id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-05-16 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2505.10058 | Toan Nguyen | Toan T. Nguyen | Remarks on Landau damping | a contribution to VIASM Lecture Notes "Resonances in the Mathematical
World" in honor of Dang Duc Trong, a mentor to many Vietnamese mathematicians | null | null | null | math.AP math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We provide few remarks on nonlinear Landau damping that concerns decay of the
electric field in the classical Vlasov-Poisson system near spatially homogenous
equilibria. In particular, this includes the analyticity framework, \`a la
Grenier-Nguyen-Rodnianski, for non specialists, treating the analytic case
studied by Mouhot-Villani, among other remarks for plasmas confined on a torus
and in the whole space.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:02:14 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Nguyen",
"Toan T.",
""
]
] |
2505.10059 | Ahmad Taha | MirSaleh Bahavarnia and Muhammad Nadeem and Ahmad F. Taha | Improving Power Systems Controllability via Edge Centrality Measures | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Improving the controllability of power networks is crucial as they are highly
complex networks operating in synchrony; even minor perturbations can cause
desynchronization and instability. To that end, one needs to assess the
criticality of key network components (buses and lines) in terms of their
impact on system performance. Traditional methods to identify the key
nodes/edges in power networks often rely on static centrality measures based on
the network's topological structure ignoring the network's dynamic behavior. In
this paper, using multi-machine power network models and a new
control-theoretic edge centrality matrix (ECM) approach, we: (i) quantify the
influence of edges (i.e., the line susceptances) in terms of controllability
performance metrics, (ii) identify the most influential lines, and (iii)
compute near-optimal edge modifications that improve the power network
controllability. Employing various IEEE power network benchmarks, we validate
the effectiveness of the ECM-based algorithm and demonstrate improvements in
system reachability, control, and damping performance.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:03:56 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bahavarnia",
"MirSaleh",
""
],
[
"Nadeem",
"Muhammad",
""
],
[
"Taha",
"Ahmad F.",
""
]
] |
2505.10060 | Thomas Benz | Philippe Sauter, Thomas Benz, Paul Scheffler, Martin Povi\v{s}er,
Frank K. G\"urkaynak, Luca Benini | Basilisk: A 34 mm2 End-to-End Open-Source 64-bit Linux-Capable RISC-V
SoC in 130nm BiCMOS | 2 pages, 5 figures, accepted at Hot Chips 2025 as a poster submission | null | null | null | cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | End-to-end open-source electronic design automation (OSEDA) enables a
collaborative approach to chip design conducive to supply chain diversification
and zero-trust step-by-step design verification. However, existing end-to-end
OSEDA flows have mostly been demonstrated on small designs and have not yet
enabled large, industry-grade chips such as Linux-capable systems-on-chip
(SoCs). This work presents Basilisk, the largest end-to-end open-source SoC to
date. Basilisk's 34 mm2, 2.7 MGE design features a 64-bit Linux-capable RISC-V
core, a lightweight 124 MB/s DRAM controller, and extensive IO, including a USB
1.1 host, a video output, and a fully digital 62 Mb/s chip-to-chip (C2C) link.
We implement Basilisk in IHP's open 130 nm BiCMOS technology, significantly
improving on the state-of-the-art (SoA) OSEDA flow. Our enhancements of the
Yosys-based synthesis flow improve design timing and area by 2.3x and 1.6x,
respectively, while consuming significantly less system resources. By tuning
OpenROAD place and route (P&R) to our design and technology, we decrease the
die size by 12%. The fabricated Basilisk chip reaches 62 MHz at its nominal 1.2
V core voltage and up to 102 MHz at 1.64 V. It achieves a peak energy
efficiency of 18.9 DP MFLOP/s/W at 0.88 V.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:04:39 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sauter",
"Philippe",
""
],
[
"Benz",
"Thomas",
""
],
[
"Scheffler",
"Paul",
""
],
[
"Povišer",
"Martin",
""
],
[
"Gürkaynak",
"Frank K.",
""
],
[
"Benini",
"Luca",
""
]
] |
2505.10061 | Philippe Jaming | Philippe Jaming (IMB), Karim Kellay (IMB), Rolando Perez Iii (UP
Diliman) | On Wiener's Lemma on locally compact abelian groups | null | null | null | null | math.CA math.CV math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish a general form of Wiener's lemma for measures on locally compact
abelian (LCA) groups by using Fourier analysis and the theory of F{{\o}}lner
sequences. Our approach provides a unified framework that that encompasses both
the discrete and continuous cases. We also show a version of Wiener's lemma for
Bochner-Riesz means on both R^d and T^d . Mathematics Subject Classification
(2010). 43A25.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:04:40 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Jaming",
"Philippe",
"",
"IMB"
],
[
"Kellay",
"Karim",
"",
"IMB"
],
[
"Perez",
"Rolando",
"Iii",
"UP\n Diliman"
]
] |
2505.10062 | Antonio Sannia | Antonio Sannia, Gian Luca Giorgi, Roberta Zambrini | Exponential concentration and symmetries in Quantum Reservoir Computing | Comments are welcome | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum reservoir computing (QRC) is an emerging framework for near-term
quantum machine learning that offers in-memory processing, platform versatility
across analogue and digital systems, and avoids typical trainability challenges
such as barren plateaus and local minima. The exponential number of independent
features of quantum reservoirs opens the way to a potential performance
improvement compared to classical settings. However, this exponential scaling
can be hindered by exponential concentration, where finite-ensemble noise in
quantum measurements requires exponentially many samples to extract meaningful
outputs, a common issue in quantum machine learning. In this work, we go beyond
static quantum machine learning tasks and address concentration in QRC for
time-series processing using quantum-scrambling reservoirs. Beyond discussing
how concentration effects can constrain QRC performance, we demonstrate that
leveraging Hamiltonian symmetries significantly suppresses concentration,
enabling robust and scalable QRC implementations. The proposed strategy is
illustrated in examples, including a well-known QRC design.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:05:06 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sannia",
"Antonio",
""
],
[
"Giorgi",
"Gian Luca",
""
],
[
"Zambrini",
"Roberta",
""
]
] |
2505.10063 | Han Peng | Han Peng, Jinhao Jiang, Zican Dong, Wayne Xin Zhao, Lei Fang | CAFE: Retrieval Head-based Coarse-to-Fine Information Seeking to Enhance
Multi-Document QA Capability | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Advancements in Large Language Models (LLMs) have extended their input
context length, yet they still struggle with retrieval and reasoning in
long-context inputs. Existing methods propose to utilize the prompt strategy
and retrieval head to alleviate this limitation. However, they still face
challenges in balancing retrieval precision and recall, impacting their
efficacy in answering questions. To address this, we introduce $\textbf{CAFE}$,
a two-stage coarse-to-fine method to enhance multi-document question-answering
capacities. By gradually eliminating the negative impacts of background and
distracting documents, CAFE makes the responses more reliant on the evidence
documents. Initially, a coarse-grained filtering method leverages retrieval
heads to identify and rank relevant documents. Then, a fine-grained steering
method guides attention to the most relevant content. Experiments across
benchmarks show CAFE outperforms baselines, achieving up to 22.1% and 13.7%
SubEM improvement over SFT and RAG methods on the Mistral model, respectively.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:05:12 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Peng",
"Han",
""
],
[
"Jiang",
"Jinhao",
""
],
[
"Dong",
"Zican",
""
],
[
"Zhao",
"Wayne Xin",
""
],
[
"Fang",
"Lei",
""
]
] |
2505.10064 | Jonathan LABBE | Jonathan Labb\'e (CEREFIGE, IAE Nancy, UL), Typhaine Leb\`egue
(VALLOREM, UT), Abdel Malik Ola (VALLOREM, UT) | Choices or constraints: decoding financial empowerment among women
entrepreneurs in France | null | Revue Internationale PME, 2025, 38 (1), pp.73-93 | 27E8;10.7202/1117705ar | null | q-fin.GN | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This research examines the empowerment of women entrepreneurs in the context
of entrepreneurial financing in France. It explores the factors that allow some
women entrepreneurs to access certain categories of external finance more
easily. The theoretical framework used is based on the concept of empowerment,
explored through its personal and relational dimensions. The study relies on a
quantitative approach, using data from a representative of women entrepreneurs.
The results show that the status of a founder affects access to external
finance in different ways: it increases the chances of successful fundraising,
but reduces the chances of obtaining bank finance. This finding highlights the
importance of empowerment dynamics, which vary according to the type of
financing. In addition, characteristics such as the presence of a spouse in the
business, high income, membership of a professional network and the diversity
of this network complete the analysis of inequalities in access. This study,
the first of its kind in France, suggests ways of enriching our understanding
of the diversity of situations experienced by female founders, thus helping to
deconstruct the homogeneous image of women's entrepreneurship.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:06:09 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Labbé",
"Jonathan",
"",
"CEREFIGE, IAE Nancy, UL"
],
[
"Lebègue",
"Typhaine",
"",
"VALLOREM, UT"
],
[
"Ola",
"Abdel Malik",
"",
"VALLOREM, UT"
]
] |
2505.10065 | Eni Musta | Eni Musta and Martina Vittorietti | A mover-stayer model with time-dependent stayer fraction | null | null | null | null | stat.ME stat.AP | http://creativecommons.org/licenses/by/4.0/ | Mover-stayer models are used in social sciences and economics to model
heterogeneous population dynamics in which some individuals never experience
the event of interest ("stayers"), while others transition between states over
time ("movers"). Conventionally, the mover-stayer status is determined at
baseline and time-dependent covariates are only incorporated in the movers'
transition probabilities. In this paper, we present a novel dynamic version of
the mover-stayer model, allowing potential movers to become stayers over time
based on time-varying circumstances. Using a multinomial logistic framework,
our model incorporates both time-fixed and exogenous time-varying covariates to
estimate transition probabilities among the states of potential movers, movers,
and stayers. Both the initial state and transitions to the stayer state are
treated as latent. The introduction of this new model is motivated by the study
of student mobility. Specifically focusing on panel data on the
inter-university mobility of Italian students, factors such as the students'
change of course and university size are considered as time-varying covariates
in modelling their probability of moving or becoming stayers; sex and age at
enrolment as time-fixed covariates. We propose a maximum likelihood estimation
approach and investigate its finite-sample performance through simulations,
comparing it to established models in the literature.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:06:30 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Musta",
"Eni",
""
],
[
"Vittorietti",
"Martina",
""
]
] |
2505.10066 | Lior Rokach | Michael Fire, Yitzhak Elbazis, Adi Wasenstein, Lior Rokach | Dark LLMs: The Growing Threat of Unaligned AI Models | null | null | null | null | cs.CL cs.AI cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) rapidly reshape modern life, advancing fields
from healthcare to education and beyond. However, alongside their remarkable
capabilities lies a significant threat: the susceptibility of these models to
jailbreaking. The fundamental vulnerability of LLMs to jailbreak attacks stems
from the very data they learn from. As long as this training data includes
unfiltered, problematic, or 'dark' content, the models can inherently learn
undesirable patterns or weaknesses that allow users to circumvent their
intended safety controls. Our research identifies the growing threat posed by
dark LLMs models deliberately designed without ethical guardrails or modified
through jailbreak techniques. In our research, we uncovered a universal
jailbreak attack that effectively compromises multiple state-of-the-art models,
enabling them to answer almost any question and produce harmful outputs upon
request. The main idea of our attack was published online over seven months
ago. However, many of the tested LLMs were still vulnerable to this attack.
Despite our responsible disclosure efforts, responses from major LLM providers
were often inadequate, highlighting a concerning gap in industry practices
regarding AI safety. As model training becomes more accessible and cheaper, and
as open-source LLMs proliferate, the risk of widespread misuse escalates.
Without decisive intervention, LLMs may continue democratizing access to
dangerous knowledge, posing greater risks than anticipated.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:07:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Fire",
"Michael",
""
],
[
"Elbazis",
"Yitzhak",
""
],
[
"Wasenstein",
"Adi",
""
],
[
"Rokach",
"Lior",
""
]
] |
2505.10067 | Henrique de Oliveira | H. P. de Oliveira | Distorted black holes: nonextensivity in the gravitational wave emission | This essay received an Honorable Mention in the 2025 Essay
Competition of the Gravity Research Foundation | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the interaction between a non-rotating black hole and incoming
gravitational waves using the characteristic formulation of the Einstein field
equations, framed as a Bondi problem. By adopting retarded time as the null
coordinate and recognizing that the final state is invariably a black hole, we
show that an apparent horizon forms once sufficient mass accretes onto the
black hole. We derive the evolution of the Bondi mass and compute its final
value, enabling us to quantify the fraction of the incident mass absorbed by
the black hole. Additionally, we establish a nonextensive relation for the
absorbed mass as a function of initial parameters, such as the amplitude of the
gravitational wave data.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:07:20 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"de Oliveira",
"H. P.",
""
]
] |
2505.10068 | Seyma Bodur | \c{S}eyma Bodur, Fernando Hernando, Edgar Mart\'inez-Moro, Diego Ruano | The Schur product of evaluation codes and its application to CSS-T
quantum codes and private information retrieval | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this work, we study the componentwise (Schur) product of
monomial-Cartesian codes by exploiting its correspondence with the Minkowski
sum of their defining exponent sets. We show that $ J$-affine variety codes are
well suited for such products, generalizing earlier results for cyclic,
Reed-Muller, hyperbolic, and toric codes. Using this correspondence, we
construct CSS-T quantum codes from weighted Reed-Muller codes and from binary
subfield-subcodes of $ J$-affine variety codes, leading to codes with better
parameters than previously known. Finally, we present Private Information
Retrieval (PIR) constructions for multiple colluding servers based on
hyperbolic codes and subfield-subcodes of $ J$-affine variety codes, and show
that they outperform existing PIR schemes.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:09:21 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bodur",
"Şeyma",
""
],
[
"Hernando",
"Fernando",
""
],
[
"Martínez-Moro",
"Edgar",
""
],
[
"Ruano",
"Diego",
""
]
] |
2505.10069 | Mohamed Chatti | Qurat Ul Ain, Mohamed Amine Chatti, Amr Shakhshir, Jean Qussa, Rawaa
Alatrash, Shoeb Joarder | Top-Down vs. Bottom-Up Approaches for Automatic Educational Knowledge
Graph Construction in CourseMapper | Accepted at EMOOCs 2025 | null | null | null | cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The automatic construction of Educational Knowledge Graphs (EduKGs) is
crucial for modeling domain knowledge in digital learning environments,
particularly in Massive Open Online Courses (MOOCs). However, identifying the
most effective approach for constructing accurate EduKGs remains a challenge.
This study compares Top-down and Bottom-up approaches for automatic EduKG
construction, evaluating their effectiveness in capturing and structuring
knowledge concepts from learning materials in our MOOC platform CourseMapper.
