id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2502.21239 | Xiaomin Li | Xiaomin Li, Zhou Yu, Ziji Zhang, Yingying Zhuang, Swair Shah,
Narayanan Sadagopan, Anurag Beniwal | Semantic Volume: Quantifying and Detecting both External and Internal
Uncertainty in LLMs | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) have demonstrated remarkable performance across
diverse tasks by encoding vast amounts of factual knowledge. However, they are
still prone to hallucinations, generating incorrect or misleading information,
often accompanied by high uncertainty. Existing methods for hallucination
detection primarily focus on quantifying internal uncertainty, which arises
from missing or conflicting knowledge within the model. However, hallucinations
can also stem from external uncertainty, where ambiguous user queries lead to
multiple possible interpretations. In this work, we introduce Semantic Volume,
a novel mathematical measure for quantifying both external and internal
uncertainty in LLMs. Our approach perturbs queries and responses, embeds them
in a semantic space, and computes the determinant of the Gram matrix of the
embedding vectors, capturing their dispersion as a measure of uncertainty. Our
framework provides a generalizable and unsupervised uncertainty detection
method without requiring internal access to LLMs. We conduct extensive
experiments on both external and internal uncertainty detection, demonstrating
that our Semantic Volume method consistently outperforms existing baselines in
both tasks. Additionally, we provide theoretical insights linking our measure
to differential entropy, unifying and extending previous sampling-based
uncertainty measures such as the semantic entropy. Semantic Volume is shown to
be a robust and interpretable approach to improving the reliability of LLMs by
systematically detecting uncertainty in both user queries and model responses.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 17:09:08 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Mar 2025 17:31:25 GMT"
},
{
"version": "v3",
"created": "Sat, 29 Mar 2025 01:44:49 GMT"
},
{
"version": "v4",
"created": "Mon, 5 May 2025 16:13:30 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Li",
"Xiaomin",
""
],
[
"Yu",
"Zhou",
""
],
[
"Zhang",
"Ziji",
""
],
[
"Zhuang",
"Yingying",
""
],
[
"Shah",
"Swair",
""
],
[
"Sadagopan",
"Narayanan",
""
],
[
"Beniwal",
"Anurag",
""
]
] |
2503.00532 | Caterina Trimarelli | L. Burmistrov, S. Davarpanah, M. Heller, T. Montaruli, C. Trimarelli,
R. Aloisio, F.C.T. Barbato, I. De Mitri, A. Di Giovanni, G. Fontanella, P.
Savina, C. T\"onnis, E. Moretti, M. Ruzzarin, J. Swako\'n, D. Wr\'obel | Performance and radiation damage mitigation strategy for silicon
photomultipliers on LEO space missions | null | null | null | null | hep-ex physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Space missions require lightweight, low-power consuming, radiation-tolerant
components. Silicon photomultipliers are increasingly used for detecting
near-UV, optical, and infrared light in space due to their compact design, low
cost, low power consumption, robustness, and high photo-detection efficiency,
which makes them sensitive to single photons. Although SiPMs outperform
traditional photomultiplier tubes in many areas, concerns about their radiation
tolerance and noise remain. In this study, we estimate the radiation effects on
a satellite in sun-synchronous low Earth orbit (LEO) at an altitude of 550~km
during the declining phase of solar cycle 25 (2026-2029). We evaluated silicon
photomultipliers produced by the Foundation Bruno Kessler (FBK) using
front-side illuminated technology with metal trenches (NUV-HD-MT), assessing
their response to a 50~MeV proton beam and exposure to a $\beta$-radioactive
source (strontium-90). Simulations with SPENVIS and Geant4 were used to
validate the experimental results. Based on our findings, we propose a
photosensor annealing strategy for space-based instruments.
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 15:17:13 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Mar 2025 09:23:41 GMT"
},
{
"version": "v3",
"created": "Sat, 3 May 2025 20:00:16 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Burmistrov",
"L.",
""
],
[
"Davarpanah",
"S.",
""
],
[
"Heller",
"M.",
""
],
[
"Montaruli",
"T.",
""
],
[
"Trimarelli",
"C.",
""
],
[
"Aloisio",
"R.",
""
],
[
"Barbato",
"F. C. T.",
""
],
[
"De Mitri",
"I.",
""
],
[
"Di Giovanni",
"A.",
""
],
[
"Fontanella",
"G.",
""
],
[
"Savina",
"P.",
""
],
[
"Tönnis",
"C.",
""
],
[
"Moretti",
"E.",
""
],
[
"Ruzzarin",
"M.",
""
],
[
"Swakoń",
"J.",
""
],
[
"Wróbel",
"D.",
""
]
] |
2503.00696 | Wojciech Tralle | Andrei S. Rapinchuk and Wojciech Tralle | On Almost Strong Approximation in Reductive Algebraic Groups | 22 pages | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate a slight weakening of the classical property of strong
approximation, which we call almost strong approximation, for connected
reductive algebraic group over global fields with respect to special sets of
valuations. While nonsimply connected groups (in particular, all algebraic
tori) always fail to have strong approximation - and even almost strong
approximation -- with respect to any finite set of valuations, we show that
under appropriate assumptions they do have almost strong approximation with
respect to (infinite) tractable sets of valuations, i.e. those sets that
contain all archimedean valuations and a generalized arithmetic progression
minus a set having Dirichlet density zero. Almost strong approximation is
likely to have a variety of applications, and as an example we use almost
strong approximation for tori to extend the essential part of the result of
Radhika and Raghunathan on the congruence subgroup problem for inner forms of
type $\textsf{A}_n$ to all absolutely almost simple simply connected groups.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 02:32:51 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 18:44:38 GMT"
},
{
"version": "v3",
"created": "Sun, 4 May 2025 16:02:48 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Rapinchuk",
"Andrei S.",
""
],
[
"Tralle",
"Wojciech",
""
]
] |
2503.01111 | Ning Xia | Ning Xia, Xiaolei Zhao, Yimin Yang, Yixuan Li, Yucong Li | Exploration on Real World Assets and Tokenization | 9 pages | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study delves into the tokenization of real-world assets (RWAs) on the
blockchain with the objective of augmenting liquidity and refining asset
management practices. By conducting an exhaustive analysis of the technical
procedures implicated and scrutinizing case studies of existing deployments,
this research evaluates the advantages, hurdles, and prospective advancements
of blockchain technology in reshaping conventional asset management paradigms.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 02:37:16 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 04:18:09 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Xia",
"Ning",
""
],
[
"Zhao",
"Xiaolei",
""
],
[
"Yang",
"Yimin",
""
],
[
"Li",
"Yixuan",
""
],
[
"Li",
"Yucong",
""
]
] |
2503.01687 | Miika Tuominen | Miika Tuominen | Completions and DK-equivalences of ${\Theta_n}$-spaces | 40 pages | null | null | null | math.CT math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish Rezk completion functors for $\Theta_n$ spaces with respect to
each and all of the completeness conditions. As a consequence, we obtain a
characterization of Dwyer-Kan equivalences between Segal $\Theta_n$ spaces.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 16:02:23 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Mar 2025 15:10:30 GMT"
},
{
"version": "v3",
"created": "Sat, 3 May 2025 17:55:55 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Tuominen",
"Miika",
""
]
] |
2503.01706 | Sergiy Akkelin | Sergii V. Akkelin, Dirk H. Rischke | Gradient corrections to the local-equilibrium energy-momentum tensor in
the Zubarev approach | 16 pages, title changed, clarifications added | null | null | null | hep-ph hep-th nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We compute the expectation value of the energy-momentum tensor of a real
scalar field in an approximation which accounts for space-time gradients of the
hydrodynamical variables in local thermodynamical equilibrium. We show that the
energy-momentum tensor receives corrections with respect to the standard
local-equilibrium result. Notably, the relation between the energy density and
pressure, i.e., the equation of state, is modified with respect to the one in
global equilibrium. The obtained corrections might be relevant for systems
created in relativistic hadron and heavy-ion collisions.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 16:19:49 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 12:42:51 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Akkelin",
"Sergii V.",
""
],
[
"Rischke",
"Dirk H.",
""
]
] |
2503.01764 | Yuvaraj Elangovan | Yuvaraj Elangovan, B. Satyanarayana, Ravindra Shinde, Mandar Saraf,
Pathaleswar, S. Thoi Thoi, Gobinda Majumder, S.R. Joshi, Piyush Verma, Honey
Khindri and Umesh L | Design and Development of Portable RPC-Based Cosmic Muon Tracker | null | null | null | null | hep-ex physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Primary cosmic rays when interact with our atmosphere, produce a cascade of
lighter secondary particles namely pion, kaon, neutrons, muons, electrons,
positrons and neutrinos. Muons are one of the most abundant and easily
detectable particles at the ground surface using a large variety of particle
detectors. Resistive Plate Chambers (RPCs) of 2m x 2m in dimension were
developed to be used in large scale as the active detector elements in the Iron
Calorimeter (ICAL) which was planned to be built by the India-based Neutrino
Observatory (INO). As a spin-off of this work, a portable stack of eight, one
square foot RPC detectors is developed named as Cosmic Muon Tracker (CMT). It
could be used to conduct small-scale particle detector experiments as well as
training Students. We will discuss design, integration, characterisation and
some of the applications of this detector in this paper.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 17:38:15 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 15:08:41 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Elangovan",
"Yuvaraj",
""
],
[
"Satyanarayana",
"B.",
""
],
[
"Shinde",
"Ravindra",
""
],
[
"Saraf",
"Mandar",
""
],
[
"Pathaleswar",
"",
""
],
[
"Thoi",
"S. Thoi",
""
],
[
"Majumder",
"Gobinda",
""
],
[
"Joshi",
"S. R.",
""
],
[
"Verma",
"Piyush",
""
],
[
"Khindri",
"Honey",
""
],
[
"L",
"Umesh",
""
]
] |
2503.01813 | Luyi Yang | Zhixuan Bi, Yunfei Bai, Ying Shi, Tao Sun, Heng Wu, Haochen Zhang,
Yubin Wang, Miao-Ling Lin, Yaxian Wang, Ping-Heng Tan, Sheng Meng, Qihua
Xiong, and Luyi Yang | Intrinsic exciton transport and recombination in single-crystal lead
bromide perovskite | null | ACS Nano (2025) | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Photogenerated carrier transport and recombination in metal halide
perovskites are critical to device performance. Despite considerable efforts,
sample quality issues and measurement techniques have limited the access to
their intrinsic physics. Here, by utilizing high-purity CsPbBr3 single crystals
and contact-free transient grating spectroscopy, we directly monitor exciton
diffusive transport from 26 to 300 K. As the temperature (T) increases, the
carrier mobility ({\mu}) decreases rapidly below 100 K wtih a {\mu}~T^{-3.0}
scaling, and then follows a more gradual {\mu}~T^{-1.7} trend at higher
temperatures. First-principles calculations perfectly reproduce this
experimental trend and reveal that optical phonon scattering governs carrier
mobility shifts over the entire temperature range, with a single longitudinal
optical mode dominating room-temperature transport. Time-resolved
photoluminescence further identifies a substantial increase in exciton
radiative lifetime with temperature, attributed to increased exciton population
in momentum-dark states caused by phonon scattering. Our findings unambiguously
resolve previous theory-experiment discrepancies, providing benchmarks for
future optoelectronic design.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 18:41:55 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Bi",
"Zhixuan",
""
],
[
"Bai",
"Yunfei",
""
],
[
"Shi",
"Ying",
""
],
[
"Sun",
"Tao",
""
],
[
"Wu",
"Heng",
""
],
[
"Zhang",
"Haochen",
""
],
[
"Wang",
"Yubin",
""
],
[
"Lin",
"Miao-Ling",
""
],
[
"Wang",
"Yaxian",
""
],
[
"Tan",
"Ping-Heng",
""
],
[
"Meng",
"Sheng",
""
],
[
"Xiong",
"Qihua",
""
],
[
"Yang",
"Luyi",
""
]
] |
2503.02147 | Ju Li | Chia-Wei Hsu, Nien-Ti Tsou, Yu-Cheng Chen, Yang Jeong Park, Ju Li | Frankenstein Optimizer: Harnessing the Potential by Revisiting
Optimization Tricks | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Gradient-based optimization drives the unprecedented performance of modern
deep neural network models across diverse applications. Adaptive algorithms
have accelerated neural network training due to their rapid convergence rates;
however, they struggle to find ``flat minima" reliably, resulting in suboptimal
generalization compared to stochastic gradient descent (SGD). By revisiting
various adaptive algorithms' mechanisms, we propose the Frankenstein optimizer,
which combines their advantages. The proposed Frankenstein dynamically adjusts
first- and second-momentum coefficients according to the optimizer's current
state to directly maintain consistent learning dynamics and immediately reflect
sudden gradient changes. Extensive experiments across several research domains
such as computer vision, natural language processing, few-shot learning, and
scientific simulations show that Frankenstein surpasses existing adaptive
algorithms and SGD empirically regarding convergence speed and generalization
performance. Furthermore, this research deepens our understanding of adaptive
algorithms through centered kernel alignment analysis and loss landscape
visualization during the learning process. Code is available at
https://github.com/acctouhou/Frankenstein_optimizer
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 00:25:54 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 21:10:48 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Hsu",
"Chia-Wei",
""
],
[
"Tsou",
"Nien-Ti",
""
],
[
"Chen",
"Yu-Cheng",
""
],
[
"Park",
"Yang Jeong",
""
],
[
"Li",
"Ju",
""
]
] |
2503.02440 | Helin Gong | Ruizhi Zhang, Shengfeng Zhu, Kan Wang, Ding She, Jean-Philippe Argaud,
Bertrand Bouriquet, Qing Li, Helin Gong | Artificial Intelligence in Reactor Physics: Current Status and Future
Prospects | 33 pages, 6 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reactor physics is the study of neutron properties, focusing on using models
to examine the interactions between neutrons and materials in nuclear reactors.
Artificial intelligence (AI) has made significant contributions to reactor
physics, e.g., in operational simulations, safety design, real-time monitoring,
core management and maintenance. This paper presents a comprehensive review of
AI approaches in reactor physics, especially considering the category of
Machine Learning (ML), with the aim of describing the application scenarios,
frontier topics, unsolved challenges and future research directions. From
equation solving and state parameter prediction to nuclear industry
applications, this paper provides a step-by-step overview of ML methods applied
to steady-state, transient and combustion problems. Most literature works
achieve industry-demanded models by enhancing the efficiency of deterministic
methods or correcting uncertainty methods, which leads to successful
applications. However, research on ML methods in reactor physics is somewhat
fragmented, and the ability to generalize models needs to be strengthened.
Progress is still possible, especially in addressing theoretical challenges and
enhancing industrial applications such as building surrogate models and digital
twins.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 09:36:58 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 08:35:01 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Zhang",
"Ruizhi",
""
],
[
"Zhu",
"Shengfeng",
""
],
[
"Wang",
"Kan",
""
],
[
"She",
"Ding",
""
],
[
"Argaud",
"Jean-Philippe",
""
],
[
"Bouriquet",
"Bertrand",
""
],
[
"Li",
"Qing",
""
],
[
"Gong",
"Helin",
""
]
] |
2503.02498 | Milin Patel | Milin Patel, Rolf Jung, Marzana Khatun | A Systematic Literature Review on Safety of the Intended Functionality
for Automated Driving Systems | Scheduled to be published in SAE journal as technical paper as a part
of non-technical event and will be available as open access in 2025 | SAE Technical Paper 2025-01-5030, 2025 | 10.4271/2025-01-5030 | null | eess.SY cs.LG cs.SY | http://creativecommons.org/licenses/by/4.0/ | In the automobile industry, ensuring the safety of automated vehicles
equipped with the Automated Driving System (ADS) is becoming a significant
focus due to the increasing development and deployment of automated driving.
Automated driving depends on sensing both the external and internal
environments of a vehicle, utilizing perception sensors and algorithms, and
Electrical/Electronic (E/E) systems for situational awareness and response. ISO
21448 is the standard for Safety of the Intended Functionality (SOTIF) that
aims to ensure that the ADS operate safely within their intended functionality.
SOTIF focuses on preventing or mitigating potential hazards that may arise from
the limitations or failures of the ADS, including hazards due to
insufficiencies of specification, or performance insufficiencies, as well as
foreseeable misuse of the intended functionality. However, the challenge lies
in ensuring the safety of vehicles despite the limited availability of
extensive and systematic literature on SOTIF. To address this challenge, a
Systematic Literature Review (SLR) on SOTIF for the ADS is performed following
the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
guidelines. The objective is to methodically gather and analyze the existing
literature on SOTIF. The major contributions of this paper are: (i) presenting
a summary of the literature by synthesizing and organizing the collective
findings, methodologies, and insights into distinct thematic groups, and (ii)
summarizing and categorizing the acknowledged limitations based on data
extracted from an SLR of 51 research papers published between 2018 and 2023.
Furthermore, research gaps are determined, and future research directions are
proposed.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 11:04:36 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 07:19:50 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Patel",
"Milin",
""
],
[
"Jung",
"Rolf",
""
],
[
"Khatun",
"Marzana",
""
]
] |
2503.02636 | Yeganeh Farahzadi | Yeganeh Farahzadi, Morteza Ansarinia, Zoltan Kekecs | YARE-GAN: Yet Another Resting State EEG-GAN | null | null | null | null | q-bio.NC cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | In this study, we implement a Wasserstein GAN with Gradient Penalty (WGAN-GP)
to generate multi-channel resting-state EEG data and assess the quality of the
synthesized signals through both visual and feature-based evaluations. Our
results indicate that the model effectively captures the statistical and
spectral characteristics of real EEG data, although challenges remain in
replicating high-frequency oscillations in the frontal region. Additionally, we
demonstrate that the Critic's learned representations can be reused for gender
classification task, achieving an out-of-sample accuracy, significantly better
than a shuffled-label baseline and a model trained directly on EEG data. These
findings suggest that generative models can serve not only as EEG data
generators but also as unsupervised feature extractors, reducing the need for
manual feature engineering. This study highlights the potential of GAN-based
unsupervised learning for EEG analysis, suggesting avenues for more
data-efficient deep learning applications in neuroscience.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 14:01:10 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 11:07:54 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 15:16:31 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Farahzadi",
"Yeganeh",
""
],
[
"Ansarinia",
"Morteza",
""
],
[
"Kekecs",
"Zoltan",
""
]
] |
2503.02650 | William Brach | William Brach, Kristi\'an Ko\v{s}\v{t}\'al, Michal Ries | The Effectiveness of Large Language Models in Transforming Unstructured
Text to Standardized Formats | null | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | The exponential growth of unstructured text data presents a fundamental
challenge in modern data management and information retrieval. While Large
Language Models (LLMs) have shown remarkable capabilities in natural language
processing, their potential to transform unstructured text into standardized,
structured formats remains largely unexplored - a capability that could
revolutionize data processing workflows across industries. This study breaks
new ground by systematically evaluating LLMs' ability to convert unstructured
recipe text into the structured Cooklang format. Through comprehensive testing
of four models (GPT-4o, GPT-4o-mini, Llama3.1:70b, and Llama3.1:8b), an
innovative evaluation approach is introduced that combines traditional metrics
(WER, ROUGE-L, TER) with specialized metrics for semantic element
identification. Our experiments reveal that GPT-4o with few-shot prompting
achieves breakthrough performance (ROUGE-L: 0.9722, WER: 0.0730), demonstrating
for the first time that LLMs can reliably transform domain-specific
unstructured text into structured formats without extensive training. Although
model performance generally scales with size, we uncover surprising potential
in smaller models like Llama3.1:8b for optimization through targeted
fine-tuning. These findings open new possibilities for automated structured
data generation across various domains, from medical records to technical
documentation, potentially transforming the way organizations process and
utilize unstructured information.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 14:14:28 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 12:25:44 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Brach",
"William",
""
],
[
"Košťál",
"Kristián",
""
],
[
"Ries",
"Michal",
""
]
] |
2503.02910 | Wenqi Guo | Wenqi Guo, Yiyang Du, Shan Du | LangGas: Introducing Language in Selective Zero-Shot Background
Subtraction for Semi-Transparent Gas Leak Detection with a New Dataset | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gas leakage poses a significant hazard that requires prevention.
