid",
""
],
[
"Chen",
"Shuaijun",
""
],
[
"Thilakarathna",
"Kanchana",
""
],
[
"Seneviratne",
"Suranga",
""
],
[
"Toosi",
"Adel Nadjaran",
""
],
[
"Zomaya",
"Albert Y.",
""
]
] |
2504.08873 | Alessandra Lingua | Alessandra Lingua, Antoine Sanner, Fran\c{c}ois Hild, and David S.
Kammer | Breaking better: Imperfections increase fracture resistance in
architected lattices | null | null | null | null | cond-mat.soft cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Architected materials offer unique opportunities to tailor fracture
properties through local structural modifications. In this study, we
investigate how the failure of architected materials with triangular lattice
topology is affected by the removal of individual struts, which represent
well-controlled and localized imperfections. Using a combination of macroscopic
mechanical testing and digital image correlation (DIC), we analyze both global
response and local crack propagation. We observe that the designed
imperfections do not alter the failure initiation site nor the peak tensile
load but significantly increase the work to failure. DIC-based tracking reveals
that this increase correlates with deviations in the crack path and may also
involve mechanisms such as crack bridging or temporary pinning near defects.
These results demonstrate that small, well-characterized imperfections, when
properly mastered, can be harnessed to improve failure resistance and expand
the design space of architected materials beyond regular, periodic structures.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 12:20:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lingua",
"Alessandra",
""
],
[
"Sanner",
"Antoine",
""
],
[
"Hild",
"François",
""
],
[
"Kammer",
"David S.",
""
]
] |
2504.08874 | Yasser Jangjou | Roshan Patel, Saeed Moayedpour, Louis De Lescure, Lorenzo
Kogler-Anele, Alan Cherney, Sven Jager and Yasser Jangjou | Distilling and exploiting quantitative insights from Large Language
Models for enhanced Bayesian optimization of chemical reactions | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Machine learning and Bayesian optimization (BO) algorithms can significantly
accelerate the optimization of chemical reactions. Transfer learning can
bolster the effectiveness of BO algorithms in low-data regimes by leveraging
pre-existing chemical information or data outside the direct optimization task
(i.e., source data). Large language models (LLMs) have demonstrated that
chemical information present in foundation training data can give them utility
for processing chemical data. Furthermore, they can be augmented with and help
synthesize potentially multiple modalities of source chemical data germane to
the optimization task. In this work, we examine how chemical information from
LLMs can be elicited and used for transfer learning to accelerate the BO of
reaction conditions to maximize yield. Specifically, we show that a survey-like
prompting scheme and preference learning can be used to infer a utility
function which models prior chemical information embedded in LLMs over a
chemical parameter space; we find that the utility function shows modest
correlation to true experimental measurements (yield) over the parameter space
despite operating in a zero-shot setting. Furthermore, we show that the utility
function can be leveraged to focus BO efforts in promising regions of the
parameter space, improving the yield of the initial BO query and enhancing
optimization in 4 of the 6 datasets studied. Overall, we view this work as a
step towards bridging the gap between the chemistry knowledge embedded in LLMs
and the capabilities of principled BO methods to accelerate reaction
optimization.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 12:45:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Patel",
"Roshan",
""
],
[
"Moayedpour",
"Saeed",
""
],
[
"De Lescure",
"Louis",
""
],
[
"Kogler-Anele",
"Lorenzo",
""
],
[
"Cherney",
"Alan",
""
],
[
"Jager",
"Sven",
""
],
[
"Jangjou",
"Yasser",
""
]
] |
2504.08875 | Xijin Ge | Xijin Ge | DataMap: A Portable Application for Visualizing High-Dimensional Data | null | null | null | null | q-bio.QM cs.HC cs.LG stat.AP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Motivation: The visualization and analysis of high-dimensional data are
essential in biomedical research. There is a need for secure, scalable, and
reproducible tools to facilitate data exploration and interpretation. Results:
We introduce DataMap, a browser-based application for visualization of
high-dimensional data using heatmaps, principal component analysis (PCA), and
t-distributed stochastic neighbor embedding (t-SNE). DataMap runs in the web
browser, ensuring data privacy while eliminating the need for installation or a
server. The application has an intuitive user interface for data
transformation, annotation, and generation of reproducible R code. Availability
and Implementation: Freely available as a GitHub page
https://gexijin.github.io/datamap/. The source code can be found at
https://github.com/gexijin/datamap, and can also be installed as an R package.
Contact: [email protected]
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:53:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ge",
"Xijin",
""
]
] |
2504.08877 | Gabriele Civitarese Dr. | Gabriele Civitarese, Michele Fiori, Andrea Arighi, Daniela Galimberti,
Graziana Florio, Claudio Bettini | The SERENADE project: Sensor-Based Explainable Detection of Cognitive
Decline | null | null | null | null | cs.LG cs.CY | http://creativecommons.org/licenses/by/4.0/ | Mild Cognitive Impairment (MCI) affects 12-18% of individuals over 60. MCI
patients exhibit cognitive dysfunctions without significant daily functional
loss. While MCI may progress to dementia, predicting this transition remains a
clinical challenge due to limited and unreliable indicators. Behavioral
changes, like in the execution of Activities of Daily Living (ADLs), can signal
such progression. Sensorized smart homes and wearable devices offer an
innovative solution for continuous, non-intrusive monitoring ADLs for MCI
patients. However, current machine learning models for detecting behavioral
changes lack transparency, hindering clinicians' trust. This paper introduces
the SERENADE project, a European Union-funded initiative that aims to detect
and explain behavioral changes associated with cognitive decline using
explainable AI methods. SERENADE aims at collecting one year of data from 30
MCI patients living alone, leveraging AI to support clinical decision-making
and offering a new approach to early dementia detection.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:36:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Civitarese",
"Gabriele",
""
],
[
"Fiori",
"Michele",
""
],
[
"Arighi",
"Andrea",
""
],
[
"Galimberti",
"Daniela",
""
],
[
"Florio",
"Graziana",
""
],
[
"Bettini",
"Claudio",
""
]
] |
2504.08878 | Mircea Andrecut Dr | M. Andrecut | Entropically Driven Agents | 16 pages, 9 figures | International Journal of Modern Physics C, 2025 | 10.1142/S0129183125500998 | null | cond-mat.stat-mech nlin.AO physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Populations of agents often exhibit surprising collective behavior emerging
from simple local interactions. The common belief is that the agents must
posses a certain level of cognitive abilities for such an emerging collective
behavior to occur. However, contrary to this assumption, it is also well known
that even noncognitive agents are capable of displaying nontrivial behavior.
Here we consider an intermediate case, where the agents borrow a little bit
from both extremes. We assume a population of agents performing random-walk in
a bounded environment, on a square lattice. The agents can sense their
immediate neighborhood, and they will attempt to move into a randomly selected
empty site, by avoiding collisions.Also, the agents will temporary stop moving
when they are in contact with at least two other agents. We show that
surprisingly, such a rudimentary population of agents undergoes a percolation
phase transition and self-organizes in a large polymer like structure, as a
consequence of an attractive entropic force emerging from their
restricted-valence and local spatial arrangement.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:41:55 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Andrecut",
"M.",
""
]
] |
2504.08879 | Zachary Pierce Scofield | Zachary P. Scofield, M. James Jee, Sangjun Cha, Hyosun Park | Is Earendel a Star?: Investigating the Sunrise Arc Using JWST Strong and
Weak Gravitational Lensing Analyses | 23 pages, 13 figures | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The galaxy cluster WHL J013719.8-08284 at $z = 0.566$ exhibits a
strong-lensing feature known as the Sunrise Arc, which hosts the candidate star
Earendel at $z \approx 6.2$, the most distant star candidate observed to date.