Through a user study and expert validation using Simple Random Sampling (SRS),
results indicate that the Bottom-up approach outperforms the Top-down approach
in accurately identifying and mapping key knowledge concepts. To further
enhance EduKG accuracy, we integrate a Human-in-the-Loop approach, allowing
course moderators to review and refine the EduKG before publication. This
structured comparison provides a scalable framework for improving knowledge
representation in MOOCs, ultimately supporting more personalized and adaptive
learning experiences.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:11:48 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ain",
"Qurat Ul",
""
],
[
"Chatti",
"Mohamed Amine",
""
],
[
"Shakhshir",
"Amr",
""
],
[
"Qussa",
"Jean",
""
],
[
"Alatrash",
"Rawaa",
""
],
[
"Joarder",
"Shoeb",
""
]
] |
2505.10070 | Hong Zhang | Zuo-Tang Liang, Qian Yang, Hong Zhang, Jiaxing Zhao, Pengfei Zhuang | Searching for True Muonium in Relativistic Heavy Ion Collisions | 6 pages, 5 figures | null | null | null | nucl-th hep-ph | http://creativecommons.org/licenses/by/4.0/ | We investigate the production of the as-yet-undetected true muonium within
the quark-gluon plasma formed in relativistic heavy-ion collisions, employing a
relativistic Boltzmann transport framework coupled to viscous hydrodynamic
simulations. The obtained effective cross sections for central collisions are
1.23~$\mu b$ in AuAu collisions with $\sqrt{s_{\rm NN}}=200$~GeV and 14.2~$\mu
b$ in PbPb collisions with $\sqrt{s_{\rm NN}}=5.02$~TeV, resulting in a yield
of $\mathcal{O}(10^4)$ and $\mathcal{O}(10^5)$ true muonium per billion $AA$
collisions at RHIC and the LHC, respectively. This establishes heavy-ion
collisions as a promising process for detecting true muonium.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:15:34 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Liang",
"Zuo-Tang",
""
],
[
"Yang",
"Qian",
""
],
[
"Zhang",
"Hong",
""
],
[
"Zhao",
"Jiaxing",
""
],
[
"Zhuang",
"Pengfei",
""
]
] |
2505.10071 | Eric Goubault | Eric Goubault and Bernardo Hummes Flores and Roman Kniazev and Jeremy
Ledent and Sergio Rajsbaum | A categorical and logical framework for iterated protocols | null | null | null | null | cs.LO cs.DC | http://creativecommons.org/licenses/by/4.0/ | In this article, we show that the now classical protocol complex approach to
distributed task solvability of Herlihy et al. can be understood in standard
categorical terms. First, protocol complexes are functors, from chromatic
(semi-) simplicial sets to chromatic simplicial sets, that naturally give rise
to algebras. These algebras describe the next state operator for the
corresponding distributed systems. This is constructed for semi-synchronous
distributed systems with general patterns of communication for which we show
that these functors are always Yoneda extensions of simpler functors, implying
a number of interesting properties. Furthermore, for these protocol complex
functors, we prove the existence of a free algebra on any initial chromatic
simplicial complex, modeling iterated protocol complexes. Under this
categorical formalization, protocol complexes are seen as transition systems,
where states are structured as chromatic simplicial sets. We exploit the
epistemic interpretation of chromatic simplicial sets and the underlying
transition system (or algebra) structure to introduce a temporal-epistemic
logic and its semantics on all free algebras on chromatic simplicial sets. We
end up by giving hints on how to extend this framework to more general dynamic
network graphs and state-dependent protocols, and give example in
fault-tolerant distributed systems and mobile robotics.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:15:42 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Goubault",
"Eric",
""
],
[
"Flores",
"Bernardo Hummes",
""
],
[
"Kniazev",
"Roman",
""
],
[
"Ledent",
"Jeremy",
""
],
[
"Rajsbaum",
"Sergio",
""
]
] |
2505.10072 | RuiYang Ju | Rui-Yang Ju, Sheng-Yen Huang, Yi-Ping Hung | ToonifyGB: StyleGAN-based Gaussian Blendshapes for 3D Stylized Head
Avatars | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The introduction of 3D Gaussian blendshapes has enabled the real-time
reconstruction of animatable head avatars from monocular video. Toonify, a
StyleGAN-based framework, has become widely used for facial image stylization.
To extend Toonify for synthesizing diverse stylized 3D head avatars using
Gaussian blendshapes, we propose an efficient two-stage framework, ToonifyGB.
In Stage 1 (stylized video generation), we employ an improved StyleGAN to
generate the stylized video from the input video frames, which addresses the
limitation of cropping aligned faces at a fixed resolution as preprocessing for
normal StyleGAN. This process provides a more stable video, which enables
Gaussian blendshapes to better capture the high-frequency details of the video
frames, and efficiently generate high-quality animation in the next stage. In
Stage 2 (Gaussian blendshapes synthesis), we learn a stylized neutral head
model and a set of expression blendshapes from the generated video. By
combining the neutral head model with expression blendshapes, ToonifyGB can
efficiently render stylized avatars with arbitrary expressions. We validate the
effectiveness of ToonifyGB on the benchmark dataset using two styles: Arcane
and Pixar.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:16:12 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ju",
"Rui-Yang",
""
],
[
"Huang",
"Sheng-Yen",
""
],
[
"Hung",
"Yi-Ping",
""
]
] |
2505.10073 | Rathin Chandra Shit | Rathin Chandra Shit, Sharmila Subudhi | Multi-Robot Task Allocation for Homogeneous Tasks with Collision
Avoidance via Spatial Clustering | 5 pages, 4 figures, Scheduled for presentation at an upcoming
conference | null | null | null | cs.RO cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a novel framework is presented that achieves a combined
solution based on Multi-Robot Task Allocation (MRTA) and collision avoidance
with respect to homogeneous measurement tasks taking place in industrial
environments. The spatial clustering we propose offers to simultaneously solve
the task allocation problem and deal with collision risks by cutting the
workspace into distinguishable operational zones for each robot. To divide task
sites and to schedule robot routes within corresponding clusters, we use
K-means clustering and the 2-Opt algorithm. The presented framework shows
satisfactory performance, where up to 93\% time reduction (1.24s against
17.62s) with a solution quality improvement of up to 7\% compared to the best
performing method is demonstrated. Our method also completely eliminates
collision points that persist in comparative methods in a most significant
sense. Theoretical analysis agrees with the claim that spatial partitioning
unifies the apparently disjoint tasks allocation and collision avoidance
problems under conditions of many identical tasks to be distributed over sparse
geographical areas. Ultimately, the findings in this work are of substantial
importance for real world applications where both computational efficiency and
operation free from collisions is of paramount importance.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:20:57 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Shit",
"Rathin Chandra",
""
],
[
"Subudhi",
"Sharmila",
""
]
] |
2505.10074 | Mohamed Chatti | Mohamed Abdelmagied and Mohamed Amine Chatti and Shoeb Joarder and
Qurat Ul Ain and Rawaa Alatrash | Leveraging Graph Retrieval-Augmented Generation to Support Learners'
Understanding of Knowledge Concepts in MOOCs | Accepted at EMOOCs 2025 | null | null | null | cs.AI cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Massive Open Online Courses (MOOCs) lack direct interaction between learners
and instructors, making it challenging for learners to understand new knowledge
concepts. Recently, learners have increasingly used Large Language Models
(LLMs) to support them in acquiring new knowledge. However, LLMs are prone to
hallucinations which limits their reliability. Retrieval-Augmented Generation
(RAG) addresses this issue by retrieving relevant documents before generating a
response. However, the application of RAG across different MOOCs is limited by
unstructured learning material. Furthermore, current RAG systems do not
actively guide learners toward their learning needs. To address these
challenges, we propose a Graph RAG pipeline that leverages Educational
Knowledge Graphs (EduKGs) and Personal Knowledge Graphs (PKGs) to guide
learners to understand knowledge concepts in the MOOC platform CourseMapper.
Specifically, we implement (1) a PKG-based Question Generation method to
recommend personalized questions for learners in context, and (2) an
EduKG-based Question Answering method that leverages the relationships between
knowledge concepts in the EduKG to answer learner selected questions. To
evaluate both methods, we conducted a study with 3 expert instructors on 3
different MOOCs in the MOOC platform CourseMapper. The results of the
evaluation show the potential of Graph RAG to empower learners to understand
new knowledge concepts in a personalized learning experience.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:24:47 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Abdelmagied",
"Mohamed",
""
],
[
"Chatti",
"Mohamed Amine",
""
],
[
"Joarder",
"Shoeb",
""
],
[
"Ain",
"Qurat Ul",
""
],
[
"Alatrash",
"Rawaa",
""
]
] |
2505.10075 | Jun Guo | Jun Guo, Xiaojian Ma, Yikai Wang, Min Yang, Huaping Liu, Qing Li | FlowDreamer: A RGB-D World Model with Flow-based Motion Representations
for Robot Manipulation | Project page: see https://sharinka0715.github.io/FlowDreamer/ | null | null | null | cs.RO cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper investigates training better visual world models for robot
manipulation, i.e., models that can predict future visual observations by
conditioning on past frames and robot actions. Specifically, we consider world
models that operate on RGB-D frames (RGB-D world models). As opposed to
canonical approaches that handle dynamics prediction mostly implicitly and
reconcile it with visual rendering in a single model, we introduce FlowDreamer,
which adopts 3D scene flow as explicit motion representations. FlowDreamer
first predicts 3D scene flow from past frame and action conditions with a
U-Net, and then a diffusion model will predict the future frame utilizing the
scene flow. FlowDreamer is trained end-to-end despite its modularized nature.
We conduct experiments on 4 different benchmarks, covering both video
prediction and visual planning tasks. The results demonstrate that FlowDreamer
achieves better performance compared to other baseline RGB-D world models by 7%
on semantic similarity, 11% on pixel quality, and 6% on success rate in various
robot manipulation domains.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:27:16 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guo",
"Jun",
""
],
[
"Ma",
"Xiaojian",
""
],
[
"Wang",
"Yikai",
""
],
[
"Yang",
"Min",
""
],
[
"Liu",
"Huaping",
""
],
[
"Li",
"Qing",
""
]
] |
2505.10076 | Ryutaro Iwakura | Ryutaro Iwakura, Yoko Funato and Junichiro Makino | A new approach for predicting the stability of hierarchical triple
systems -- I. Coplanar Cases | 10 pages, 9 figures | null | null | null | astro-ph.SR astro-ph.EP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hierarchical triple systems play a crucial role in various astrophysical
contexts, and therefore the understanding of their stability is important.
Traditional empirical stability criteria rely on a threshold value of $Q$, the
ratio between the outer orbit's pericenter distance and the inner orbit's
semi-major axis. However, determining a single critical value of $Q$ is
impossible because there is a range of the value of $Q$ for which both stable
and unstable systems exist, referred to as the mixed-region. In this study, we
introduce a novel method to assess the stability of triple systems within this
mixed-region. We numerically integrate equal-mass, coplanar hierarchical
triples within the mixed-region. By performing Fourier analysis of the time
evolution of the semi-major axes ratio during the first 1000 inner orbital
periods of the systems, we find notable features in stable systems: if the main
peaks are periodically spaced in the frequency domain and the continuous
components and irregularly spaced peaks are small, the system tends to be
stable. This observation indicates that the evolution of stable triples is more
periodic than that of unstable ones. We quantified the periodicity of the
triples and investigated the correlation between the Fourier power distribution
and the system's lifetime. Using this correlation, we show that it is possible
to determine if a triple system in the mixed-region is stable or not with very
high accuracy. These findings suggest that periodicity in orbital evolution can
serve as a robust indicator of stability for hierarchical triples.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:29:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Iwakura",
"Ryutaro",
""
],
[
"Funato",
"Yoko",
""
],
[
"Makino",
"Junichiro",
""
]
] |
2505.10077 | Christian Bernert | Christian Bernert, Ulrich Derenthal | Integral points of bounded height on quintic del Pezzo surfaces over
number fields | 25 pages | null | null | null | math.NT math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove an asymptotic formula for the number of integral points of bounded
log-anticanonical height on split smooth quintic del Pezzo surfaces over number
fields, with respect to one of the lines as the boundary divisor.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:31:42 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bernert",
"Christian",
""
],
[
"Derenthal",
"Ulrich",
""
]
] |
2505.10078 | Liuze Long | Liuze Long, Yulong Gao, Qiusheng Gu, Yong Shi, Xin Li, Can Xu, Yifei
Jin, Zhiyuan Zheng, Jing Dou, Fuyan Bian, Xiaoling Yu | Clumpy Starburst in a Local Dwarf Galaxy, NGC 1522 | 15 pages, 10 figures, Accepted for publication in ApJ | null | 10.3847/1538-4357/add0b7 | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To investigate the star-forming process in nearby dwarf galaxies, we present
Integral Field Units (IFU) observation of the star-forming dwarf galaxy, NGC
1522, with the Very Large Telescope (VLT)/Multi Unit Spectroscopic Explorer
(MUSE) as a part of Dwarf Galaxy Integral Survey (DGIS). Our observation
reveals the presence of a star-forming clumpy ring in its central region. We
identify nine distinct star-forming clumps based on extinction-corrected
H$\alpha$ emission-line map, with the total star formation rate (SFR) of about
0.1 $M_\odot$ yr$^{-1}$. The nine clumps are considered to be starbursts, which
represent an extreme case in the local universe, without invoking major
merging. We investigate the properties of ionized gas using the strong emission
lines and `BPT' diagrams, in conjunction with the velocity mapping. Our
analysis unveils intriguing patterns, including the positive metallicity
gradient and low N/O abundance ratio. This peculiar distribution of metallicity
may signify external gas accretion. Our results suggest that the ongoing star
formation in NGC 1522 might be triggered and sustained by the inflow of
external metal-poor gas.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:32:30 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Long",
"Liuze",
""
],
[
"Gao",
"Yulong",
""
],
[
"Gu",
"Qiusheng",
""
],
[
"Shi",
"Yong",
""
],
[
"Li",
"Xin",
""
],
[
"Xu",
"Can",
""
],
[
"Jin",
"Yifei",
""
],
[
"Zheng",
"Zhiyuan",
""
],
[
"Dou",
"Jing",
""
],
[
"Bian",
"Fuyan",
""
],
[
"Yu",
"Xiaoling",
""
]
] |
2505.10079 | Deung-Jang Choi | Deung-Jang Choi, Soo-hyon Phark, Andreas J. Heinrich, and Nicol\'as
Lorente | Electron spin resonance with scanning tunneling microscopy: a tool for
an on-surface quantum platform of identical qubits | null | null | null | null | cond-mat.mes-hall quant-ph | http://creativecommons.org/licenses/by/4.0/ | Integration of electron spin resonance (ESR) in a scanning tunneling
microscope (STM) has enabled an all-electrical control of atomic and molecular
spins on solid surfaces with atomic-scale precision and energy resolution
beyond thermal limitations. Further, coherent manipulation and detection of
individual spins in an ESR-STM establishes a powerful quantum platform,
allowing for the implementation of fundamental quantum logic operations to
on-surface identical qubits. In this review, we introduce recent advances of
ESR-STM, focusing on its application to atomic-scale qubits and extension to
molecular qubit systems. We discuss the principles underlying ESR-STM, followed
by single-spin addressability, coherent control via Rabi oscillations, and
quantum state readout through frequency-resolved detection. We further
demonstrate multi-qubit control architectures enabled by atom manipulation and
local magnetic field engineering, culminating in the realization of multi-qubit
logic gates such as the Controlled-NOT and Toffoli gates. These implementations
highlight the specialty of ESR-STM towards atomic-scale quantum circuits.