Traditionally, human inspection has been used for detection, a slow and
labour-intensive process. Recent research has applied machine learning
techniques to this problem, yet there remains a shortage of high-quality,
publicly available datasets. This paper introduces a synthetic dataset, SimGas,
featuring diverse backgrounds, interfering foreground objects, diverse leak
locations, and precise segmentation ground truth. We propose a zero-shot method
that combines background subtraction, zero-shot object detection, filtering,
and segmentation to leverage this dataset. Experimental results indicate that
our approach significantly outperforms baseline methods based solely on
background subtraction and zero-shot object detection with segmentation,
reaching an IoU of 69%. We also present an analysis of various prompt
configurations and threshold settings to provide deeper insights into the
performance of our method. Finally, we qualitatively (because of the lack of
ground truth) tested our performance on GasVid and reached decent results on
the real-world dataset. The dataset, code, and full qualitative results are
available at https://github.com/weathon/Lang-Gas.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 06:17:17 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 05:19:44 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Apr 2025 01:35:57 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Guo",
"Wenqi",
""
],
[
"Du",
"Yiyang",
""
],
[
"Du",
"Shan",
""
]
] |
2503.02952 | Kavya Ravichandran | Avrim Blum, Emily Diana, Kavya Ravichandran, Alexander Williams
Tolbert | A Theoretical Model for Grit in Pursuing Ambitious Ends | null | null | null | null | cs.CY cs.DS cs.GT stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ambition and risk-taking have been heralded as important ways for
marginalized communities to get out of cycles of poverty. As a result,
educational messaging often encourages individuals to strengthen their personal
resolve and develop characteristics such as discipline and grit to succeed in
ambitious ends. However, recent work in philosophy and sociology highlights
that this messaging often does more harm than good for students in these
situations. We study similar questions using a different epistemic approach and
in simple theoretical models -- we provide a quantitative model of
decision-making between stable and risky choices in the improving multi-armed
bandits framework. We use this model to first study how individuals'
"strategies" are affected by their level of grittiness and how this affects
their accrued rewards. Then, we study the impact of various interventions, such
as increasing grit or providing a financial safety net. Our investigation of
rational decision making involves two different formal models of rationality,
the competitive ratio between the accrued reward and the optimal reward and
Bayesian quantification of uncertainty.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 19:17:42 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 20:10:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Blum",
"Avrim",
""
],
[
"Diana",
"Emily",
""
],
[
"Ravichandran",
"Kavya",
""
],
[
"Tolbert",
"Alexander Williams",
""
]
] |
2503.03269 | Saurabh Kumar | Saurabh Kumar, Jacob Buckman, Carles Gelada, Sean Zhang | Conformal Transformations for Symmetric Power Transformers | SCOPE Workshop at ICLR 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Transformers with linear attention offer significant computational advantages
over softmax-based transformers but often suffer from degraded performance. The
symmetric power (sympow) transformer, a particular type of linear transformer,
addresses some of this performance gap by leveraging symmetric tensor
embeddings, achieving comparable performance to softmax transformers. However,
the finite capacity of the recurrent state in sympow transformers limits their
ability to retain information, leading to performance degradation when scaling
the training or evaluation context length. To address this issue, we propose
the conformal-sympow transformer, which dynamically frees up capacity using
data-dependent multiplicative gating and adaptively stores information using
data-dependent rotary embeddings. Preliminary experiments on the LongCrawl64
dataset demonstrate that conformal-sympow overcomes the limitations of sympow
transformers, achieving robust performance across scaled training and
evaluation contexts.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 08:50:53 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 05:24:08 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Kumar",
"Saurabh",
""
],
[
"Buckman",
"Jacob",
""
],
[
"Gelada",
"Carles",
""
],
[
"Zhang",
"Sean",
""
]
] |
2503.03317 | Wenting Wang | Wenting Wang, Xiaohu Yang, Yipeng Jing, Ashley J. Ross, Malgorzata
Siudek, John Moustakas, Samuel G. Moore, Shaun Cole, Carlos Frenk, Jiaxi Yu,
Sergey E. Koposov, Jiaxin Han, Zhenlin Tan, Kun Xu, Yizhou Gu, Yirong Wang,
Oleg Y. Gnedin, Jessica Nicole Aguilar, Steven Ahlen, Davide Bianchi, David
Brooks, Todd Claybaugh, Axel de la Macorra, Arjun Dey, Peter Doel, Jaime E.
Forero-Romero, Enrique Gaztanaga, Satya Gontcho A Gontcho, Gaston Gutierrez,
Klaus Honscheid, Mustapha Ishak, Theodore Kisner, Martin Landriau, Laurent Le
Guillou, Marc Manera, Aaron Meisner, Ramon Miquel, Seshadri Nadathur, Claire
Poppett, Francisco Prada, Ignasi P\'erez-R\`afols, Graziano Rossi, Eusebio
Sanchez, David Schlegel, Hee-Jong Seo, Joseph Harry Silber, David Sprayberry,
Gregory Tarle, Benjamin Alan Weaver, Hu Zou | Luminosity and stellar mass functions of faint photometric satellites
around spectroscopic central galaxies from DESI Year-1 Bright Galaxy Survey | accepted by ApJ | null | null | null | astro-ph.GA astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We measure the luminosity functions (LFs) and stellar mass functions (SMFs)
of photometric satellite galaxies around spectroscopically identified isolated
central galaxies (ICGs). The photometric satellites are from the DESI Legacy
Imaging Surveys (DR9), while the spectroscopic ICGs are selected from the DESI
Year-1 BGS sample. We can measure satellite LFs down to $r$-band absolute
magnitudes of $M_{r,\mathrm{sat}}\sim-7$, around ICGs as small as
$7.1<\log_{10}M_{\ast,\mathrm{ICG}}/\mathrm{M_\odot}<7.8$, with the stellar
mass of ICGs measured by the DESI Fastspecfit pipeline. The satellite SMF can
be measured down to $\log_{10}M_{\ast,\mathrm{sat}}/\mathrm{M_\odot}\sim 5.5$.
Interestingly, we discover that the faint/low-mass end slopes of satellite
LFs/SMFs become steeper with the decrease in the stellar masses of host ICGs,
with smaller and nearby host ICGs capable of being used to probe their fainter
satellites.. The steepest slopes are $-2.298\pm0.656$ and $-$2.888$\pm$0.916
for satellite LF and SMF, respectively. Detailed comparisons are performed
between the satellite LFs around ICGs selected from DESI BGS or from the SDSS
NYU-VAGC spectroscopic Main galaxies over
$7.1<\log_{10}M_{\ast,\mathrm{ICG}}/\mathrm{M_\odot}<11.7$, showing reasonable
agreements, but we show that the differences between DESI and SDSS stellar
masses for ICGs play a role to affect the results. We also compare measurements
based on DESI Fastspecfit and Cigale stellar masses used to bin ICGs, with the
latter including the modeling of AGN based on WISE photometry, and we find good
agreements in the measured satellite LFs by using either of the DESI stellar
mass catalogs.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 09:52:06 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 10:33:36 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Wang",
"Wenting",
""
],
[
"Yang",
"Xiaohu",
""
],
[
"Jing",
"Yipeng",
""
],
[
"Ross",
"Ashley J.",
""
],
[
"Siudek",
"Malgorzata",
""
],
[
"Moustakas",
"John",
""
],
[
"Moore",
"Samuel G.",
""
],
[
"Cole",
"Shaun",
""
],
[
"Frenk",
"Carlos",
""
],
[
"Yu",
"Jiaxi",
""
],
[
"Koposov",
"Sergey E.",
""
],
[
"Han",
"Jiaxin",
""
],
[
"Tan",
"Zhenlin",
""
],
[
"Xu",
"Kun",
""
],
[
"Gu",
"Yizhou",
""
],
[
"Wang",
"Yirong",
""
],
[
"Gnedin",
"Oleg Y.",
""
],
[
"Aguilar",
"Jessica Nicole",
""
],
[
"Ahlen",
"Steven",
""
],
[
"Bianchi",
"Davide",
""
],
[
"Brooks",
"David",
""
],
[
"Claybaugh",
"Todd",
""
],
[
"de la Macorra",
"Axel",
""
],
[
"Dey",
"Arjun",
""
],
[
"Doel",
"Peter",
""
],
[
"Forero-Romero",
"Jaime E.",
""
],
[
"Gaztanaga",
"Enrique",
""
],
[
"Gontcho",
"Satya Gontcho A",
""
],
[
"Gutierrez",
"Gaston",
""
],
[
"Honscheid",
"Klaus",
""
],
[
"Ishak",
"Mustapha",
""
],
[
"Kisner",
"Theodore",
""
],
[
"Landriau",
"Martin",
""
],
[
"Guillou",
"Laurent Le",
""
],
[
"Manera",
"Marc",
""
],
[
"Meisner",
"Aaron",
""
],
[
"Miquel",
"Ramon",
""
],
[
"Nadathur",
"Seshadri",
""
],
[
"Poppett",
"Claire",
""
],
[
"Prada",
"Francisco",
""
],
[
"Pérez-Ràfols",
"Ignasi",
""
],
[
"Rossi",
"Graziano",
""
],
[
"Sanchez",
"Eusebio",
""
],
[
"Schlegel",
"David",
""
],
[
"Seo",
"Hee-Jong",
""
],
[
"Silber",
"Joseph Harry",
""
],
[
"Sprayberry",
"David",
""
],
[
"Tarle",
"Gregory",
""
],
[
"Weaver",
"Benjamin Alan",
""
],
[
"Zou",
"Hu",
""
]
] |
2503.03560 | Jiayi Yao | Jiayi Yao and Shuowen Zhang | Optimal Beamforming for Multi-Target Multi-User ISAC Exploiting Prior
Information: How Many Sensing Beams Are Needed? | This is the longer version of a paper submitted for possible journal
publication | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper studies a multi-target multi-user integrated sensing and
communication (ISAC) system where a multi-antenna base station (BS)
communicates with multiple single-antenna users in the downlink and senses the
unknown and random angle information of multiple targets based on their
reflected echo signals at the BS receiver as well as their prior probability
information. We focus on a general beamforming structure with both
communication beams and dedicated sensing beams, whose design is highly
non-trivial as more sensing beams provide more flexibility in sensing, but
introduce extra interference to communication. To resolve this trade-off, we
first characterize the periodic posterior Cram\'er-Rao bound (PCRB) as a lower
bound of the mean-cyclic error (MCE) in multi-target sensing. Then, we optimize
the beamforming to minimize the maximum periodic PCRB among all targets to
ensure fairness, subject to individual communication rate constraints at
multiple users. Despite the non-convexity of this problem, we propose a general
construction method for the optimal solution by leveraging semi-definite
relaxation (SDR), and derive a general bound on the number of sensing beams
needed. Moreover, we unveil specific structures of the optimal solution in
various cases, where tighter bounds on the number of sensing beams needed are
derived (e.g., no or at most one sensing beam is needed under stringent rate
constraints or with homogeneous targets). Next, we study the beamforming
optimization to minimize the sum periodic PCRB under user rate constraints. By
applying SDR, we propose a general construction method for the optimal solution
and its specific structures which yield lower computational complexities. We
derive a general bound and various tighter bounds on the number of sensing
beams needed. Numerical results validate our analysis and effectiveness of our
proposed beamforming designs.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 14:46:33 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 14:01:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Yao",
"Jiayi",
""
],
[
"Zhang",
"Shuowen",
""
]
] |
2503.03738 | Feliks Przytycki | Feliks Przytycki | There are not many periodic orbits in bunches for iteration of complex
quadratic polynomials of one variable | v2: Section 4 "Cremer periodic orbits" has been rewritten and
simplified | null | null | null | math.DS | http://creativecommons.org/licenses/by/4.0/ | It is proved that for every complex quadratic polynomial $f$ with Cremer's
fixed point $z_0$ (or periodic orbit) for every $\delta>0$, there is at most
one periodic orbit of minimal period $n$ for all $n$ large enough, entirely in
the disc (ball) $B(z_0, \exp -\delta n)$ (at most $p$ for a periodic Cremer
orbit of period $p$). Next, it is proved that the number of periodic orbits of
period $n$ in a bunch $P_n$, that is for all $x,y\in P_n$, $|f^j(x)- f^j(y)|\le
\exp -\delta n$ for all $j=0,...,n-1$, does not exceed $\exp \delta n$. We
conclude that the geometric pressure defined with the use of periodic points
coincides with the one defined with the use of preimages of an arbitrary
typical point. I. Binder, K. Makarov and S. Smirnov (Duke Math. J. 2003) proved
this for all polynomials but assuming all periodic orbits were hyperbolic, and
asked about general situations. We prove here a positive answer for all
quadratic polynomials.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 18:48:05 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 12:34:18 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Przytycki",
"Feliks",
""
]
] |
2503.04168 | Itai Epstein | Michael Klein, Yonatan Gershuni, Alisa Perutski, Jean-Paul Hugonin,
and Itai Epstein | Nanometer-scale Cavities for Mid-infrared Radiation via Image Phonon
Polariton Resonators | null | null | null | null | physics.optics physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Surface-polaritons play a pivotal role in strong light-matter interactions at
the nanoscale due to their ability to confine light to deep subwavelength
dimensions. A promising class of materials exhibiting such polaritonic response
are polar dielectrics, which support surface phonon polaritons (SPhPs). While
SPhPs offer significantly lower losses compared to other polaritons, their
potential has been underutilized due to limited ability to reach large
continent factors. Here, we demonstrate a system composed of silver nanocubes
deposited on a SiC polar dielectric, which experimentally realizes the
antisymmetric-image-phonon-polaritons (AIPhPs) mode: a hybridized SPhPs mode
that can confine mid-infrared radiation to extremely small mode-volumes, almost
a billion times smaller than their free-space volume, with quality-factors an
order-of-magnitude greater than those of surface-plasmons or graphene-plasmons,
surpassing values of 180. Our method is general, scalable, and applicable to
any polar dielectric, opening the path for controlling and manipulating strong
light-matter interactions at the nanoscale in the long-wavelength range.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 07:31:29 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 03:58:58 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Klein",
"Michael",
""
],
[
"Gershuni",
"Yonatan",
""
],
[
"Perutski",
"Alisa",
""
],
[
"Hugonin",
"Jean-Paul",
""
],
[
"Epstein",
"Itai",
""
]
] |
2503.04349 | Christoph Steinacker | Christoph Steinacker, Mads Paulsen, Malte Schr\"oder, Jeppe Rich | Robust design of bicycle infrastructure networks | null | null | 10.1038/s41598-025-99976-9 | null | physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | Promoting active mobility like cycling relies on the availability of
well-connected, high-quality bicycle networks. However, expanding these
networks over an extended planning horizon presents one of the most complex
challenges in transport science. This complexity arises from the intricate
interactions between infrastructure availability and usage, such as network
spillover effects and mode choice substitutions. In this paper, we approach the
problem from two perspectives: direct optimization methods, which generate
near-optimal solutions using operations research techniques, and conceptual
heuristics, which offer intuitive and scalable algorithms grounded in network
science. Specifically, we compare direct welfare optimization with an inverse
network percolation approach to planning cycle superhighway extensions in
Copenhagen. Interestingly, while the more complex optimization models yield
better overall welfare results, the improvements over simpler methods are
small. More importantly, we demonstrate that the increased complexity of
planning approaches generally makes them more vulnerable to input uncertainty,
reflecting the bias-variance tradeoff. This issue is particularly relevant in
the context of long-term planning, where conditions change during the
implementation of the planned infrastructure expansions. Therefore, while
planning bicycle infrastructure is important and renders exceptionally high
benefit-cost ratios, considerations of robustness and ease of implementation
may justify the use of more straightforward network-based methods.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 11:45:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Steinacker",
"Christoph",
""
],
[
"Paulsen",
"Mads",
""
],
[
"Schröder",
"Malte",
""
],
[
"Rich",
"Jeppe",
""
]
] |
2503.04429 | Narmeen Oozeer | Narmeen Oozeer, Dhruv Nathawani, Nirmalendu Prakash, Michael Lan, Abir
Harrasse, Amirali Abdullah | Activation Space Interventions Can Be Transferred Between Large Language
Models | 68 pages | null | null | null | cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | The study of representation universality in AI models reveals growing
convergence across domains, modalities, and architectures. However, the
practical applications of representation universality remain largely
unexplored. We bridge this gap by demonstrating that safety interventions can
be transferred between models through learned mappings of their shared
activation spaces. We demonstrate this approach on two well-established AI
safety tasks: backdoor removal and refusal of harmful prompts, showing
successful transfer of steering vectors that alter the models' outputs in a
predictable way. Additionally, we propose a new task, \textit{corrupted
capabilities}, where models are fine-tuned to embed knowledge tied to a
backdoor. This tests their ability to separate useful skills from backdoors,
reflecting real-world challenges. Extensive experiments across Llama, Qwen and
Gemma model families show that our method enables using smaller models to
efficiently align larger ones. Furthermore, we demonstrate that autoencoder
mappings between base and fine-tuned models can serve as reliable ``lightweight
safety switches", allowing dynamic toggling between model behaviors.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 13:38:44 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 16:39:22 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Oozeer",
"Narmeen",
""
],
[
"Nathawani",
"Dhruv",
""
],
[
"Prakash",
"Nirmalendu",
""
],
[
"Lan",
"Michael",
""
],
[
"Harrasse",
"Abir",
""
],
[
"Abdullah",
"Amirali",
""
]
] |
2503.04785 | Jos\'e Antonio Siqueira De Cerqueira | Jos\'e Siqueira de Cerqueira, Kai-Kristian Kemell, Rebekah Rousi,
Nannan Xi, Juho Hamari, Pekka Abrahamsson | Mapping Trustworthiness in Large Language Models: A Bibliometric
Analysis Bridging Theory to Practice | null | null | null | null | cs.CL cs.CY | http://creativecommons.org/licenses/by/4.0/ | The rapid proliferation of Large Language Models (LLMs) has raised
significant trustworthiness and ethical concerns. Despite the widespread
adoption of LLMs across domains, there is still no clear consensus on how to
define and operationalise trustworthiness. This study aims to bridge the gap
between theoretical discussion and practical implementation by analysing
research trends, definitions of trustworthiness, and practical techniques. We
conducted a bibliometric mapping analysis of 2,006 publications from Web of
Science (2019-2025) using the Bibliometrix, and manually reviewed 68 papers. We
found a shift from traditional AI ethics discussion to LLM trustworthiness
frameworks. We identified 18 different definitions of trust/trustworthiness,
with transparency, explainability and reliability emerging as the most common
dimensions. We identified 20 strategies to enhance LLM trustworthiness, with
fine-tuning and retrieval-augmented generation (RAG) being the most prominent.