If this object is a star, or a system of a few stars, its apparent magnitude
implies both extreme gravitational lensing magnification and unusually high
luminosity. This study revisits Earendel's magnification, which, in previous
literature, exhibits significant uncertainty across various lens models ($2\mu
= 4{,}000$-$35{,}000$). We present an improved cluster mass reconstruction and
a tighter constraint on Earendel's magnification using a joint strong- and
weak-lensing analysis with JWST data. Our strong-lensing mass model,
incorporating newly identified multiple-image systems from JWST imaging data
and modifying the existing multiple-image assignment scheme, produces a
root-mean-square (RMS) lens-plane scatter of less than $0.''3$. Additionally,
our weak-lensing catalog achieves a source density of $\sim 100$ galaxies
arcmin$^{-2}$, providing constraints on the mass profile beyond the
strong-lensing regime. In our best-fit model, we estimate the magnification of
Earendel to be $\mu = 43$-$67$, significantly lower than previously proposed
and thus calling into question its classification as a star.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Scofield",
"Zachary P.",
""
],
[
"Jee",
"M. James",
""
],
[
"Cha",
"Sangjun",
""
],
[
"Park",
"Hyosun",
""
]
] |
2504.08880 | Joris Josiek | Joris Josiek, Andreas A. C. Sander, Matheus Bernini-Peron, Sylvia
Ekstr\"om, Gemma Gonz\'alez-Tor\`a, Roel R. Lefever, Nicolas Moens, Varsha
Ramachandran, Elisa C. Sch\"osser | Exploring the connection between atmosphere models and evolution models
of very massive stars | Accepted for publication in A&A, 12(+1) pages, 10(+1) figures | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Very massive stars (VMS) dominate the light of young stellar populations and
are sources of intense stellar feedback. Their evolution is mainly driven by
strong wind mass loss, yet current evolution models make simplistic assumptions
on their atmospheric physics which are incompatible with the nature of VMS. In
this work, we aim to understand VMS atmospheres throughout their evolution by
supplementing structure models (computed with GENEC) with detailed atmosphere
models (computed with PoWR) capable of capturing the physics of a
radially-expanding medium in non-LTE. An important aspect is the computation of
atmosphere models reaching into deeper layers of the star, notably including
the iron-opacity peak as an important source of radiative driving. In this
study, we compute atmosphere models at 16 snapshots along the main sequence of
a 150 $M_\odot$ star. For each snapshot, we compute two atmosphere models
connected to the underlying structure model at different depths (below and
above the hot iron bump). We perform a detailed spectroscopic and structural
comparison of the two sequences of model atmospheres, and present a generalized
method for the correction of the effective temperature in evolution models with
strong winds. The choice of connection point between structure and atmosphere
models has a severe influence on the predicted spectral appearance, which
constitutes a previously unexplored source of uncertainty in quantitative
spectroscopy. The simplified atmosphere treatment of current stellar structure
codes likely leads to an overestimation of the spatial extension of very
massive stars, caused by opacity-induced sub-surface inflation. This inflation
does not occur in our deep atmosphere models, resulting in a discrepancy in
predicted effective temperatures of up to 20 kK. Future improvements with
turbulence and dynamically-consistent models may resolve these discrepancies.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Josiek",
"Joris",
""
],
[
"Sander",
"Andreas A. C.",
""
],
[
"Bernini-Peron",
"Matheus",
""
],
[
"Ekström",
"Sylvia",
""
],
[
"González-Torà",
"Gemma",
""
],
[
"Lefever",
"Roel R.",
""
],
[
"Moens",
"Nicolas",
""
],
[
"Ramachandran",
"Varsha",
""
],
[
"Schösser",
"Elisa C.",
""
]
] |
2504.08881 | Simon Weng | Simon Weng, Elaine M. Sadler, Emily Kerrison, Victoria Bollo, C\'eline
P\'eroux, Martin Zwaan, Elizabeth K. Mahony, James R. Allison, Jianhang Chen,
Roland Szakacs, Hyein Yoon | ALMACAL XIV: X-Shooter Spectroscopy, Infrared Properties and Radio SEDs
of Calibrators | Accepted for publication in MNRAS. 11 pages, 9 figures. 70 spectra in
appendix | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The ALMACAL-22 survey includes over 2700 hours of observations of ALMA phase
and amplitude calibrators, spanning frequencies from 84 to 950 GHz across bands
3 to 10. In total, 687 out of the 1,047 calibrators have redshifts confirmed
with spectroscopy and we find an additional 50 featureless blazars. The
redshift distribution of the ALMACAL-22 sample peaks at $z \approx 1$ and spans
a wide range, from the nuclei of nearby galaxies at $z \ll 0.01$ to quasars at
$z = 3.742$. 70 new VLT/X-Shooter spectra of these sources covering UV to NIR
wavelengths are also presented, which will be used in future stacking
experiments to search for cold gas in the circumgalactic medium. Infrared
magnitudes from WISE indicate that the majority of the sources are consistent
with being quasars or blazars. After fitting the radio spectral energy
distributions of the calibrators, we find that most ALMA calibrators exhibit
peaked spectra or are re-triggered which is surprising given the large number
of blazars in the sample. The peak frequencies span three orders of magnitude
from 100 MHz to 170 GHz, corresponding to linear sizes ranging from sub-pc to
$>$ 10 kpc. In the future, when combined with high-resolution radio imaging,
these results will offer valuable constraints on the molecular gas content of
the CGM, as well as the ages and duty cycles of AGN jets. The ever-growing
ALMACAL data set will remain an indispensable resource for studying the various
aspects of galaxy formation and evolution.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Weng",
"Simon",
""
],
[
"Sadler",
"Elaine M.",
""
],
[
"Kerrison",
"Emily",
""
],
[
"Bollo",
"Victoria",
""
],
[
"Péroux",
"Céline",
""
],
[
"Zwaan",
"Martin",
""
],
[
"Mahony",
"Elizabeth K.",
""
],
[
"Allison",
"James R.",
""
],
[
"Chen",
"Jianhang",
""
],
[
"Szakacs",
"Roland",
""
],
[
"Yoon",
"Hyein",
""
]
] |
2504.08882 | Tohid Farajollahpour | T. Farajollahpour, R. Ganesh and K. V. Samokhin | Unveiling Berry curvature contributions to Hall current in $C_4K$
materials | 6 pages + 10 pages supplementary materials | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | We identify a new contribution to the conventional Hall effect that emerges
in materials with $C_4K$ symmetry. This contribution originates from the
modification of phase space density due to the Berry curvature, as we
demonstrate using semiclassical equations of motion for band electrons. As an
illustration, we build a minimal two-band tight-binding model with
altermagnetic order that breaks $C_4$ and $K$ symmetries while preserving
$C_4K$. The resulting Hall conductivity shows a square-root feature at the
altermagnetic phase transition, which is due to the novel Berry
curvature-driven contribution emerging below the critical temperature. This
effect may offer a simple transport-based signature of an altermagnetic phase
transition.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Farajollahpour",
"T.",
""
],
[
"Ganesh",
"R.",
""
],
[
"Samokhin",
"K. V.",
""
]
] |
2504.08883 | Xiaofei Yu | Xiaofei Yu, Evan J. Villafranca, Stella Wang, Jessica C. Jones, Mouzhe
Xie, Jonah Nagura, Ignacio Chi-Dur\'an, Nazar Delegan, Alex B. F. Martinson,
Michael E. Flatt\'e, Denis R. Candido, Giulia Galli, and Peter C. Maurer | Engineering Dark Spin-Free Diamond Interfaces | Main text: 7 pages, 5 figures | null | null | null | quant-ph cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Nitrogen-vacancy (NV) centers in diamond are extensively utilized as quantum
sensors for imaging fields at the nanoscale. The ultra-high sensitivity of NV
magnetometers has enabled the detection and spectroscopy of individual electron
spins, with potentially far-reaching applications in condensed matter physics,
spintronics, and molecular biology. However, the surfaces of these diamond
sensors naturally contain electron spins, which create a background signal that
can be hard to differentiate from the signal of the target spins. In this
study, we develop a surface modification approach that eliminates the unwanted
signal of these so-called dark electron spins. Our surface passivation
technique, based on coating diamond surfaces with a thin titanium oxide (TiO2)
layer, reduces the dark spin density. The observed reduction in dark spin
density aligns with our findings on the electronic structure of the
diamond-TiO2 interface. The reduction, from a typical value of
$2,000$~$\mu$m$^{-2}$ to a value below that set by the detection limit of our
NV sensors ($200$~$\mu$m$^{-2}$), results in a two-fold increase in spin echo
coherence time of near surface NV centers. Furthermore, we derive a
comprehensive spin model that connects dark spin depolarization with NV
coherence, providing additional insights into the mechanisms behind the
observed spin dynamics. Our findings are directly transferable to other quantum
platforms, including nanoscale solid state qubits and superconducting qubits.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yu",
"Xiaofei",
""
],
[
"Villafranca",
"Evan J.",
""
],
[
"Wang",
"Stella",
""
],
[
"Jones",
"Jessica C.",
""
],
[
"Xie",
"Mouzhe",
""
],
[
"Nagura",
"Jonah",
""
],
[
"Chi-Durán",
"Ignacio",
""
],
[
"Delegan",
"Nazar",
""
],
[
"Martinson",
"Alex B. F.",
""
],
[
"Flatté",
"Michael E.",
""
],
[
"Candido",
"Denis R.",
""
],
[
"Galli",
"Giulia",
""
],
[
"Maurer",
"Peter C.",
""
]
] |
2504.08884 | Jes\'us David Carrillo Santamar\'ia Mr | Jes\'us Carrillo-Santamar\'ia, Diego L\'opez-C\'amara, Fabio De Colle,
Enrique Moreno M\'endez and Javier S\'anchez-Salcedo | The role of wind asphericity in dynamical friction | 12 pages, 10 figures, 4 tablas, submitted to MNRAS | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Dynamical friction (DF) may affect the dynamics of stars moving through dense
media. This is the case for stars and compact objects (COs) crossing active
galactic nuclei (AGN) discs, stellar clusters, and common envelopes (CE),
driving stellar migration. DF may decelerate the moving stellar object and may
also, under certain conditions, produce an acceleration. In this paper, we
study the DF and its effects in the interaction between a star and the ambient
gaseous medium through a set of two-dimensional, hydrodynamical numerical
simulations using a wind tunnel configuration. Three different stellar wind
configurations are considered: isotropic, polar, and equatorial. We confirm
that the DF can decelerate and accelerate the star and find the critical value
of the normalized velocity ($u_c$) that marks the transition between these
regimes, for the three wind profiles. The value of $u_c$ for the isotropic wind
differs slightly from that obtained in the thin shell approximation; for an
aspherical wind, it may either be larger or smaller. Aspherical winds with
small $u$ values produce larger accelerations than isotropic winds, while at
high $u$ values, they lead to greater deceleration than the isotropic case. The
timescale for DF to substantially affect the velocity of a stellar object is
calculated. It is shown to be relevant in AGN discs and CEs.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Carrillo-Santamaría",
"Jesús",
""
],
[
"López-Cámara",
"Diego",
""
],
[
"De Colle",
"Fabio",
""
],
[
"Méndez",
"Enrique Moreno",
""
],
[
"Sánchez-Salcedo",
"Javier",
""
]
] |
2504.08885 | Sze Ching Audrey Fung | Audrey Fung, Thalles Lucas, Levente Balogh, Matthew Leybourne, Aaron
C. Vincent | Refining the sensitivity of new physics searches with ancient minerals | null | null | null | null | hep-ph astro-ph.CO physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Paleodetection has been proposed as a competitive method for detecting dark
matter and other new physics interactions, complementing conventional direct
detection experiments. In this work, we utilise TRIM simulations to improve the
modelling of track length distributions. Our findings suggest that previous
studies have overestimated the number of tracks caused by weakly interacting
particles, and that the lowest observable dark matter mass should be higher
than previously predicted. These differences are mainly attributed to the fact
that (a) the recoil energy-track length relation is not one-to-one, (b) at low
recoil energies, a substantial fraction of recoils do not yield any tracks, and
(c) at high energies, electronic stopping becomes dominant, resulting in a
track length barrier at $\sim200$ nm. In addition to WIMPs, we also modelled
tracks from generalised coherent elastic neutrino nucleus scattering
(CE$\nu$NS) via new light mediators and estimated the projected sensitivity for
these interactions.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fung",
"Audrey",
""
],
[
"Lucas",
"Thalles",
""
],
[
"Balogh",
"Levente",
""
],
[
"Leybourne",
"Matthew",
""
],
[
"Vincent",
"Aaron C.",
""
]
] |
2504.08887 | Zijian Liang | Zijian Liang, Jens Niklas Eberhardt, Yu-An Chen | Planar quantum low-density parity-check codes with open boundaries | 32 pages, 21 figures, 10 tables | null | null | null | quant-ph cond-mat.str-el math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct high-performance planar quantum low-density parity-check (qLDPC)
codes with open boundaries, demonstrating substantially improved resource
efficiency compared to the surface code. We present planar code families with
logical dimensions ranging from $k=6$ to $k=13$ (e.g., $[[79, 6, 6]]$, $[[107,
7, 7]]$, $[[173, 8, 9]]$, $[[268, 8, 12]]$, $[[405, 9, 15]]$, $[[374, 10,
13]]$, $[[409, 11, 13]]$, $[[386, 12, 12]]$, $[[362, 13, 11]]$), all using
local stabilizers of weight 6 or lower. These codes achieve an efficiency
metric ($kd^2/n$) that is an order of magnitude greater than that of the
surface code. They can be interpreted as planar bivariate bicycle codes,
adapted from the original design based on a torus that is challenging to
implement physically. Our construction method, which combines boundary anyon
condensation with a novel "lattice grafting" optimization, circumvents this
difficulty and produces codes featuring only local low-weight stabilizers
suitable for 2D planar hardware architectures. Furthermore, we observe fractal
logical operators in the form of Sierpinski triangles, with the code distances
scaling proportionally to the area of the truncated fractal in finite systems.