Indeed, ESR-STM can be an excellent tool to perform and evaluate quantum
operations in molecular qubits. The results reviewed in this collection
establish ESR-STM as a versatile tool for advancing quantum coherent science at
the atomic and molecular level in solid-state environments.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:34:53 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Choi",
"Deung-Jang",
""
],
[
"Phark",
"Soo-hyon",
""
],
[
"Heinrich",
"Andreas J.",
""
],
[
"Lorente",
"Nicolás",
""
]
] |
2505.10080 | Weijie Xiong | Weijie Xiong, Zo\"e Holmes, Armando Angrisani, Yudai Suzuki, Thiparat
Chotibut, Supanut Thanasilp | Role of scrambling and noise in temporal information processing with
quantum systems | 14+35 pages, 6+5 figures, 1 table | null | null | null | quant-ph cond-mat.dis-nn cs.LG cs.NE stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scrambling quantum systems have been demonstrated as effective substrates for
temporal information processing. While their role in providing rich feature
maps has been widely studied, a theoretical understanding of their performance
in temporal tasks is still lacking. Here we consider a general quantum
reservoir processing framework that captures a broad range of physical
computing models with quantum systems. We examine the scalability and memory
retention of the model with scrambling reservoirs modelled by high-order
unitary designs in both noiseless and noisy settings. In the former regime, we
show that measurement readouts become exponentially concentrated with
increasing reservoir size, yet strikingly do not worsen with the reservoir
iterations. Thus, while repeatedly reusing a small scrambling reservoir with
quantum data might be viable, scaling up the problem size deteriorates
generalization unless one can afford an exponential shot overhead. In contrast,
the memory of early inputs and initial states decays exponentially in both
reservoir size and reservoir iterations. In the noisy regime, we also prove
exponential memory decays with iterations for local noisy channels. Proving
these results required us to introduce new proof techniques for bounding
concentration in temporal quantum learning models.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:35:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Xiong",
"Weijie",
""
],
[
"Holmes",
"Zoë",
""
],
[
"Angrisani",
"Armando",
""
],
[
"Suzuki",
"Yudai",
""
],
[
"Chotibut",
"Thiparat",
""
],
[
"Thanasilp",
"Supanut",
""
]
] |
2505.10081 | Francois Meyer | Wisdom Aduah and Francois Meyer | Designing and Contextualising Probes for African Languages | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pretrained language models (PLMs) for African languages are continually
improving, but the reasons behind these advances remain unclear. This paper
presents the first systematic investigation into probing PLMs for linguistic
knowledge about African languages. We train layer-wise probes for six
typologically diverse African languages to analyse how linguistic features are
distributed. We also design control tasks, a way to interpret probe
performance, for the MasakhaPOS dataset. We find PLMs adapted for African
languages to encode more linguistic information about target languages than
massively multilingual PLMs. Our results reaffirm previous findings that
token-level syntactic information concentrates in middle-to-last layers, while
sentence-level semantic information is distributed across all layers. Through
control tasks and probing baselines, we confirm that performance reflects the
internal knowledge of PLMs rather than probe memorisation. Our study applies
established interpretability techniques to African-language PLMs. In doing so,
we highlight the internal mechanisms underlying the success of strategies like
active learning and multilingual adaptation.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:35:14 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Aduah",
"Wisdom",
""
],
[
"Meyer",
"Francois",
""
]
] |
2505.10082 | Danish Kashaev | Danish Kashaev | Price of Anarchy for Congestion and Scheduling Games via Vector Fitting | null | null | null | null | cs.GT cs.DS | http://creativecommons.org/licenses/by/4.0/ | We provide a dual fitting technique on a semidefinite program yielding simple
proofs of tight bounds for the robust price of anarchy of several congestion
and scheduling games under the sum of weighted completion times objective. The
same approach also allows to bound the approximation ratio of local search
algorithms for the scheduling problem $R || \sum w_j C_j$. All of our results
are obtained through a simple unified dual fitting argument on the same
semidefinite programming relaxation, which can essentially be obtained through
the first round of the Lasserre/Sum of Squares hierarchy.
As our main application, we show that the known coordination ratio bounds of
respectively $4, (3 + \sqrt{5})/2 \approx 2.618,$ and $32/15 \approx 2.133$ for
the scheduling game $R || \sum w_j C_j$ under the coordination mechanisms
Smith's Rule, Proportional Sharing and Rand (STOC 2011) can be extended to
congestion games and obtained through this approach. For the natural
restriction where the weight of each player is proportional to its processing
time on every resource, we show that the last bound can be improved from 2.133
to 2. This improvement can also be made for general instances when considering
the price of anarchy of the game, rather than the coordination ratio. As a
further application of the technique, we show that it recovers the tight bound
of $(3 + \sqrt{5})/2$ for the price of anarchy of weighted affine congestion
games and the Kawaguchi-Kyan bound of $(1+ \sqrt{2})/2$ for the pure price of
anarchy of $P || \sum w_j C_j$. In addition, this approach recovers the known
tight approximation ratio of $(3 + \sqrt{5})/2 \approx 2.618$ for a natural
local search algorithm for $R || \sum w_j C_j$, as well as the best currently
known combinatorial approximation algorithm for this problem achieving an
approximation ratio of $(5 + \sqrt{5})/4 + \varepsilon \approx 1.809 +
\varepsilon$.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:37:15 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Kashaev",
"Danish",
""
]
] |
2505.10083 | Chengsen Wang | Chengsen Wang, Qi Qi, Zhongwen Rao, Lujia Pan, Jingyu Wang, Jianxin
Liao | ChronoSteer: Bridging Large Language Model and Time Series Foundation
Model via Synthetic Data | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Conventional forecasting methods rely on unimodal time series data, limiting
their ability to exploit rich textual information. Recently, large language
models (LLMs) and time series foundation models (TSFMs) have demonstrated
powerful capability in textual reasoning and temporal modeling, respectively.
Integrating the strengths of both to construct a multimodal model that
concurrently leverages both temporal and textual information for future
inference has emerged as a critical research challenge. To address the scarcity
of event-series paired data, we propose a decoupled framework: an LLM is
employed to transform textual events into revision instructions, which are then
used to steer the output of TSFM. To implement this framework, we introduce
ChronoSteer, a multimodal TSFM that can be steered through textual revision
instructions, effectively bridging LLM and TSFM. Moreover, to mitigate the
shortage of cross-modal instruction-series paired data, we devise a two-stage
training strategy based on synthetic data. In addition, we also construct a
high-quality multimodal time series forecasting benchmark to address the
information leakage concerns during evaluation. After integrating with an LLM,
ChronoSteer, which is trained exclusively on synthetic data, achieves a 25.7%
improvement in prediction accuracy compared to the unimodal backbone and a
22.5% gain over the previous state-of-the-art multimodal method.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:37:23 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Chengsen",
""
],
[
"Qi",
"Qi",
""
],
[
"Rao",
"Zhongwen",
""
],
[
"Pan",
"Lujia",
""
],
[
"Wang",
"Jingyu",
""
],
[
"Liao",
"Jianxin",
""
]
] |
2505.10084 | Delfina G\'omez | Pablo Benavent-Ocejo, Delfina G\'omez, Mar\'ia-Eugenia
P\'erez-Mart\'inez | Longitudinal oscillations for eigenfunctions in rod like structures | null | null | null | null | math.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We consider the spectrum of the Laplace operator on 3D rod structures, with a
small cross section depending on a small parameter $\varepsilon$. The boundary
conditions are of Dirichlet type on the basis of this structure and Neumann on
the lateral boundary. We focus on the low frequencies. We study the asymptotic
behavior of the eigenvalues and associated eigenfunctions, which are approached
as $\varepsilon\to 0$ by those of a 1D model with Dirichlet boundary
conditions, but which takes into account the geometry of the domain. Explicit
and numerical computations enlighten the interest of this study, when the
parameter becomes smaller. At the same time they show that in order to capture
oscillations in the transverse direction we need to deal with the high
frequencies. For prism like domains, we show the different asymptotic behavior
of the spectrum depending on the boundary conditions.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:38:32 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Benavent-Ocejo",
"Pablo",
""
],
[
"Gómez",
"Delfina",
""
],
[
"Pérez-Martínez",
"María-Eugenia",
""
]
] |
2505.10085 | Stephan Zieger | Stephan Zieger and Hannah Richta | DB InfraGO's Automated Dispatching Assistant ADA-PMB | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | As railway infrastructure manager, DB InfraGO AG is faced with the challenge
of offering fluid and punctual operation despite rising demand and increased
construction activity. The high capacity utilisation, especially in the core
network sections, causes delays to be propagated quickly and widely across the
entire network. Up to now, conflicts between train runs can be identified
automatically, but dispatching measures have been based on past human
experience.
An automated dispatching assistance system is currently being piloted to
provide support for train dispatchers in their work. The aim is to offer them
helpful dispatching recommendations, particularly in stressful situations with
a high conflict density in the network section under consideration, in order to
ensure the most efficient operation of the system.
The recommendations are currently displayed separately alongside the central
control system. In future, they will be integrated into the central control
system, which will significantly simplify communication between the train
dispatcher and signal setter. Further development steps for the integration
process are also presented and discussed.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:39:28 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zieger",
"Stephan",
""
],
[
"Richta",
"Hannah",
""
]
] |
2505.10086 | Kui Ji | Xie yufang, Ji shanshan, Xu jing and Ji Kui | On the quasi-similarity of operators with flag structure | null | null | null | null | math.FA | http://creativecommons.org/licenses/by/4.0/ | Let $\mathcal{A}$ denote by the operator class with some proper conditions.
%satisfying that for any two operators $T,\tilde{T}$ in $\mathcal{A}$, the
non-zero intertwining operator between $T$ and $\tilde{T}$ has dense range.
Then by taking the operators in $\mathcal{A}$ as atoms and using the flag
structure as bondings, we introduce a new operator class. For operators with
certain properties in this new class, we prove that the operator matrix of the
intertwining operator is of the upper-triangular form. According to this
critical result, we firstly show that the strongly irreducible operators in the
new class preserve strong irreducibility under quasi-similarity, which gives a
partial answer to the question proposed by C.L. Jiang. Also, when $\mathcal{A}$
is backward weighted shift operators class, we prove that the quasi-similarity
between operators in the new class implies the similarity relation, which
partially answers the question proposed by D.A. Herrero. Finally, we describe
some properties of intertwining operators in term of geometric language.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:39:52 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"yufang",
"Xie",
""
],
[
"shanshan",
"Ji",
""
],
[
"jing",
"Xu",
""
],
[
"Kui",
"Ji",
""
]
] |
2505.10087 | Dawid Strzelczyk | Dawid Strzelczyk and Gregor Kosec and Maciej Matyka | The role of porosity in the transition to inertial regime in porous
media flows | null | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | In this work we investigate the fundamental physical mechanism of the
transition from Darcy to inertial (Darcy-Forchheimer) regime in the flow
through porous media from the perspective of the vortex formation and growth.
We focus on investigating the influence of porosity on the tortuosity--Reynolds
number relation for periodic stochastic and simple cubic systems by numerically
solving the Navier-Stokes equations in the pore-scale. We observe that the
tortuosity defined by integrals over the fluid volume exhibits similar behavior
in both types of systems. We also show that the tortuosity based on averaging
of the streamlines length diverges from the volume-integrated one when the
inertia onset takes place. We show that the discrepancy between those two
tortuosities at increasing Reynolds number carries information about the
geometrical confinement within the porous sample and is largely governed by the
growth of the volumes occupied by vortices and their kinetic energy. Our
results shed new light on the fundamental mechanisms of the transition between
Darcy and the inertial regime and bring a better understanding of the role of
the porous media geometry in this transition.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:40:44 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Strzelczyk",
"Dawid",
""
],
[
"Kosec",
"Gregor",
""
],
[
"Matyka",
"Maciej",
""
]
] |
2505.10088 | YunCheng Guo | Yuncheng Guo, Xiaodong Gu | MMRL++: Parameter-Efficient and Interaction-Aware Representation
Learning for Vision-Language Models | Due to the limitation "The abstract field cannot be longer than 1,920
characters", the abstract appearing here is slightly shorter than that in the
PDF file | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large-scale pre-trained Vision-Language Models (VLMs) have significantly
advanced transfer learning across diverse tasks. However, adapting these models
with limited few-shot data often leads to overfitting, undermining their
ability to generalize to new tasks. To address this, we propose Multi-Modal
Representation Learning (MMRL), which introduces a shared, learnable,
modality-agnostic representation space. MMRL generates space tokens projected
into both text and image encoders as representation tokens, enabling more
effective cross-modal interactions. Unlike prior methods that mainly optimize
class token features, MMRL inserts representation tokens into higher encoder
layers--where task-specific features are more prominent--while preserving
general knowledge in the lower layers. During training, both class and
representation features are jointly optimized: a trainable projection layer is
applied to representation tokens for task adaptation, while the projection
layer for class token remains frozen to retain pre-trained knowledge. To
further promote generalization, we introduce a regularization term aligning
class and text features with the frozen VLM's zero-shot features. At inference,
a decoupling strategy uses both class and representation features for base
tasks, but only class features for novel tasks due to their stronger
generalization. Building upon this, we propose MMRL++, a parameter-efficient
and interaction-aware extension that significantly reduces trainable parameters
and enhances intra-modal interactions--particularly across the layers of
representation tokens--allowing gradient sharing and instance-specific
information to propagate more effectively through the network. Extensive
experiments on 15 datasets demonstrate that MMRL and MMRL++ consistently
outperform state-of-the-art methods, achieving a strong balance between
task-specific adaptation and generalization.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:43:53 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guo",
"Yuncheng",
""
],
[
"Gu",
"Xiaodong",
""
]
] |
2505.10089 | Wei Liu | Wei Liu, Sony Trenous, Leonardo F. R. Ribeiro, Bill Byrne, Felix
Hieber | XRAG: Cross-lingual Retrieval-Augmented Generation | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose XRAG, a novel benchmark designed to evaluate the generation
abilities of LLMs in cross-lingual Retrieval-Augmented Generation (RAG)
settings where the user language does not match the retrieval results. XRAG is
constructed from recent news articles to ensure that its questions require
external knowledge to be answered. It covers the real-world scenarios of
monolingual and multilingual retrieval, and provides relevancy annotations for
each retrieved document. Our novel dataset construction pipeline results in
questions that require complex reasoning, as evidenced by the significant gap
between human and LLM performance. Consequently, XRAG serves as a valuable
benchmark for studying LLM reasoning abilities, even before considering the
additional cross-lingual complexity. Experimental results on five LLMs uncover
two previously unreported challenges in cross-lingual RAG: 1) in the
monolingual retrieval setting, all evaluated models struggle with response
language correctness; 2) in the multilingual retrieval setting, the main
challenge lies in reasoning over retrieved information across languages rather
than generation of non-English text.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:47:55 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Liu",
"Wei",
""
],
[
"Trenous",
"Sony",
""
],
[
"Ribeiro",
"Leonardo F. R.",
""
],
[
"Byrne",
"Bill",
""
],
[
"Hieber",
"Felix",
""
]
] |
2505.10090 | Reza Nikandish | M. Badie, R. Nikandish, M. Pirniian | Sombor index of clean graphs | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | Let $G = (V, E)$ be a graph with the vertex set $V (G)$ and edge set $E(G)$.