Most of the strategies are developer-driven and applied during the
post-training phase. Several authors propose fragmented terminologies rather
than unified frameworks, leading to the risks of "ethics washing," where
ethical discourse is adopted without a genuine regulatory commitment. Our
findings highlight: persistent gaps between theoretical taxonomies and
practical implementation, the crucial role of the developer in operationalising
trust, and call for standardised frameworks and stronger regulatory measures to
enable trustworthy and ethical deployment of LLMs.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 14:02:33 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Apr 2025 15:04:55 GMT"
},
{
"version": "v3",
"created": "Sun, 4 May 2025 23:37:27 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"de Cerqueira",
"José Siqueira",
""
],
[
"Kemell",
"Kai-Kristian",
""
],
[
"Rousi",
"Rebekah",
""
],
[
"Xi",
"Nannan",
""
],
[
"Hamari",
"Juho",
""
],
[
"Abrahamsson",
"Pekka",
""
]
] |
2503.05136 | Ronny Ko | Ronny Ko | The Beginner's Textbook for Fully Homomorphic Encryption | null | null | null | null | cs.CR cs.DM | http://creativecommons.org/licenses/by/4.0/ | Fully Homomorphic Encryption (FHE) is a cryptographic scheme that enables
computations to be performed directly on encrypted data, as if the data were in
plaintext. After all computations are performed on the encrypted data, it can
be decrypted to reveal the result. The decrypted value matches the result that
would have been obtained if the same computations were applied to the plaintext
data.
FHE supports basic operations such as addition and multiplication on
encrypted numbers. Using these fundamental operations, more complex
computations can be constructed, including subtraction, division, logic gates
(e.g., AND, OR, XOR, NAND, MUX), and even advanced mathematical functions such
as ReLU, sigmoid, and trigonometric functions (e.g., sin, cos). These functions
can be implemented either as exact formulas or as approximations, depending on
the trade-off between computational efficiency and accuracy.
Fully Homomorphic Encryption (FHE) enables privacy-preserving machine
learning by allowing a server to process the client's data in its encrypted
form through an ML model. With FHE, the server learns neither the plaintext
version of the input features nor the inference results. Only the client, using
their secret key, can decrypt and access the results at the end of the service
protocol.FHE can also be applied to confidential blockchain services, ensuring
that sensitive data in smart contracts remains encrypted and confidential while
maintaining the transparency and integrity of the execution process. Other
applications of FHE include secure outsourcing of data analytics, encrypted
database queries, privacy-preserving searches, efficient multi-party
computation for digital signatures, and more.
This article is designed to help the reader understand how FHE works from the
mathematical level.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 04:29:11 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 15:18:50 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 03:22:13 GMT"
},
{
"version": "v4",
"created": "Sun, 13 Apr 2025 13:14:01 GMT"
},
{
"version": "v5",
"created": "Sat, 26 Apr 2025 18:20:16 GMT"
},
{
"version": "v6",
"created": "Sun, 4 May 2025 15:31:10 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ko",
"Ronny",
""
]
] |
2503.05153 | Yunhao Luo | Yunhao Luo, Utkarsh A. Mishra, Yilun Du, Danfei Xu | Generative Trajectory Stitching through Diffusion Composition | Project page: https://comp-diffuser.github.io/ | null | null | null | cs.RO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Effective trajectory stitching for long-horizon planning is a significant
challenge in robotic decision-making. While diffusion models have shown promise
in planning, they are limited to solving tasks similar to those seen in their
training data. We propose CompDiffuser, a novel generative approach that can
solve new tasks by learning to compositionally stitch together shorter
trajectory chunks from previously seen tasks. Our key insight is modeling the
trajectory distribution by subdividing it into overlapping chunks and learning
their conditional relationships through a single bidirectional diffusion model.
This allows information to propagate between segments during generation,
ensuring physically consistent connections. We conduct experiments on benchmark
tasks of various difficulties, covering different environment sizes, agent
state dimension, trajectory types, training data quality, and show that
CompDiffuser significantly outperforms existing methods.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 05:22:52 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 16:26:30 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Luo",
"Yunhao",
""
],
[
"Mishra",
"Utkarsh A.",
""
],
[
"Du",
"Yilun",
""
],
[
"Xu",
"Danfei",
""
]
] |
2503.05440 | Jian-Rong Li | Xiao-Juan An, Jian-Rong Li, Yan-Feng Luo, and Wen-Ting Zhang | A path description for $\varepsilon$-characters of representations of
type $A$ restricted quantum loop algebras at roots of unity | null | null | null | null | math.QA | http://creativecommons.org/licenses/by/4.0/ | Fix $\varepsilon^{2\ell}=1$ with $\ell \geq 2$. In this paper, we show that
all finite-dimensional simple modules of any restricted quantum loop algebra
$U_{\varepsilon}^{\rm res}({L\mathfrak{sl}_{n+1}})$ in a certain category can
be transformed into snake modules. We obtain an effective and concrete path
description for $\varepsilon$-characters of any simple module with highest
$l$-weight of degree two and any Kirillov-Reshetikhin module of
$U_{\varepsilon}^{\rm res}({L\mathfrak{sl}_{n+1}})$. As an application of our
path description, we obtain a necessary and sufficient condition for the tensor
product of two fundamental representations of $U_{\varepsilon}^{\rm
res}({L\mathfrak{sl}_{n+1}})$ to be irreducible. Additionally, we obtain a
necessary condition for the tensor product of two or more fundamental
representations of $U_{\varepsilon}^{\rm res}({L\mathfrak{sl}_{n+1}})$ to be
irreducible.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 14:11:23 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 18:19:37 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"An",
"Xiao-Juan",
""
],
[
"Li",
"Jian-Rong",
""
],
[
"Luo",
"Yan-Feng",
""
],
[
"Zhang",
"Wen-Ting",
""
]
] |
2503.05824 | Minda Ma | Minda Ma, Shufan Zhang, Junhong Liu, Ran Yan, Weiguang Cai, Nan Zhou,
Jinyue Yan | Building floorspace and stock measurement: A review of global efforts,
knowledge gaps, and research priorities | null | null | null | null | physics.soc-ph econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | Despite a substantial body of research-evidenced by our analysis of 2,628
peer-reviewed papers-global building floorspace data remain fragmented,
inconsistent, and methodologically diverse. The lack of high-quality and openly
accessible datasets poses major challenges to accurately assessing building
carbon neutrality. This review focuses on global building floorspace,
especially its nexus with energy and emissions. The key research areas include
energy modeling, emissions analysis, building retrofits, and life cycle
assessments. Each measurement approach-top-down, bottom-up, and hybrid-has its
own limitations: top-down methods provide broad estimates but low accuracy,
whereas bottom-up approaches are more precise but data intensive. Our
simulations reveal a surge in floorspace growth across emerging economies-most
notably in India, Indonesia, and Africa-with India's per capita floorspace
projected to triple by 2070. We emphasize the need for a high-resolution global
floorspace imagery database to compare energy efficiency, track decarbonization
progress, and assess renovation impacts while promoting building sufficiency
and accelerating the transition to net-zero building systems.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 10:29:12 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 12:59:51 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 01:51:06 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ma",
"Minda",
""
],
[
"Zhang",
"Shufan",
""
],
[
"Liu",
"Junhong",
""
],
[
"Yan",
"Ran",
""
],
[
"Cai",
"Weiguang",
""
],
[
"Zhou",
"Nan",
""
],
[
"Yan",
"Jinyue",
""
]
] |
2503.06222 | Meng Wang | Meng Wang, Fan Wu, Yunchuan Qin, Ruihui Li, Zhuo Tang, Kenli Li | Vision-based 3D Semantic Scene Completion via Capture Dynamic
Representations | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The vision-based semantic scene completion task aims to predict dense
geometric and semantic 3D scene representations from 2D images. However, the
presence of dynamic objects in the scene seriously affects the accuracy of the
model inferring 3D structures from 2D images. Existing methods simply stack
multiple frames of image input to increase dense scene semantic information,
but ignore the fact that dynamic objects and non-texture areas violate
multi-view consistency and matching reliability. To address these issues, we
propose a novel method, CDScene: Vision-based Robust Semantic Scene Completion
via Capturing Dynamic Representations. First, we leverage a multimodal
large-scale model to extract 2D explicit semantics and align them into 3D
space. Second, we exploit the characteristics of monocular and stereo depth to
decouple scene information into dynamic and static features. The dynamic
features contain structural relationships around dynamic objects, and the
static features contain dense contextual spatial information. Finally, we
design a dynamic-static adaptive fusion module to effectively extract and
aggregate complementary features, achieving robust and accurate semantic scene
completion in autonomous driving scenarios. Extensive experimental results on
the SemanticKITTI, SSCBench-KITTI360, and SemanticKITTI-C datasets demonstrate
the superiority and robustness of CDScene over existing state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 13:49:43 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 02:33:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Wang",
"Meng",
""
],
[
"Wu",
"Fan",
""
],
[
"Qin",
"Yunchuan",
""
],
[
"Li",
"Ruihui",
""
],
[
"Tang",
"Zhuo",
""
],
[
"Li",
"Kenli",
""
]
] |
2503.06337 | Gopeshh Raaj Subbaraj | Mohit Pandey, Gopeshh Subbaraj, Artem Cherkasov, Martin Ester,
Emmanuel Bengio | Pretraining Generative Flow Networks with Inexpensive Rewards for
Molecular Graph Generation | arXiv admin note: text overlap with arXiv:2409.09702 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Generative Flow Networks (GFlowNets) have recently emerged as a suitable
framework for generating diverse and high-quality molecular structures by
learning from rewards treated as unnormalized distributions. Previous works in
this framework often restrict exploration by using predefined molecular
fragments as building blocks, limiting the chemical space that can be accessed.
In this work, we introduce Atomic GFlowNets (A-GFNs), a foundational generative
model leveraging individual atoms as building blocks to explore drug-like
chemical space more comprehensively. We propose an unsupervised pre-training
approach using drug-like molecule datasets, which teaches A-GFNs about
inexpensive yet informative molecular descriptors such as drug-likeliness,
topological polar surface area, and synthetic accessibility scores. These
properties serve as proxy rewards, guiding A-GFNs towards regions of chemical
space that exhibit desirable pharmacological properties. We further implement a
goal-conditioned finetuning process, which adapts A-GFNs to optimize for
specific target properties. In this work, we pretrain A-GFN on a subset of ZINC
dataset, and by employing robust evaluation metrics we show the effectiveness
of our approach when compared to other relevant baseline methods for a wide
range of drug design tasks.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 20:41:07 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Mar 2025 19:56:33 GMT"
},
{
"version": "v3",
"created": "Fri, 2 May 2025 18:37:20 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Pandey",
"Mohit",
""
],
[
"Subbaraj",
"Gopeshh",
""
],
[
"Cherkasov",
"Artem",
""
],
[
"Ester",
"Martin",
""
],
[
"Bengio",
"Emmanuel",
""
]
] |
2503.06457 | Yanbiao Ma | Yanbiao Ma, Wei Dai, Wenke Huang, Jiayi Chen | Geometric Knowledge-Guided Localized Global Distribution Alignment for
Federated Learning | Accepted by CVPR Oral 2025 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Data heterogeneity in federated learning, characterized by a significant
misalignment between local and global distributions, leads to divergent local
optimization directions and hinders global model training. Existing studies
mainly focus on optimizing local updates or global aggregation, but these
indirect approaches demonstrate instability when handling highly heterogeneous
data distributions, especially in scenarios where label skew and domain skew
coexist. To address this, we propose a geometry-guided data generation method
that centers on simulating the global embedding distribution locally. We first
introduce the concept of the geometric shape of an embedding distribution and
then address the challenge of obtaining global geometric shapes under privacy
constraints. Subsequently, we propose GGEUR, which leverages global geometric
shapes to guide the generation of new samples, enabling a closer approximation
to the ideal global distribution. In single-domain scenarios, we augment
samples based on global geometric shapes to enhance model generalization; in
multi-domain scenarios, we further employ class prototypes to simulate the
global distribution across domains. Extensive experimental results demonstrate
that our method significantly enhances the performance of existing approaches
in handling highly heterogeneous data, including scenarios with label skew,
domain skew, and their coexistence. Code published at:
https://github.com/WeiDai-David/2025CVPR_GGEUR
| [
{
"version": "v1",
"created": "Sun, 9 Mar 2025 05:30:28 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 11:38:40 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ma",
"Yanbiao",
""
],
[
"Dai",
"Wei",
""
],
[
"Huang",
"Wenke",
""
],
[
"Chen",
"Jiayi",
""
]
] |
2503.06504 | Nancy Ghangas | Nancy Ghangas, Ghanasyam Remesh, Venu Gopal Achanta, Shubhrangshu
Dasgupta | Dynamics of Light Localization via Coherent Control: The Interplay of
Transmission, Absorption and Disorder in Photonic Crystals | 13 pages, 5 figures | Volume 587, August 2025, 131876 | 10.1016/j.optcom.2025.131876 | null | physics.optics | http://creativecommons.org/licenses/by-sa/4.0/ | This study investigates the interplay between structural disorder,
absorption, and Lyapunov exponent dynamics to exploit localization phenomena in
photonic crystals with engineered defect layers. We generate disorder by
introducing random refractive index variations in one of the bilayers, while
the application of a control field to $\Lambda$-type atoms within a central
defect layer enables dynamic tuning of the effective refractive index of
crystal. We have employed traditional transfer matrix method to demonstrate
transmission, Lyapunov exponents and absorption in the crystal. Through
coherent control, we dynamically tune absorption, revealing sharp contrasts in
band gap and band edge regions. while Lyapunov exponents, quantifying
localization lengths, exhibit a consistent scaling across both band gap and
band edge frequencies, and this behavior remains robust even in the presence of
disorder. Hence, distinct localization mechanisms emerge at bandgap and
band-edge frequencies. Bandgap localization arises from optical mode
confinement and resonant alignment of atomic transitions with the probe field
while band edge localization stems from a synergy of loss-difference-induced
trapping and Anderson like disorder effects. Notably, while disorder weakens
confinement localization in the band gap, it actually strengthens localization
at the band edges. These results deepen the understanding of light-matter
coupling in disordered photonic systems and provide a framework for designing
reconfigurable optical devices with tailored localization properties.
| [
{
"version": "v1",
"created": "Sun, 9 Mar 2025 08:09:53 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 17:06:18 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ghangas",
"Nancy",
""
],
[
"Remesh",
"Ghanasyam",
""
],
[
"Achanta",
"Venu Gopal",
""
],
[
"Dasgupta",
"Shubhrangshu",
""
]
] |
2503.06685 | Zhaowei Chen | Zhaowei Chen, Borui Zhao, Yuchen Ge, Yuhao Chen, Renjie Song, Jiajun
Liang | Asymmetric Decision-Making in Online Knowledge Distillation:Unifying
Consensus and Divergence | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Online Knowledge Distillation (OKD) methods streamline the distillation
training process into a single stage, eliminating the need for knowledge
transfer from a pretrained teacher network to a more compact student network.
This paper presents an innovative approach to leverage intermediate spatial
representations. Our analysis of the intermediate features from both teacher
and student models reveals two pivotal insights: (1) the similar features
between students and teachers are predominantly focused on foreground objects.
(2) teacher models emphasize foreground objects more than students. Building on
these findings, we propose Asymmetric Decision-Making (ADM) to enhance feature
consensus learning for student models while continuously promoting feature
diversity in teacher models. Specifically, Consensus Learning for student
models prioritizes spatial features with high consensus relative to teacher
models. Conversely, Divergence Learning for teacher models highlights spatial
features with lower similarity compared to student models, indicating superior
performance by teacher models in these regions. Consequently, ADM facilitates
the student models to catch up with the feature learning process of the teacher
models. Extensive experiments demonstrate that ADM consistently surpasses
existing OKD methods across various online knowledge distillation settings and
also achieves superior results when applied to offline knowledge distillation,
semantic segmentation and diffusion distillation tasks.
| [
{
"version": "v1",
"created": "Sun, 9 Mar 2025 16:32:25 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Chen",
"Zhaowei",
""
],
[
"Zhao",
"Borui",
""
],
[
"Ge",
"Yuchen",
""
],
[
"Chen",
"Yuhao",
""
],
[
"Song",
"Renjie",
""
],
[
"Liang",
"Jiajun",
""
]
] |
2503.06705 | Apostolos Angelis | Apostolos Angelis and George Kousiouris | A Survey on the Landscape of Self-adaptive Cloud Design and Operations
Patterns: Goals, Strategies, Tooling, Evaluation and Dataset Perspectives | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cloud-native applications have significantly advanced the development and
scalability of online services through the use of microservices and modular
architectures. However, achieving adaptability, resilience, and efficient
performance management within cloud environments remains a key challenge. This
survey provides an overview of self-adaptive cloud design and operations
patterns published over the last seven years, focusing on a taxonomy of their
objectives, scope of control, decision-making mechanisms approach, automation
level and validation methodologies. Overall, 96 papers have been taken under
consideration, indicating a significant increase in the years since 2023 in the
produced output. The analysis highlights the prevalence of feedback loop
structures, with both reactive and proactive implementations, and underscores
the increasing role of machine learning techniques in predictive management,
especially when it comes to resource provisioning and management of the
executed applications. On the other hand, adaptive application architectures
through direct application-level pattern-based management seem significantly
underrepresented in the current field of research, thus serving as an
uninvestigated area for future research. Furthermore, the current work
highlights practical aspects such as validation datasets per category
(application, resource, network, etc.), tools, technologies and frameworks
usage during the experimentation, in order to guide researchers in the
validation process for comparative and robust experimentation.
| [
{
"version": "v1",
"created": "Sun, 9 Mar 2025 17:41:47 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 20:18:56 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Angelis",
"Apostolos",
""
],
[
"Kousiouris",
"George",
""
]
] |
2503.06846 | Claire Miller Dr | Claire M. Miller, Michael J. Lydeamore, Lee Berger, Lee F. Skerratt,
Jennifer A. Flegg, Anthony W. Waddle, and Patricia Therese Campbell | Sunlight-heated refugia protect frogs from chytridiomycosis: a
mathematical modelling study | null | null | null | null | q-bio.PE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The fungal disease Chytridiomycosis poses a threat to frog populations
worldwide. It has driven over 90 amphibian species to extinction and severely
affected hundreds more. Difficulties in disease management have shown a need
for novel conservation approaches.
We present a novel mathematical model for chytridiomycosis transmission in
frogs that includes the natural history of infection, to test the hypothesis
that sunlight-heated refugia reduce transmission. The model is fit using
approximate Bayesian computation to experimental data where a cohort of frogs,
a fixed subset of which had cleared a prior infection, were provided access to
either sunlight-heated or shaded refugia. Using our model, we can estimate the
extent to which prior chytridiomycosis infection protects against subsequent
infection, and quantify the effect of sunlight-heating of refugia.
Results estimate a 40% reduction in chytridiomycosis transmission when frogs
have access to sunlight-heated refugia, compared to shaded refugia. This
strongly supports the hypothesis that the sunlight-heated refugia reduce
disease transmission. Frogs that were infected and recovered were estimated to
have a reduction in susceptibility of approximately 97% compared to frogs with
no prior infection.