We anticipate that our codes and construction methods offer a promising pathway
toward realizing near-term fault-tolerant quantum computers.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:05 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liang",
"Zijian",
""
],
[
"Eberhardt",
"Jens Niklas",
""
],
[
"Chen",
"Yu-An",
""
]
] |
2504.08888 | Sun Woo P. Kim | Sun Woo P. Kim, Curt von Keyserlingk, Austen Lamacraft | Measurement-induced phase transitions in quantum inference problems and
quantum hidden Markov models | 24 pages of main text, 23 pages of appendix, 9 figures | null | null | null | cond-mat.stat-mech cond-mat.dis-nn quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Recently, there is interest in coincident 'sharpening' and 'learnability'
transitions in monitored quantum systems. In the latter, an outside observer's
ability to infer properties of a quantum system from measurements undergoes a
phase transition. Such transitions appear to be related to the decodability
transition in quantum error correction, but the precise connection is not
clear. Here, we study these problems under one framework we call the general
quantum inference problem. In cases as above where the system has a Markov
structure, we say that the inference is on a quantum hidden Markov model. We
show a formal connection to classical hidden Markov models and that they
coincide for certain setups. For example, we prove this for those involving
Haar-random unitaries and measurements. We introduce the notion of Bayes
non-optimality, where parameters used for inference differs from true ones.
This allows us to expand the phase diagrams of above models. At Bayes
optimality, we obtain an explicit relation between 'sharpening' and
'learnability' order parameters, explicitly showing that the two transitions
coincide. Next, we study concrete examples. We review quantum error correction
on the toric and repetition code and their mapping to 2D random-bond Ising
model (RBIM) through our framework. We study the Haar-random U(1)-symmetric
monitored quantum circuit and tree, mapping each to inference models that we
call the planted SSEP and planted XOR, respectively, and expanding the phase
diagram to Bayes non-optimality. For the circuit, we deduce the phase boundary
numerically and analytically argue that it is of a single universality class.
For the tree, we present an exact solution of the entire phase boundary, which
displays re-entrance as does the 2D RBIM. We discuss these phase diagrams, with
their interpretations for quantum inference problems and rigorous arguments on
their shapes.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kim",
"Sun Woo P.",
""
],
[
"von Keyserlingk",
"Curt",
""
],
[
"Lamacraft",
"Austen",
""
]
] |
2504.08890 | Fazeel Mahmood Khan Dr. | Fazeel Mahmood Khan, \'Angel Rodr\'iguez, Andrea V. Macci\`o and
Smarika Sharma | AGN Feedback-Induced Stellar Density Expansion in the Inner Regions of
Early-Type Galaxies | Submitted to ApJ | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Observations indicate that early-type galaxies exhibit varying slopes in the
relation between their central stellar surface density and stellar mass
($\Sigma_1 - M_{\star}$). Low-mass galaxies tend to follow a steep slope, close
to one, while the slope flattens for high-mass early type galaxies. In our
study, we investigate the $\Sigma_1 - M_{\star}$ scaling relation and its
evolution using the NIHAO suite of cosmological simulations and compare our
findings with recent results from the MaNGA survey. Our analysis shows that
NIHAO galaxies successfully reproduce the observed scaling relation based on
MANGA survey. Our analysis suggests that AGN feedback plays a critical role in
flattening the $\Sigma_1$ slope by expelling gas from galactic centers, leading
to a decrease in both stellar and dark matter density as the gravitational
potential becomes shallower. To further support our findings, we conducted
high-resolution N-body simulations, which confirmed that
({\it sudden}) gas removal does substantially alter the stellar density in
the central region, consistent with results from NIHAO. Furthermore, our
numerical experiments show that even if the same amount of gas is re-accreted
on a typical ({\it longer}) free-fall time, it is not able to restore the
original stellar density. Our study concludes that AGN feedback assisted gas
removal presents a plausible explanation for the decline in central stellar
surface density as observed in massive elliptical galaxies.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:00:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Khan",
"Fazeel Mahmood",
""
],
[
"Rodríguez",
"Ángel",
""
],
[
"Macciò",
"Andrea V.",
""
],
[
"Sharma",
"Smarika",
""
]
] |
2504.08891 | \'Elie Gouzien | Hugo Jacinto and \'Elie Gouzien and Nicolas Sangouard | Network Requirements for Distributed Quantum Computation | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Physical constraints and engineering challenges, including wafer dimensions,
classical control cabling, and refrigeration volumes, impose significant
limitations on the scalability of quantum computing units. As a result, a
modular quantum computing architecture, comprising small processors
interconnected by quantum links, is emerging as a promising approach to
fault-tolerant quantum computing. However, the requirements that the network
must fulfill to enable distributed quantum computation remain largely
unexplored. We consider an architecture tailored for qubits with
nearest-neighbor physical connectivity, leveraging the surface code for error
correction and enabling fault-tolerant operations through lattice surgery and
magic state distillation. We propose measurement teleportation as a tool to
extend lattice surgery techniques to qubits located on different computing
units interconnected via Bell pairs. Through memory simulations, we build an
error model for logical operations and deduce an end-to-end resource estimation
of Shor's algorithm over a minimalist distributed architecture. Concretely, for
a characteristic physical gate error rate of 1e-3, a processor cycle time of 1
microsecond, factoring a 2048-bit RSA integer is shown to be possible with 379
computing processors, each made with 89781 qubits, with negligible space and
time overhead with respect to a monolithic approach without parallelization, if
70 Bell pairs are available per cycle time between each processor with a
fidelity exceeding 98.4 percent.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:01:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jacinto",
"Hugo",
""
],
[
"Gouzien",
"Élie",
""
],
[
"Sangouard",
"Nicolas",
""
]
] |
2504.08892 | Ashley Chrimes | A. A. Chrimes, A. J. Levan, J. D. Lyman, A. Borghese, V. S. Dhillon,
P. Esposito, M. Fraser, A. S. Fruchter, D. Gotz, R. A. Hounsell, G. L.
Israel, C. Kouveliotou, S. Mereghetti, R. P. Mignani, R. Perna, N. Rea, I.
Skillen, D. Steeghs, N. R. Tanvir, K. Wiersema, N. J. Wright, S. Zane | The infrared counterpart and proper motion of magnetar SGR0501+4516 | Accepted for publication in A&A. 8 figures, 1 table | null | 10.1051/0004-6361/202453479 | null | astro-ph.HE astro-ph.GA astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Soft gamma repeaters (SGRs) are highly magnetised neutron stars (magnetars)
notable for their gamma-ray and X-ray outbursts. In this paper, we use
near-infrared (NIR) imaging of SGR 0501+4516 in the days, weeks, and years
after its 2008 outburst to characterise the multi-wavelength emission, and to
obtain a proper motion from our long temporal baseline observations. Unlike
most magnetars, the source has only moderate foreground extinction with minimal
crowding. Our observations began only 2 hours after the first activation of SGR
0501+4516 in August 2008, and continued for 4 years, including two epochs of
Hubble Space Telescope (HST) imaging. The proper motion constraint is improved
by a third HST epoch 10 years later. The near-infrared and X-rays faded slowly
during the first week, thereafter following a steeper power-law decay. The
behaviour is satisfactorily fit by a broken power-law. Three epochs of HST
imaging with a 10-year baseline allow us to determine a quiescent level, and to
measure a proper motion of 5.4+/-0.6 mas/yr. This corresponds to a low
transverse peculiar velocity of 51+/-14 km/s (at 2 kpc). The magnitude and
direction of the proper motion rules out supernova remnant HB9 as the
birth-site. We can find no other supernova remnants or groups of massive stars
within the region traversed by SGR 0501+4516 during its characteristic lifetime
(20 kyr). Our observations of SGR 0501+4516 suggest that some magnetars may be
either significantly older than expected, that their progenitors produce low
supernova ejecta masses, or alternatively that they can be formed through
accretion-induced collapse (AIC) or low-mass neutron star mergers. Although the
progenitor of SGR 0501+4516 remains unclear, we propose that SGR 0501+4516 is
the best Galactic candidate for a magnetar formed through a mechanism other
than massive star core-collapse.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:01:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chrimes",
"A. A.",
""
],
[
"Levan",
"A. J.",
""
],
[
"Lyman",
"J. D.",
""
],
[
"Borghese",
"A.",
""
],
[
"Dhillon",
"V. S.",
""
],
[
"Esposito",
"P.",
""
],
[
"Fraser",
"M.",
""
],
[
"Fruchter",
"A. S.",
""
],
[
"Gotz",
"D.",
""
],
[
"Hounsell",
"R. A.",
""
],
[
"Israel",
"G. L.",
""
],
[
"Kouveliotou",
"C.",
""
],
[
"Mereghetti",
"S.",
""
],
[
"Mignani",
"R. P.",
""
],
[
"Perna",
"R.",
""
],
[
"Rea",
"N.",
""
],
[
"Skillen",
"I.",
""
],
[
"Steeghs",
"D.",
""
],
[
"Tanvir",
"N. R.",
""
],
[
"Wiersema",
"K.",
""
],
[
"Wright",
"N. J.",
""
],
[
"Zane",
"S.",
""
]
] |
2504.08893 | Jakub Tomczak | Jasper Linders, Jakub M. Tomczak | Knowledge Graph-extended Retrieval Augmented Generation for Question
Answering | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) and Knowledge Graphs (KGs) offer a promising
approach to robust and explainable Question Answering (QA). While LLMs excel at
natural language understanding, they suffer from knowledge gaps and
hallucinations. KGs provide structured knowledge but lack natural language
interaction. Ideally, an AI system should be both robust to missing facts as
well as easy to communicate with. This paper proposes such a system that
integrates LLMs and KGs without requiring training, ensuring adaptability
across different KGs with minimal human effort. The resulting approach can be
classified as a specific form of a Retrieval Augmented Generation (RAG) with a
KG, thus, it is dubbed Knowledge Graph-extended Retrieval Augmented Generation
(KG-RAG). It includes a question decomposition module to enhance multi-hop
information retrieval and answer explainability. Using In-Context Learning
(ICL) and Chain-of-Thought (CoT) prompting, it generates explicit reasoning
chains processed separately to improve truthfulness. Experiments on the MetaQA
benchmark show increased accuracy for multi-hop questions, though with a slight
trade-off in single-hop performance compared to LLM with KG baselines. These
findings demonstrate KG-RAG's potential to improve transparency in QA by
bridging unstructured language understanding with structured knowledge
retrieval.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:03:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Linders",
"Jasper",
""
],
[
"Tomczak",
"Jakub M.",
""
]
] |
2504.08894 | Gayathri Gururajan | G. Gururajan, O. Cucciati, B. C. Lemaux, M. Talia, G. Zamorani, F.