The Sombor index of $G$, $SO(G)$, is defined as $\sum_{uv\in E(G)}
\sqrt{deg(u)^2 + deg(v)^2}$, where $deg(u)$ is the degree of vertex $u$ in $V
(G)$. The clean graph of a ring R, denoted by $Cl(R)$, is a graph with vertex
set $\{(e, u) : e \in Id(R), u \in U(R)\}$ and two distinct vertices $(e, u)$
and$(f, v)$ are adjacent if and only if $ef = 0$ or $uv = 1$ ($Id(R)$ and
$U(R)$ are the sets of idempotents and unit elements of R, respectively). The
induced subgraph on $\{(e, u) : e \in Id^{*}(R), u \in U(R)\}$ is denoted by
$Cl_2(R)$. In this paper, $SO(Cl2(\mathbb{Z}_n))$, for different values of the
positive integer $n$, is investigated.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:48:35 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Badie",
"M.",
""
],
[
"Nikandish",
"R.",
""
],
[
"Pirniian",
"M.",
""
]
] |
2505.10091 | Colin Meulblok | Colin M. Meulblok, Hadrien Bense, M. Caelen, Martin van Hecke | Accelerated snapping of slender beams under lateral forcing | 10 pages, 8 figures | null | null | null | cond-mat.soft | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The hysteretic snapping under lateral forcing of a compressed, buckled beam
is fundamental for many devices and mechanical metamaterials. For a single-tip
lateral pusher, an important limitation is that snapping requires the pusher to
cross the centerline of the beam. Here, we show that dual-tip pushers allow
accelerated snapping, where the beam snaps before the pusher reaches the
centerline. As a consequence, we show that when a buckled beam under increased
compression comes in contact with a dual-tip pusher, it can snap to the
opposite direction -- this is impossible with a single-tip pusher.
Additionally, we reveal a novel two-step snapping regime, in which the beam
sequentially loses contact with the two tips of the dual-tip pusher. To
characterize this class of snapping instabilities, we employ a systematic modal
expansion of the beam shape. This expansion allows us to capture and analyze
the transition from one-step to two-step snapping geometrically. Finally we
demonstrate how to maximize the distance between the pusher and the beam's
centerline at the moment of snapping. Together, our work opens up a new avenue
for quantitatively and qualitatively controlling and modifying the snapping of
buckled beams, with potential applications in mechanical sensors, actuators,
and metamaterials.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:49:23 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Meulblok",
"Colin M.",
""
],
[
"Bense",
"Hadrien",
""
],
[
"Caelen",
"M.",
""
],
[
"van Hecke",
"Martin",
""
]
] |
2505.10092 | Sylvain de L\'es\'eleuc | Vincent Lienhard, Romain Martin, Yuki Torii Chew, Takafumi Tomita,
Kenji Ohmori and Sylvain de L\'es\'eleuc | Generation of Motional Squeezed States for Neutral Atoms in Optical
Tweezers | null | null | null | null | physics.atom-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Optical tweezers are a powerful tool for the precise positioning of a variety
of small objects, including single neutral atoms. Once trapped, atoms can be
cooled to the motional ground state of the tweezers. For a more advanced
control of their spatial wavefunction, we report here a simple method to
squeeze their motion, and the protocol to measure the squeezing factor based on
momentum spreading estimation. We explore the limitations set by the technical
imperfections of the tweezers, as well as the more fundamental limit set by
their anharmonicity, and finally demonstrate a squeezing of 5.8 dB. The
implementation of motional squeezing allows to push back the limit set by the
position quantum noise and thus to explore more extreme situations requiring
atoms positioned with nanometric precision.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:51:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Lienhard",
"Vincent",
""
],
[
"Martin",
"Romain",
""
],
[
"Chew",
"Yuki Torii",
""
],
[
"Tomita",
"Takafumi",
""
],
[
"Ohmori",
"Kenji",
""
],
[
"de Léséleuc",
"Sylvain",
""
]
] |
2505.10093 | Hsuan Lei Shao | Hsuan-Lei Shao | From Text to Network: Constructing a Knowledge Graph of Taiwan-Based
China Studies Using Generative AI | 4 pages, 4 figures | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Taiwanese China Studies (CS) has developed into a rich, interdisciplinary
research field shaped by the unique geopolitical position and long standing
academic engagement with Mainland China. This study responds to the growing
need to systematically revisit and reorganize decades of Taiwan based CS
scholarship by proposing an AI assisted approach that transforms unstructured
academic texts into structured, interactive knowledge representations. We apply
generative AI (GAI) techniques and large language models (LLMs) to extract and
standardize entity relation triples from 1,367 peer reviewed CS articles
published between 1996 and 2019. These triples are then visualized through a
lightweight D3.js based system, forming the foundation of a domain specific
knowledge graph and vector database for the field. This infrastructure allows
users to explore conceptual nodes and semantic relationships across the corpus,
revealing previously uncharted intellectual trajectories, thematic clusters,
and research gaps. By decomposing textual content into graph structured
knowledge units, our system enables a paradigm shift from linear text
consumption to network based knowledge navigation. In doing so, it enhances
scholarly access to CS literature while offering a scalable, data driven
alternative to traditional ontology construction. This work not only
demonstrates how generative AI can augment area studies and digital humanities
but also highlights its potential to support a reimagined scholarly
infrastructure for regional knowledge systems.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:51:53 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Shao",
"Hsuan-Lei",
""
]
] |
2505.10094 | Sofia Gallego | Sofia G. Gallego, Simona Mei, Christopher Martin, Donal O'Sullivan,
Emanuele Daddi, Dominika Wylezalek, Nicholas Seymour | Coevolution of halo and quasar properties in dense environments: CARLA
J1017+6116 at z=2.8 | 15 pages, 14 figures. Accepted in A&A | null | null | null | astro-ph.GA astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radio-loud active galactic nuclei, in particular radio-loud quasars, are
fueled by accretion onto supermassive black holes and are among the most
energetic sources in the Universe. While their impact on their surroundings -
from the interstellar medium to the circumgalactic medium - is well recognized,
the specific mechanisms remain uncertain. In this study we analyze deep Keck
Cosmic Web Imager observations of the Lyman-alpha (Lya) halo surrounding the
radio-loud quasar at the center of the cluster CARLA J1017+6116 at redshift z =
2.8. As is known from previous observations, the cluster hosts a high fraction
of early-type galaxies, and the star formation of its spectroscopically
confirmed cluster members is typical of or higher than that of galaxies on the
main sequence. We find that the Lya halo extends at least 16 arcsec (128 pkpc)
down to a surface brightness level of 1e-19 erg/s/cm^2/arcsec^2, with a total
observed Lya luminosity of log10(L/Lsun) = 43.35 +- 0.05. The halo has distinct
kinematic regions with asymmetries suggestive of complex interactions between
the quasar and the intracluster medium, possibly driven by a combination of
biconical feedback and episodic activity. Despite the quasar classification,
our reanalysis of very long baseline interferometry data finds no evidence of
extended jet structures; we instead find compact and variable radio emission
that could indicate episodic jet activity or suppression by the dense
interstellar medium. Combining these observations with imaging obtained with
the Hubble Space Telescope, we identified one Lya-emitting source within the
quasar halo. While mechanical feedback from a jet appears limited or episodic,
radiative feedback likely plays a dominant role in shaping the extended Lya
halo, highlighting the complex interplay between quasar-driven processes and
the surrounding dense environment.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:54:08 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Gallego",
"Sofia G.",
""
],
[
"Mei",
"Simona",
""
],
[
"Martin",
"Christopher",
""
],
[
"O'Sullivan",
"Donal",
""
],
[
"Daddi",
"Emanuele",
""
],
[
"Wylezalek",
"Dominika",
""
],
[
"Seymour",
"Nicholas",
""
]
] |
2505.10095 | Philipp Zilk | Thomas Apel, Philipp Zilk | Error Estimates and Graded Mesh Refinement for Isogeometric Analysis on
Polar Domains with Corners | 44 pages, 11 figures | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Isogeometric analysis (IGA) enables exact representations of computational
geometries and higher-order approximation of PDEs. In non-smooth domains,
however, singularities near corners limit the effectiveness of IGA, since
standard methods typically fail to achieve optimal convergence rates. These
constraints can be addressed through local mesh refinement, but existing
approaches require breaking the tensor-product structure of splines, which
leads to increased implementation complexity.
This work introduces a novel local refinement strategy based on a polar
parameterization, in which one edge of the parametric square is collapsed into
the corner. By grading the standard mesh toward the collapsing edge, the
desired locality near the singularity is obtained while maintaining the
tensor-product structure. A mathematical analysis and numerical tests show that
the resulting isogeometric approximation achieves optimal convergence rates
with suitable grading parameters.
Polar parameterizations, however, suffer from a lack of regularity at the
polar point, making existing standard isogeometric approximation theory
inapplicable. To address this, a new framework is developed for deriving error
estimates on polar domains with corners. This involves the construction of
polar function spaces on the parametric domain and a modified projection
operator onto the space of $C^0$-smooth polar splines. The theoretical results
are verified by numerical experiments confirming both the accuracy and
efficiency of the proposed approach.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:54:13 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Apel",
"Thomas",
""
],
[
"Zilk",
"Philipp",
""
]
] |
2505.10096 | A. Jain | A. Jain, M. Imran, and G. Saxena | Theoretical Investigation of {\alpha}-decay in Heavy and Superheavy
Isomers | 24 pages, 6 figures, 10 tables | null | null | null | nucl-th | http://creativecommons.org/publicdomain/zero/1.0/ | The heavy and superheavy elements of the periodic table predominately
disintegrate by {\alpha}-decay, facilitating transitions mainly between ground
states and occasionally involving isomeric states. This study focuses on
estimating the half-lives of {\alpha}-transitions both from and to isomeric
states, using a recently refined formula which shows excellent agreement with
experimental data when isospin of parent nucleus as well as angular momentum
taken away by the {\alpha} particle are incorporated. These findings provide
valuable insights for upcoming experimental investigations of isomeric states.
Additionally, the study predicts potential {\alpha}-decay in several
yet-unobserved isomeric nuclei, contributing to a deeper understanding of
nuclear structure in heavy and superheavy elements.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:58:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Jain",
"A.",
""
],
[
"Imran",
"M.",
""
],
[
"Saxena",
"G.",
""
]
] |
2505.10097 | Lujia Wang | Meirun Chen, Reza Naserasr, Lujia Wang, Sanming Zhou | Odd Hadwiger's conjecture for the complements of Kneser graphs | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | A generalization of the four-color theorem, Hadwiger's conjecture is
considered as one of the most important and challenging problems in graph
theory, and odd Hadwiger's conjecture is a strengthening of Hadwiger's
conjecture by way of signed graphs. In this paper, we prove that odd Hadwiger's
conjecture is true for the complements $\overline{K}(n,k)$ of the Kneser graphs
$K(n,k)$, where $n\geq 2k \ge 4$. This improves a result of G. Xu and S. Zhou
(2017) which states that Hadwiger's conjecture is true for this family of
graphs. Moreover, we prove that $\overline{K}(n,k)$ contains a 1-shallow
complete minor of a special type with order no less than the chromatic number
$\chi(\overline{K}(n,k))$, and in the case when $7 \le 2k+1 \le n \le 3k-1$ the
gap between the odd Hadwiger number and chromatic number of $\overline{K}(n,k)$
is $\Omega(1.5^{k})$.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 08:58:13 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Chen",
"Meirun",
""
],
[
"Naserasr",
"Reza",
""
],
[
"Wang",
"Lujia",
""
],
[
"Zhou",
"Sanming",
""
]
] |
2505.10098 | Anja Heim | Anja Heim and Thomas Lang and Christoph Heinzl | Exploring Large Quantities of Secondary Data from High-Resolution
Synchrotron X-ray Computed Tomography Scans Using AccuStripes | in Review at Journal of Nondestructive Evaluation | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The analysis of secondary quantitative data extracted from high-resolution
synchrotron X-ray computed tomography scans represents a significant challenge
for users. While a number of methods have been introduced for processing large
three-dimensional images in order to generate secondary data, there are only a
few techniques available for simple and intuitive visualization of such data in
their entirety. This work employs the AccuStripes visualization technique for
that purpose, which enables the visual analysis of secondary data represented
by an ensemble of univariate distributions. It supports different schemes for
adaptive histogram binnings in combination with several ways of rendering
aggregated data and it allows the interactive selection of optimal visual
representations depending on the data and the use case. We demonstrate the
usability of AccuStripes on a high-resolution synchrotron scan of a
particle-reinforced metal matrix composite sample, containing more than 20
million particles. Through AccuStripes, detailed insights are facilitated into
distributions of derived particle characteristics of the entire sample.
Furthermore, research questions such as how the overall shape of the particles
is or how homogeneously they are distributed across the sample can be answered.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:00:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Heim",
"Anja",
""
],
[
"Lang",
"Thomas",
""
],
[
"Heinzl",
"Christoph",
""
]
] |
2505.10099 | Matias Quiroz | Sarat Moka, Matias Quiroz, Vali Asimit, Samuel Muller | A Scalable Gradient-Based Optimization Framework for Sparse
Minimum-Variance Portfolio Selection | null | null | null | null | stat.ML cs.LG math.OC q-fin.PM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Portfolio optimization involves selecting asset weights to minimize a
risk-reward objective, such as the portfolio variance in the classical
minimum-variance framework. Sparse portfolio selection extends this by imposing
a cardinality constraint: only $k$ assets from a universe of $p$ may be
included. The standard approach models this problem as a mixed-integer
quadratic program and relies on commercial solvers to find the optimal
solution. However, the computational costs of such methods increase
exponentially with $k$ and $p$, making them too slow for problems of even
moderate size. We propose a fast and scalable gradient-based approach that
transforms the combinatorial sparse selection problem into a constrained
continuous optimization task via Boolean relaxation, while preserving
equivalence with the original problem on the set of binary points. Our
algorithm employs a tunable parameter that transmutes the auxiliary objective
from a convex to a concave function. This allows a stable convex starting
point, followed by a controlled path toward a sparse binary solution as the
tuning parameter increases and the objective moves toward concavity. In
practice, our method matches commercial solvers in asset selection for most
instances and, in rare instances, the solution differs by a few assets whilst
showing a negligible error in portfolio variance.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:01:07 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Moka",
"Sarat",
""
],
[
"Quiroz",
"Matias",
""
],
[
"Asimit",
"Vali",
""
],
[
"Muller",
"Samuel",
""
]
] |
2505.10100 | Joachim K\"onig | Joachim K\"onig | Unramified extensions of quadratic number fields with Galois group
$2.A_n$ | null | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | We realize infinitely many covering groups $2.A_n$ (where $A_n$ is the
alternating group) as the Galois group of everywhere unramified Galois
extensions over infinitely many quadratic number fields. After several
predecessor works investigating special cases or proving conditional results in
this direction, these are the first unramified realizations of infinitely many
of these groups.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:02:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"König",
"Joachim",
""
]
] |
2505.10101 | Jongmin Jung | Jongmin Jung and Dasaem Jeong | LAV: Audio-Driven Dynamic Visual Generation with Neural Compression and
StyleGAN2 | Paper accepted at ISEA 2025, The 30th International Symposium on
Electronic/Emerging Art, Seoul, Republic of Korea, 23 - 29 May 2025 | null | null | null | cs.SD cs.AI cs.GR cs.MM eess.AS | http://creativecommons.org/licenses/by/4.0/ | This paper introduces LAV (Latent Audio-Visual), a system that integrates
EnCodec's neural audio compression with StyleGAN2's generative capabilities to
produce visually dynamic outputs driven by pre-recorded audio. Unlike previous
works that rely on explicit feature mappings, LAV uses EnCodec embeddings as
latent representations, directly transformed into StyleGAN2's style latent
space via randomly initialized linear mapping. This approach preserves semantic
richness in the transformation, enabling nuanced and semantically coherent
audio-visual translations. The framework demonstrates the potential of using
pretrained audio compression models for artistic and computational
applications.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:04:12 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Jung",
"Jongmin",
""
],
[
"Jeong",
"Dasaem",
""
]
] |
2505.10102 | Elio Marconi | Elio Marconi, Laura V. Spinolo | Nonlocal Generalized Aw-Rascle-Zhang model: well-posedness and singular
limit | 25 pages | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We discuss a nonlocal version of the Generalized Aw-Rascle-Zhang model, a
second-order vehicular traffic model where the empty road velocity is a
Lagrangian marker governed by a transport equation. The evolution of the car
density is described by a continuity equation where the drivers' velocity
depends on both the empty road velocity and the convolution of the car density
with an anisotropic kernel. We establish existence and uniqueness results. When
the convolution kernel is replaced by a Dirac Delta, the nonlocal model
formally boils down to the classical (local) Generalized Aw-Rascle-Zhang model,
which consists of a conservation law coupled with a transport equation. In the
case of exponential kernels, we establish convergence in the nonlocal-to-local
limit by proving an Oleinik-type estimate for the convolution term. To the best
of our knowledge, this is the first nonlocal-to-local limit result for a system
of two non-decoupling equations with a nonlocal flux function.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:04:19 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Marconi",
"Elio",
""
],
[
"Spinolo",
"Laura V.",
""
]
] |
2505.10103 | Janine Birnbaum | J. Birnbaum, J. Schauroth, J. Weaver, J. E. Kendrick, A. Lamur, Y.