This research provides quantitative evidence supporting sunlight-heated
refugia as an effective disease management tool for chytridiomycosis in frog
populations. By estimating both the impact of refugia and the protective
effects of prior infection, the model provides an evidence base for
implementing sunlight-heated refugia as part of amphibian conservation
strategies. This work represents an important first step in using mathematical
modelling to inform policy on the design and implementation of habitat-based
interventions to support amphibian population recovery and long-term
sustainability.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 02:15:51 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 04:43:09 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 01:40:47 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Miller",
"Claire M.",
""
],
[
"Lydeamore",
"Michael J.",
""
],
[
"Berger",
"Lee",
""
],
[
"Skerratt",
"Lee F.",
""
],
[
"Flegg",
"Jennifer A.",
""
],
[
"Waddle",
"Anthony W.",
""
],
[
"Campbell",
"Patricia Therese",
""
]
] |
2503.06957 | David Darrow | David Darrow and George Stepaniants | A Spectral Theory of Scalar Volterra Equations | 93 pages, 18 figures. Submitted to Communications of the AMS. Updated
to include several new results, new numerical experiments, and general
revisions | null | null | null | math.CA math.DS math.SP physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | This work aims to bridge the gap between pure and applied research on scalar,
linear Volterra equations by examining five major classes: integral and
integro-differential equations with completely monotone kernels (viscoelastic
models); equations with positive definite kernels (partially observed quantum
systems); difference equations with discrete, positive definite kernels; a
generalized class of delay differential equations; and a generalized class of
fractional differential equations. We develop a general, spectral theory that
provides a system of correspondences between these disparate domains. As a
result, we see how `interconversion' (operator inversion) arises as a natural,
continuous involution within each class, yielding a plethora of novel formulas
for analytical solutions of such equations. This spectral theory unifies and
extends existing results in viscoelasticity, signal processing, and analysis.
It also introduces a geometric construction of the regularized Hilbert
transform, revealing a fundamental connection to fractional and delay
differential equations. Finally, it offers a practical toolbox for Volterra
equations of all classes, reducing many to pen-and-paper calculations and
introducing a powerful spectral approach for general Volterra equations based
on rational approximation.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 06:10:09 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 22:17:34 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Darrow",
"David",
""
],
[
"Stepaniants",
"George",
""
]
] |
2503.06961 | Wonwoo Lee | Hyeong-Chan Kim (Korea Natl. U. Trans.) and Wonwoo Lee (CQUeST, Seoul) | Dressing rotating black holes with anisotropic matter | 18 pages, 3 figures, title is changed, Ch. 2 is reduced, and
reference papers are added | null | null | CQUeST-2025-0755 | gr-qc astro-ph.GA hep-th | http://creativecommons.org/licenses/by/4.0/ | We present a new rotating black hole solution to the Einstein equations as an
extension of the Kerr spacetime. To derive this solution, we use the
Newman-Janis algorithm and results from the non-complexification procedure.
Interestingly, the solution we find may not be uniquely characterized by
asymptotic parameters such as mass, angular momentum, and charge, thereby
challenging the no-hair theorem. We also analyze in detail how this additional
characteristics (``hair") affects the thermodynamic properties of the black
hole.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 06:20:09 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 05:00:36 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Kim",
"Hyeong-Chan",
"",
"Korea Natl. U. Trans."
],
[
"Lee",
"Wonwoo",
"",
"CQUeST, Seoul"
]
] |
2503.07379 | Kentaro Inoue | Kentaro Inoue | Log Prismatic Dieudonn\'{e} theory and its application to Shimura
varieties | 55 pages. Some part of the original article has been detached is now
included in arXiv:2505.01084 | null | null | null | math.AG math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the log version of the prismatic Dieudonn\'{e} theory established by
Ansch\"{u}tz-Le Bras. By applying this result to the integral toroidal
compactification of a Shimura variety of Hodge type, we extend the prismatic
realization, originally constructed by Imai-Kato-Youcis, to the
compactification. This extension enables us to prove Lovering's conjecture on
$p$-adic comparison isomorphisms for Shimura varieties.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 14:31:42 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 05:33:14 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Inoue",
"Kentaro",
""
]
] |
2503.07749 | Yingtong Shen | Yukun Wang, Yingtong Shen, Zhichao Zhang and Linchun Wan | RBM-Based Simulated Quantum Annealing for Graph Isomorphism Problems | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The graph isomorphism problem remains a fundamental challenge in computer
science, driving the search for efficient decision algorithms. Due to its
ambiguous computational complexity, heuristic approaches such as simulated
annealing are frequently used, achieving high solution probabilities while
avoiding exhaustive enumeration. However, traditional simulated annealing
usually struggles with low sampling efficiency and reduced solution-finding
probability in complex or large graph problems. In this study, we integrate the
principles of quantum technology to address the graph isomorphism problem. By
mapping the solution space to a quantum many-body system, we developed a
parameterized model for variational simulated annealing. This approach
emphasizes the regions of the solution space that are most likely to contain
the optimal solution, thereby enhancing the search accuracy.Artificial neural
networks were utilized to parameterize the quantum many-body system, leveraging
their capacity for efficient function approximation to perform accurate
sampling in the intricate energy landscapes of large graph problems.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 18:03:22 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 10:30:30 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Wang",
"Yukun",
""
],
[
"Shen",
"Yingtong",
""
],
[
"Zhang",
"Zhichao",
""
],
[
"Wan",
"Linchun",
""
]
] |
2503.08555 | Jannis L\"ubsen | Jannis O. Luebsen, Annika Eichler | An Analysis of Safety Guarantees in Multi-Task Bayesian Optimization | null | null | null | null | cs.LG cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper addresses the integration of additional information sources into a
Bayesian optimization framework while ensuring that safety constraints are
satisfied. The interdependencies between these information sources are modeled
using an unknown correlation matrix. We explore how uniform error bounds must
be adjusted to maintain constraint satisfaction throughout the optimization
process, considering both Bayesian and frequentist statistical perspectives.
This is achieved by appropriately scaling the error bounds based on a
confidence interval that can be estimated from the data. Furthermore, the
efficacy of the proposed approach is demonstrated through experiments on two
benchmark functions and a controller parameter optimization problem. Our
results highlight a significant improvement in sample efficiency, demonstrating
the methods suitability for optimizing expensive-to-evaluate functions.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 15:45:37 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Mar 2025 10:10:28 GMT"
},
{
"version": "v3",
"created": "Sat, 3 May 2025 07:50:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Luebsen",
"Jannis O.",
""
],
[
"Eichler",
"Annika",
""
]
] |
2503.09240 | Timothy Miller | Timothy C. Miller | Vertex models for the product of a permuted-basement Demazure atom and a
Schur polynomial | 38 pages, 3 figures | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present the first positive combinatorial rule for expanding the product of
a permuted-basement Demazure atom and a Schur polynomial. Special cases of
permuted-basement Demazure atoms include Demazure atoms and characters. These
cases have known tableau formulas for their expansions when multiplied by a
Schur polynomial, due to Haglund, Luoto, Mason and van Willigenburg. We find a
vertex model formula, giving a new rule even in these special cases, extending
a technique introduced by Zinn-Justin for calculating Littlewood-Richardson
coefficients.
We derive a coloured vertex model for permuted-basement Demazure atoms,
inspired by Borodin and Wheeler's model for non-symmetric Macdonald
polynomials. We make this model compatible with an uncoloured vertex model for
Schur polynomials, putting them in a single framework. Unlike previous work on
structure coefficients via vertex models, a remarkable feature of our
construction is that it relies on a Yang-Baxter equation that only holds for
certain boundary conditions. However, this restricted Yang-Baxter equation is
sufficient to show our result.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 10:35:08 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 00:34:34 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Miller",
"Timothy C.",
""
]
] |
2503.09333 | Bilal Canturk | Bilal Canturk | Unification of stochastic matrices and quantum operations for N-level
systems | 12 pages, 1 figure, research paper | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The time evolution of the one-point probability vector of stochastic
processes and quantum processes for $N$-level systems have been unified. Hence,
quantum states and quantum operations can be regarded as generalizations of the
one-point probability vectors and stochastic matrices, respectively. More
essentially, based on the unification, it has been proven that completely
positive divisibility (CP-divisibility) for quantum operations is the natural
extension of the Chapman-Kolmogorov equation. It is thus shown that
CP-divisibility is a necessary but insufficient condition for a quantum process
to be specified as Markovian. The main results have been illustrated through a
dichotomic Markov process.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:27:39 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 13:33:58 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Canturk",
"Bilal",
""
]
] |
2503.09805 | Jordan Taylor | Jordan Taylor, Joel Mire, Franchesca Spektor, Alicia DeVrio, Maarten
Sap, Haiyi Zhu, Sarah Fox | Un-Straightening Generative AI: How Queer Artists Surface and Challenge
the Normativity of Generative AI Models | null | null | 10.1145/3715275.3732061 | null | cs.HC cs.AI | http://creativecommons.org/licenses/by/4.0/ | Queer people are often discussed as targets of bias, harm, or discrimination
in research on generative AI. However, the specific ways that queer people
engage with generative AI, and thus possible uses that support queer people,
have yet to be explored. We conducted a workshop study with 13 queer artists,
during which we gave participants access to GPT-4 and DALL-E 3 and facilitated
group sensemaking activities. We found our participants struggled to use these
models due to various normative values embedded in their designs, such as
hyper-positivity and anti-sexuality. We describe various strategies our
participants developed to overcome these models' limitations and how,
nevertheless, our participants found value in these highly-normative
technologies. Drawing on queer feminist theory, we discuss implications for the
conceptualization of "state-of-the-art" models and consider how FAccT
researchers might support queer alternatives.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 20:16:38 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 17:07:07 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Taylor",
"Jordan",
""
],
[
"Mire",
"Joel",
""
],
[
"Spektor",
"Franchesca",
""
],
[
"DeVrio",
"Alicia",
""
],
[
"Sap",
"Maarten",
""
],
[
"Zhu",
"Haiyi",
""
],
[
"Fox",
"Sarah",
""
]
] |
2503.10067 | Luling Jin | Zhikun Zou, Gan Guo, Meng Wen, Bin Liu, Xue Yan, Yangji\'e Liu, and
Luling Jin | Enhanced MVA of Polarized Proton Beams via PW Laser-Driven Plasma Bubble | 11 pages, 9 figures | Matter Radiat. Extremes 10, 037202 (2025) | 10.1063/5.0249082 | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The significance of laser-driven polarized beam acceleration has been
increasingly recognized in recent years. We propose an efficient method for
generating polarized proton beams from a pre-polarized hydrogen halide gas jet,
utilizing magnetic vortex acceleration enhanced by a laser-driven plasma
bubble. When a petawatt laser pulse passes through a pre-polarized gas jet, a
bubble-like ultra-nonlinear plasma wave is formed. As part of the wave
particles, background protons are swept by the acceleration field of the bubble
and oscillate significantly along the laser propagation axis. Some of the
pre-accelerated protons in the plasma wave are trapped by the acceleration
field at the rear side of the target. This acceleration field is intensified by
the transverse expansion of the laser-driven magnetic vortex, resulting in
energetic polarized proton beams. The spin of energetic protons is determined
by their precession within the electromagnetic field, as described by the
Thomas-Bargmann-Michel-Telegdi equation in analytical models and
particle-in-cell simulations. Multidimensional simulations reveal that
monoenergetic proton beams with hundreds of MeV in energy, a beam charge of
hundreds of pC, and a beam polarization of tens of percent can be produced at
laser powers of several petawatts. Laser-driven polarized proton beams offer
promising potential for application in polarized beam colliders, where they can
be utilized to investigate particle interactions and to explore the properties
of matter under unique conditions.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 05:30:34 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Zou",
"Zhikun",
""
],
[
"Guo",
"Gan",
""
],
[
"Wen",
"Meng",
""
],
[
"Liu",
"Bin",
""
],
[
"Yan",
"Xue",
""
],
[
"Liu",
"Yangjié",
""
],
[
"Jin",
"Luling",
""
]
] |
2503.10822 | Johannes Buchner | Johannes Buchner | Reinforcement Learning and Life Cycle Assessment for a Circular Economy
-- Towards Progressive Computer Science | Shortened conference paper version, with a new section on Life Cycle
Assessment, and thus a new title | null | null | null | cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The aim of this paper is to discuss the potential of using methods from
Reinforcement Learning for Life Cycle Assessment in a circular economy, and to
present some new ideas in this direction. To give some context, we explain how
Reinforcement Learning was successfully applied in computer chess (and beyond).
As computer chess was historically called the "drosophila of AI", we start by
describing a method for the board representation called 'rotated bitboards'
that can potentially also be applied in the context of sustainability. In the
first part of this paper, the concepts of the bitboard-representation and the
advantages of (rotated) bitboards in move generation are explained. In order to
illustrate those ideas practice, the concrete implementation of the
move-generator in FUSc# (a chess engine developed at FU Berlin in C# some years
ago) is described. In addition, rotated binary neural networks are discussed
briefly.
The second part deals with reinforcement learning in computer chess (and
beyond). We exemplify the progress that has been made in this field in the last
15-20 years by comparing the "state of the art" from 2002-2008, when FUSc# was
developed, with the ground-breaking innovations connected to "AlphaZero". We
review some application of the ideas developed in AlphaZero in other domains,
e.g. the "other Alphas" like AlphaFold, AlphaTensor, AlphaGeometry and
AlphaProof. In the final part of the paper, we discuss the computer-science
related challenges that changing the economic paradigm towards (absolute)
sustainability poses and in how far what we call 'progressive computer science'
needs to contribute. Concrete challenges include the closing of material loops
in a circular economy with Life Cycle Assessment in order to optimize for
(absolute) sustainability, and we present some new ideas in this direction.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 19:13:51 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Apr 2025 19:57:41 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 17:53:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Buchner",
"Johannes",
""
]
] |
2503.11494 | Alessandro Ciani | Marco Schumann and Tobias Stollenwerk and Alessandro Ciani | Bridging wire and gate cutting with ZX-calculus | 14 pages, 8 figures, 5 tables. Added new example in Appendix B. Fixed
typo in Table VI and added scalar factors to Fig. 10 | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum circuit cutting refers to a series of techniques that allow one to
partition a quantum computation on a large quantum computer into several
quantum computations on smaller devices. This usually comes at the price of a
sampling overhead, that is quantified by the $1$-norm of the associated
decomposition. The applicability of these techniques relies on the possibility
of finding decompositions of the ideal, global unitaries into quantum
operations that can be simulated onto each sub-register, which should ideally
minimize the $1$-norm. In this work, we show how these decompositions can be
obtained diagrammatically using ZX-calculus expanding on the work of Ufrecht et
al. [arXiv:2302.00387]. The central idea of our work is that since in
ZX-calculus only connectivity matters, it should be possible to cut wires in
ZX-diagrams by inserting known decompositions of the identity in standard
quantum circuits. We show how, using this basic idea, many of the gate
decompositions known in the literature can be re-interpreted as an instance of
wire cuts in ZX-diagrams. Furthermore, we obtain improved decompositions for
multi-qubit controlled-Z (MCZ) gates with $1$-norm equal to $3$ for any number
of qubits and any partition, which we argue to be optimal. Our work gives new
ways of thinking about circuit cutting that can be particularly valuable for
finding decompositions of large unitary gates. Besides, it sheds light on the
question of why exploiting classical communication decreases the 1-norm of a
wire cut but does not do so for certain gate decompositions. In particular,
using wire cuts with classical communication, we obtain gate decompositions
that do not require classical communication.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 15:20:47 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 16:39:43 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 12:24:44 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Schumann",
"Marco",
""
],
[
"Stollenwerk",
"Tobias",
""
],
[
"Ciani",
"Alessandro",
""
]
] |
2503.11653 | Wilhelm Kerle-Malcharek | Wilhelm Kerle-Malcharek and Niklas Hann-von-Weyhern and Ulf Hailer and
Steffen Diefenbach and Stefan P Feyer and Karsten Klein and Falk Schreiber | The Junction of Immersive Analytics and Virtual Reconstructions -- A
Case Study on the Mausoleum of Emperor Maxentius | null | null | null | null | cs.OH | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Virtual Archaeology has significantly evolved over the last decades through
advancements in data acquisition and representation by, e.g., improved data
recording technologies and virtual reality devices. Immersive environments
provide novel ways to present historical events or objects with high visual
quality for both, the general public as well as researchers. Here, we examine
how the emerging field of Immersive Analytics can contribute to enhancing the
understanding and exploration of archaeological data and explore the junction
of Virtual Archaeology and Immersive Analytics, utilising reconstructions of
the mausoleum of the late Roman emperor Maxentius in Rome for argumentation.
Based on our work, we advocate the value of combining historical and computer
science expertise for virtual reconstructions and immersive environments to
facilitate deeper understanding and interactive exploration of archaeological
data.
| [
{
"version": "v1",
"created": "Wed, 15 Jan 2025 11:42:36 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 08:36:22 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Kerle-Malcharek",
"Wilhelm",
""
],
[
"Hann-von-Weyhern",
"Niklas",
""
],
[
"Hailer",
"Ulf",
""
],
[
"Diefenbach",
"Steffen",
""
],
[
"Feyer",
"Stefan P",
""
],
[
"Klein",
"Karsten",
""
],
[
"Schreiber",
"Falk",
""
]
] |
2503.11676 | Yuchen Ding | Yuchen Ding and Honghu Liu | Note on a theorem of Birch and Erd\H{o}s | In this new version, we obtain a few more results | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | Let $p,q>1$ be two relatively prime integers and $\mathbb{N}$ the set of
nonnegative integers. Let $f_{p,q}(n)$ be the number of different expressions
of $n$ written as a sum of distinct terms taken from
$\{p^{\alpha}q^{\beta}:\alpha,\beta\in \mathbb{N}\}$. Erd\H os conjectured and
then Birch proved that $f_{p,q}(n)\ge 1$ provided that $n$ is sufficiently
large. In this note, for all sufficiently large number $n$ we prove $$
f_{p,q}(n)=2^{\frac{(\log n)^2}{2\log p\log q}\big(1+O(\log\log n/\log
n)\big)}. $$ We also prove that $f_{2,q}(n+1)\ge f_{2,q}(n)$ for any $n\ge 1$
and furthermore classify all the situations of $n$ such that
$f_{2,q}(n+1)=f_{2,q}(n)$. Additionally, we will show that $$
\lim_{n\rightarrow\infty}f_{2,q}(n+1)/f_{2,q}(n)=1. $$
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 12:30:12 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 14:45:15 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ding",
"Yuchen",
""
],
[
"Liu",
"Honghu",
""
]
] |
2503.11929 | Lingyang Liu | Lingyang Liu and Hang Gao | Local controllability of a free-boundary problem for 1D degenerate
parabolic equations | null | null | null | null | math.OC math.AP | http://creativecommons.org/licenses/by/4.0/ | This paper deals with the local controllability of a free-boundary problem
for the 1D boundary-degenerate parabolic equation with distributed controls,
locally supported in space. We prove that, if the final time T is fixed and the
initial state is sufficiently small, there exist controls that drive the state
exactly to rest at time t = T. The proof is based on Schauder's fixed point
theorem, combined with appropriate estimates for solutions to degenerate
parabolic equations and for the control function.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 00:27:02 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 09:51:41 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Liu",
"Lingyang",
""
],
[
"Gao",
"Hang",
""
]
] |
2503.12117 | William Cook | William Cook | The Resonance Bias Framework: Resonance, Structure, and Arithmetic in
Quadrature Error | 21 pages, 6 figures | null | null | null | math.GM | http://creativecommons.org/licenses/by/4.0/ | We study the trapezoidal rule for periodic functions on uniform grids and
show that the quadrature error exhibits a rich deterministic structure, beyond
traditional asymptotic or statistical interpretations. Focusing on the
prototype function f(x) = sin^2(2 pi k x), we derive an analytical expression
for the error governed by a resonance function chi_P(y), closely related to the
Dirichlet kernel, roots of unity, and discrete Fourier analysis on the group
Z/PZ. This function acts as a spectral filter, connecting the integration error
to arithmetic properties such as k/P and geometric phase cancellation,
visualized as vector averaging on the unit circle. We introduce the Resonance
Bias Framework (RBF), a generalization to arbitrary smooth periodic functions,
leading to the error representation B_P[f] = sum_{k != 0} c_k chi_P(k/P).