Pozzi, R. Decarli, B. Forrest, L. Shen, G. De Lucia, F. Fontanot, S.
Bardelli, D. C. Baxter, L. P. Cassar\`a, E. Golden-Marx, D. Sikorski, E. A.
Shah, R. R. Gal, M. Giavalisco, F. Giddings, N. P. Hathi, D. Hung, A. M.
Koekemoer, V. Le Brun, L. M. Lubin, L. A. M. Tasca, L. Tresse, D. Vergani,
and E. Zucca | Gas properties as a function of environment in the proto-supercluster
Hyperion at z ~ 2.45 | Accepted for publication in A&A. 15 pages, 6 figures, 2 tables | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The cosmic star-formation rate density, molecular gas density and the AGN
activity of the Universe peak at z~ 2-3, showing the Universe is most active at
this epoch. The nature of the galaxies at these redshifts and their properties
as a function of their environment are particularly interesting to understand
the mechanisms driving their star-formation and quenching. At z~ 2.5, a massive
(~ 4.8 X 10^15 Msun) proto-supercluster, Hyperion, was identified Cucciati et
al. 2018, consisting of 7 groups/peaks and extending over a comoving volume of
60 X 60 X 150 Mpc^3, providing an excellent laboratory to probe the properties
and evolution of galaxies as a function of their environments. We use a large
compilation of photometric (optical to radio wavelengths, COSMOS2020,
COSMOS-Super-deblended, and, A3COSMOS) and spectroscopic (C3VO, HST-Hyperion,
VUDS, zCOSMOS, DEIMOS10K, MAGAZ3NE) data to assign membership and study the
relation between the local environment and the molecular gas mass, the
star-formation rate (SFR), gas depletion timescales, and quenching mechanisms.
We find that the depletion timescales and the molecular gas fractions decrease
and SFR increases in denser environments at the ~ 2 sigma level, suggesting
accelerated evolution in the densest regions of this proto-supercluster
resulting from gas stripping, over-consumption, and/or cessation of cold flows.
Dedicated observations at sub-millimeter wavelengths enabling further
spectroscopic confirmation and better coverage in the sub-millimetric (sub-mm)
wavelengths can provide more conclusive results on the environmental
implications on gas reservoirs of galaxies in Hyperion.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:04:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gururajan",
"G.",
""
],
[
"Cucciati",
"O.",
""
],
[
"Lemaux",
"B. C.",
""
],
[
"Talia",
"M.",
""
],
[
"Zamorani",
"G.",
""
],
[
"Pozzi",
"F.",
""
],
[
"Decarli",
"R.",
""
],
[
"Forrest",
"B.",
""
],
[
"Shen",
"L.",
""
],
[
"De Lucia",
"G.",
""
],
[
"Fontanot",
"F.",
""
],
[
"Bardelli",
"S.",
""
],
[
"Baxter",
"D. C.",
""
],
[
"Cassarà",
"L. P.",
""
],
[
"Golden-Marx",
"E.",
""
],
[
"Sikorski",
"D.",
""
],
[
"Shah",
"E. A.",
""
],
[
"Gal",
"R. R.",
""
],
[
"Giavalisco",
"M.",
""
],
[
"Giddings",
"F.",
""
],
[
"Hathi",
"N. P.",
""
],
[
"Hung",
"D.",
""
],
[
"Koekemoer",
"A. M.",
""
],
[
"Brun",
"V. Le",
""
],
[
"Lubin",
"L. M.",
""
],
[
"Tasca",
"L. A. M.",
""
],
[
"Tresse",
"L.",
""
],
[
"Vergani",
"D.",
""
],
[
"Zucca",
"E.",
""
]
] |
2504.08895 | Hao Chen | Hao Chen, Dan Mao, Andrea Kouta Dagnino, Glenn Wagner, Mark H.
Fischer, Juraj Hasik, Eun-Ah Kim, Titus Neupert | How quantum fluctuations freeze a classical liquid and then melt it into
a topological one | 14 pages, 10 figures | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Topologically ordered quantum liquids are highly sought-after quantum phases
of matter and recently, fractional Chern insulators (FCIs) joined the few
experimental realizations of such phases. Here, we ask whether a gapped
classical, highly degenerate liquid can be the birthplace of FCIs upon the
addition of suitable quantum fluctuations. Two competing tendencies can be
anticipated: (i) following the quantum order-by-disorder paradigm, quantum
fluctuations could induce symmetry-breaking (charge) order, or (ii) the
classical liquid builds up long-range entanglement and turns into a quantum
liquid. We study spinless fermions on a honeycomb lattice subject to
cluster-charging interactions and introduce quantumness through a Haldane
kinetic term. Based on extensive exact diagonalization calculations and
high-order perturbation theory, we find that neither scenario (i) or (ii)
prevails, but (i) and (ii) manifest sequentially as the kinetic energy is
increased. We demonstrate how the gradual lifting of kinematic constraints
gives rise to this sequence of phases. Our results relate to the regime of
intermediate-scale interactions present in moir\'e systems, where band
projections are not suitable to model FCIs and competing charge-ordered phases
have been identified.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:05:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Hao",
""
],
[
"Mao",
"Dan",
""
],
[
"Dagnino",
"Andrea Kouta",
""
],
[
"Wagner",
"Glenn",
""
],
[
"Fischer",
"Mark H.",
""
],
[
"Hasik",
"Juraj",
""
],
[
"Kim",
"Eun-Ah",
""
],
[
"Neupert",
"Titus",
""
]
] |
2504.08896 | Neil He | Neil He, Jiahong Liu, Buze Zhang, Ngoc Bui, Ali Maatouk, Menglin Yang,
Irwin King, Melanie Weber, Rex Ying | Position: Beyond Euclidean -- Foundation Models Should Embrace
Non-Euclidean Geometries | 22 pages, 4 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In the era of foundation models and Large Language Models (LLMs), Euclidean
space has been the de facto geometric setting for machine learning
architectures. However, recent literature has demonstrated that this choice
comes with fundamental limitations. At a large scale, real-world data often
exhibit inherently non-Euclidean structures, such as multi-way relationships,
hierarchies, symmetries, and non-isotropic scaling, in a variety of domains,
such as languages, vision, and the natural sciences. It is challenging to
effectively capture these structures within the constraints of Euclidean
spaces. This position paper argues that moving beyond Euclidean geometry is not
merely an optional enhancement but a necessity to maintain the scaling law for
the next-generation of foundation models. By adopting these geometries,
foundation models could more efficiently leverage the aforementioned
structures. Task-aware adaptability that dynamically reconfigures embeddings to
match the geometry of downstream applications could further enhance efficiency
and expressivity. Our position is supported by a series of theoretical and
empirical investigations of prevalent foundation models.Finally, we outline a
roadmap for integrating non-Euclidean geometries into foundation models,
including strategies for building geometric foundation models via fine-tuning,
training from scratch, and hybrid approaches.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:07:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Neil",
""
],
[
"Liu",
"Jiahong",
""
],
[
"Zhang",
"Buze",
""
],
[
"Bui",
"Ngoc",
""
],
[
"Maatouk",
"Ali",
""
],
[
"Yang",
"Menglin",
""
],
[
"King",
"Irwin",
""
],
[
"Weber",
"Melanie",
""
],
[
"Ying",
"Rex",
""
]
] |
2504.08897 | Abhronil Sengupta | Jiaqi Lin, Abhronil Sengupta | Toward Spiking Neural Network Local Learning Modules Resistant to
Adversarial Attacks | null | null | null | null | cs.LG cs.CR cs.NE | http://creativecommons.org/licenses/by/4.0/ | Recent research has shown the vulnerability of Spiking Neural Networks (SNNs)
under adversarial examples that are nearly indistinguishable from clean data in
the context of frame-based and event-based information. The majority of these
studies are constrained in generating adversarial examples using
Backpropagation Through Time (BPTT), a gradient-based method which lacks
biological plausibility. In contrast, local learning methods, which relax many
of BPTT's constraints, remain under-explored in the context of adversarial
attacks. To address this problem, we examine adversarial robustness in SNNs
through the framework of four types of training algorithms. We provide an
in-depth analysis of the ineffectiveness of gradient-based adversarial attacks
to generate adversarial instances in this scenario. To overcome these
limitations, we introduce a hybrid adversarial attack paradigm that leverages
the transferability of adversarial instances. The proposed hybrid approach
demonstrates superior performance, outperforming existing adversarial attack
methods. Furthermore, the generalizability of the method is assessed under
multi-step adversarial attacks, adversarial attacks in black-box FGSM
scenarios, and within the non-spiking domain.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:07:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lin",
"Jiaqi",
""
],
[
"Sengupta",
"Abhronil",
""
]
] |
2504.08898 | Xueyue Zhang | Xueyue Zhang, Niccolo Fiaschi, Lukasz Komza, Hanbin Song, Thomas
Schenkel, Alp Sipahigil | Laser-induced spectral diffusion of T centers in silicon nanophotonic
devices | null | null | null | null | quant-ph cond-mat.mtrl-sci physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Color centers in silicon are emerging as spin-photon interfaces operating at
telecommunication wavelengths. The nanophotonic device integration of silicon
color centers via ion implantation leads to significant optical linewidth
broadening, which makes indistinguishable photon generation challenging. Here,
we study the optical spectral diffusion of T centers in a silicon photonic
crystal cavity. We investigate the linewidth broadening timescales and origins
by measuring the temporal correlations of the resonance frequency under
different conditions. Spectral hole burning measurements reveal no spectral
broadening at short timescales from 102 ns to 725 ns. We probe broadening at
longer timescales using a check pulse to herald the T center frequency and a
probe pulse to measure frequency after a wait time. The optical resonance