Lavall\'ee | Shear-enhanced permeability development of magma vesiculating in
cylindrical conduits | null | null | null | null | physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | During magma vesiculation, permeability is established when growing bubbles
begin to form connected networks, which allow fluids to percolate. This
percolation threshold controls the relative rates between magma ascent and
volatile exsolution, which in turn dictate eruptive style. Percolation is
controlled primarily by total vesicularity and shear conditions. We performed
vesiculation experiments on rhyolitic glass in a spatially confined,
cylindrical, conduit-like geometry. The amount of shear experienced by the
sample is controlled by varying the sample and confining diameters to allow for
various degrees of free (isotropic) followed by confined (anisotropic)
expansion. Pore anisotropy develops sub-parallel to the flow direction. We
measure the total and connected porosity and permeability of the vesiculated
samples. We observe two regimes of behavior for samples dominated by 1)
isotropic expansion, in which the onset of percolation corresponds with the
beginning of shear deformation and 2) anisotropic expansion which shows low
percolation thresholds (<20%) with a near-constant permeability even with
increasing porosity. We find a good correlation between connected porosity and
permeability, with distinct trends for sheared and unsheared samples. The
development of anisotropy in vesiculating magmas is ubiquitous in magmatic
networks and our data highlight the importance of considering the in situ shear
conditions when investigating the percolation threshold.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:06:40 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Birnbaum",
"J.",
""
],
[
"Schauroth",
"J.",
""
],
[
"Weaver",
"J.",
""
],
[
"Kendrick",
"J. E.",
""
],
[
"Lamur",
"A.",
""
],
[
"Lavallée",
"Y.",
""
]
] |
2505.10104 | Elio Marconi | Elio Marconi, Laura V. Spinolo | Well-posedness results for the Generalized Aw-Rascle-Zhang model | 13 pages | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We establish existence, uniqueness and stability results for the so-called
Generalized Aw-Rascle-Zhang model, a second order traffic model introduced by
Fan, Herty and Seibold. Our analysis is motivated by the companion paper
'Nonlocal Generalized Aw-Rascle-Zhang model: well-posedness and singular
limit'.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:12:08 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Marconi",
"Elio",
""
],
[
"Spinolo",
"Laura V.",
""
]
] |
2505.10105 | Zibin Dong | Zibin Dong, Fei Ni, Yifu Yuan, Yinchuan Li, Jianye Hao | EmbodiedMAE: A Unified 3D Multi-Modal Representation for Robot
Manipulation | null | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present EmbodiedMAE, a unified 3D multi-modal representation for robot
manipulation. Current approaches suffer from significant domain gaps between
training datasets and robot manipulation tasks, while also lacking model
architectures that can effectively incorporate 3D information. To overcome
these limitations, we enhance the DROID dataset with high-quality depth maps
and point clouds, constructing DROID-3D as a valuable supplement for 3D
embodied vision research. Then we develop EmbodiedMAE, a multi-modal masked
autoencoder that simultaneously learns representations across RGB, depth, and
point cloud modalities through stochastic masking and cross-modal fusion.
Trained on DROID-3D, EmbodiedMAE consistently outperforms state-of-the-art
vision foundation models (VFMs) in both training efficiency and final
performance across 70 simulation tasks and 20 real-world robot manipulation
tasks on two robot platforms. The model exhibits strong scaling behavior with
size and promotes effective policy learning from 3D inputs. Experimental
results establish EmbodiedMAE as a reliable unified 3D multi-modal VFM for
embodied AI systems, particularly in precise tabletop manipulation settings
where spatial perception is critical.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:12:17 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Dong",
"Zibin",
""
],
[
"Ni",
"Fei",
""
],
[
"Yuan",
"Yifu",
""
],
[
"Li",
"Yinchuan",
""
],
[
"Hao",
"Jianye",
""
]
] |
2505.10106 | Radouan Boukharfane | Radouan Boukharfane | Numerical Analysis of Pure and Blended Fuel Sonic Jets in a Mach 2
Crossflow | 19 | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | The injection of transverse jets into supersonic compressible crossflows
represents a fundamental configuration relevant to a spectrum of high-speed
applications. The intricate interactions arising between the crossflow and the
injected jet induce complex flow phenomena, including shock waves and vortical
structures, the characteristics of which are significantly contingent upon the
thermophysical properties of the injected fuel. While prior investigations have
addressed the influence of various fuels, a knowledge gap persists concerning
the behaviour of alternative and synthetic multicomponent fuels within this
flow regime. The present work employs high-fidelity large-eddy simulations
(LES) to examine the impact of ten distinct fuels-hydrogen, methane, ethylene,
ammonia, syngas mixture, and a synthetic blend, alongside several NH3/H2/N2
mixtures-on the macroscopic flow structures and mixing attributes within a
transverse sonic jet immersed in a Mach 2 crossflow. By maintaining a uniform
momentum flux ratio across the investigated cases, the study aims to isolate
the influence of the unique thermophysical properties of each fuel on the
windward mixing layer, a region critically important for initial entrainment
processes. The investigation quantifies the effects of molecular weight, heat
capacity ratio, and density on the development and evolution of coherent
structures through a detailed examination of instantaneous flow fields, vortex
dynamics, scalar distributions, and turbulence statistics. The results are
expected to provide pertinent insights into fuel-dependent mixing mechanisms in
supersonic flows, thereby contributing to the advancement of more efficient and
versatile propulsion systems.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:12:35 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Boukharfane",
"Radouan",
""
]
] |
2505.10107 | Luise Poley | Luise Poley, Ian Dyckes, Javier Fernandez-Tejero, Celeste Fleta,
Dennis Sperlich, Miguel Ullan | Investigating the impact of the diode edge geometry on the effective
active area of silicon sensors using AREA-X | 10 pages, 7 figures, published in Journal of Instrumentation | 2025 JINST 20 P05010 | 10.1088/1748-0221/20/05/P05010 10.1088/1748-0221/20/05/P05010
10.1088/1748-0221/20/05/P05010 | null | physics.ins-det hep-ex | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Sensors for particle tracking detectors are required to provide a maximum
active area in addition to fulfilling performance criteria concerning radiation
hardness, charge collection and operating conditions (e.g. leakage current,
depletion voltage and breakdown voltage). While the requirement for optimised
coverage within the tracking detector necessitates a slim sensor edge between
active region and physical sensor edge, wider edge regions were found to be
beneficial for the sensor performance during an early prototyping phase. In
order to study the impact of differently sized edge regions, test structures
were used to compare their individual active regions.
Measurements of each diode were performed using a micro-focused X-ray beam to
map its respective active area. This paper presents measurements of these test
structures using AREA-X showing that the active area of a silicon particle
tracking sensor does not only depend on the size of its bias ring, but also the
size and configuration of its edge structure.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:14:51 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Poley",
"Luise",
""
],
[
"Dyckes",
"Ian",
""
],
[
"Fernandez-Tejero",
"Javier",
""
],
[
"Fleta",
"Celeste",
""
],
[
"Sperlich",
"Dennis",
""
],
[
"Ullan",
"Miguel",
""
]
] |
2505.10108 | Xiangxian Luo | Lei Li, Xiangxian Luo, Yinchen Luo | A generalized discontinuous Hamilton Monte Carlo for transdimensional
sampling | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a discontinuous Hamilton Monte Carlo (DHMC) to
sample from dimensional varying distributions, and particularly the grand
canonical ensemble. The DHMC was proposed in [Biometrika, 107(2)] for
discontinuous potential where the variable has a fixed dimension. When the
dimension changes, there is no clear explanation of the volume-preserving
property, and the conservation of energy is also not necessary. We use a random
sampling for the extra dimensions, which corresponds to a measure transform. We
show that when the energy is corrected suitably for the trans-dimensional
Hamiltonian dynamics, the detailed balance condition is then satisfied. For the
grand canonical ensemble, such a procedure can be explained very naturally to
be the extra free energy change brought by the newly added particles, which
justifies the rationality of our approach. To sample the grand canonical
ensemble for interacting particle systems, the DHMC is then combined with the
random batch method to yield an efficient sampling method. In experiments, we
show that the proposed DHMC combined with the random batch method generates
samples with much less correlation when compared with the traditional
Metropolis-Hastings method.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:21:54 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Li",
"Lei",
""
],
[
"Luo",
"Xiangxian",
""
],
[
"Luo",
"Yinchen",
""
]
] |
2505.10109 | Christopher Wareing | Christopher J. Wareing, Alasdair T. Roy, Matthew Golden, Roman O.
Grigoriev and Steven M. Tobias | Data-driven discovery of the equations of turbulent convection | 29 pages, 8 figures, 5 tables, accepted for publication in GAFD.
Associated data and methods available at https://doi.org/10.5518/1577 | null | null | null | astro-ph.SR physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | We compare the efficiency and ease-of-use of the Sparse Identification of
Nonlinear Dynamics (SINDy) algorithm and Sparse Physics-Informed Discovery of
Empirical Relations (SPIDER) framework in recovering the relevant governing
equations and boundary conditions from data generated by direct numerical
simulations (DNS) of turbulent convective flows. In the former case, a
weak-form implementation pySINDy is used. Time-dependent data for two- (2D) and
three-dimensional (3D) DNS simulation of Rayleigh-Benard convection and
convective plane Couette flow is generated using the Dedalus PDE framework for
spectrally solving differential equations. Using pySINDy we are able to recover
the governing equations of 2D models of Rayleigh-Benard convection at Rayleigh
numbers, R, from laminar, through transitional to moderately turbulent flow
conditions, albeit with increasing difficulty with larger Rayleigh number,
especially in recovery of the diffusive terms (with coefficient magnitude
proportional to 1/R^0.5). SPIDER requires a much smaller library of terms and
we are able to recover more easily the governing equations for a wider range of
R in 2D and 3D convection and plane flow models and go on to recover
constraints (the incompressibility condition) and boundary conditions,
demonstrating the benefits and capabilities of SPIDER to go beyond pySINDy for
these fluid problems governed by second-order PDEs. [We] demonstrat[e] the
potential of machine-learning methods to validate numerical solvers and
solutions for such flow problems. We also find that properties of the flow,
specifically the correlation time and spatial scales, should inform the initial
selection of spatiotemporal subdomain sizes [abbreviated]
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:23:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wareing",
"Christopher J.",
""
],
[
"Roy",
"Alasdair T.",
""
],
[
"Golden",
"Matthew",
""
],
[
"Grigoriev",
"Roman O.",
""
],
[
"Tobias",
"Steven M.",
""
]
] |
2505.10110 | Salvatore F.E. Oliviero | Lorenzo Leone, Salvatore F.E. Oliviero, Alioscia Hamma, Jens Eisert,
Lennart Bittel | The non-Clifford cost of random unitaries | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent years have enjoyed a strong interest in exploring properties and
applications of random quantum circuits. In this work, we explore the ensemble
of $t$-doped Clifford circuits on $n$ qubits, consisting of Clifford circuits
interspersed with $t$ single-qubit non-Clifford gates. We establish rigorous
convergence bounds towards unitary $k$-designs, revealing the intrinsic cost in
terms of non-Clifford resources in various flavors. First, we analyze the
$k$-th order frame potential, which quantifies how well the ensemble of doped
Clifford circuits is spread within the unitary group. We prove that a quadratic
doping level, $t = \tilde{\Theta}(k^2)$, is both necessary and sufficient to
approximate the frame potential of the full unitary group. As a consequence, we
refine existing upper bounds on the convergence of the ensemble towards state
$k$-designs. Second, we derive tight bounds on the convergence of $t$-doped
Clifford circuits towards relative-error $k$-designs, showing that $t =
\tilde{\Theta}(nk)$ is both necessary and sufficient for the ensemble to form a
relative $\varepsilon$-approximate $k$-design. Similarly, $t =
\tilde{\Theta}(n)$ is required to generate pseudo-random unitaries. All these
results highlight that generating random unitaries is extremely costly in terms
of non-Clifford resources, and that such ensembles fundamentally lie beyond the
classical simulability barrier. Additionally, we introduce doped-Clifford
Weingarten functions to derive analytic expressions for the twirling operator
over the ensemble of random doped Clifford circuits, and we establish their
asymptotic behavior in relevant regimes.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:28:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Leone",
"Lorenzo",
""
],
[
"Oliviero",
"Salvatore F. E.",
""
],
[
"Hamma",
"Alioscia",
""
],
[
"Eisert",
"Jens",
""
],
[
"Bittel",
"Lennart",
""
]
] |
2505.10111 | Jeonghyun Woo | Jeonghyun Woo, Joyce Qu, Gururaj Saileshwar, Prashant J. Nair | When Mitigations Backfire: Timing Channel Attacks and Defense for
PRAC-Based RowHammer Mitigations | This paper was originally submitted to ISCA 2025
(https://iscaconf.org/isca2025/) on November 22, 2024, accepted on March 21,
2025, and will be presented on June 24, 2025. It is 19 pages long, including
references and an appendix. The artifact is available at:
https://github.com/STAR-Laboratory/PRAC_TC_ISCA25 | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Per Row Activation Counting (PRAC) has emerged as a robust framework for
mitigating RowHammer (RH) vulnerabilities in modern DRAM systems. However, we
uncover a critical vulnerability: a timing channel introduced by the Alert
Back-Off (ABO) protocol and Refresh Management (RFM) commands. We present
PRACLeak, a novel attack that exploits these timing differences to leak
sensitive information, such as secret keys from vulnerable AES implementations,
by monitoring memory access latencies.