Although this is mathematically equivalent to the classical aliasing sum, it
reveals a deeper mechanism: the quadrature error arises from structured
resonance rather than random aliasing noise. The RBF thus provides an
interpretable framework for understanding integration errors at finite
resolution, grounded in number theory and geometry.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 12:54:00 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Apr 2025 00:27:53 GMT"
},
{
"version": "v3",
"created": "Sat, 3 May 2025 15:33:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Cook",
"William",
""
]
] |
2503.12313 | Yang Su | Yang Su, Shiyu Zhang, Yan Sun, Ji Yang, Fujun Du, Min Fang, Qing-Zeng
Yan, Shaobo Zhang, Zhiwei Chen, Xuepeng Chen, Xin Zhou, Lixia Yuan, and
Yuehui Ma | Gas Transfer Between the Inner 3-kpc Disk and the Galactic Central
Molecular Zone | Published in the ApJ, 984, 109 (2025) | ApJ, 984, 109 (2025) | 10.3847/1538-4357/adc38e | null | astro-ph.GA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We uncovered a more tilted molecular gas structure with highly negative
velocities located near the dust lane. Our observations also show that the
approaching gas flows from the overshoot process are captured by the bar
gravitational and then flows towards the Galactic central molecular zone (CMZ)
through the bar channel. The recycling gas from the overshoot effect, in
conjunction with freshly accreted gas from the inner 3-kpc disk, accumulates
significantly near R_GC~1/2R_bar and R_GC~2/3R_bar regions by adopting a bar
length of ~3.2--3.4kpc. Importantly, within these regions, there are frequent
collisions and substantial angular momentum exchanges between gas flows with
different trajectories. In this scenario, the DISSIPATION processes arising
from interactions between colliding flows, together with the varying torques
induced by the nonaxisymmetric bar, effectively transfer the angular momentum
of viscous gas outward, thereby driving the molecular gas to settle into the
CMZ within ~3 orbital periods. A long-term gas inflow with an average rate of
>1.1Msun/yr, coupled with intense transient accretion events that exceed the
average rate by several times due to the overshoot effect, significantly
regulates the gas distribution, physical properties, and dynamical evolution of
the CMZ. These findings provide robust observational evidence for elucidating
the intricate dynamics of molecular gas flows towards the CMZ. Our results show
that gas dynamics has a significant impact on the secular evolution of both the
Milky Way and the extragalactic gas-rich galaxies.
| [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 01:41:20 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 02:57:13 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Su",
"Yang",
""
],
[
"Zhang",
"Shiyu",
""
],
[
"Sun",
"Yan",
""
],
[
"Yang",
"Ji",
""
],
[
"Du",
"Fujun",
""
],
[
"Fang",
"Min",
""
],
[
"Yan",
"Qing-Zeng",
""
],
[
"Zhang",
"Shaobo",
""
],
[
"Chen",
"Zhiwei",
""
],
[
"Chen",
"Xuepeng",
""
],
[
"Zhou",
"Xin",
""
],
[
"Yuan",
"Lixia",
""
],
[
"Ma",
"Yuehui",
""
]
] |
2503.12850 | Alessandro Catania | Alessandro Catania, Michele Bertozzi, Nikita J. Greenidge, Benjamin
Calme, Gabriele Bandini, Christian Sbrana, Roberto Cecchi, Alice Buffi,
Massimo Macucci, Sebastiano Strangio, Pietro Valdastri, Giuseppe Iannaccone | In vivo validation of Wireless Power Transfer System for Magnetically
Controlled Robotic Capsule Endoscopy | 9 pages, 9 figures, regular paper | null | null | null | cs.RO cs.SY eess.SY physics.med-ph | http://creativecommons.org/licenses/by/4.0/ | This paper presents the in vivo validation of an inductive wireless power
transfer (WPT) system integrated for the first time into a magnetically
controlled robotic capsule endoscopy platform. The proposed system enables
continuous power delivery to the capsule without the need for onboard
batteries, thus extending operational time and reducing size constraints. The
WPT system operates through a resonant inductive coupling mechanism, based on a
transmitting coil mounted on the end effector of a robotic arm that also houses
an external permanent magnet and a localization coil for precise capsule
manipulation. To ensure robust and stable power transmission in the presence of
coil misalignment and rotation, a 3D receiving coil is integrated within the
capsule. Additionally, a closed-loop adaptive control system, based on
load-shift keying (LSK) modulation, dynamically adjusts the transmitted power
to optimize efficiency while maintaining compliance with specific absorption
rate (SAR) safety limits. The system has been extensively characterized in
laboratory settings and validated through in vivo experiments using a porcine
model, demonstrating reliable power transfer and effective robotic navigation
in realistic gastrointestinal conditions: the average received power was 110 mW
at a distance of 9 cm between the coils, with variable capsule rotation angles.
The results confirm the feasibility of the proposed WPT approach for
autonomous, battery-free robotic capsule endoscopy, paving the way for enhanced
diagnostic in gastrointestinal medicine.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 06:03:49 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 13:48:35 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Catania",
"Alessandro",
""
],
[
"Bertozzi",
"Michele",
""
],
[
"Greenidge",
"Nikita J.",
""
],
[
"Calme",
"Benjamin",
""
],
[
"Bandini",
"Gabriele",
""
],
[
"Sbrana",
"Christian",
""
],
[
"Cecchi",
"Roberto",
""
],
[
"Buffi",
"Alice",
""
],
[
"Macucci",
"Massimo",
""
],
[
"Strangio",
"Sebastiano",
""
],
[
"Valdastri",
"Pietro",
""
],
[
"Iannaccone",
"Giuseppe",
""
]
] |
2503.13955 | Bing-Sui Lu | Bing-Sui Lu | Nonresonant Casimir-Polder repulsion with a monolayer topological
insulator | Invited contribution to EPL's Focus Issue on "Casimir Effect and Its
Role in Modern Physics"; 7 pages, 4 figures | null | 10.1209/0295-5075/add36e | null | cond-mat.mes-hall cond-mat.mtrl-sci quant-ph | http://creativecommons.org/licenses/by/4.0/ | We investigate the behavior of the nonresonant Casimir-Polder force acting on
a metastable polarized state of a two-level atomic system with a
right-circularly polarized electric dipole transition in the presence of a
monolayer topological insulator inhabiting an anomalous quantum Hall insulator
state with a negative Chern number, finding that the force can be repulsive for
a certain range within the far-field region. By considering stanene irradiated
by circularly polarized monochromatic laser light as a material example of a
monolayer topological insulator, we study the effect of the frequency
dispersion in the conductivity tensor on the force behavior, finding that the
dispersion generally leads to a slight downward shift of the force relative to
that predicted by a nondispersive approximation, as well as a slight reduction
of the range over which the force is repulsive.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 06:47:48 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 00:31:14 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Lu",
"Bing-Sui",
""
]
] |
2503.14179 | Hisaaki Shinkai | Rika Shimomura, Yuuichi Tabe, Hisaaki Shinkai | Gravitational-wave Extraction using Independent Component Analysis | 19 pages, 7 figures. Revised version (mainly section 2), added
Appendix | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Independent component analysis (ICA) is a method to extract a set of
time-series data using ``statistical independency" of each component. We
applied ICA to extract gravitational wave (GW) signals directly from the
detector data. Our idea is to extract a coherent signal that is included in
multiple detectors and find it by shifting the data set around its arrival
time. In this article, we report several tests using injected signals, and show
that this method works for inspiral-wave events with a signal-to-noise ratio of
$>$ 15. We then applied the method to actual LIGO-Virgo-KAGRA O1-O3 events, and
showed that the identification of the arrival time can be estimated more
precisely than previously reported. This approach does not require templates of
waveform, therefore it is attractive for testing theories of gravity, and for
finding unknown GW.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 11:54:50 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 07:28:43 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Shimomura",
"Rika",
""
],
[
"Tabe",
"Yuuichi",
""
],
[
"Shinkai",
"Hisaaki",
""
]
] |
2503.14594 | Banibrata Mukhopadhyay | Mayusree Das (IISc), Armen Sedrakian (UWr, FIAS), Banibrata
Mukhopadhyay (IISc) | Superconductivity in magnetars: Exploring type-I and type-II states in
toroidal magnetic fields | 7 pages including 5 figures; version published in Physical Review D
Letters | Physical Review D 111, L081307 (2025) | 10.1103/PhysRevD.111.L081307 | null | astro-ph.SR astro-ph.HE gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a first two-dimensional general-relativistic analysis of
superconducting regions in axially symmetric highly magnetized neutron star
(magnetar) models with toroidal magnetic fields. We investigate the topology
and distribution of type-II and type-I superconducting regions for varying
toroidal magnetic field strengths and stellar masses by solving the
Einstein-Maxwell equations using the XNS code. Our results reveal that the
outer cores of low- to intermediate-mass magnetars sustain superconductivity
over larger regions compared to higher-mass stars with non-trivial distribution
of type-II and type-I regions. Consistent with previous one-dimensional (1D)
models, we find that regardless of the gravitational mass, the inner cores of
magnetars with toroidal magnetic fields are devoid of $S$-wave proton
superconductivity. Furthermore, these models contain non-superconducting,
torus-shaped regions - a novel feature absent in previous 1D studies. Finally,
we speculate on the potential indirect effects of superconductivity on
continuous gravitational wave emissions from millisecond pulsars, such as PSR
J1843-1113, highlighting their relevance for future gravitational wave
detectors.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 18:00:07 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 12:23:10 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Das",
"Mayusree",
"",
"IISc"
],
[
"Sedrakian",
"Armen",
"",
"UWr, FIAS"
],
[
"Mukhopadhyay",
"Banibrata",
"",
"IISc"
]
] |
2503.14634 | Xanda Kolesnikow | Xanda C Kolesnikow and Thomas B Smith and Felix Thomsen and Abhijeet
Alase and Andrew C Doherty | Protected phase gate for the $0$-$\pi$ qubit using its internal modes | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Protected superconducting qubits such as the $0$-$\pi$ qubit promise to
substantially reduce physical error rates through a multi-mode encoding. This
protection comes at the cost of controllability, as standard techniques for
quantum gates are ineffective. We propose a protected phase gate for the
$0$-$\pi$ qubit that utilises an internal mode of the circuit as an ancilla.
The gate is achieved by varying the qubit-ancilla coupling via a tunable
Josephson element. Our scheme is a modified version of a protected gate
proposed by Brooks, Kitaev and Preskill that uses an external oscillator as an
ancilla. We find that our scheme is compatible with the protected regime of the
$0$-$\pi$ qubit, and does not suffer from spurious coupling to additional modes
of the $0$-$\pi$ circuit. Through numerical simulations, we study how the gate
error scales with the circuit parameters of the $0$-$\pi$ qubit and the tunable
Josephson element that enacts the gate.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 18:35:38 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 01:29:02 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Kolesnikow",
"Xanda C",
""
],
[
"Smith",
"Thomas B",
""
],
[
"Thomsen",
"Felix",
""
],
[
"Alase",
"Abhijeet",
""
],
[
"Doherty",
"Andrew C",
""
]
] |
2503.16702 | Abdou Oussama Benabida | Abdou Oussama Benabida | Asymptotics for resolutions and smoothings of Calabi-Yau conifolds | 45 pages. Minor changes | null | null | null | math.DG math-ph math.AG math.AP math.MP | http://creativecommons.org/licenses/by/4.0/ | We show that the Calabi-Yau metrics with isolated conical singularities of
Hein-Sun admit polyhomogeneous expansions near their singularities. Moreover,
we show that, under certain generic assumptions, natural families of smooth
Calabi-Yau metrics on crepant resolutions and on polarized smoothings of
conical Calabi-Yau manifolds degenerating to the initial conical Calabi-Yau
metric admit polyhomogeneous expansions where the singularities are forming.
The construction proceeds by performing weighted Melrose-type blow-ups and then
gluing conical and scaled asymptotically conical Calabi-Yau metrics on the
fibers, close to the blow-up's front face without compromising polyhomogeneity.
This yields a polyhomogeneous family of K\"ahler metrics that are approximately
Calabi-Yau. Solving formally a complex Monge-Amp\`ere equation, we obtain a
polyhomogeneous family of K\"ahler metrics with Ricci potential converging
rapidly to zero as the family is degenerating. We can then conclude that the
corresponding family of degenerating Calabi-Yau metrics is polyhomogeneous by
using a fixed point argument.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 20:51:23 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 18:16:27 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Benabida",
"Abdou Oussama",
""
]
] |
2503.16831 | Tigran V. Shahbazyan | Tigran V. Shahbazyan | Non-Lorentzian model for strong exciton-plasmon coupling | 10 pages, 4 figures | null | null | null | cond-mat.mes-hall physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a non-Lorentzian analytical model for quantum emitters (QE)
resonantly coupled to localized surface plasmons (LSP) in metal-dielectric
structures. Using the explicit form of LSP Green function, we derive
non-Lorentzian version of semiclassical Maxwell-Bloch equations that describe
LSPs directly in terms of metal complex dielectric function rather than via
Lorentzian resonances. For a single QE resonantly coupled to an LSP, we obtain
an analytical expression for effective optical polarizability of the hybrid
system which, in the Lorentzian approximation, recovers the results of the
classical coupled oscillators model. We demonstrate that non-Lorentzian effects
originating from temporal dispersion of the metal dielectric function affect
significantly the optical spectra as the hybrid system transitions to the
strong coupling regime. Specifically, in contrast to results of Lorentzian
models, the main spectral weight in the system scattering spectra is shifted
toward the lower energy polaritonic band, consistent with the experiment.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 03:56:58 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 20:19:41 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Shahbazyan",
"Tigran V.",
""
]
] |
2503.17003 | Jian Guan | Jian Guan, Junfei Wu, Jia-Nan Li, Chuanqi Cheng, Wei Wu | A Survey on Personalized Alignment -- The Missing Piece for Large
Language Models in Real-World Applications | Survey paper; 11 pages; Literature reviewed up to ICLR 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have demonstrated remarkable capabilities, yet
their transition to real-world applications reveals a critical limitation: the
inability to adapt to individual preferences while maintaining alignment with
universal human values. Current alignment techniques adopt a one-size-fits-all
approach that fails to accommodate users' diverse backgrounds and needs. This
paper presents the first comprehensive survey of personalized alignment-a
paradigm that enables LLMs to adapt their behavior within ethical boundaries
based on individual preferences. We propose a unified framework comprising
preference memory management, personalized generation, and feedback-based
alignment, systematically analyzing implementation approaches and evaluating
their effectiveness across various scenarios. By examining current techniques,
potential risks, and future challenges, this survey provides a structured
foundation for developing more adaptable and ethically-aligned LLMs.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 10:09:16 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Mar 2025 02:58:20 GMT"
},
{
"version": "v3",
"created": "Tue, 1 Apr 2025 09:33:19 GMT"
},
{
"version": "v4",
"created": "Mon, 5 May 2025 01:21:53 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Guan",
"Jian",
""
],
[
"Wu",
"Junfei",
""
],
[
"Li",
"Jia-Nan",
""
],
[
"Cheng",
"Chuanqi",
""
],
[
"Wu",
"Wei",
""
]
] |
2503.17611 | Jhixon Mac\'ias | Jhixon Mac\'ias | On continuous polynomials of the Mac\'ias space | null | null | null | null | math.GN | http://creativecommons.org/licenses/by/4.0/ | Let $\mathbb{N}$ be the set of natural numbers. The Mac\'ias space
$M(\mathbb{N})$ is the topological space $(\mathbb{N},\tau_M)$ where $\tau_M$
is generated by the collection of sets $\sigma_n := \{ m \in \mathbb{N} :
\gcd(n, m) = 1 \}$. In this paper, we characterize the continuity of
polynomials over $ M(\mathbb{N})$ and prove that the only continuous
polynomials are monomials
| [
{
"version": "v1",
"created": "Sat, 22 Mar 2025 02:06:11 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 02:02:41 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Macías",
"Jhixon",
""
]
] |
2503.18236 | Rohitash Chandra | Hardik A. Jain and Rohitash Chandra | Research impact evaluation based on effective authorship contribution
sensitivity: h-leadership index | null | null | null | null | cs.DL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The evaluation of a researcher's performance has traditionally relied on
various bibliometric measures, with the h-index being one of the most
prominent. However, the h-index only accounts for the number of citations
received in a publication and does not account for other factors such as the
number of authors or their specific contributions in collaborative works.
Therefore, the h-index has been placed on scrutiny as it has motivated academic
integrity issues where non-contributing authors get authorship merely for
raising their h-index. In this study, we comprehensively evaluate existing
metrics in their ability to account for authorship contribution by their
position and introduce a novel variant of the h-index, known as the
h-leadership index. The h-leadership index aims to advance the fair evaluation
of academic contributions in multi-authored publications by giving importance
to authorship position beyond the first and last authors, motivated by
Stanford's ranking of the top 2 \% of world scientists. We assign weighted
citations based on a modified complementary unit Gaussian curve, ensuring that
the contributions of middle authors are appropriately recognised. We apply the
h-leadership index to analyse the top 50 researchers across the Group of 8
(Go8) universities in Australia, demonstrating its potential to provide a more
balanced assessment of research performance. We provide open-source software
for extending the work further.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 23:06:50 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 23:21:46 GMT"
},
{
"version": "v3",
"created": "Sun, 27 Apr 2025 08:28:44 GMT"
},
{
"version": "v4",
"created": "Sat, 3 May 2025 09:49:40 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Jain",
"Hardik A.",
""
],
[
"Chandra",
"Rohitash",
""
]
] |
2503.18772 | Alberto Hijano | Alberto Hijano, Henri Lyyra, Juha T. Muhonen and Tero T. Heikkil\"a | Characterization of a quantum bus between two driven qubits | 13 pages, 5 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the use of driven qubits coupled to a harmonic oscillator to
implement a $\sqrt{i\mathrm{SWAP}}$-gate. By dressing the qubits through an
external driving field, the qubits and the harmonic oscillator can be
selectively coupled, leading to effective qubit-qubit interactions. We analyze
a qubit readout mechanism based on the detection of a shift of the harmonic
oscillator's resonance frequency, and demonstrate that when coupled to
low-frequency resonators, dressed qubits provide a more robust readout than
bare qubits in the presence of damping and thermal effects. Furthermore, we
study the impact of various system parameters on the fidelity of the two-qubit
gate, identifying an optimal range for quantum computation. Our findings guide
the implementation of high-fidelity quantum gates in experimental setups, for
example those employing nanoscale mechanical resonators.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 15:20:22 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:25:23 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 14:52:25 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Hijano",
"Alberto",
""
],
[
"Lyyra",
"Henri",
""
],
[
"Muhonen",
"Juha T.",
""
],
[
"Heikkilä",
"Tero T.",
""
]
] |
2503.19049 | Vyacheslav M. Abramov | Vyacheslav M. Abramov | Classes of functions in the set of circulant matrices and their
characterization | 10 pages | null | null | null | math.CA math.AC math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we introduce new classes of functions that extend the known
classes of functions of complex variable, such as entire functions, meromorphic
functions, rational functions and polynomial functions and take values in the
set of circulant matrices with complex entries. For these new classes of
functions, we extend the recently obtained characterization theorems in [Li, B.