frequency is stable up to 3 ms in the dark. Laser pulses below the silicon band
gap applied during the wait time leads to linewidth broadening. Our
observations establish laser-induced processes as the dominant spectral
diffusion mechanism for T centers in devices, and inform materials and feedback
strategies for indistinguishable photon generation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:08:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Xueyue",
""
],
[
"Fiaschi",
"Niccolo",
""
],
[
"Komza",
"Lukasz",
""
],
[
"Song",
"Hanbin",
""
],
[
"Schenkel",
"Thomas",
""
],
[
"Sipahigil",
"Alp",
""
]
] |
2504.08899 | Bozo Vareskic | Bozo Vareskic, Finn G. Kennedy, Takashi Taniguchi, Kenji Watanabe,
Kenji Yasuda, Daniel C. Ralph | Gate-tunable electroresistance in a sliding ferroelectric tunnel
junction | null | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | We fabricate and measure electrically-gated tunnel junctions in which the
insulating barrier is a sliding van der Waals ferroelectric made from
parallel-stacked bilayer hexagonal boron nitride and the electrodes are
single-layer graphene. Despite the nominally-symmetric tunnel-junction
structure, these devices can exhibit substantial electroresistance upon
reversing the ferroelectric polarization. The magnitude and sign of tunneling
electroresistance are tunable by bias and gate voltage. We show that this
behavior can be understood within a simple tunneling model that takes into
account the quantum capacitance of the graphene electrodes, so that the
tunneling densities of states in the electrodes are separately modified as a
function of bias and gate voltage.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:09:21 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Vareskic",
"Bozo",
""
],
[
"Kennedy",
"Finn G.",
""
],
[
"Taniguchi",
"Takashi",
""
],
[
"Watanabe",
"Kenji",
""
],
[
"Yasuda",
"Kenji",
""
],
[
"Ralph",
"Daniel C.",
""
]
] |
2504.08900 | Yang Yang | Zhe Liu, Zehua Chen, and Yang Yang | Nonadiabatic Dynamics with Constrained Nuclear-Electronic Orbital Theory | null | null | null | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Incorporating nuclear quantum effects into nonadiabatic dynamics remains a
significant challenge. Herein we introduce new nonadiabatic dynamics approaches
based on the recently developed constrained nuclear-electronic orbital (CNEO)
theory. The CNEO approach integrates nuclear quantum effects, particularly
quantum nuclear delocalization effects, into effective potential energy
surfaces. When combined with Ehrenfest dynamics and surface hopping, it
efficiently captures both nonadiabaticity and quantum nuclear delocalization
effects. We apply these new approaches to a one-dimensional proton-coupled
electron transfer model and find that they outperform conventional Ehrenfest
dynamics and surface hopping in predicting proton transfer dynamics and proton
transmission probabilities, especially in the low momentum regime.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:09:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Zhe",
""
],
[
"Chen",
"Zehua",
""
],
[
"Yang",
"Yang",
""
]
] |
2504.08901 | Panos Papadopoulos | Asterios Reppas, Grigorios-Aris Cheimariotis, Panos K. Papadopoulos,
Panagiotis Frasiolas, Dimitrios Zarpalas | HAL-NeRF: High Accuracy Localization Leveraging Neural Radiance Fields | 8 pages, 4 figures | null | 10.1109/AIxVR63409.2025.00024 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Precise camera localization is a critical task in XR applications and
robotics. Using only the camera captures as input to a system is an inexpensive
option that enables localization in large indoor and outdoor environments, but
it presents challenges in achieving high accuracy. Specifically, camera
relocalization methods, such as Absolute Pose Regression (APR), can localize
cameras with a median translation error of more than $0.5m$ in outdoor scenes.
This paper presents HAL-NeRF, a high-accuracy localization method that combines
a CNN pose regressor with a refinement module based on a Monte Carlo particle
filter. The Nerfacto model, an implementation of Neural Radiance Fields
(NeRFs), is used to augment the data for training the pose regressor and to
measure photometric loss in the particle filter refinement module. HAL-NeRF
leverages Nerfacto's ability to synthesize high-quality novel views,
significantly improving the performance of the localization pipeline. HAL-NeRF
achieves state-of-the-art results that are conventionally measured as the
average of the median per scene errors. The translation error was $0.025m$ and
the rotation error was $0.59$ degrees and 0.04m and 0.58 degrees on the
7-Scenes dataset and Cambridge Landmarks datasets respectively, with the
trade-off of increased computational time. This work highlights the potential
of combining APR with NeRF-based refinement techniques to advance monocular
camera relocalization accuracy.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:10:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Reppas",
"Asterios",
""
],
[
"Cheimariotis",
"Grigorios-Aris",
""
],
[
"Papadopoulos",
"Panos K.",
""
],
[
"Frasiolas",
"Panagiotis",
""
],
[
"Zarpalas",
"Dimitrios",
""
]
] |
2504.08902 | Pascal Chang | Pascal Chang, Sergio Sancho, Jingwei Tang, Markus Gross, Vinicius C.
Azevedo | LookingGlass: Generative Anamorphoses via Laplacian Pyramid Warping | Accepted at CVPR 2025 (Oral) | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anamorphosis refers to a category of images that are intentionally distorted,
making them unrecognizable when viewed directly. Their true form only reveals
itself when seen from a specific viewpoint, which can be through some
catadioptric device like a mirror or a lens. While the construction of these
mathematical devices can be traced back to as early as the 17th century, they
are only interpretable when viewed from a specific vantage point and tend to
lose meaning when seen normally. In this paper, we revisit these famous optical
illusions with a generative twist. With the help of latent rectified flow
models, we propose a method to create anamorphic images that still retain a
valid interpretation when viewed directly. To this end, we introduce Laplacian
Pyramid Warping, a frequency-aware image warping technique key to generating
high-quality visuals. Our work extends Visual Anagrams (arXiv:2311.17919) to
latent space models and to a wider range of spatial transforms, enabling the
creation of novel generative perceptual illusions.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:12:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chang",
"Pascal",
""
],
[
"Sancho",
"Sergio",
""
],
[
"Tang",
"Jingwei",
""
],
[
"Gross",
"Markus",
""
],
[
"Azevedo",
"Vinicius C.",
""
]
] |
2504.08903 | Rongrong Liu | Rongrong Liu, Le-Chris Wang, Zafar Rustamkulov, David K. Sing | Unveiling the atmosphere of the super-Jupiter HAT-P-14 b with JWST
NIRISS and NIRSpec | null | null | null | null | astro-ph.EP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report the combined JWST NIRSpec/G395H and NIRISS/SOSS transmission
spectrum of the transiting super-Jupiter HAT-P-14 b, from 0.60 $\mu m$ to 5.14
$\mu m$. Initial analysis of these data reported a near-featureless spectrum at
NIRSpec wavelengths range (2.87 $\mu m$ to 5.14 $\mu m$) consistent with the
small atmospheric scale height of the planet and unexplained bumps and wiggles
at NIRISS wavelengths range (0.6 $\mu m$ to 2.8 $\mu m$). Here, we produce a
self-consistent spectrum of HAT-P-14 b's atmosphere with an up-to-date
reduction. We detect H$_2$O (3.09 $\sigma$) both across NIRISS/SOSS wavelengths
range and at the bluest end of NIRSpc/G395H as well as a gray cloud deck (1.90
$\sigma$). We constrain the atmospheric metallicity of HAT-P-14 b to be roughly
Solar, with [Fe/H] $= -0.08^{+0.89}_{-0.98}$, consistent with the planet
mass-metallicity relationship. The differences compared to previous works are
likely due to the improved STScI jwst pipeline, which highlights the need to
reanalyze the early NIRISS/SOSS transiting exoplanet targets with the latest
methods. As HAT-P-14 b is placed as the 805th best target for transmission
spectroscopy according to Transmission Spectroscopy Metrics (TSM), our results
showcase JWST's unparalleled photometric precision which can easily
characterize a thousand exoplanets' atmospheres through transmission
spectroscopy.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:12:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Rongrong",
""
],
[
"Wang",
"Le-Chris",
""
],
[
"Rustamkulov",
"Zafar",
""
],
[
"Sing",
"David K.",
""
]
] |
2504.08904 | Winter Allen | Winter Allen and N. Sanjay Rebello | Exploring the Feasibility of Employing a Hybrid Machine Learning Method
to Unpack Student Reasoning Patterns in Physics Essays | null | null | null | null | physics.ed-ph | http://creativecommons.org/licenses/by/4.0/ | We propose a novel clustering pipeline that combines two classic clustering
algorithms to better understand student problem-solving strategies. This
unsupervised machine learning method helps uncover patterns in reasoning
without pre-defined labels. We applied it to essays written for an online
multiple-choice quiz, the resulting clusters showed strong statistical
alignment with students' selected answers. We also report on the resulting
clusters of the hybrid pipeline compared to that of K-Means (MacQueen, 1967)
and Hierarchal Density-Based Spatial Clustering of Application with Noise
(HDBSCAN) (McInnes, Healy, and Astels, 2017) by analyzing the Scatter Plots,
Silhouette Scores (Rousseeuw, 1987), and Davies Bouldin Index (Davies and
Bouldin, 1979).