To counter this, we propose Timing-Safe PRAC (TPRAC), a defense that
eliminates PRAC-induced timing channels without compromising RH mitigation
efficacy. TPRAC uses Timing-Based RFMs, issued periodically and independent of
memory activity. It requires only a single-entry in-DRAM mitigation queue per
DRAM bank and is compatible with existing DRAM standards. Our evaluations
demonstrate that TPRAC closes timing channels while incurring only 3.4%
performance overhead at the RH threshold of 1024.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:28:46 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Woo",
"Jeonghyun",
""
],
[
"Qu",
"Joyce",
""
],
[
"Saileshwar",
"Gururaj",
""
],
[
"Nair",
"Prashant J.",
""
]
] |
2505.10112 | Jayesh Goswami | Jayesh Goswami, Shataneek Banerjee, Snigdhadev Chakraborty, Srestha
Roy, Atanu Ghosh, Mrutyunjaya Rath, Agniva Das, Mukul Sagar, Krishna Kumari
Swain, Prasanta Pal, and Basudev Roy | A label-free sub-diffractive technique for 3D intracellular tomography
using thermally induced convection currents | null | null | null | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conventionally, 3-dimensional cellular tomography can be done with light
sheet or multi-angle observations. Recently, a new technique was introduced
where the cell was rotated using convection currents to visualize the outer
periphery (Liu et al., Nano Lett., 2023, 23, 5148). However, the work falls
short of actually observing intracellular objects like organelles etc. In this
manuscript, we modify the technique by relying on computer vision algorithm
called Contrast Limited Adaptive Histogram Equalisation (CLAHE) to improve the
contrast for better detection of intra-cellular points, and then use optical
flow detection technique to extract the in-plane speed of the point. This then
is used to extract the vertical location, knowing that at the bottom part of
the sphere, the point would be moving in one direction, close to the center
there would be much less motion, while in the top portion of the sphere, the
point would be moving in the reverse direction than the bottom. The velocity
allows the exact localisation of the point in the vertical direction. This
process allows for sub-diffractive intracellular tomography. This technique can
further allow high-resolution detection of fluorescent molecules inside the
cell also, when combined with convective flows.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:33:32 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Goswami",
"Jayesh",
""
],
[
"Banerjee",
"Shataneek",
""
],
[
"Chakraborty",
"Snigdhadev",
""
],
[
"Roy",
"Srestha",
""
],
[
"Ghosh",
"Atanu",
""
],
[
"Rath",
"Mrutyunjaya",
""
],
[
"Das",
"Agniva",
""
],
[
"Sagar",
"Mukul",
""
],
[
"Swain",
"Krishna Kumari",
""
],
[
"Pal",
"Prasanta",
""
],
[
"Roy",
"Basudev",
""
]
] |
2505.10113 | Xinlan Yan | Xinlan Yan, Di Wu, Yibin Lei, Christof Monz, Iacer Calixto | What Does Neuro Mean to Cardio? Investigating the Role of Clinical
Specialty Data in Medical LLMs | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | In this paper, we introduce S-MedQA, an English medical question-answering
(QA) dataset for benchmarking large language models in fine-grained clinical
specialties. We use S-MedQA to check the applicability of a popular hypothesis
related to knowledge injection in the knowledge-intense scenario of medical QA,
and show that: 1) training on data from a speciality does not necessarily lead
to best performance on that specialty and 2) regardless of the specialty
fine-tuned on, token probabilities of clinically relevant terms for all
specialties increase consistently. Thus, we believe improvement gains come
mostly from domain shifting (e.g., general to medical) rather than knowledge
injection and suggest rethinking the role of fine-tuning data in the medical
domain. We release S-MedQA and all code needed to reproduce all our experiments
to the research community.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:35:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Yan",
"Xinlan",
""
],
[
"Wu",
"Di",
""
],
[
"Lei",
"Yibin",
""
],
[
"Monz",
"Christof",
""
],
[
"Calixto",
"Iacer",
""
]
] |
2505.10114 | Seungjae Lee | Seungjae Lee and Lennart J. Kuklinski and Marc Timme | Extreme Synchronization Transitions | null | Nature Communications 16, 4505 (2025) | 10.1038/s41467-025-59729-8 | null | nlin.AO cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | Across natural and human-made systems, transition points mark sudden changes
of order and are thus key to understanding overarching system features.
Motivated by recent experimental observations, we here uncover an intriguing
class of transitions in coupled oscillators, extreme synchronization
transitions, from asynchronous disordered states to synchronous states with
almost completely ordered phases. Whereas such a transition appears like
discontinuous or explosive phase transitions, it exhibits markedly distinct
features. First, the transition occurs already in finite systems of $N$ units
and so constitutes an intriguing bifurcation of multi-dimensional systems
rather than a genuine phase transition that emerges in the thermodynamic limit
$N\rightarrow \infty$ only. Second, the synchronization order parameter jumps
from moderate values of the order of $N^{-1/2}$ to values extremely close to
$1$, its theoretical maximum, immediately upon crossing a critical coupling
strength. We analytically explain the mechanisms underlying such extreme
transitions in coupled complexified Kuramoto oscillators. Extreme transitions
may similarly occur across other systems of coupled oscillators as well as in
certain percolation processes. In applications, their occurrence impacts our
ability of ensuring or preventing strong forms of ordering, for instance in
biological and engineered systems.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:36:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Lee",
"Seungjae",
""
],
[
"Kuklinski",
"Lennart J.",
""
],
[
"Timme",
"Marc",
""
]
] |
2505.10115 | Ivor Kresic | Marin {\DH}uji\'c, Mateo Kruljac, Lovre Kardum, Neven \v{S}anti\'c,
Damir Aumiler, Ivor Kre\v{s}i\'c and Ticijana Ban | Collective light shifts of many cavity modes probed by an optical
frequency comb | null | null | null | null | quant-ph physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | Optical cavities can be used to enhance the interaction of light with atoms
or molecules. In a regime where cavity resonances are very far from the atomic
or molecular transition frequencies, photon absorption can be neglected and the
dispersive part of the interaction can induce spectacular phenomena such as
squeezing, cooling, and self-organization. So far, this dispersive interaction
regime has only been explored using a single or few longitudinal cavity modes
at a time. We experimentally study the dispersive interaction between a cold
atom cloud and many longitudinal modes of a Fabry-Perot cavity, by detecting
collective light shifts in the transmission spectrum of an optical frequency
comb probe. Notably, using a high quality resonator coupled to more than $10^5$
intracavity atoms, we detect appreciable shifts of $\sim 100$ cavity modes
simultaneously. These results constitute a direct demonstration of dispersive
coupling of matter with many longitudinal modes of a linear resonator, and pave
the way towards deeper exploration of atom-light dynamics in the multifrequency
domain.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:36:17 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ðujić",
"Marin",
""
],
[
"Kruljac",
"Mateo",
""
],
[
"Kardum",
"Lovre",
""
],
[
"Šantić",
"Neven",
""
],
[
"Aumiler",
"Damir",
""
],
[
"Krešić",
"Ivor",
""
],
[
"Ban",
"Ticijana",
""
]
] |
2505.10116 | Andrey Polyakov | Andrey Polyakov | Discontinuous integro-differential control systems with sliding modes | null | null | null | null | math.OC cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | The paper deals with analysis and design sliding mode control systems modeled
by integro-differential equations. Filippov method and equivalent control
approach are extended to a class of nonlinear discontinuous
integro-differential equations. Sliding mode control algorithm is designed for
a control system with distributed input delay. The obtained results are
illustrated by numerical example.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:40:32 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Polyakov",
"Andrey",
""
]
] |
2505.10117 | JieHao Wu | JieHao Wu, Ziwei Wang, Junjie Sheng, Wenhao Li, Xiangfei Wang, Jun Luo | Learning Virtual Machine Scheduling in Cloud Computing through Language
Agents | null | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In cloud services, virtual machine (VM) scheduling is a typical Online
Dynamic Multidimensional Bin Packing (ODMBP) problem, characterized by
large-scale complexity and fluctuating demands. Traditional optimization
methods struggle to adapt to real-time changes, domain-expert-designed
heuristic approaches suffer from rigid strategies, and existing learning-based
methods often lack generalizability and interpretability. To address these
limitations, this paper proposes a hierarchical language agent framework named
MiCo, which provides a large language model (LLM)-driven heuristic design
paradigm for solving ODMBP. Specifically, ODMBP is formulated as a Semi-Markov
Decision Process with Options (SMDP-Option), enabling dynamic scheduling
through a two-stage architecture, i.e., Option Miner and Option Composer.
Option Miner utilizes LLMs to discover diverse and useful non-context-aware
strategies by interacting with constructed environments. Option Composer
employs LLMs to discover a composing strategy that integrates the
non-context-aware strategies with the contextual ones. Extensive experiments on
real-world enterprise datasets demonstrate that MiCo achieves a 96.9\%
competitive ratio in large-scale scenarios involving more than 10,000 virtual
machines. It maintains high performance even under nonstationary request flows
and diverse configurations, thus validating its effectiveness in complex and
large-scale cloud environments.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:42:11 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wu",
"JieHao",
""
],
[
"Wang",
"Ziwei",
""
],
[
"Sheng",
"Junjie",
""
],
[
"Li",
"Wenhao",
""
],
[
"Wang",
"Xiangfei",
""
],
[
"Luo",
"Jun",
""
]
] |
2505.10118 | Yangfu Li | Yangfu Li, Hongjian Zhan, Tianyi Chen, Qi Liu, Yue Lu | Why 1 + 1 < 1 in Visual Token Pruning: Beyond Naive Integration via
Multi-Objective Balanced Covering | 31 pages,9 figures,conference | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by/4.0/ | Existing visual token pruning methods target prompt alignment and visual
preservation with static strategies, overlooking the varying relative
importance of these objectives across tasks, which leads to inconsistent
performance. To address this, we derive the first closed-form error bound for
visual token pruning based on the Hausdorff distance, uniformly characterizing
the contributions of both objectives. Moreover, leveraging $\epsilon$-covering
theory, we reveal an intrinsic trade-off between these objectives and quantify
their optimal attainment levels under a fixed budget. To practically handle
this trade-off, we propose Multi-Objective Balanced Covering (MoB), which
reformulates visual token pruning as a bi-objective covering problem. In this
framework, the attainment trade-off reduces to budget allocation via greedy
radius trading. MoB offers a provable performance bound and linear scalability
with respect to the number of input visual tokens, enabling adaptation to
challenging pruning scenarios. Extensive experiments show that MoB preserves
96.4% of performance for LLaVA-1.5-7B using only 11.1% of the original visual
tokens and accelerates LLaVA-Next-7B by 1.3-1.5$\times$ with negligible
performance loss. Additionally, evaluations on Qwen2-VL and Video-LLaVA confirm
that MoB integrates seamlessly into advanced MLLMs and diverse vision-language
tasks.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:43:28 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Li",
"Yangfu",
""
],
[
"Zhan",
"Hongjian",
""
],
[
"Chen",
"Tianyi",
""
],
[
"Liu",
"Qi",
""
],
[
"Lu",
"Yue",
""
]
] |
2505.10119 | Joachim K\"onig | Joachim K\"onig | A note on monogenic even polynomials | null | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | We extend several predecessor works on even sextic monogenic polynomials. In
particular, we prove a conjecture of Lenny Jones, thereby classifying even
sextic monogenic polynomials with cyclic Galois group. This result is key to
completing previous partial results on existence or non-existence of infinite
families of even sextic monogenic polynomials with a prescribed Galois group.
Some of the underlying ideas are relevant for investigation of more general
families of even polynomials $f(X^2)$, or power-compositional polynomials
$f(X^\ell)$.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:43:37 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"König",
"Joachim",
""
]
] |
2505.10120 | Guillaume Godin | Guillaume Godin | All You Need Is Synthetic Task Augmentation | 14 pages, 3 Figures, 6 tables | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Injecting rule-based models like Random Forests into differentiable neural
network frameworks remains an open challenge in machine learning. Recent
advancements have demonstrated that pretrained models can generate efficient
molecular embeddings. However, these approaches often require extensive
pretraining and additional techniques, such as incorporating posterior
probabilities, to boost performance. In our study, we propose a novel strategy
that jointly trains a single Graph Transformer neural network on both sparse
multitask molecular property experimental targets and synthetic targets derived
from XGBoost models trained on Osmordred molecular descriptors. These synthetic
tasks serve as independent auxiliary tasks. Our results show consistent and
significant performance improvement across all 19 molecular property prediction
tasks. For 16 out of 19 targets, the multitask Graph Transformer outperforms
the XGBoost single-task learner. This demonstrates that synthetic task
augmentation is an effective method for enhancing neural model performance in
multitask molecular property prediction without the need for feature injection
or pretraining.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:46:27 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Godin",
"Guillaume",
""
]
] |
2505.10121 | Mohamed Meguebel | Mohamed Meguebel, Maxime Federico, Simone Felicetti, Nadia Belabas,
Nicolas Fabre | Generation of frequency entanglement with an effective quantum
dot-waveguide two-photon quadratic interaction | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Light-matter interactions with quantum dots have been extensively studied to
harness key quantum properties of photons, such as indistinguishability and
entanglement. In this theoretical work, we exploit the atomic-like four-level
structure of a quantum dot coupled to a waveguide to model a shaping frequency
entangling gate (ShaFrEnGa) for single photons. Our approach is based on the
identification of input frequencies and an atomic level structure for which
frequency-dependent one-photon transitions are adiabatically eliminated, while
frequency-dependent two-photon transitions are resonantly enhanced. The
frequency entanglement performance of the gate is analyzed using a Schmidt
decomposition for continuous variables, revealing a trade-off between
entanglement generation efficiency and entanglement quality. We further
demonstrate the use of the ShaFrEnGa for the generation of entangled frequency
qudit states.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:47:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Meguebel",
"Mohamed",
""
],
[
"Federico",
"Maxime",
""
],
[
"Felicetti",
"Simone",
""
],
[
"Belabas",
"Nadia",
""
],
[
"Fabre",
"Nicolas",
""
]
] |
2505.10122 | Syed Luqman Shah | Syed Luqman Shah, Ziaul Haq Abbas, Ghulam Abbas, and Nurul Huda
Mahmood | Energy-Efficient and Reliable Data Collection in Receiver-Initiated
Wake-up Radio Enabled IoT Networks | Already submitted for potential publication | null | null | null | cs.NI eess.SP | http://creativecommons.org/licenses/by/4.0/ | In unmanned aerial vehicle (UAV)-assisted wake-up radio (WuR)-enabled
internet of things (IoT) networks, UAVs can instantly activate the main radios
(MRs) of the sensor nodes (SNs) with a wake-up call (WuC) for efficient data
collection in mission-driven data collection scenarios. However, the
spontaneous response of numerous SNs to the UAV's WuC can lead to significant
packet loss and collisions, as WuR does not exhibit its superiority for
high-traffic loads. To address this challenge, we propose an innovative
receiver-initiated WuR UAV-assisted clustering (RI-WuR-UAC) medium access
control (MAC) protocol to achieve low latency and high reliability in ultra-low
power consumption applications. We model the proposed protocol using the
$M/G/1/2$ queuing framework and derive expressions for key performance metrics,
i.e., channel busyness probability, probability of successful clustering,
average SN energy consumption, and average transmission delay. The RI-WuR-UAC
protocol employs three distinct data flow models, tailored to different network
traffic conditions, which perform three MAC mechanisms: channel assessment
(CCA) clustering for light traffic loads, backoff plus CCA clustering for dense
and heavy traffic, and adaptive clustering for variable traffic loads.
Simulation results demonstrate that the RI-WuR-UAC protocol significantly
outperforms the benchmark sub-carrier modulation clustering protocol. By
varying the network load, we capture the trade-offs among the performance
metrics, showcasing the superior efficiency and reliability of the RI-WuR-UAC
protocol.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:48:39 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Shah",
"Syed Luqman",
""
],
[
"Abbas",
"Ziaul Haq",
""
],
[
"Abbas",
"Ghulam",
""
],
[
"Mahmood",
"Nurul Huda",
""
]
] |
2505.10123 | Luca Naponiello Dr. | L. Naponiello, S. Vissapragada, A. S. Bonomo, M.-L. Steinmeyer, S.
Filomeno, V. D'Orazi, C. Dorn, A. Sozzetti, L. Mancini, A. F. Lanza, K.