Q. Two elementary properties of entire functions and their applications. Amer.
Math. Monthly 122(2) (2015), 169--172] and [Li, B. Q. A characterization of
rational functions. Amer. Math. Monthly. 132(3) (2025), 269--271] to an
algebraic structure of circulant matrices that includes several complex
variables. Our characterization theorems lead to the generalization of the
variant of the fundamental theorem of algebra established recently in [Abramov,
V. M. Extension of the fundamental theorem of algebra for polynomial matrix
equation with circulant matrices. Amer. Math. Monthly. DOI:
10.1080/00029890.2024.2440298].
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 18:20:58 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 09:12:04 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Abramov",
"Vyacheslav M.",
""
]
] |
2503.19071 | Karl Bergson Hallberg | Karl Bergson Hallberg, Erik Wegner Hodt, Jacob Linder | Visualization of spin-splitter effect in altermagnets via
non-equilibrium Green functions on a lattice | 14 pages, 13 figures | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | When a charge current is injected into an altermagnet along a suitable
crystallographic direction, a transverse spin current can be generated. This
so-called spin-splitter effect does not rely on spin-orbit coupling, and is
thus distinct from the spin Hall effect. The spin-splitter effect was predicted
by \textit{ab initio} calculations and has been experimentally confirmed. To
utilize the spin-splitter effect for practical purposes in spintronic devices,
it is important to understand (i) how the system parameters affect the
transverse spin current, such as filling fraction, altermagnetic strength,
interface parameters, spin-orbit interactions, and impurities and (ii)
determine the properties of any associated spin accumulation, which is the
measurable quantity. Here, we determine the answer to these questions and
provide a real-space visualization of the spin flow and spin accumulation due
to the spin-splitter effect. We utilize the non-equilibrium Keldysh Green
function method on a 2D square lattice to this end. We find that the presence
of edges induces oscillations in the spin accumulation and strongly modify the
signal for small samples. At half-filling, the spin accumulation acquires an
anomalous pattern and the spin-splitter effect vanishes. We prove analytically
that this follows from a combined particle-hole and spin-reversal symmetry of
the model used for the altermagnetic state. Increasing the altermagnetic
strength leads to a larger spin accumulation, as expected. However, when adding
Rashba spin-orbit interaction, providing an additional spin Hall signal, we
find that the spin accumulation is not simply the sum of the spin Hall and
spin-splitter contribution. Finally, we show that the spin-splitter effect is
robust towards moderate impurity scattering with a potential of the same order
as the hopping parameter, which facilitates its observation in real materials.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 19:00:00 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 12:38:49 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Hallberg",
"Karl Bergson",
""
],
[
"Hodt",
"Erik Wegner",
""
],
[
"Linder",
"Jacob",
""
]
] |
2503.19105 | Denis Grebenkov | Denis S. Grebenkov, Ralf Metzler, and Gleb Oshanin | Fastest first-passage time statistics for time-dependent particle
injection | null | null | null | null | cond-mat.stat-mech physics.bio-ph physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | A common scenario in a variety of biological systems is that multiple
particles are searching in parallel for an immobile target located in a bounded
domain, and the fastest among them that arrives to the target first triggers a
given desirable or detrimental process. The statistics of such extreme events
-- the \textit{fastest\/} first-passage to the target -- is well-understood by
now through a series of theoretical analyses, but exclusively under the
assumption that all $N$ particles start \textit{simultaneously\/}, i.e., all
are introduced into the domain instantly, by $\delta$-function-like pulses.
However, in many practically important situations this is not the case: in
order to start their search, the particles often have to enter first into a
bounded domain, e.g., a cell or its nucleus, penetrating through gated channels
or nuclear pores. This entrance process has a random duration so that the
particles appear in the domain sequentially and with a time delay. Here we
focus on the effect of such an extended-in-time injection of multiple particles
on the fastest first-passage time (fFPT) and its statistics. We derive the full
probability density function $H_N(t)$ of the fFPT with an arbitrary
time-dependent injection intensity of $N$ particles. Under rather general
assumptions on the survival probability of a single particle and on the
injection intensity, we derive the large-$N$ asymptotic formula for the mean
fFPT, which is quite different from that obtained for the instantaneous
$\delta$-pulse injection. The extended injection is also shown to considerably
slow down the convergence of $H_N(t)$ to the large-$N$ limit -- the Gumbel
distribution -- so that the latter may be inapplicable in the most relevant
settings with few tens to few thousands of particles.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 19:51:14 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 14:57:06 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Grebenkov",
"Denis S.",
""
],
[
"Metzler",
"Ralf",
""
],
[
"Oshanin",
"Gleb",
""
]
] |
2503.19343 | Victor Vassiliev | V.A.Vassiliev | Equilevel algebras | null | null | null | null | math.AG math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An equilevel algebra is a subalgebra of the space of smooth functions $f: M
\to {\mathbb R}$ distinguished in this space by finitely many linear conditions
of the type $f(x_i) = f(\tilde x_i)$, $x_i \neq \tilde x_i \in M$, or
approximated by such subalgebras. For $M = S^1$ or ${\mathbb R}^1$, the regular
points of the variety of all equilevel subalgebras of codimension $k$ in
$C^\infty(M, {\mathbb R})$ are known in knot theory as $k$-chord diagrams. The
whole of this variety completes the space of chord diagrams in the same way as
the spaces of ideals of given codimensions in complete the configuration
spaces. We describe cell structures of the varieties of all equilevel algebras
up to the codimension three in $C^\infty(S^1, {\mathbb R})$ and compute their
homology rings and characteristic classes of canonical vector bundles. We also
give lower estimates of these rings for arbitrary $k$.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 04:35:48 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 09:43:02 GMT"
},
{
"version": "v3",
"created": "Sun, 20 Apr 2025 03:07:21 GMT"
},
{
"version": "v4",
"created": "Sun, 27 Apr 2025 15:54:51 GMT"
},
{
"version": "v5",
"created": "Sat, 3 May 2025 11:41:47 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Vassiliev",
"V. A.",
""
]
] |
2503.20669 | Jos\'e Galindo Jim\'enez | Jos\'e Galindo Jim\'enez | Kulikov-Persson-Pinkham theorem via smoothing of dlt models | 12 pages original in English, 12 pages Catalan translation. V2:
Restructured due to a missing major reference being pointed out | null | null | null | math.CV math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give an alternative proof of the Kulikov-Persson-Pinkham Theorem for a
projective degeneration of K-trivial smooth surfaces. After running the Minimal
Model Program, the obtained minimal dlt model has mild singularities which we
resolve via Brieskorn's simultaneous resolutions and toric resolutions.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 16:02:07 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 15:26:10 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Jiménez",
"José Galindo",
""
]
] |
2503.20778 | Yun-Long Zhang | Hong Su, Baoyu Xu, Ju Chen, Chang Liu, Yun-Long Zhang | Detectability of the chiral gravitational wave background from audible
axions with the LISA-Taiji network | 15 pages, 10 figures | Commun. Theor. Phys.(2025) | 10.1088/1572-9494/add1b9 | null | astro-ph.CO gr-qc hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The chiral gravitational wave background (GWB) can be produced by axion-like
fields in the early universe. We perform parameter estimation for two types of
chiral GWB with the LISA-Taiji network: axion-dark photon coupling and
axion-Nieh-Yan coupling. We estimate the spectral parameters of these two
mechanisms induced by axion and determine the normalized model parameters using
the Fisher information matrix. For highly chiral GWB signals that we choose to
analyze in the mHz band, the normalized model parameters are constrained with a
relative error less than $6.7\%$ (dark photon coupling) and $2.2\%$ (Nieh-Yan
coupling) at the one-sigma confidence level. The circular polarization
parameters are constrained with a relative error around $21\%$ (dark photon
coupling) and $6.2\%$ (Nieh-Yan coupling) at the one-sigma confidence level.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 17:56:38 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Su",
"Hong",
""
],
[
"Xu",
"Baoyu",
""
],
[
"Chen",
"Ju",
""
],
[
"Liu",
"Chang",
""
],
[
"Zhang",
"Yun-Long",
""
]
] |
2503.21499 | Suman Pramanick | Suman Pramanick, Somnath Bharadwaj, Khandakar Md Asif Elahi and Rajesh
Mondal | The Evolving Power Spectrum for the light cone Epoch of Reionization
(EoR) 21-cm signal | 12 pages, 9 figures, Accepted for publication in ApJ | The Astrophysical Journal, Volume 984, Number 2, 2025 | 10.3847/1538-4357/adc678 | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid evolution of the cosmological neutral hydrogen (HI) distribution
during the EoR is imprinted along the line of sight (LoS) in the redshifted
21-cm signal due to the light cone (LC) effect. The LC EoR 21-cm signal ceases
to be ergodic along the LoS, and the Fourier transform-based three-dimensional
power spectrum (PS) fails to capture the full two-point statistics. Several
earlier studies have used the multi-frequency angular power spectrum (MAPS)
$\mathcal{C}_\ell(\nu_1,\nu_2)$ to overcome this limitation. However, we do not
have a simple interpretation of $\mathcal{C}_\ell(\nu_1,\nu_2)$ in terms of
comoving length scale, and the data volume is large. Here we introduce the
evolving power spectrum (ePS) to quantify the two-point statistics of the LC
EoR 21-cm signal. This has a simple interpretation in terms of redshift
evolution and comoving length scales, and the binned ePS reduces the data
volume by several orders of magnitude compared to MAPS. Considering
simulations, we study the first three even angular multipoles of ePS to
quantify the LoS anisotropy of the signal. We find that as reionization
progresses, at large $k$ ($ \ge 0.6 \, {\rm Mpc}^{-1}$), $P_{e\,0}(k,z)$ the
monopole moment decreases as $\propto \bar{x}_{\rm H I}$ the mean neutral HI
fraction, which, in principle, can be used to observationally determine the
reionization history. Furthermore, $P_{e\,2}(k,z)$ the quadrupole moment is
negative at small $k$ and positive at large $k$. We propose the binned ePS,
which captures the entire information contained in MAPS, to quantify the full
two-point statistics of the LC EoR 21-cm signal.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 13:37:16 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Pramanick",
"Suman",
""
],
[
"Bharadwaj",
"Somnath",
""
],
[
"Elahi",
"Khandakar Md Asif",
""
],
[
"Mondal",
"Rajesh",
""
]
] |
2503.22602 | Francesco Marzioni | Francesco Marzioni, Riccardo Natali, Michele Bonaldi, Antonio
Borrielli, Enrico Serra, Bruno Morana, Francesco Marin, Francesco Marino,
Nicola Malossi, David Vitali, Giovanni Di Giuseppe, Paolo Piergentili | Mechanical characterization of a membrane with an on-chip loss shield in
a cryogenic environment | 7 pages, 4 figures. In this version some typos have been corrected
and the text of the manuscript has been improved | Applied Physics Letters, 28 April 2025, 126 (17): 174002 | 10.1063/5.0268805 | null | quant-ph physics.app-ph physics.optics | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The quantum transduction of an rf/microwave signal to the optical domain, and
vice versa, paves the way for technologies that exploit the advantages of each
domain to perform quantum operations. Since electro-optomechanical devices
implement a simultaneous coupling of a mechanical oscillator to both an
rf/microwave field and an optical field, they are suitable for the realization
of a quantum transducer. The membrane-in-the-middle setup is a possible
solution, once its vibrational mode is cooled down to ultra cryogenic
temperature for achieving quantum operation. This work is focused on the
mechanical characterization via an optical interferometric probe, down to
T=18mK, of a loss-shielded metalized membrane designed for this purpose. A
stroboscopic technique has been exploited for revealing a mechanical quality
factor up to 64 millions at the lowest temperature. In fact, with continuous
illumination and a cryostat temperature below 1K, the heat due to optical
absorption is not efficiently dissipated anymore, and the membrane remains
hotter than its environment.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 16:45:02 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 18:24:00 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Marzioni",
"Francesco",
""
],
[
"Natali",
"Riccardo",
""
],
[
"Bonaldi",
"Michele",
""
],
[
"Borrielli",
"Antonio",
""
],
[
"Serra",
"Enrico",
""
],
[
"Morana",
"Bruno",
""
],
[
"Marin",
"Francesco",
""
],
[
"Marino",
"Francesco",
""
],
[
"Malossi",
"Nicola",
""
],
[
"Vitali",
"David",
""
],
[
"Di Giuseppe",
"Giovanni",
""
],
[
"Piergentili",
"Paolo",
""
]
] |
2503.22707 | Amit Maji | Kritika Babbar and Amit Maji | Power partial isometries | 28 pages. Revised. Some new references added | null | null | null | math.FA math.CV math.OA | http://creativecommons.org/licenses/by/4.0/ | In this paper we obtain a complete characterization of reducing, invariant,
and hyperinvariant subspaces for the completely non-unitary component of a
power partial isometry. In particular, precise characterization of reducing,
invariant, and hyperinvariant subspaces of a truncated shift operator has been
achieved.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 13:27:31 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 11:51:13 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Babbar",
"Kritika",
""
],
[
"Maji",
"Amit",
""
]
] |
2503.23211 | Sepideh Mosaferi | Sepideh Mosaferi, Abolfazl Safikhani, Peiliang Bai | Optimal Change Point Detection and Inference in the Spectral Density of
General Time Series Models | Fixed minor typos in V1 | null | null | null | stat.ME math.ST stat.AP stat.TH | http://creativecommons.org/licenses/by/4.0/ | This paper addresses the problem of detecting change points in the spectral
density of time series, motivated by EEG analysis of seizure patients. Seizures
disrupt coherence and functional connectivity, necessitating precise detection.
Departing from traditional parametric approaches, we utilize the Wold
decomposition, representing general time series as autoregressive processes
with infinite lags, which are truncated and estimated around the change point.
Our detection procedure employs an initial estimator that systematically
searches across time points. We examine the localization error and its
dependence on time series properties and sample size. To enhance accuracy, we
introduce an optimal rate method with an asymptotic distribution, facilitating
the construction of confidence intervals. The proposed method effectively
identifies seizure onset in EEG data and extends to event detection in video
data. Comprehensive numerical experiments demonstrate its superior performance
compared to existing techniques.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 20:19:39 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 12:38:05 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Mosaferi",
"Sepideh",
""
],
[
"Safikhani",
"Abolfazl",
""
],
[
"Bai",
"Peiliang",
""
]
] |
2503.23236 | Ismael Zighed | Isma\"el Zighed, Nicolas Thome, Patrick Gallinari and Taraneh Sayadi | UP-dROM : Uncertainty-Aware and Parametrised dynamic Reduced-Order
Model, application to unsteady flows | null | null | null | null | cs.LG physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Reduced order models (ROMs) play a critical role in fluid mechanics by
providing low-cost predictions, making them an attractive tool for engineering
applications. However, for ROMs to be widely applicable, they must not only
generalise well across different regimes, but also provide a measure of
confidence in their predictions. While recent data-driven approaches have begun
to address nonlinear reduction techniques to improve predictions in transient
environments, challenges remain in terms of robustness and parametrisation. In
this work, we present a nonlinear reduction strategy specifically designed for
transient flows that incorporates parametrisation and uncertainty
quantification. Our reduction strategy features a variational auto-encoder
(VAE) that uses variational inference for confidence measurement. We use a
latent space transformer that incorporates recent advances in attention
mechanisms to predict dynamical systems. Attention's versatility in learning
sequences and capturing their dependence on external parameters enhances
generalisation across a wide range of dynamics. Prediction, coupled with
confidence, enables more informed decision making and addresses the need for
more robust models. In addition, this confidence is used to cost-effectively
sample the parameter space, improving model performance a priori across the
entire parameter space without requiring evaluation data for the entire domain.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 22:17:36 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Apr 2025 15:51:44 GMT"
},
{
"version": "v3",
"created": "Sat, 3 May 2025 10:20:59 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Zighed",
"Ismaël",
""
],
[
"Thome",
"Nicolas",
""
],
[
"Gallinari",
"Patrick",
""
],
[
"Sayadi",
"Taraneh",
""
]
] |
2503.23418 | Fenton Doolan Mr | Fenton John Doolan | The Two-component Model of the 'spokes' in Saturn's Rings | 43 pages, 15 figures , 13 tables | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | The spokes observed in Saturn's rings have been a subject of scientific
debate since their discovery by Stephen J. O'Meara in the 1970s and their
confirmation by the Voyager flybys in the early 1980s (Smith et al., 1982).
While the Cassini spacecraft confirmed that the spokes are linked to Saturn's
magnetosphere, their exact formation mechanism remains uncertain. This paper
proposes that the spokes in Saturn's rings consist of two distinct components:
(1) carbonaceous materials, namely pyrolytic carbon with diamagnetic properties
and potentially other various forms of carbon-bearing compounds, which persist
over longer timescales, and (2) rapidly forming and dissipating diamagnetic ice
grains, which interact with Saturn's magnetosphere on much shorter timescales.
We suggest that the spokes consist of diamagnetic pyrolytic carbon that has
coated silicates through the process of high-temperature Chemical Vapour
Deposition (CVD) during the formation of Saturn's protoplanetary disk.
Additionally, the spokes consist of diamagnetic ice particles that can
disappear within minutes to hours due to sublimation. The photoelectric effect
causes the pyrolytic carbon grains to lose electrons, thus becoming
paramagnetic, which results in their attraction back to the main B ring
structure. We suggest that Saturn's rings are charged by the solar wind,
generating a magnetic field that emanates orthogonally above and below the B
ring plane due to the movement of charged particles in the rings. Statistical
analysis of Cassini data reveals significant correlations between spoke
activity and both Saturn's magnetospheric rotation and solar elevation angle,
providing strong support for an electromagnetic mechanism in spoke formation.
This hypothesis suggests that the visibility of the spokes is dictated by
illumination conditions and electromagnetic effects.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 12:21:24 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 14:15:22 GMT"
},
{
"version": "v3",
"created": "Tue, 8 Apr 2025 09:21:06 GMT"
},
{
"version": "v4",
"created": "Mon, 5 May 2025 12:41:54 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Doolan",
"Fenton John",
""
]
] |
2503.23436 | Sheng Lu | Sheng Lu and Mingxi Ge and Jiuyi Zhang and Wanli Zhu and Guanjin Li
and Fangming Gu | Filtering with Time-frequency Analysis: An Adaptive and Lightweight
Model for Sequential Recommender Systems Based on Discrete Wavelet Transform | 17pages, accepted by ICIC 2025 oral | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential Recommender Systems (SRS) aim to model sequential behaviors of
users to capture their interests which usually evolve over time.