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:15:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Allen",
"Winter",
""
],
[
"Rebello",
"N. Sanjay",
""
]
] |
2504.08905 | Yunfan Zhang | Yunfan Zhang, Kathleen McKeown, Smaranda Muresan | Forecasting Communication Derailments Through Conversation Generation | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Forecasting communication derailment can be useful in real-world settings
such as online content moderation, conflict resolution, and business
negotiations. However, despite language models' success at identifying
offensive speech present in conversations, they struggle to forecast future
communication derailments. In contrast to prior work that predicts conversation
outcomes solely based on the past conversation history, our approach samples
multiple future conversation trajectories conditioned on existing conversation
history using a fine-tuned LLM. It predicts the communication outcome based on
the consensus of these trajectories. We also experimented with leveraging
socio-linguistic attributes, which reflect turn-level conversation dynamics, as
guidance when generating future conversations. Our method of future
conversation trajectories surpasses state-of-the-art results on English
communication derailment prediction benchmarks and demonstrates significant
accuracy gains in ablation studies.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:15:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Yunfan",
""
],
[
"McKeown",
"Kathleen",
""
],
[
"Muresan",
"Smaranda",
""
]
] |
2504.08906 | Jiahuan Long | Jiahuan Long, Zhengqin Xu, Tingsong Jiang, Wen Yao, Shuai Jia, Chao
Ma, Xiaoqian Chen | Robust SAM: On the Adversarial Robustness of Vision Foundation Models | Accepted by AAAI2025 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Segment Anything Model (SAM) is a widely used vision foundation model
with diverse applications, including image segmentation, detection, and
tracking. Given SAM's wide applications, understanding its robustness against
adversarial attacks is crucial for real-world deployment. However, research on
SAM's robustness is still in its early stages. Existing attacks often overlook
the role of prompts in evaluating SAM's robustness, and there has been
insufficient exploration of defense methods to balance the robustness and
accuracy. To address these gaps, this paper proposes an adversarial robustness
framework designed to evaluate and enhance the robustness of SAM. Specifically,
we introduce a cross-prompt attack method to enhance the attack transferability
across different prompt types. Besides attacking, we propose a few-parameter
adaptation strategy to defend SAM against various adversarial attacks. To
balance robustness and accuracy, we use the singular value decomposition (SVD)
to constrain the space of trainable parameters, where only singular values are
adaptable. Experiments demonstrate that our cross-prompt attack method
outperforms previous approaches in terms of attack success rate on both SAM and
SAM 2. By adapting only 512 parameters, we achieve at least a 15\% improvement
in mean intersection over union (mIoU) against various adversarial attacks.
Compared to previous defense methods, our approach enhances the robustness of
SAM while maximally maintaining its original performance.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:17:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Long",
"Jiahuan",
""
],
[
"Xu",
"Zhengqin",
""
],
[
"Jiang",
"Tingsong",
""
],
[
"Yao",
"Wen",
""
],
[
"Jia",
"Shuai",
""
],
[
"Ma",
"Chao",
""
],
[
"Chen",
"Xiaoqian",
""
]
] |
2504.08907 | Yang Bai | Ayushi Mishra, Yang Bai, Priyadarshan Narayanasamy, Nakul Garg,
Nirupam Roy | Spatial Audio Processing with Large Language Model on Wearable Devices | null | null | null | null | cs.SD cs.CL eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Integrating spatial context into large language models (LLMs) has the
potential to revolutionize human-computer interaction, particularly in wearable
devices. In this work, we present a novel system architecture that incorporates
spatial speech understanding into LLMs, enabling contextually aware and
adaptive applications for wearable technologies. Our approach leverages
microstructure-based spatial sensing to extract precise Direction of Arrival
(DoA) information using a monaural microphone. To address the lack of existing
dataset for microstructure-assisted speech recordings, we synthetically create
a dataset called OmniTalk by using the LibriSpeech dataset. This spatial
information is fused with linguistic embeddings from OpenAI's Whisper model,
allowing each modality to learn complementary contextual representations. The
fused embeddings are aligned with the input space of LLaMA-3.2 3B model and
fine-tuned with lightweight adaptation technique LoRA to optimize for on-device
processing. SING supports spatially-aware automatic speech recognition (ASR),
achieving a mean error of $25.72^\circ$-a substantial improvement compared to
the 88.52$^\circ$ median error in existing work-with a word error rate (WER) of
5.3. SING also supports soundscaping, for example, inference how many people
were talking and their directions, with up to 5 people and a median DoA error
of 16$^\circ$. Our system demonstrates superior performance in spatial speech
understanding while addressing the challenges of power efficiency, privacy, and
hardware constraints, paving the way for advanced applications in augmented
reality, accessibility, and immersive experiences.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:19:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mishra",
"Ayushi",
""
],
[
"Bai",
"Yang",
""
],
[
"Narayanasamy",
"Priyadarshan",
""
],
[
"Garg",
"Nakul",
""
],
[
"Roy",
"Nirupam",
""
]
] |
2504.08908 | Ahmad Talafha | Ahmad Talafha | Finite Mixture Cox Model for Heterogeneous Time-dependent Right-Censored
Data | null | null | null | null | stat.ME stat.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, we address the challenge of survival analysis within
heterogeneous patient populations, where traditional reliance on a single
regression model such as the Cox proportional hazards (Cox PH) model often
falls short. Recognizing that such populations frequently exhibit varying
covariate effects, resulting in distinct subgroups, we argue for the necessity
of using separate regression models for each subgroup to avoid the biases and
inaccuracies inherent in a uniform model. To address subgroup identification
and component selection in survival analysis, we propose a novel approach that
integrates the Cox PH model with dynamic penalty functions, specifically the
smoothly clipped absolute deviation (SCAD) and the minimax concave penalty
(MCP). These modifications provide a more flexible and theoretically sound
method for determining the optimal number of mixture components, which is
crucial for accurately modeling heterogeneous datasets. Through a modified
expectation--maximization (EM) algorithm for parameter estimation and component
selection, supported by simulation studies and two real data analyses, our
method demonstrates improved precision in risk prediction.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:21:13 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Talafha",
"Ahmad",
""
]
] |
2504.08909 | Islam Mansour | Islam Mansour, Georg Fischer, Ronny Haensch, Irena Hajnsek | Hybrid AI-Physical Modeling for Penetration Bias Correction in X-band
InSAR DEMs: A Greenland Case Study | 8 pages | null | null | null | cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Digital elevation models derived from Interferometric Synthetic Aperture
Radar (InSAR) data over glacial and snow-covered regions often exhibit
systematic elevation errors, commonly termed "penetration bias." We leverage
existing physics-based models and propose an integrated correction framework
that combines parametric physical modeling with machine learning. We evaluate
the approach across three distinct training scenarios - each defined by a
different set of acquisition parameters - to assess overall performance and the
model's ability to generalize. Our experiments on Greenland's ice sheet using
TanDEM-X data show that the proposed hybrid model corrections significantly
reduce the mean and standard deviation of DEM errors compared to a purely
physical modeling baseline. The hybrid framework also achieves significantly
improved generalization than a pure ML approach when trained on data with
limited diversity in acquisition parameters.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:24:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mansour",
"Islam",
""
],
[
"Fischer",
"Georg",
""
],
[
"Haensch",
"Ronny",
""
],
[
"Hajnsek",
"Irena",
""
]
] |
2504.08910 | Winter Allen | Winter Allen and Carina M. Rebello and N. Sanjay Rebello | Assessing Physics Students' Scientific Argumentation using Natural
Language Processing | null | null | null | null | physics.ed-ph | http://creativecommons.org/licenses/by/4.0/ | Scientific argumentation is an important science and engineering practice and
a necessary 21st Century workforce skill. Due to the nature of large enrollment
classes, it is difficult to individually assess students and provide feedback
on their argumentation. The recent developments in Natural Language Processing
(NLP) and Machine Learning (ML) may provide a solution. In this study we
investigate methods using NLP and ML to assess and understand students'
argumentation. Specifically, we investigate the use of topic modeling to
analyze student essays of argumentation after solving a problem in the
recitation section of an introductory calculus-based physics course four
semesters. We report on the emergent themes present in each semester.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:25:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Allen",
"Winter",
""
],
[
"Rebello",
"Carina M.",
""
],
[
"Rebello",
"N. Sanjay",
""
]
] |
2504.08911 | Yuhuai Zhou | Felix R\"ohrich, Yuhuai Zhou | Low-Rank Tensor Recovery via Theta Nuclear p-Norm | null | null | null | null | math.OC math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the low-rank tensor recovery problem using a relaxation of the
nuclear p-norm by theta bodies.
We provide algebraic descriptions of the norms and compute their Gr\"obner
bases.
Moreover, we develop geometric properties of these bodies.
Finally, our numerical results suggest that for
$n\times\cdots\times n$ tensors,
$m\geq O(n)$ measurements should be sufficient to recover low-rank tensors
via theta body relaxation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:26:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Röhrich",
"Felix",
""
],
[
"Zhou",
"Yuhuai",
""
]
] |
2504.08912 | Neil He | Neil He, Menglin Yang, Rex Ying | HyperCore: The Core Framework for Building Hyperbolic Foundation Models
with Comprehensive Modules | 11 pages, 4 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Hyperbolic neural networks have emerged as a powerful tool for modeling
hierarchical data across diverse modalities. Recent studies show that token
distributions in foundation models exhibit scale-free properties, suggesting
that hyperbolic space is a more suitable ambient space than Euclidean space for
many pre-training and downstream tasks. However, existing tools lack essential
components for building hyperbolic foundation models, making it difficult to
leverage recent advancements. We introduce HyperCore, a comprehensive
open-source framework that provides core modules for constructing hyperbolic
foundation models across multiple modalities. HyperCore's modules can be
effortlessly combined to develop novel hyperbolic foundation models,
eliminating the need to extensively modify Euclidean modules from scratch and
possible redundant research efforts. To demonstrate its versatility, we build
and test the first fully hyperbolic vision transformers (LViT) with a
fine-tuning pipeline, the first fully hyperbolic multimodal CLIP model
(L-CLIP), and a hybrid Graph RAG with a hyperbolic graph encoder. Our
experiments demonstrate that LViT outperforms its Euclidean counterpart.