Biazzo, C. N. Watkins, G. H\'ebrard, J. Lissauer, S. B. Howell, D. R. Ciardi,
G. Mantovan, D. Baker, V. Bourrier, L. A. Buchhave, C. A. Clark, K. A.
Collins, R. Cosentino, M. Damasso, X. Dumusque, A. Fiorenzano, T. Forveille,
N. Heidari, D. W. Latham, C. Littlefield, M. L\'opez-Morales, M. B. Lund, L.
Malavolta, F. Manni, D. Nardiello, M. Pinamonti, S. W. Yee, R. Zambelli, C.
Ziegler and T. Zingales | The Hot-Neptune Initiative (HONEI) I. Two hot sub-Neptunes on a
close-in, eccentric orbit (TOI-5800 b) and a farther-out, circular orbit
(TOI-5817 b) | Paper submitted to A&A on 2025-05-15 | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Neptune-sized exoplanets are key targets for atmospheric studies, yet their
formation and evolution remain poorly understood due to their diverse
characteristics and limited sample size. The so-called "Neptune desert", a
region of parameter space with a dearth of short-period sub- to super-Neptunes,
is a critical testbed for theories of atmospheric escape and migration. The
HONEI program aims to confirm and characterize the best Neptune-sized
candidates for composition, atmospheric and population studies. By measuring
planetary masses with high precision, we want to provide the community with
optimal targets whose atmosphere can be effectively explored with the JWST or
by ground-based high-resolution spectroscopy. For this purpose, we started a
radial velocity follow-up campaign, using the twin high-precision spectrographs
HARPS and HARPS-N, to measure the masses of TESS Neptune-sized candidates and
confirm their planetary nature. In this first paper of the series, we confirm
the planetary nature of two candidates: TOI-5800b and TOI-5817b. TOI-5800b is a
hot sub-Neptune ($R_p=2.44\pm0.29$ $R_\oplus$, $M_p=9.4\pm1.8$ $M_\oplus$,
$T_{eq} = 1108\pm20$ K) located at the lower edges of the Neptune desert
($P=2.628$ days) and is the most eccentric planet ($e\sim0.3$) ever found
within $P<3$ d. TOI-5800b is expected to be still in the tidal migration phase
with its parent star, a K3 V dwarf ($V=9.6$ mag), although its eccentricity
could arise from interactions with another object in the system. Having a
high-transmission spectroscopy metric ($TSM\sim103$), it represents a prime
target for future atmospheric characterization. TOI-5817b is a relatively hot
sub-Neptune ($R_p=3.08\pm0.14$ $R_\oplus$, $M_p=10.3\pm1.4$ $M_\oplus$,
$T_{eq}=950\pm21$ K) located in the Neptune savanna ($P=15.610$ d) [...]
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:50:39 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Naponiello",
"L.",
""
],
[
"Vissapragada",
"S.",
""
],
[
"Bonomo",
"A. S.",
""
],
[
"Steinmeyer",
"M. -L.",
""
],
[
"Filomeno",
"S.",
""
],
[
"D'Orazi",
"V.",
""
],
[
"Dorn",
"C.",
""
],
[
"Sozzetti",
"A.",
""
],
[
"Mancini",
"L.",
""
],
[
"Lanza",
"A. F.",
""
],
[
"Biazzo",
"K.",
""
],
[
"Watkins",
"C. N.",
""
],
[
"Hébrard",
"G.",
""
],
[
"Lissauer",
"J.",
""
],
[
"Howell",
"S. B.",
""
],
[
"Ciardi",
"D. R.",
""
],
[
"Mantovan",
"G.",
""
],
[
"Baker",
"D.",
""
],
[
"Bourrier",
"V.",
""
],
[
"Buchhave",
"L. A.",
""
],
[
"Clark",
"C. A.",
""
],
[
"Collins",
"K. A.",
""
],
[
"Cosentino",
"R.",
""
],
[
"Damasso",
"M.",
""
],
[
"Dumusque",
"X.",
""
],
[
"Fiorenzano",
"A.",
""
],
[
"Forveille",
"T.",
""
],
[
"Heidari",
"N.",
""
],
[
"Latham",
"D. W.",
""
],
[
"Littlefield",
"C.",
""
],
[
"López-Morales",
"M.",
""
],
[
"Lund",
"M. B.",
""
],
[
"Malavolta",
"L.",
""
],
[
"Manni",
"F.",
""
],
[
"Nardiello",
"D.",
""
],
[
"Pinamonti",
"M.",
""
],
[
"Yee",
"S. W.",
""
],
[
"Zambelli",
"R.",
""
],
[
"Ziegler",
"C.",
""
],
[
"Zingales",
"T.",
""
]
] |
2505.10124 | Ziad Kheil | Ziad Kheil, Lucas Robinet, Laurent Risser, Soleakhena Ken | IMITATE: Image Registration with Context for unknown time frame recovery | IEEE ISBI 2025 | 2025 IEEE 22nd International Symposium on Biomedical Imaging
(ISBI), Houston, TX, USA, 2025, pp. 01-05 | 10.1109/ISBI60581.2025.10981086 | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we formulate a novel image registration formalism dedicated to
the estimation of unknown condition-related images, based on two or more known
images and their associated conditions. We show how to practically model this
formalism by using a new conditional U-Net architecture, which fully takes into
account the conditional information and does not need any fixed image. Our
formalism is then applied to image moving tumors for radiotherapy treatment at
different breathing amplitude using 4D-CT (3D+t) scans in thoracoabdominal
regions. This driving application is particularly complex as it requires to
stitch a collection of sequential 2D slices into several 3D volumes at
different organ positions. Movement interpolation with standard methods then
generates well known reconstruction artefacts in the assembled volumes due to
irregular patient breathing, hysteresis and poor correlation of breathing
signal to internal motion. Results obtained on 4D-CT clinical data showcase
artefact-free volumes achieved through real-time latencies. The code is
publicly available at https://github.com/Kheil-Z/IMITATE .
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:51:05 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Kheil",
"Ziad",
""
],
[
"Robinet",
"Lucas",
""
],
[
"Risser",
"Laurent",
""
],
[
"Ken",
"Soleakhena",
""
]
] |
2505.10125 | Wujun Zhou | Wujun Zhou, Shu Ding, ZeLin Li and Wei Wang | Enhancing the Performance of Global Model by Improving the Adaptability
of Local Models in Federated Learning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Federated learning enables the clients to collaboratively train a global
model, which is aggregated from local models. Due to the heterogeneous data
distributions over clients and data privacy in federated learning, it is
difficult to train local models to achieve a well-performed global model. In
this paper, we introduce the adaptability of local models, i.e., the average
performance of local models on data distributions over clients, and enhance the
performance of the global model by improving the adaptability of local models.
Since each client does not know the data distributions over other clients, the
adaptability of the local model cannot be directly optimized. First, we provide
the property of an appropriate local model which has good adaptability on the
data distributions over clients. Then, we formalize the property into the local
training objective with a constraint and propose a feasible solution to train
the local model. Extensive experiments on federated learning benchmarks
demonstrate that our method significantly improves the adaptability of local
models and achieves a well-performed global model that consistently outperforms
the baseline methods.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:51:47 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zhou",
"Wujun",
""
],
[
"Ding",
"Shu",
""
],
[
"Li",
"ZeLin",
""
],
[
"Wang",
"Wei",
""
]
] |
2505.10126 | Xin Guo | Xin Guo and Xin Wen | Nonstationary nonzero-sum Markov games under a probability criterion | null | null | null | null | math.PR math.OC | http://creativecommons.org/licenses/by/4.0/ | This paper deals with N-person nonzero-sum discrete-time Markov games under a
probability criterion, in which the transition probabilities and reward
functions are allowed to vary with time. Differing from the existing works on
the expected reward criteria, our concern here is to maximize the probabilities
that the accumulated rewards until the first passage time to any target set
exceed a given goal, which represent the reliability of the players income.
Under a mild condition, by developing a comparison theorem for the probability
criterion, we prove the existence of a Nash equilibrium over history-dependent
policies. Moreover, we provide an efficient algorithm for computing
epsilon-Nash equilibria. Finally, we illustrate our main results by a
nonstationary energy management model and take a numerical experiment.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:52:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guo",
"Xin",
""
],
[
"Wen",
"Xin",
""
]
] |
2505.10127 | Laura Ratcliff | Kritam Thapa, Jennifer I. Jones, Laura E. Ratcliff | A High Throughput Virtual Screening Approach for Identifying Thermally
Activated Delayed Fluorescence-Based Emitters | null | null | null | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Thermally activated delayed fluorescence (TADF) offers the promise of highly
efficient organic light emitting diodes (OLEDs), without the heavy metals
requirement of the previous generation of OLEDs. However, the design of new
TADF emitters is complicated by competing requirements, which require opposing
design strategies. High throughput virtual screening (HTVS) approaches,
however, offer the possibility of identifying new TADF emitters without
necessarily relying on existing design rules. In this work the STONED algorithm
[A. Nigam et al., Chem. Sci., 2021, 12, 7079] is used to impose random
structural mutations starting from a set of twenty parent molecules, composed
of both traditional donor-acceptor and multiresonant TADF emitters. Following
this, successive filters are applied based on features of the atomic structure
through to time-dependent density functional theory calculations. Although the
randomised approach proves to be ill-suited to rediscovering existing TADF
emitters, the resulting workflow leads to the identification of a number of
molecules with promising properties for TADF, across a range of emission
colours.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:52:44 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Thapa",
"Kritam",
""
],
[
"Jones",
"Jennifer I.",
""
],
[
"Ratcliff",
"Laura E.",
""
]
] |
2505.10128 | Huy Le Quang | Huy Q. Le, Latif U. Khan and Choong Seon Hong | Robust Federated Learning on Edge Devices with Domain Heterogeneity | IWCMC 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | Federated Learning (FL) allows collaborative training while ensuring data
privacy across distributed edge devices, making it a popular solution for
privacy-sensitive applications. However, FL faces significant challenges due to
statistical heterogeneity, particularly domain heterogeneity, which impedes the
global mode's convergence. In this study, we introduce a new framework to
address this challenge by improving the generalization ability of the FL global
model under domain heterogeneity, using prototype augmentation. Specifically,
we introduce FedAPC (Federated Augmented Prototype Contrastive Learning), a
prototype-based FL framework designed to enhance feature diversity and model
robustness. FedAPC leverages prototypes derived from the mean features of
augmented data to capture richer representations. By aligning local features
with global prototypes, we enable the model to learn meaningful semantic
features while reducing overfitting to any specific domain. Experimental
results on the Office-10 and Digits datasets illustrate that our framework
outperforms SOTA baselines, demonstrating superior performance.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:53:14 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Le",
"Huy Q.",
""
],
[
"Khan",
"Latif U.",
""
],
[
"Hong",
"Choong Seon",
""
]
] |
2505.10129 | Borja Genoves Guzman | Borja Genoves Guzman, Maximo Morales-Cespedes, Ana Garcia Armada,
Maite Brandt-Pearce | Angle diversity receiver as a key enabler for reliable ORIS-based
Visible Light Communication | null | null | null | null | eess.SP cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visible Light Communication (VLC) offers a promising solution to satisfy the
increasing demand for wireless data. However, link blockages remain a
significant challenge. This paper addresses this issue by investigating the
combined use of angle diversity receivers (ADRs) and optical reconfigurable
intelligent surfaces (ORISs) in multiuser VLC systems. We consider ORIS
elements as small movable mirrors. We demonstrate the complementarity of ADR
and ORIS in mitigating link blockages, as well as the advantages of using a
larger number of ORIS elements due to the increased field-of-view (FoV) at the
receiver enabled by the ADR. An optimization algorithm is proposed to maximize
the minimum signal-to-noise power ratio (SNR) to deploy a fair communication
network. Numerical results show that integrating ADR and ORIS significantly
enhances VLC communication performance, achieving an SNR gain of up to 30 dB
compared to a system without ORIS, and mitigating communication outages
produced by link blockages or out-of-FoV received signals. We also prove that
an ADR with a single tier of photodiodes is sufficient to complement
ORIS-assisted VLC.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:56:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guzman",
"Borja Genoves",
""
],
[
"Morales-Cespedes",
"Maximo",
""
],
[
"Armada",
"Ana Garcia",
""
],
[
"Brandt-Pearce",
"Maite",
""
]
] |
2505.10130 | Iryna Timchenko | I.S. Timchenko, O.S. Deiev, S.M. Olejnik, S.M. Potin, V.A. Kushnir,
V.V. Mytrochenko, S.A. Perezhogin, and A. Herz\'a\v{n} | Cross-section of the ${^{\rm nat}\rm{Ni}}(\gamma,{\rm p} x\rm
n)^{58}\rm{Co}$ reaction at the bremsstrahlung end-point energy of 35-94 MeV | 10 pages, 5 figures, 2 tables | null | null | null | nucl-ex hep-ex | http://creativecommons.org/licenses/by/4.0/ | The photoproduction of the $^{58}\rm{Co}$ nuclei on ${^{\rm nat}\rm{Ni}}$ was
studied using the induced $\gamma$-activity method and off-line $\gamma$-ray
spectrometric technique. The experiment was performed at the electron linear
accelerator LUE-40 NSC KIPT, Ukraine. The total flux-averaged cross-section
$\langle{\sigma(E_{\rm{\gamma max}})}\rangle$ for the ${^{\rm
nat}\rm{Ni}}(\gamma,{\rm p} x\rm n)^{58}\rm{Co}$ reaction has been measured in
the range of bremsstrahlung end-point energies $E_{\rm{\gamma max}}$ = 35-94
MeV. The obtained $\langle{\sigma(E_{\rm{\gamma max}})}\rangle$ were compared
with theoretical estimates. The theoretical values of
$\langle{\sigma(E_{\rm{\gamma max}})}\rangle_{\rm th}$ were calculated using
the partial cross-sections $\sigma(E)$ from the TALYS1.96 code for different
level density models and gamma strength functions.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:56:43 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Timchenko",
"I. S.",
""
],
[
"Deiev",
"O. S.",
""
],
[
"Olejnik",
"S. M.",
""
],
[
"Potin",
"S. M.",
""
],
[
"Kushnir",
"V. A.",
""
],
[
"Mytrochenko",
"V. V.",
""
],
[
"Perezhogin",
"S. A.",
""
],
[
"Herzáň",
"A.",
""
]
] |
2505.10131 | Asim Kumar Ghosh | Rittwik Chatterjee and Asim Kumar Ghosh | Floquet topological phases of higher winding numbers in extended
Su-Schrieffer-Heeger model under quenched drive | 12 pages and 13 figures | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | In this study topological properties of static and dynamic
Su-Schrieffer-Heeger models with staggered further neighbor hopping terms
of different extents are investigated. Topological characterization
of the static chiral models is established in terms of
conventional winding number while Floquet topological character is
studied by a pair of winding numbers.
With the increase of extent of further neighbor terms topological phases
with higher winding numbers are found to emerge in both static and dynamic
systems. Topological phase diagrams of static models for four different
extents of further neighbor terms are presented, which has been generalized
for arbitrary extent afterwards. Similarly, Floquet topological
phase diagrams of four such dynamic models have been presented.
For every model four different parametrizations of hopping terms
are introduced which
exhibits different patterns of topological phase diagrams.