Transformer-based SRS have achieved distinguished successes recently. However,
studies reveal self-attention mechanism in Transformer-based models is
essentially a low-pass filter and ignores high frequency information
potentially including meaningful user interest patterns. This motivates us to
seek better filtering technologies for SRS, and finally we find Discrete
Wavelet Transform (DWT), a famous time-frequency analysis technique from
digital signal processing field, can effectively process both low-frequency and
high-frequency information. We design an adaptive time-frequency filter with
DWT technique, which decomposes user interests into multiple signals with
different frequency and time, and can automatically learn weights of these
signals. Furthermore, we develop DWTRec, a model for sequential recommendation
all based on the adaptive time-frequency filter. Thanks to fast DWT technique,
DWTRec has a lower time complexity and space complexity theoretically, and is
Proficient in modeling long sequences. Experiments show that our model
outperforms state-of-the-art baseline models in datasets with different
domains, sparsity levels and average sequence lengths. Especially, our model
shows great performance increase in contrast with previous models when the
sequence grows longer, which demonstrates another advantage of our model.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 13:28:42 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Apr 2025 04:11:54 GMT"
},
{
"version": "v3",
"created": "Wed, 30 Apr 2025 04:04:16 GMT"
},
{
"version": "v4",
"created": "Thu, 1 May 2025 06:04:46 GMT"
},
{
"version": "v5",
"created": "Sun, 4 May 2025 05:03:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Lu",
"Sheng",
""
],
[
"Ge",
"Mingxi",
""
],
[
"Zhang",
"Jiuyi",
""
],
[
"Zhu",
"Wanli",
""
],
[
"Li",
"Guanjin",
""
],
[
"Gu",
"Fangming",
""
]
] |
2503.23792 | Takeshi Fukasawa | Takeshi Fukasawa | When do firms sell high durability products? The case of light bulb
industry | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | This study empirically investigates firms' incentives on the choice of
product durability, and its social optimality, by developing a dynamic
structural model of durable goods with forward-looking consumers and
oligopolistic multi-product firms. Based on the observations of the light bulb
market, it specifies a model where firms produce multiple products with
different durability levels and set product prices based on dynamic incentives.
It proposes and applies novel estimation algorithms that alleviate the
computational burden and data requirement for estimating demand and marginal
cost parameters of dynamic demand models. Using light bulb market data in
Japan, structural parameters are estimated. This study obtains the following
results. First, large firms have incentives to collude to eliminate high
durability incandescent lamps, though it is profitable to sell them for each
firm. In contrast, when they can collude on prices, they don't have incentives
to eliminate high durability bulbs. Second, eliminating high durability
incandescent lamps leads to larger producer and total surplus, though it leads
to lower consumer surplus.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 07:13:44 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 00:50:49 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Fukasawa",
"Takeshi",
""
]
] |
2503.24235 | Fuyuan Lyu | Qiyuan Zhang, Fuyuan Lyu, Zexu Sun, Lei Wang, Weixu Zhang, Wenyue Hua,
Haolun Wu, Zhihan Guo, Yufei Wang, Niklas Muennighoff, Irwin King, Xue Liu,
Chen Ma | A Survey on Test-Time Scaling in Large Language Models: What, How,
Where, and How Well? | v3: Expand Agentic and SFT Chapters. Build Website for better
visualization | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As enthusiasm for scaling computation (data and parameters) in the
pretraining era gradually diminished, test-time scaling (TTS), also referred to
as ``test-time computing'' has emerged as a prominent research focus. Recent
studies demonstrate that TTS can further elicit the problem-solving
capabilities of large language models (LLMs), enabling significant
breakthroughs not only in specialized reasoning tasks, such as mathematics and
coding, but also in general tasks like open-ended Q&A. However, despite the
explosion of recent efforts in this area, there remains an urgent need for a
comprehensive survey offering a systemic understanding. To fill this gap, we
propose a unified, multidimensional framework structured along four core
dimensions of TTS research: what to scale, how to scale, where to scale, and
how well to scale. Building upon this taxonomy, we conduct an extensive review
of methods, application scenarios, and assessment aspects, and present an
organized decomposition that highlights the unique functional roles of
individual techniques within the broader TTS landscape. From this analysis, we
distill the major developmental trajectories of TTS to date and offer hands-on
guidelines for practical deployment. Furthermore, we identify several open
challenges and offer insights into promising future directions, including
further scaling, clarifying the functional essence of techniques, generalizing
to more tasks, and more attributions. Our repository is available on
https://github.com/testtimescaling/testtimescaling.github.io/
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 15:46:15 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 04:32:24 GMT"
},
{
"version": "v3",
"created": "Sun, 4 May 2025 15:48:08 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Zhang",
"Qiyuan",
""
],
[
"Lyu",
"Fuyuan",
""
],
[
"Sun",
"Zexu",
""
],
[
"Wang",
"Lei",
""
],
[
"Zhang",
"Weixu",
""
],
[
"Hua",
"Wenyue",
""
],
[
"Wu",
"Haolun",
""
],
[
"Guo",
"Zhihan",
""
],
[
"Wang",
"Yufei",
""
],
[
"Muennighoff",
"Niklas",
""
],
[
"King",
"Irwin",
""
],
[
"Liu",
"Xue",
""
],
[
"Ma",
"Chen",
""
]
] |
2504.00099 | Diptarka Das | Diptarka Das, Sumit R. Das, Arnab Kundu, Krishnendu Sengupta | Dynamical Phases of Higher Dimensional Floquet CFTs | 46 pages, 6 figures, references and some discussions updated | null | null | null | hep-th cond-mat.stat-mech cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the dynamical phases of Floquet Conformal Field
Theories (CFTs) in space-time dimensions greater than two. Building upon our
previous work [1] which introduced quaternionic representations for studying
Floquet dynamics in higher-dimensional CFTs, we now explore more general square
pulse drive protocols that go beyond a single SU(1,1) subgroup. We demonstrate
that, for multi-step drive protocols, the system exhibits distinct dynamical
phases characterized by the nature of the eigenvalues of the quaternionic
matrix representing time evolution in a single cycle, leading to different
stroboscopic responses. Our analysis establishes a fundamental geometric
interpretation where these dynamical phases directly correspond to the presence
or absence of Killing horizons in the base space of the CFT and in a higher
dimensional AdS space on which a putative dual lives. The heating phase is
associated with a non-extremal horizon, the critical phase with an extremal
horizon which disappears in the non heating phase. We develop perturbative
approaches to compute the Floquet Hamiltonians in different regimes and show,
how tuning drive parameters can lead to horizons, providing a geometric
framework for understanding heating phenomena in driven conformal systems.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 18:00:12 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 10:19:17 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Das",
"Diptarka",
""
],
[
"Das",
"Sumit R.",
""
],
[
"Kundu",
"Arnab",
""
],
[
"Sengupta",
"Krishnendu",
""
]
] |
2504.00159 | Advaith Venkatramanan Sethuraman | Advaith V. Sethuraman, Max Rucker, Onur Bagoren, Pou-Chun Kung,
Nibarkavi N.B. Amutha, Katherine A. Skinner | SonarSplat: Novel View Synthesis of Imaging Sonar via Gaussian Splatting | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we present SonarSplat, a novel Gaussian splatting framework
for imaging sonar that demonstrates realistic novel view synthesis and models
acoustic streaking phenomena. Our method represents the scene as a set of 3D
Gaussians with acoustic reflectance and saturation properties. We develop a
novel method to efficiently rasterize Gaussians to produce a range/azimuth
image that is faithful to the acoustic image formation model of imaging sonar.
In particular, we develop a novel approach to model azimuth streaking in a
Gaussian splatting framework. We evaluate SonarSplat using real-world datasets
of sonar images collected from an underwater robotic platform in a controlled
test tank and in a real-world river environment. Compared to the
state-of-the-art, SonarSplat offers improved image synthesis capabilities (+3.2
dB PSNR) and more accurate 3D reconstruction (52% lower Chamfer Distance). We
also demonstrate that SonarSplat can be leveraged for azimuth streak removal.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 19:13:45 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 20:20:33 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Sethuraman",
"Advaith V.",
""
],
[
"Rucker",
"Max",
""
],
[
"Bagoren",
"Onur",
""
],
[
"Kung",
"Pou-Chun",
""
],
[
"Amutha",
"Nibarkavi N. B.",
""
],
[
"Skinner",
"Katherine A.",
""
]
] |
2504.00418 | Naoka Karube | Naoka Karube and Yasuhiro Wakabayashi | The moduli space of dormant opers on elliptic curves | 40 pages | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A dormant oper is a specific type of principal bundle with a flat connection,
defined on an algebraic curve in positive characteristic. The moduli spaces of
dormant opers and their variants, known as dormant Miura opers, have been
studied in various contexts. This paper focuses on the case where the
underlying spaces are (possibly nodal) elliptic curves and provides a detailed
examination of the geometric structures of their moduli spaces. In particular,
we explicitly describe dormant (generic Miura) opers in terms of regular
elements in an associated Lie algebra and establish the connectedness of these
moduli spaces. We also explore generalizations to higher level and prime-power
characteristic.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 04:48:20 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 01:29:16 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Karube",
"Naoka",
""
],
[
"Wakabayashi",
"Yasuhiro",
""
]
] |
2504.00524 | Oliver Solow | Oliver Solow, Karsten Flensberg | Signatures of exceptional points in multiterminal superconductor-normal
metal junctions | 5+3 pages, 5 figures | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a non-Hermitian, multiterminal superconducting-normal system in
order to identify experimental signatures of exceptional points. We focus on a
minimal setting with a single spinful level, spin-dependent normal leads, and a
noncollinear magnetic field. This system hosts both topologically-protected, as
well as symmetry-protected exceptional points. Using an exact transport
formalism, we show that the exceptional points are visible through spectroscopy
of the Andreev states, but that they have a minor effect on the Josephson
current. We also argue that these findings hold with interactions.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 08:17:29 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 13:38:40 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Solow",
"Oliver",
""
],
[
"Flensberg",
"Karsten",
""
]
] |
2504.00548 | Jasmine Vrancken | Hugues Sana and Jasmine Vrancken | Observing Binaries | This is a pre-print of a chapter for the Encyclopedia of Astrophysics
(edited by I. Mandel, section editor F.R.N. Schneider) to be published by
Elsevier as a Reference Module | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Binary stars are dynamical systems formed by two stars that are physically
bound by the gravitational force. Binary stars are privileged laboratories,
allowing one to measure the fundamental properties of stars but also
potentially changing the way stars live and die. Because of this, binary stars
have continuously played a central role in astrophysics. In this chapter, we
focus on the observational properties of binary stars. What are the fundamental
quantities that describe binaries? How do we detect and classify them? Which
parameters can we constrain for different type of objects?
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 08:54:26 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 14:37:46 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Sana",
"Hugues",
""
],
[
"Vrancken",
"Jasmine",
""
]
] |
2504.00711 | Enjun Du | Enjun Du, Xunkai Li, Tian Jin, Zhihan Zhang, Rong-Hua Li, and Guoren
Wang | GraphMaster: Automated Graph Synthesis via LLM Agents in Data-Limited
Environments | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The era of foundation models has revolutionized AI research, yet Graph
Foundation Models (GFMs) remain constrained by the scarcity of large-scale
graph corpora. Traditional graph data synthesis techniques primarily focus on
simplistic structural operations, lacking the capacity to generate semantically
rich nodes with meaningful textual attributes: a critical limitation for
real-world applications. While large language models (LLMs) demonstrate
exceptional text generation capabilities, their direct application to graph
synthesis is impeded by context window limitations, hallucination phenomena,
and structural consistency challenges. To address these issues, we introduce
GraphMaster, the first multi-agent framework specifically designed for graph
data synthesis in data-limited environments. GraphMaster orchestrates four
specialized LLM agents (Manager, Perception, Enhancement, and Evaluation) that
collaboratively optimize the synthesis process through iterative refinement,
ensuring both semantic coherence and structural integrity. To rigorously
evaluate our approach, we create new data-limited "Sub" variants of six
standard graph benchmarks, specifically designed to test synthesis capabilities
under realistic constraints. Additionally, we develop a novel interpretability
assessment framework that combines human evaluation with a principled
Grassmannian manifold-based analysis, providing both qualitative and
quantitative measures of semantic coherence. Experimental results demonstrate
that GraphMaster significantly outperforms traditional synthesis methods across
multiple datasets, establishing a strong foundation for advancing GFMs in
data-scarce environments.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 12:21:50 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 13:57:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Du",
"Enjun",
""
],
[
"Li",
"Xunkai",
""
],
[
"Jin",
"Tian",
""
],
[
"Zhang",
"Zhihan",
""
],
[
"Li",
"Rong-Hua",
""
],
[
"Wang",
"Guoren",
""
]
] |
2504.00782 | Rafael Santos | Rafael M. Santos, Rafael C. Nunes, Jose C. N. de Araujo | Constraining Scalar Charge in Neutron Stars through Gravitational Wave
Signals from NSBH Mergers | 9 pages, 5 figures | null | null | null | gr-qc astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Neutron stars (NS) offer an exceptional opportunity to investigate
gravitational physics in extreme environments. In the context of alternative
theories of General Relativity (GR), a scalar field non-minimally coupled to
gravity can lead to "hairy" NS solutions with a non-zero scalar charge. As a
result, binary systems containing at least one "hairy" NS can serve as probes
for testing GR through observations. In this work, we analyze the gravitational
wave signals from four neutron-star-black-hole (NS-BH) merger events: GW190814,
GW200115, GW200105, and GW230529 examine the "hairy" NS hypothesis predicted by
the luminal Horndeski theories. Our analysis constrains the scalar charge to be
on the order of $10^{-3}$ to $10^{-4}$, with the most stringent limit arising
from the recent GW230529 event. In addition, we discuss the implications of our
results in relation to the scalarized NS solutions found in the literature.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 13:33:52 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Santos",
"Rafael M.",
""
],
[
"Nunes",
"Rafael C.",
""
],
[
"de Araujo",
"Jose C. N.",
""
]
] |
2504.00801 | Ikki Fukuda | Ikki Fukuda, Yoshiki Kagaya | Higher-order asymptotic expansions for Laplace's integral and their
error estimates | 7 pages. This version includes some corrections to the original paper | The 43rd JSST Annual International Conference on Simulation
Technology, Conference Proceedings (2024) 502-508 | null | null | math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We deal with the asymptotic analysis for Laplace's integral. For this
problem, the so-called Laplace's method by P.S. Laplace (1812) is well-known
and it has been developed in various forms over many years of studies. In this
paper, we derive some formulas of the higher-order asymptotic expansions for
that integral, with error estimates, which generalize previous results.
Moreover, we discuss a comparison of these asymptotic formulas with
approximations using numerical integral.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 13:55:48 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 14:05:51 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Fukuda",
"Ikki",
""
],
[
"Kagaya",
"Yoshiki",
""
]
] |
2504.01183 | Alexander Stadler | Alexander Stadler | Explicit multiplicities in the cuspidal spectrum of SU(n,1) | null | null | null | null | math.RT | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the cuspidal spectrum of the quotient of the real Lie
group $G= SU(n,1)$ and a principal congruence subgroup $\Gamma(m)$ for $m\geq
3$, focusing on the multiplicities of integrable discrete series
representations. Using the Selberg trace formula, we derive an explicit formula
for the multiplicity $m(\Gamma(m), \pi_\tau)$ of a representation $\pi_\tau$ of
integrable discrete series of $G$ within $L^2(\Gamma(m) \backslash G)$. The
formula involves the Harish-Chandra parameter $\tau$, the discriminant $D_\ell$
of the imaginary quadratic field $\ell$ over which $G$ is defined and special
values of the Dirichlet $L$-function $L_\ell$ associated to $\ell$. We apply
these results on the one hand to compute the cuspidal cohomology of locally
symmetric spaces $\Gamma(m) \backslash G / K$, where $K$ is a maximal compact
subgroup of $G$. On the other hand we use them to reprove a known rationality
result involving the values of $L_\ell$ at odd positive integers and make them
more explicit. This work extends previous studies on real and quaternionic
hyperbolic spaces to the complex hyperbolic case, contributing to the
understanding of the spectrum of $\mathbb{R}$-rank one algebraic groups.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 20:50:11 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 13:36:40 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Stadler",
"Alexander",
""
]
] |
2504.01744 | Igor Anikin | I.V. Anikin | Universal inverse Radon transforms: Inhomogeneity, angular restrictions
and boundary | 11 pages in JHEP style, 2 figures, misprints fixed | null | null | null | math.CA hep-ph hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | An alternative method to invert the Radon transforms without the use of
Courant-Hilbert's identities has been proposed and developed independently from
the space dimension. For the universal representation of inverse Radon
transform, we study the consequences of inhomogeneity of outset function
without the restrictions on the angular Radon coordinates. We show that this
inhomogeneity yields a natural evidence for the presence of the extra
contributions in the case of the full angular region. In addition, if the
outset function is well-localized in the space, we demonstrate that the
corresponding boundary conditions and the angular restrictions should be
applied for both the direct and inverse Radon transforms. Besides, we relate
the angular restrictions on the Radon variable to the boundary exclusion of
outset function and its Radon image.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 13:54:46 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 14:43:16 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 16:01:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Anikin",
"I. V.",
""
]
] |
2504.01975 | Jesus Guillera | Jes\'us Guillera | Fast formulas for the Hurwitz values $\zeta(2,a)$ and $\zeta(3,a)$ | 4 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove two fast formulas for the Hurwitz values $\zeta(2,a)$ and
$\zeta(3,a)$ respectively with the help of the WZ method. In them $(a)_n$
denotes the rising factorial or Pochhammer's symbol defined by $(a)_0=1$ and
$(a)_n=a(a+1)\cdots(a+n-1)$ for positive integers $n$. The Huwitz $\zeta$
function is defined by $\zeta(s,a)=\zeta(0,s,a)=\sum_{k=0}^{\infty}
(k+a)^{-s}$.
| [
{
"version": "v1",
"created": "Tue, 25 Mar 2025 16:17:16 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 14:56:11 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Guillera",
"Jesús",
""
]
] |
2504.02462 | Sung Mook Lee | Sungwoo Hong, Sung Mook Lee, and Qiuyue Liang | Gravitational Wave with Domain Wall Dominance | 15 pages, 5 figures, v2: minor corrections, references added | null | null | CERN-TH-2025-070 | astro-ph.CO gr-qc hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | Domain walls (DWs) can be produced when a discrete symmetry is spontaneously
broken, and long-lived DWs can dominate the energy density of the universe. In
this work, we explore the possibility that a "domain wall dominant (DWD)" phase
existed in the early universe and ended with DW decay. During the DWD phase,
the universe undergoes a power-law accelerated expansion of the scale factor
and exhibits temporal superhorizon evolution of the relevant frequency modes.