Additionally, we benchmark and reproduce experiments across hyperbolic GNNs,
CNNs, Transformers, and vision Transformers to highlight HyperCore's
advantages.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:35:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Neil",
""
],
[
"Yang",
"Menglin",
""
],
[
"Ying",
"Rex",
""
]
] |
2504.08913 | Zachary Riedel | Z. W. Riedel, W. Simeth, C. S. Kengle, S. M. Thomas, J. D. Thompson,
A. O. Scheie, F. Ronning, C. Lane, Jian-Xin Zhu, P. F. S. Rosa, E. D. Bauer | Magnetic order and physical properties of the Kagome metal UNb$_6$Sn$_6$ | 28 pages, 23 figures | null | null | null | cond-mat.str-el cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The $RM_6X_6$ family of materials ($R$ = rare-earth, $M$ = transition metal,
$X$ = Ga, Si, Ge, Sn) produces an array of emergent phenomena, such as charge
density waves, intrinsic Hall effects, and complex magnetic order, due to its
Kagome net of transition metal atoms, its local-moment magnetic anisotropies,
and its extensive chemical tunability. Here, we report a new ``166" material
containing both an actinide (uranium) and a 4$d$ transition metal (niobium) to
investigate the properties of a 5$f$-4$d$ electron 166 system. UNb$_6$Sn$_6$
crystallizes in the hexagonal $P$6/$mmm$ space group with a small degree of
disorder due to shifts in the size of the CoSn-like cages along the $c$ axis.
Upon cooling at zero magnetic field, the material undergoes two magnetic phase
transitions at $T_\mathrm{2}$ = 46 K and $T_\mathrm{N}$ = 43 K. The
low-temperature, zero-field phase is an antiferromagnet with ordered uranium
moments and a $\textbf{k}$=(0,0,1/2) propagation vector determined by neutron
diffraction. Remarkably, with a magnetic field applied along the $c$ axis, five
additional magnetic transitions occur, evidenced by magnetization and
resistivity data, before the moment saturates at 2.62 ${\mu}_{\mathrm{B}}$/U at
2 K and $\ge$13.6 T. In two magnetic phase regions, the Hall resistivity of
UNb$_6$Sn$_6$ significantly deviates from the magnetization, suggesting that
the phases have a large Berry curvature or a change in the Fermi surface. The
unknown magnetic ordering of the field-dependent phases of UNb$_6$Sn$_6$
demonstrates the complexity of the 5$f$-4$d$ 166 system and encourages further
study of its properties.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:37:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Riedel",
"Z. W.",
""
],
[
"Simeth",
"W.",
""
],
[
"Kengle",
"C. S.",
""
],
[
"Thomas",
"S. M.",
""
],
[
"Thompson",
"J. D.",
""
],
[
"Scheie",
"A. O.",
""
],
[
"Ronning",
"F.",
""
],
[
"Lane",
"C.",
""
],
[
"Zhu",
"Jian-Xin",
""
],
[
"Rosa",
"P. F. S.",
""
],
[
"Bauer",
"E. D.",
""
]
] |
2504.08914 | Austen Z. Fan | Austen Z. Fan, Paraschos Koutris, Sudeepa Roy | Circuits and Formulas for Datalog over Semirings | To appear in PODS 2025 | null | null | null | cs.DB cs.CC | http://creativecommons.org/licenses/by/4.0/ | In this paper, we study circuits and formulas for provenance polynomials of
Datalog programs. We ask the following question: given an absorptive semiring
and a fact of a Datalog program, what is the optimal depth and size of a
circuit/formula that computes its provenance polynomial? We focus on absorptive
semirings as these guarantee the existence of a polynomial-size circuit. Our
main result is a dichotomy for several classes of Datalog programs on whether
they admit a formula of polynomial size or not. We achieve this result by
showing that for these Datalog programs the optimal circuit depth is either
$\Theta(\log m)$ or $\Theta(\log^2 m)$, where $m$ is the input size. We also
show that for Datalog programs with the polynomial fringe property, we can
always construct low-depth circuits of size $O(\log^2 m)$. Finally, we give
characterizations of when Datalog programs are bounded over more general
semirings.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:44:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fan",
"Austen Z.",
""
],
[
"Koutris",
"Paraschos",
""
],
[
"Roy",
"Sudeepa",
""
]
] |
2504.08915 | Jiahuan Long | Jiahuan Long, Tingsong Jiang, Wen Yao, Yizhe Xiong, Zhengqin Xu, Shuai
Jia, Chao Ma | Parameter-Free Fine-tuning via Redundancy Elimination for Vision
Foundation Models | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision foundation models (VFMs) are large pre-trained models that form the
backbone of various vision tasks. Fine-tuning VFMs can further unlock their
potential for downstream tasks or scenarios. However, VFMs often contain
significant feature redundancy, which may limit their adaptability to new
tasks. In this paper, we investigate the redundancies in the segment anything
model (SAM) and then propose a parameter-free fine-tuning method to address
this issue. Unlike traditional fine-tuning methods that adjust parameters, our
method emphasizes selecting, reusing, and enhancing pre-trained features,
offering a new perspective on model fine-tuning. Specifically, we introduce a
channel selection algorithm based on the model's output difference to identify
redundant and effective channels. By selectively replacing the redundant
channels with more effective ones, we filter out less useful features and reuse
the more relevant features to downstream tasks, thereby enhancing the
task-specific feature representation. Experiments on both out-of-domain and
in-domain datasets demonstrate the efficiency and effectiveness of our method.
Notably, our approach can seamlessly integrate with existing fine-tuning
strategies (e.g., LoRA, Adapter), further boosting the performance of already
fine-tuned models. Moreover, since our channel selection involves only model
inference, our method significantly reduces computational and GPU memory
overhead.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:44:27 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Long",
"Jiahuan",
""
],
[
"Jiang",
"Tingsong",
""
],
[
"Yao",
"Wen",
""
],
[
"Xiong",
"Yizhe",
""
],
[
"Xu",
"Zhengqin",
""
],
[
"Jia",
"Shuai",
""
],
[
"Ma",
"Chao",
""
]
] |
2504.08916 | Louis Grenioux | Louis Grenioux, Maxence Noble, Marylou Gabri\'e | Improving the evaluation of samplers on multi-modal targets | Accepted at ICLR 2025 workshop "Frontiers in Probabilistic Inference:
Learning meets Sampling" | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Addressing multi-modality constitutes one of the major challenges of
sampling. In this reflection paper, we advocate for a more systematic
evaluation of samplers towards two sources of difficulty that are mode
separation and dimension. For this, we propose a synthetic experimental setting
that we illustrate on a selection of samplers, focusing on the challenging
criterion of recovery of the mode relative importance. These evaluations are
crucial to diagnose the potential of samplers to handle multi-modality and
therefore to drive progress in the field.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:47:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Grenioux",
"Louis",
""
],
[
"Noble",
"Maxence",
""
],
[
"Gabrié",
"Marylou",
""
]
] |
2504.08917 | Shugo Yoshii | Tatsushi Uno, Shugo Yoshii, Sotaro Mae, Ei Shigematsu, Ryo Ohshima,
Yuichiro Ando, and Masashi Shiraishi | Tunable Magnon Polaritons via Eddy-Current-Induced Dissipation in
Metallic-Banded YIG Spheres | 5 figures | null | null | null | cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We demonstrate a robust method to dynamically tune magnon dissipation in
yttrium iron garnet spheres by equipping a metallic band around the sphere's
equator, enabling precise control over magnon-photon coupling states. The
collective magnetization dynamics in the YIG sphere induce circular eddy
currents in the metallic band, whose magnitude can be systematically varied by
adjusting the angle between the metallic band plane and an external static
magnetic field. This angular dependence yields a pronounced modulation of the
ferromagnetic resonance (FMR) linewidth, facilitating seamless transitions
between the Purcell and strong coupling regimes without altering photon cavity
parameters. Systematic FMR and cavity spectroscopy measurements confirm that
eddy-current-induced losses govern the primary mechanism behind the observed
tunable damping. By achieving extensive periodic-angular dependence of magnon
relaxation rate, we precisely control the magnon-photon coupling state,
approaching the critical coupling condition. These results establish the
YIG-metallic-band platform as a versatile and practical approach for
engineering tunable magnon-polariton systems and advancing magnonic
applications, including those exploring non-Hermitian magnonics.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:48:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Uno",
"Tatsushi",
""
],
[
"Yoshii",
"Shugo",
""
],
[
"Mae",
"Sotaro",
""
],
[
"Shigematsu",
"Ei",
""
],
[
"Ohshima",
"Ryo",
""
],
[
"Ando",
"Yuichiro",
""
],
[
"Shiraishi",
"Masashi",
""
]
] |
2504.08918 | Yichen Xu | Yichen Xu and Arpit Dua | Fault-tolerant protocols through spacetime concatenation | 48 pages, 56 figures. Talk of the paper available online at
https://www.youtube.com/watch?v=xMuzOVfP5BQ | null | null | null | quant-ph cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We introduce a framework called spacetime concatenation for fault-tolerant
compilation of syndrome extraction circuits of stabilizer codes. Spacetime
concatenation enables efficient compilation of syndrome extraction circuits
into dynamical codes through structured gadget layouts and encoding matrices,
facilitating low-weight measurements while preserving logical information. Our
framework uses conditions that are sufficient for fault-tolerance of the
dynamical code, including not measuring logical operators and preserving the
spacetime distance. We construct explicit examples of dynamical codes using
this framework, including the dynamical bivariate bicycle code and a dynamical
Haah code, while illustrating their fault-tolerant properties. Furthermore, we
analyze the classification and resource trade-offs of dynamical codes,
demonstrating their adaptability to hardware constraints, including fabrication
defects and qubit dropout scenarios.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:57:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Yichen",
""
],
[
"Dua",
"Arpit",
""
]
] |
2504.08919 | Zhen Tan | Zhen Tan, Song Wang, Yifan Li, Yu Kong, Jundong Li, Tianlong Chen,
Huan Liu | Are We Merely Justifying Results ex Post Facto? Quantifying Explanatory
Inversion in Post-Hoc Model Explanations | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Post-hoc explanation methods provide interpretation by attributing
predictions to input features. Natural explanations are expected to interpret
how the inputs lead to the predictions. Thus, a fundamental question arises: Do
these explanations unintentionally reverse the natural relationship between
inputs and outputs? Specifically, are the explanations rationalizing
predictions from the output rather than reflecting the true decision process?
To investigate such explanatory inversion, we propose Inversion Quantification
(IQ), a framework that quantifies the degree to which explanations rely on
outputs and deviate from faithful input-output relationships. Using the
framework, we demonstrate on synthetic datasets that widely used methods such
as LIME and SHAP are prone to such inversion, particularly in the presence of
spurious correlations, across tabular, image, and text domains. Finally, we
propose Reproduce-by-Poking (RBP), a simple and model-agnostic enhancement to
post-hoc explanation methods that integrates forward perturbation checks. We
further show that under the IQ framework, RBP theoretically guarantees the
mitigation of explanatory inversion. Empirically, for example, on the
synthesized data, RBP can reduce the inversion by 1.8% on average across iconic
post-hoc explanation approaches and domains.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:00:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Tan",
"Zhen",
""
],
[
"Wang",
"Song",
""
],
[
"Li",
"Yifan",
""
],
[
"Kong",
"Yu",
""
],
[
"Li",
"Jundong",
""
],
[
"Chen",
"Tianlong",
""
],
[
"Liu",
"Huan",
""
]
] |
2504.08921 | Siyang Li | Siyang Li, Adam G. Riess, Gagandeep S. Anand, Daniel Scolnic, Yukei S.
Murakami, Dillon Brout, Erik R. Peterson | The Complete Sample of Available SNe Ia Luminosity Calibrations from the
TRGB Observed with either HST or JWST | 14 pages, 5 figures, 3 tables, submitted to ApJ | null | null | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distance ladders which calibrate the luminosity of Type Ia supernovae (SNe
Ia) currently provide the strongest constraints on the local value of H0.
Recent studies from the Hubble Space Telescope (HST) and James Webb Space
Telescope (JWST) show good consistency between measurements of SNe Ia host
distances. These are calibrated to NGC 4258 using different primary distance
indicators (Cepheids, Tip of the Red Giant Branch (TRGB), J-region Asymptotic
Giant Branch, and Miras). However, some sub-samples of calibrated SNe Ia
employed to measure H0 yield noteworthy differences due to small sample
statistics but also due to differences in sample selection. This issue is
particularly important for TRGB-derived calibrations owing to the smaller
volume they reach compared to Cepheids, reducing sample size and enhancing the
size of statistical fluctuations. To mitigate this issue, we compile the
largest and complete (as currently available) sample of HST or JWST
measurements of the TRGB in the hosts of normal SNe Ia for a total of N=35, 50%
larger than the previous largest. Most are present in the literature, and we
compile multiple measures when available. We also add 5 SNe Ia hosts from the
HST archive not previously published. The full sample together with the
Pantheon+ SN catalog gives H0=72.1-73.3 +/- 1.8 km/s/Mpc (depending on
methodology), in good agreement with the value of 72.5 +/- 1.5 km/s/Mpc from
HST Cepheids in hosts of 42 SNe Ia calibrated by the same anchor, NGC 4258. We
trace the difference in the result of H0=70.4 +/- 1.9 km/s/Mpc from Freedman et
al. 2025 to 11 hosts not selected for that CCHP compilation (of N=24) which
alone yield H0=74.1 km/s/Mpc, 2$\sigma$ higher than the selected sample. A
smaller increase of 0.6 km/s/Mpc comes from a commonly employed correction for
peculiar velocities.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:02:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Siyang",
""
],
[
"Riess",
"Adam G.",
""
],
[
"Anand",
"Gagandeep S.",
""
],
[
"Scolnic",
"Daniel",
""
],
[
"Murakami",
"Yukei S.",
""
],
[
"Brout",
"Dillon",
""
],
[
"Peterson",
"Erik R.",
""
]
] |
2504.08922 | Chunmei Xu | Chunmei Xu, Yi Ma, Rahim Tafazolli, Jiangzhou Wang | Data-Importance-Aware Power Allocation for Adaptive Real-Time
Communication in Computer Vision Applications | Submitted to JSAC | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Life-transformative applications such as immersive extended reality are
revolutionizing wireless communications and computer vision (CV). This paper
presents a novel framework for importance-aware adaptive data transmissions,
designed specifically for real-time CV applications where task-specific
fidelity is critical. A novel importance-weighted mean square error (IMSE)
metric is introduced as a task-oriented measure of reconstruction quality,
considering sub-pixel-level importance (SP-I) and semantic segment-level
importance (SS-I) models. To minimize IMSE under total power constraints,
data-importance-aware waterfilling approaches are proposed to optimally
allocate transmission power according to data importance and channel
conditions, prioritizing sub-streams with high importance. Simulation results
demonstrate that the proposed approaches significantly outperform
margin-adaptive waterfilling and equal power allocation strategies. The data
partitioning that combines both SP-I and SS-I models is shown to achieve the
most significant improvements, with normalized IMSE gains exceeding $7\,$dB and
$10\,$dB over the baselines at high SNRs ($>10\,$dB). These substantial gains
highlight the potential of the proposed framework to enhance data efficiency
and robustness in real-time CV applications, especially in bandwidth-limited
and resource-constrained environments.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:06:32 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Chunmei",
""
],
[
"Ma",
"Yi",
""
],
[
"Tafazolli",
"Rahim",
""
],
[
"Wang",
"Jiangzhou",
""
]
] |
2504.08923 | Vera Koponen | Vera Koponen | A convergence law for continuous logic and continuous structures with
finite domains | null | null | null | null | cs.LO cs.AI math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider continuous relational structures with finite domain $[n] := \{1,
\ldots, n\}$ and a many valued logic, $CLA$, with values in the unit interval
and which uses continuous connectives and continuous aggregation functions.
$CLA$ subsumes first-order logic on ``conventional'' finite structures. To each
relation symbol $R$ and identity constraint $ic$ on a tuple the length of which
matches the arity of $R$ we associate a continuous probability density function
$\mu_R^{ic} : [0, 1] \to [0, \infty)$.
We also consider a probability distribution on the set $\mathbf{W}_n$ of
continuous structures with domain $[n]$ which is such that for every relation
symbol $R$, identity constraint $ic$, and tuple $\bar{a}$ satisfying $ic$, the
distribution of the value of $R(\bar{a})$ is given by $\mu_R^{ic}$,
independently of the values for other relation symbols or other tuples.
In this setting we prove that every formula in $CLA$ is asymptotically
equivalent to a formula without any aggregation function. This is used to prove
a convergence law for $CLA$ which reads as follows for formulas without free
variables: If $\varphi \in CLA$ has no free variable and $I \subseteq [0, 1]$
is an interval, then there is $\alpha \in [0, 1]$ such that, as $n$ tends to
infinity, the probability that the value of $\varphi$ is in $I$ tends to
$\alpha$.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:08:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Koponen",
"Vera",
""
]
] |
2504.08924 | Dev R Sadaula | Dev R Sadaula, Timothy R Kallman, Sibasish Laha | The response of warm absorbers to the variations in the ionizing
continuum in the active galaxy NGC4051 | null | null | null | null | astro-ph.GA astro-ph.HE physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the response of warm absorbers to variations in the ionizing
continuum of the Seyfert 1 galaxy NGC 4051 using time-resolved X-ray
observations from the \textit{Neutron Star Interior Composition Explorer}
(\textit{NICER}). In this work, we have demonstrated we can perform
time-resolved spectroscopic studies of warm absorbers of about $\sim 5500$ s
time resolution using NICER data. We have extracted 15 spectra for this source,
corresponding to 15 different visits to the source, or pointings, each
separated by a longer Earth occultation. By modeling the spectral variability
of the warm absorber with the \texttt{{\sc {\sc warmabs}}} analytic model, we
detect significant variations in the ionization parameter that correlate with
changes in the ionizing flux. A time lag of approximately 5500 seconds is
observed between the flux variations and the absorber's ionization response,
suggesting that the gas is out of photoionization equilibrium during these
periods. Using this time lag, we estimate the lower limit of the gas density
$8.91 \times 10^6 \, \text{cm}^{-3}$ and constrain the location of the warm
absorber to within $7.02 \times 10^{16} \, \text{cm}$ ($\sim 0.02$ parsec) from
the central black hole. This study uses time-resolved spectral analysis to
contribute to our understanding of the physical conditions of ionized AGN
outflows, such as density and location.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:13:09 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sadaula",
"Dev R",
""
],
[
"Kallman",
"Timothy R",
""
],
[
"Laha",
"Sibasish",
""
]
] |
2504.08925 | Duff Neill | Duff Neill, Hanqing Liu | Universality of $SU(\infty)$ relaxation dynamics for $SU(n_f)$-symmetric
spin-models | LA-UR-25-20460 | null | null | null | hep-th cond-mat.dis-nn nucl-th physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spin-models, where the $N$ spins interact pairwise with a $SU(n_f)$ symmetry
preserving hamiltonian, famously simplify in the large $n_f$, $N$ limits, as
derived by Sachdev and Ye when exploring mean-field behavior of spin-glasses.
We present numerical evidence that for a large class of models, the large $n_f$
limit is not necessary: the same dynamical equations can describe the
relaxation processes at high temperatures for a set of classical models
inspired from mean-field treatments of interacting dense neutrino gases, up to
times set by the radius of convergence of the perturbation series for the
correlation function. After a simple rescaling of time, the dynamics display a
surprising universality, being identical for any value of $n_f$ as long as the
rank of the coupling matrix is small. As a corollary of our results, we find
that the direct interaction approximation originating from the study of
stochastic flows in fluid turbulence should be thought of as only a short-time
approximation for generic random coupling systems.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 19:14:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Neill",
"Duff",
""
],
[
"Liu",
"Hanqing",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.