We show that this can lead to distinct features imprinted on the stochastic
gravitational wave (GW) background. Our findings provide a comprehensive
framework for evaluating GW emission associated with DWD, leading to
distinguishable long-lived DW-induced GWs from other cosmological sources, with
significant implications for future GW observatories.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 10:28:48 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 14:52:57 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Hong",
"Sungwoo",
""
],
[
"Lee",
"Sung Mook",
""
],
[
"Liang",
"Qiuyue",
""
]
] |
2504.02702 | Yu Cheng | Yu Cheng, Shuailiang Ge | Background-Enhanced Axion Force by Axion Dark Matter | 6 pages, 2 figures | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | We investigate the influence of axion dark matter as a background on the
spin-independent axion forces between nucleons. Notably, we find that the
potential for axion forces scales from $1/r^3$ in a vacuum-only context to
$1/r$ when the background effect is considered. Also, the magnitude of the
axion force is substantially amplified in proportion to the number density of
axion DM particles. These enhancements significantly improve the constraints on
the axion decay constant by several orders of magnitude, across a broad range
of axion masses, based on the fifth-force experiments such as the Casimir-less
and torsion balance tests. This suggests that such experiments are more
effective than previously understood in detecting axions.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 15:37:46 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 03:38:56 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Cheng",
"Yu",
""
],
[
"Ge",
"Shuailiang",
""
]
] |
2504.02769 | Md Somir Khan | Md Somir Khan and Mehmet Engin Tozal | Curbing the Ramifications of Authorship Abuse in Science | null | null | null | null | cs.DL | http://creativecommons.org/licenses/by/4.0/ | Research performance is often measured using bibliometric indicators, such as
publication count, total citations, and $h$-index. These metrics influence
career advancements, salary adjustments, administrative opportunities, funding
prospects, and professional recognition. However, the reliance on these metrics
has also made them targets for manipulation, misuse, and abuse. One primary
ethical concern is authorship abuse, which includes paid, ornamental,
exploitative, cartel, and colonial authorship. These practices are prevalent
because they artificially enhance multiple bibliometric indicators all at once.
Our study confirms a significant rise in the mean and median number of authors
per publication across multiple disciplines over the last 34 years. While it is
important to identify the cases of authorship abuse, a thorough investigation
of every paper proves impractical. In this study, we propose a credit
allocation scheme based on the reciprocals of the Fibonacci numbers, designed
to adjust credit for individual contributions while systematically reducing
credit for potential authorship abuse. The proposed scheme aligns with rigorous
authorship guidelines from scientific associations, which mandate significant
contributions across most phases of a study, while accommodating more lenient
guidelines from scientific publishers, which recognize authorship for minimal
contributions. We recalibrate the traditional bibliometric indicators to
emphasize author contribution rather than participation in publications.
Additionally, we propose a new indicator, $T^{\prime}$-index, to assess
researchers' leading and contributing roles in their publications. Our proposed
credit allocation scheme mitigates the effects of authorship abuse and promotes
a more ethical scientific ecosystem.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 17:06:25 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 23:57:51 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Khan",
"Md Somir",
""
],
[
"Tozal",
"Mehmet Engin",
""
]
] |
2504.02846 | Uddhav Bhattarai | Uddhav Bhattarai, Rajkishan Arikapudi, Chen Peng, Steven A. Fennimore,
Frank N Martin, Stavros G. Vougioukas | Precision yield estimation and mapping in manual strawberry harvesting
with instrumented picking carts and a robust data processing pipeline | Submitted to Computers and Electronics in Agriculture | null | null | null | eess.SP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | High-resolution yield maps for manually harvested crops are impractical to
generate on commercial scales because yield monitors are available only for
mechanical harvesters. However, precision crop management relies on accurately
determining spatial and temporal yield variability. This study presents the
development of an integrated system for precision yield estimation and mapping
for manually harvested strawberries. Conventional strawberry picking carts were
instrumented with a Global Positioning System (GPS) receiver, an Inertial
Measurement Unit (IMU), and load cells to record real-time geo-tagged harvest
data and cart motion. Extensive data were collected in two strawberry fields in
California, USA, during a harvest season. To address the inconsistencies and
errors caused by the sensors and the manual harvesting process, a robust data
processing pipeline was developed by integrating supervised deep learning
models with unsupervised algorithms. The pipeline was used to estimate the
yield distribution and generate yield maps for season-long harvests at the
desired grid resolution. The estimated yield distributions were used to
calculate two metrics: the total mass harvested over specific row segments and
the total mass of trays harvested. The metrics were compared to ground truth
and achieved accuracies of 90.48% and 94.05%, respectively. Additionally, the
accuracy of the estimated yield based on the number of trays harvested per cart
for season-long harvest was better than 94%. It showed a strong correlation
(Pearson r = 0.99) with the actual number of counted trays in both fields. The
proposed system provides a scalable and practical solution for specialty crops,
assisting in efficient yield estimation and mapping, field management, and
labor management for sustainable crop production.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 00:10:39 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 02:15:12 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Bhattarai",
"Uddhav",
""
],
[
"Arikapudi",
"Rajkishan",
""
],
[
"Peng",
"Chen",
""
],
[
"Fennimore",
"Steven A.",
""
],
[
"Martin",
"Frank N",
""
],
[
"Vougioukas",
"Stavros G.",
""
]
] |
2504.02847 | Tribeni Prasad Banerjee Dr | Lohit Bibar, Samali bose, Tribeni Prasad Banerjee | Integrating Notch Filtering and Statistical Methods for Improved Cardiac
Diagnostics Using MATLAB | 11 | null | 10.48550/arXiv.2504.02847 | arXiv preprint arXiv:2504.02847 | eess.SP cs.IR eess.IV | http://creativecommons.org/licenses/by-sa/4.0/ | A Notch Filter is essential in ECG signal processing to eliminate narrowband
noise, especially powerline interference at 50 Hz or 60 Hz. This interference
overlaps with vital ECG signal features, affecting the accuracy of downstream
classification tasks (e.g., arrhythmia detection). A properly designed notch
filter enhances signal quality, preserves essential ECG components (P, QRS, T
waves), and improves the performance of machine learning or deep learning
models used for ECG classification.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 09:35:39 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Bibar",
"Lohit",
""
],
[
"bose",
"Samali",
""
],
[
"Banerjee",
"Tribeni Prasad",
""
]
] |
2504.03210 | Jihui Pei | Ji-Hui Pei and Christian Maes | Transfer of active motion from medium to probe via the induced friction
and noise | null | null | null | null | cond-mat.stat-mech cond-mat.soft physics.bio-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Can activity be transmitted from smaller to larger scales? We report on such
a transfer from a homogeneous active medium to a Newtonian spherical probe. The
active medium consists of faster and dilute self-propelled particles, modeled
as run-and-tumble particles in 1D or as active Brownian particles in 2D. We
derive the reduced fluctuating dynamics of the probe, valid for arbitrary probe
velocity, characterized by velocity-dependent friction and noise. In addition
to a standard passive regime, we identify peculiar active regimes where the
probe becomes self-propelled with high persistence, and its velocity
distribution begets peaks at nonzero values. These features are quantitatively
confirmed by numerical simulations of the joint probe-medium system. The
emergence of active regimes depends not only on the far-from-equilibrium nature
of the medium but also on the probe-medium coupling. Our findings reveal how,
solely via the induced friction and noise, persistence can cross different
scales to transfer active motion.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 06:49:10 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 14:21:49 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Pei",
"Ji-Hui",
""
],
[
"Maes",
"Christian",
""
]
] |
2504.03218 | Jian-Pin Wu | Wen-Juan Ai, Ruo-Ting Chen, Li-Gang Zhu, Jian-Pin Wu | Probing loop quantum effects through solar system experiments:
observational signatures and parameter constraints | 22 pages, 1 figure | null | null | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates quantum gravity effects within the framework of an
effective loop quantum gravity (LQG) black hole model parameterized by $\zeta$,
utilizing precision measurements from solar system experiments and
astrophysical observations. We analyze three classical tests of general
relativity (GR): (1) Light deflection constrained by very long baseline
interferometry (VLBI) observations of quasar radio signals, (2) Shapiro time
delay measurements from the Cassini mission, and (3) Mercury's perihelion
precession determined by MESSENGER mission data. Additionally, we extend our
analysis to Earth-orbiting LAGEOS satellites and the relativistic trajectory of
the S2 star orbiting the Galactic Center supermassive black hole Sagittarius
$\rm{A}^*$ (Sgr $\rm{A}^{*}$). Our multi-probe approach reveals that the
tightest constraint on the LQG parameter comes from Mercury's perihelion
precession, yielding an upper bound $\zeta \lesssim 10^{-2}$. These results
establish new observational benchmarks for probing quantum gravity effects.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 07:07:14 GMT"
},
{
"version": "v2",
"created": "Sun, 4 May 2025 02:14:51 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Ai",
"Wen-Juan",
""
],
[
"Chen",
"Ruo-Ting",
""
],
[
"Zhu",
"Li-Gang",
""
],
[
"Wu",
"Jian-Pin",
""
]
] |
2504.03228 | Monika Avila Marquez | Monika Avila Marquez | Weak instrumental variables due to nonlinearities in panel data: A Super
Learner Control Function estimator | null | null | null | null | econ.EM stat.ML | http://creativecommons.org/licenses/by/4.0/ | A triangular structural panel data model with additive separable
individual-specific effects is used to model the causal effect of a covariate
on an outcome variable when there are unobservable confounders with some of
them time-invariant. In this setup, a linear reduced-form equation might be
problematic when the conditional mean of the endogenous covariate and the
instrumental variables is nonlinear. The reason is that ignoring the
nonlinearity could lead to weak instruments As a solution, we propose a
triangular simultaneous equation model for panel data with additive separable
individual-specific fixed effects composed of a linear structural equation with
a nonlinear reduced form equation. The parameter of interest is the structural
parameter of the endogenous variable. The identification of this parameter is
obtained under the assumption of available exclusion restrictions and using a
control function approach. Estimating the parameter of interest is done using
an estimator that we call Super Learner Control Function estimator (SLCFE). The
estimation procedure is composed of two main steps and sample splitting. We
estimate the control function using a super learner using sample splitting. In
the following step, we use the estimated control function to control for
endogeneity in the structural equation. Sample splitting is done across the
individual dimension. We perform a Monte Carlo simulation to test the
performance of the estimators proposed. We conclude that the Super Learner
Control Function Estimators significantly outperform Within 2SLS estimators.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 07:22:18 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Apr 2025 06:57:50 GMT"
},
{
"version": "v3",
"created": "Sun, 4 May 2025 20:45:51 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Marquez",
"Monika Avila",
""
]
] |
2504.03302 | Afshin Khadangi | Afshin Khadangi, Amir Sartipi, Igor Tchappi, Ramin Bahmani | Noise Augmented Fine Tuning for Mitigating Hallucinations in Large
Language Models | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) often produce inaccurate or misleading
content-hallucinations. To address this challenge, we introduce Noise-Augmented
Fine-Tuning (NoiseFiT), a novel framework that leverages adaptive noise
injection based on the signal-to-noise ratio (SNR) to enhance model robustness.
In particular, NoiseFiT selectively perturbs layers identified as either
high-SNR (more robust) or low-SNR (potentially under-regularized) using a
dynamically scaled Gaussian noise. We further propose a hybrid loss that
combines standard cross-entropy, soft cross-entropy, and consistency
regularization to ensure stable and accurate outputs under noisy training
conditions. Our theoretical analysis shows that adaptive noise injection is
both unbiased and variance-preserving, providing strong guarantees for
convergence in expectation. Empirical results on multiple test and benchmark
datasets demonstrate that NoiseFiT significantly reduces hallucination rates,
often improving or matching baseline performance in key tasks. These findings
highlight the promise of noise-driven strategies for achieving robust,
trustworthy language modeling without incurring prohibitive computational
overhead. Given the comprehensive and detailed nature of our experiments, we
have publicly released the fine-tuning logs, benchmark evaluation artifacts,
and source code online at W&B, Hugging Face, and GitHub, respectively, to
foster further research, accessibility and reproducibility.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 09:27:19 GMT"
},
{
"version": "v2",
"created": "Sat, 3 May 2025 07:26:02 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Khadangi",
"Afshin",
""
],
[
"Sartipi",
"Amir",
""
],
[
"Tchappi",
"Igor",
""
],
[
"Bahmani",
"Ramin",
""
]
] |
2504.03471 | Xi Wang | Xi Wang, Ziqi He, Yang Zhou | Dynamic Importance in Diffusion U-Net for Enhanced Image Synthesis | Accepted to ICME 2025. Appendix & Code:
https://github.com/Hytidel/UNetReweighting | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional diffusion models typically employ a U-Net architecture. Previous
studies have unveiled the roles of attention blocks in the U-Net. However, they
overlook the dynamic evolution of their importance during the inference
process, which hinders their further exploitation to improve image
applications. In this study, we first theoretically proved that, re-weighting
the outputs of the Transformer blocks within the U-Net is a "free lunch" for
improving the signal-to-noise ratio during the sampling process. Next, we
proposed Importance Probe to uncover and quantify the dynamic shifts in
importance of the Transformer blocks throughout the denoising process. Finally,
we design an adaptive importance-based re-weighting schedule tailored to
specific image generation and editing tasks. Experimental results demonstrate
that, our approach significantly improves the efficiency of the inference
process, and enhances the aesthetic quality of the samples with identity
consistency. Our method can be seamlessly integrated into any U-Net-based
architecture. Code: https://github.com/Hytidel/UNetReweighting
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 14:23:30 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 06:32:15 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Wang",
"Xi",
""
],
[
"He",
"Ziqi",
""
],
[
"Zhou",
"Yang",
""
]
] |
2504.03601 | Akshara Prabhakar | Akshara Prabhakar, Zuxin Liu, Ming Zhu, Jianguo Zhang, Tulika
Awalgaonkar, Shiyu Wang, Zhiwei Liu, Haolin Chen, Thai Hoang, Juan Carlos
Niebles, Shelby Heinecke, Weiran Yao, Huan Wang, Silvio Savarese, Caiming
Xiong | APIGen-MT: Agentic Pipeline for Multi-Turn Data Generation via Simulated
Agent-Human Interplay | 12 pages plus references and appendices | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Training effective AI agents for multi-turn interactions requires
high-quality data that captures realistic human-agent dynamics, yet such data
is scarce and expensive to collect manually. We introduce APIGen-MT, a
two-phase framework that generates verifiable and diverse multi-turn agent
data. In the first phase, our agentic pipeline produces detailed task
blueprints with ground-truth actions, leveraging a committee of LLM reviewers
and iterative feedback loops. These blueprints are then transformed into
complete interaction trajectories through simulated human-agent interplay. We
train a family of models -- the xLAM-2-fc-r series with sizes ranging from 1B
to 70B parameters. Our models outperform frontier models such as GPT-4o and
Claude 3.5 on $\tau$-bench and BFCL benchmarks, with the smaller models
surpassing their larger counterparts, particularly in multi-turn settings,
while maintaining superior consistency across multiple trials. Comprehensive
experiments demonstrate that our verified blueprint-to-details approach yields
high-quality training data, enabling the development of more reliable,
efficient, and capable agents. We open-source 5K synthetic data trajectories
and the trained xLAM-2-fc-r models to advance research in AI agents.
Models at
https://huggingface.co/collections/Salesforce/xlam-2-67ef5be12949d8dcdae354c4;
Dataset at https://huggingface.co/datasets/Salesforce/APIGen-MT-5k and Website
at https://apigen-mt.github.io
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 17:13:57 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Apr 2025 17:46:44 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 11:54:13 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Prabhakar",
"Akshara",
""
],
[
"Liu",
"Zuxin",
""
],
[
"Zhu",
"Ming",
""
],
[
"Zhang",
"Jianguo",
""
],
[
"Awalgaonkar",
"Tulika",
""
],
[
"Wang",
"Shiyu",
""
],
[
"Liu",
"Zhiwei",
""
],
[
"Chen",
"Haolin",
""
],
[
"Hoang",
"Thai",
""
],
[
"Niebles",
"Juan Carlos",
""
],
[
"Heinecke",
"Shelby",
""
],
[
"Yao",
"Weiran",
""
],
[
"Wang",
"Huan",
""
],
[
"Savarese",
"Silvio",
""
],
[
"Xiong",
"Caiming",
""
]
] |
2504.03613 | Renbo Zhao | Renbo Zhao | Dual Averaging With Non-Strongly-Convex Prox-Functions: New Analysis and
Algorithm | 39 pages | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present new analysis and algorithm of the dual-averaging-type (DA-type)
methods for solving the composite convex optimization problem
${\min}_{x\in\mathbb{R}^n} \, f(\mathsf{A} x) + h(x)$, where $f$ is a convex
and globally Lipschitz function, $\mathsf{A}$ is a linear operator, and $h$ is
a ``simple'' and convex function that is used as the prox-function in the
DA-type methods. We open new avenues of analyzing and developing DA-type
methods, by going beyond the canonical setting where the prox-function $h$ is
assumed to be strongly convex (on its domain). To that end, we identify two new
sets of assumptions on $h$ (and also $f$ and $\mathsf{A}$) and show that they
hold broadly for many important classes of non-strongly-convex functions. Under
the first set of assumptions, we show that the original DA method still has a
$O(1/k)$ primal-dual convergence rate. Moreover, we analyze the affine
invariance of this method and its convergence rate. Under the second set of
assumptions, we develop a new DA-type method with dual monotonicity, and show
that it has a $O(1/k)$ primal-dual convergence rate. Finally, we consider the
case where $f$ is only convex and Lipschitz on
$\mathcal{C}:=\mathsf{A}(\mathsf{dom} h)$, and construct its globally convex
and Lipschitz extension based on the Pasch-Hausdorff envelope. Furthermore, we
characterize the sub-differential and Fenchel conjugate of this extension using
the convex analytic objects associated with $f$ and $\mathcal{C}$.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 17:33:45 GMT"
},
{
"version": "v2",
"created": "Mon, 5 May 2025 15:54:47 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"Zhao",
"Renbo",
""
]
] |
2504.04237 | Zhiyu He | Zhiyu He, Zhixin Ling, Jiayu Li, Zhiqiang Guo, Weizhi Ma, Xinchen Luo,
Min Zhang, Guorui Zhou | Short Video Segment-level User Dynamic Interests Modeling in
Personalized Recommendation | This paper has been accepted by SIGIR 2025 | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of short videos has necessitated effective recommender
systems to match users with content tailored to their evolving preferences.
Current video recommendation models primarily treat each video as a whole,
overlooking the dynamic nature of user preferences with specific video
segments. In contrast, our research focuses on segment-level user interest
modeling, which is crucial for understanding how users' preferences evolve
during video browsing. To capture users' dynamic segment interests, we propose
an innovative model that integrates a hybrid representation module, a
multi-modal user-video encoder, and a segment interest decoder. Our model
addresses the challenges of capturing dynamic interest patterns, missing
segment-level labels, and fusing different modalities, achieving precise
segment-level interest prediction. We present two downstream tasks to evaluate
the effectiveness of our segment interest modeling approach: video-skip
prediction and short video recommendation. Our experiments on real-world short
video datasets with diverse modalities show promising results on both tasks. It
demonstrates that segment-level interest modeling brings a deep understanding
of user engagement and enhances video recommendations. We also release a unique
dataset that includes segment-level video data and diverse user behaviors,
enabling further research in segment-level interest modeling. This work
pioneers a novel perspective on understanding user segment-level preference,
offering the potential for more personalized and engaging short video
experiences.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 17:45:32 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 02:35:55 GMT"
},
{
"version": "v3",
"created": "Mon, 5 May 2025 06:51:21 GMT"
}
] | 2025-05-06T00:00:00 | [
[
"He",
"Zhiyu",
""
],
[
"Ling",
"Zhixin",
""
],
[
"Li",
"Jiayu",
""
],
[
"Guo",
"Zhiqiang",
""
],
[
"Ma",
"Weizhi",
""
],
[
"Luo",
"Xinchen",
""
],
[
"Zhang",
"Min",
""
],
[
"Zhou",
"Guorui",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.