In each case emergence of `0' and `$\pi$' energy edge states
is noted and they are found to consistent to the
bulk-boundary correspondence rule applicable for chiral topological systems.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:58:08 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Chatterjee",
"Rittwik",
""
],
[
"Ghosh",
"Asim Kumar",
""
]
] |
2505.10132 | Kaushal Thakkar | Kinjal Patel, Kaushal Thakkar | Electromagnetic and weak decay of singly Heavy Baryons (Qqq) | null | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | The heavy-to-heavy exclusive semileptonic transitions of singly heavy baryons
(SHBs) are investigated within the framework of the Hypercentral Constituent
Quark Model (hCQM). The six-dimensional hyperradial Schr\"{o}dinger equation is
solved in the variational approach to calculate the ground state masses of
bottom and charmed baryons. The transition magnetic moments and radiative $M1$
decay widths are calculated using the spin-flavour wave function and the
effective quark masses of constituent baryon. The Isgur-Wise function (IWF) is
determined at zero recoil to compute the $b \rightarrow c$ semileptonic decay.
Additionally, the branching ratios, as well as the slope and convexity
parameters of IWF are evaluated and compared with results from other studies.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 09:59:17 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Patel",
"Kinjal",
""
],
[
"Thakkar",
"Kaushal",
""
]
] |
2505.10133 | Ziyue Zheng | Zi-Yue Zheng, Ting-ting Sun, Huan Chen, Jin-Biao Wei, Xiao-Ping Zheng,
H.-J. Schulze, G. F. Burgio | $f$-mode oscillations of protoneutron stars | 15 pages, 11 figures | null | null | null | astro-ph.HE gr-qc | http://creativecommons.org/licenses/by/4.0/ | We investigate nonradial $f$-mode oscillations of protoneutron stars in full
general relativity, employing equations of state described by the
Brueckner-Hartree-Fock theory or the relativistic mean field model, while
assuming isentropy and fixed lepton fractions for the internal structure. The
validity of various universal relations for cold neutron stars involving
$f$-mode characteristics and macroscopic properties of the star is confirmed
for those isentropic protoneutron stars. Prospects of observations are also
discussed. According to simulation results, we then model details of the
thermal and trapping profiles in a PNS with the canonical mass. The
corresponding $f$-mode frequencies and gravitational-wave strain amplitudes are
presented. The validity of the universal relations during the evolution to the
formation of a cold neutron star is confirmed.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:01:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zheng",
"Zi-Yue",
""
],
[
"Sun",
"Ting-ting",
""
],
[
"Chen",
"Huan",
""
],
[
"Wei",
"Jin-Biao",
""
],
[
"Zheng",
"Xiao-Ping",
""
],
[
"Schulze",
"H. -J.",
""
],
[
"Burgio",
"G. F.",
""
]
] |
2505.10134 | Guangjin Pan | Guangjin Pan, Kaixuan Huang, Hui Chen, Shunqing Zhang, Christian
H\"ager, Henk Wymeersch | Large Wireless Localization Model (LWLM): A Foundation Model for
Positioning in 6G Networks | 13 pages,16 figures.This work has been submitted to the IEEE for
possible publication | null | null | null | eess.SP cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate and robust localization is a critical enabler for emerging 5G and 6G
applications, including autonomous driving, extended reality (XR), and smart
manufacturing. While data-driven approaches have shown promise, most existing
models require large amounts of labeled data and struggle to generalize across
deployment scenarios and wireless configurations. To address these limitations,
we propose a foundation-model-based solution tailored for wireless
localization. We first analyze how different self-supervised learning (SSL)
tasks acquire general-purpose and task-specific semantic features based on
information bottleneck (IB) theory. Building on this foundation, we design a
pretraining methodology for the proposed Large Wireless Localization Model
(LWLM). Specifically, we propose an SSL framework that jointly optimizes three
complementary objectives: (i) spatial-frequency masked channel modeling
(SF-MCM), (ii) domain-transformation invariance (DTI), and (iii)
position-invariant contrastive learning (PICL). These objectives jointly
capture the underlying semantics of wireless channel from multiple
perspectives. We further design lightweight decoders for key downstream tasks,
including time-of-arrival (ToA) estimation, angle-of-arrival (AoA) estimation,
single base station (BS) localization, and multiple BS localization.
Comprehensive experimental results confirm that LWLM consistently surpasses
both model-based and supervised learning baselines across all localization
tasks. In particular, LWLM achieves 26.0%--87.5% improvement over transformer
models without pretraining, and exhibits strong generalization under
label-limited fine-tuning and unseen BS configurations, confirming its
potential as a foundation model for wireless localization.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:04:44 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Pan",
"Guangjin",
""
],
[
"Huang",
"Kaixuan",
""
],
[
"Chen",
"Hui",
""
],
[
"Zhang",
"Shunqing",
""
],
[
"Häger",
"Christian",
""
],
[
"Wymeersch",
"Henk",
""
]
] |
2505.10135 | Cl\'emence Gourv\`es | C. Gourv\`es, S.N. Breton, A. Dyrek, A.F. Lanza, R.A. Garc\'ia, S.
Mathur, \^A.R.G. Santos, A. Strugarek | Non-transiting exoplanets as a means to understand star-planet
interactions in close-in systems | 21 pages, 12 figures. Accepted in A&A | null | null | null | astro-ph.EP astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Previous studies showed evidence of a dearth of close-in exoplanets around
fast rotators, which can be explained by the combined action of intense tidal
and magnetic interactions between planet and their host star. Detecting more
exoplanets experiencing such interactions, with orbits evolving on short
timescales, is therefore crucial to improve our understanding of the underlying
physical mechanisms. For this purpose, we perform a new search for close-in
non-transiting substellar companions in the Kepler data, focusing on orbital
periods below 2.3 days. We focus on main-sequence solar-type stars and subgiant
stars for which a surface rotation period was measured. For each star, we look
for an excess in the power spectral density of the light curve, that could
correspond to the signature of a close-in non-transiting companion. We compare
our candidates with existing catalogues to eliminate potential contaminants in
our sample, and we visually inspect the phase-folded light curve and its
wavelet decomposition. We identify 88 stars, exhibiting a signature consistent
with the presence of a close non-transiting substellar companion. We show that
the objects in our sample are located mostly within the dearth zone,
emphasising the importance of performing follow-up of such systems in order to
gather observational evidence of star-planet interactions.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:05:05 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Gourvès",
"C.",
""
],
[
"Breton",
"S. N.",
""
],
[
"Dyrek",
"A.",
""
],
[
"Lanza",
"A. F.",
""
],
[
"García",
"R. A.",
""
],
[
"Mathur",
"S.",
""
],
[
"Santos",
"Â. R. G.",
""
],
[
"Strugarek",
"A.",
""
]
] |
2505.10136 | Philipp Pfeffer | Philipp Pfeffer, Peter Brearley, Sylvain Laizet and J\"org Schumacher | Spectral quantum algorithm for passive scalar transport in shear flows | 13 pages, 6 figures | null | null | null | quant-ph physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | The mixing of scalar substances in fluid flows by stirring and diffusion is
ubiquitous in natural flows, chemical engineering, and microfluidic drug
delivery. Here, we present a spectral quantum algorithm for scalar mixing by
solving the advection-diffusion equation in a quantum computational fluid
dynamics framework. We derive exact gate decompositions of the advection and
diffusion operators in spectral space. For all but the simplest one-dimensional
flows, these operators do not commute. Therefore, we use operator splitting and
construct quantum circuits capable of simulating arbitrary polynomial velocity
profiles, such as the Blasius profile of a laminar boundary layer. Periodic,
Neumann, and Dirichlet boundary conditions can be imposed with the appropriate
quantum spectral transform plus additional constraints on the Fourier
expansion. We evaluate our approach in statevector simulations of a Couette
flow, plane Poiseuille flow, and a polynomial Blasius profile approximation to
demonstrate its potential and versatility for scalar mixing in shear flows. The
number of gates grows with, at most, the cubed logarithm of the number of grid
points. This evaluation shows that spectral accuracy allows comparably large
time steps even though the operator splitting limits the temporal order.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:09:52 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Pfeffer",
"Philipp",
""
],
[
"Brearley",
"Peter",
""
],
[
"Laizet",
"Sylvain",
""
],
[
"Schumacher",
"Jörg",
""
]
] |
2505.10137 | Vladimir Vatutin Alekseevich | Vladimir Vatutin, Elena Dyakonova, Yakubdjan Khusanbaev | Small deviations for critical Galton-Watson processes with infinite
variance | 30 pages | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We study the asymptotic behavior of small deviation probabilities for the
critical Galton-Watson processes with infinite variance of the offspring sizes
of particles and apply the obtained result to investigate the structure of a
reduced critical Galton-Watson process.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:11:33 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Vatutin",
"Vladimir",
""
],
[
"Dyakonova",
"Elena",
""
],
[
"Khusanbaev",
"Yakubdjan",
""
]
] |
2505.10138 | Anjali Rawat | Anjali Rawat, Bhanu Kumar, Aaron J. Rosengren, Shane D. Ross | Cislunar Mean-Motion Resonances: Definitions, Widths, and Comparisons
with Resonant Satellites | Submitted to The Journal of Guidance, Control, and Dynamics | null | null | null | astro-ph.EP math.DS | http://creativecommons.org/licenses/by/4.0/ | Lunar mean-motion resonances (MMRs) significantly shape cislunar dynamics
beyond GEO, forming stable-unstable orbit pairs with corresponding intermingled
chaotic and regular regions. The resonance zone is rigorously defined using the
separatrix of unstable resonant periodic orbits surrounding stable
quasi-periodic regions. Our study leverages the planar, circular, restricted
three-body problem (PCR3BP) to estimate the (stable) resonance widths and
(unstable) chaotic resonance zones of influence of the 2:1 and 3:1 MMRs across
various Jacobi constants, employing a Poincar\'e map at perigee and presenting
findings in easily interpretable geocentric orbital elements. An analysis of
the semi-major axis versus eccentricity plane reveals broader regions of
resonance influence than those predicted by semi-analytical models based on the
perturbed Kepler problem. A comparison with high-fidelity 3-dimensional
ephemeris propagation of several spacecraft - TESS, IBEX, and Spektr-R - in
these regions is made, which shows good agreement with the simplified CR3BP
model.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:12:13 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Rawat",
"Anjali",
""
],
[
"Kumar",
"Bhanu",
""
],
[
"Rosengren",
"Aaron J.",
""
],
[
"Ross",
"Shane D.",
""
]
] |
2505.10139 | Lorenz Vaitl | Lorenz Vaitl, Leon Klein | Path Gradients after Flow Matching | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Boltzmann Generators have emerged as a promising machine learning tool for
generating samples from equilibrium distributions of molecular systems using
Normalizing Flows and importance weighting. Recently, Flow Matching has helped
speed up Continuous Normalizing Flows (CNFs), scale them to more complex
molecular systems, and minimize the length of the flow integration
trajectories. We investigate the benefits of using path gradients to fine-tune
CNFs initially trained by Flow Matching, in the setting where a target energy
is known. Our experiments show that this hybrid approach yields up to a
threefold increase in sampling efficiency for molecular systems, all while
using the same model, a similar computational budget and without the need for
additional sampling. Furthermore, by measuring the length of the flow
trajectories during fine-tuning, we show that path gradients largely preserve
the learned structure of the flow.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:13:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Vaitl",
"Lorenz",
""
],
[
"Klein",
"Leon",
""
]
] |
2505.10140 | Mateusz Homenda | Mateusz Homenda, Pawel Jakubczyk, and Hiroyuki Yamase | Quantum criticality and non-Fermi liquids: the nonperturbative
renormalization group perspective | null | null | null | null | cond-mat.str-el hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a thorough theoretical framework based on the nonperturvative
renormalization group (RG) a la Wetterich to tackle the interplay of coupled
fermionic and order-parameter fluctuations at metallic quantum critical points
with ordering wavevectors $\vec{Q}=\vec{0}$. We consistently treat the
dynamical emergence of the Landau damping of the bosonic mode and non-Fermi
liquid scaling of fermions upon lowering the cutoff scale. The loop integrals
of the present theory involve only contributions from fluctuations above the
cutoff scale, which drive the system to a non-Fermi liquid RG fixed point of
different scaling properties from those obtained within the random phase
approximation (RPA) or expansions around it. In particular the scaling exponent
for the Fermi self-energy acquires the value $\alpha\approx 0.50$ rather than
the anticipated $\alpha\approx 0.66$, while the bosonic dynamical exponent
$z\approx 2$. We demonstrate how results characteristic for the RPA-type
fixed-point scaling are recovered in our framework by a questionable procedure
of removing the fermionic cutoff much faster than the bosonic one.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:14:03 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Homenda",
"Mateusz",
""
],
[
"Jakubczyk",
"Pawel",
""
],
[
"Yamase",
"Hiroyuki",
""
]
] |
2505.10141 | Abhimanyu Susobhanan | Abhimanyu Susobhanan | Bayesian pulsar timing and noise analysis with Vela.jl: the wideband
paradigm | Submitted to MNRAS | null | null | null | astro-ph.IM astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Vela.jl is a package for performing Bayesian pulsar timing & noise analysis
written in Julia and Python. In the wideband paradigm of pulsar timing,
simultaneous time of arrival and dispersion measure measurements are derived
from a radio observation using frequency-resolved integrated pulse profiles and
templates without splitting the observation into multiple frequency sub-bands.
We describe the implementation of the wideband timing paradigm in Vela.jl, and
demonstrate its usage using the NANOGrav 12.5-year wideband data of PSR
J1923+2515. Vela.jl is the first software package to provide this
functionality.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:16:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Susobhanan",
"Abhimanyu",
""
]
] |
2505.10142 | Tim Wittenborg | Tim Wittenborg, Ildar Baimuratov, Ludvig Kn\"o\"os Franz\'en, Ingo
Staack, Ulrich R\"omer and S\"oren Auer | Knowledge-Based Aerospace Engineering -- A Systematic Literature Review | 40 pages, 14 figures, submitted to Advanced Engineering Informatics | null | null | null | cs.CE | http://creativecommons.org/licenses/by/4.0/ | The aerospace industry operates at the frontier of technological innovation
while maintaining high standards regarding safety and reliability. In this
environment, with an enormous potential for re-use and adaptation of existing
solutions and methods, Knowledge-Based Engineering (KBE) has been applied for
decades. The objective of this study is to identify and examine
state-of-the-art knowledge management practices in the field of aerospace
engineering. Our contributions include: 1) A SWARM-SLR of over 1,000 articles
with qualitative analysis of 164 selected articles, supported by two aerospace
engineering domain expert surveys. 2) A knowledge graph of over 700
knowledge-based aerospace engineering processes, software, and data, formalized
in the interoperable Web Ontology Language (OWL) and mapped to Wikidata entries
where possible. The knowledge graph is represented on the Open Research
Knowledge Graph (ORKG), and an aerospace Wikibase, for reuse and continuation
of structuring aerospace engineering knowledge exchange. 3) Our resulting
intermediate and final artifacts of the knowledge synthesis, available as a
Zenodo dataset. This review sets a precedent for structured, semantic-based
approaches to managing aerospace engineering knowledge. By advancing these
principles, research, and industry can achieve more efficient design processes,
enhanced collaboration, and a stronger commitment to sustainable aviation.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:16:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wittenborg",
"Tim",
""
],
[
"Baimuratov",
"Ildar",
""
],
[
"Franzén",
"Ludvig Knöös",
""
],
[
"Staack",
"Ingo",
""
],
[
"Römer",
"Ulrich",
""
],
[
"Auer",
"Sören",
""
]
] |
2505.10143 | Longchao Da | Longchao Da, Parth Mitesh Shah, Kuan-Ru Liou, Jiaxing Zhang, Hua Wei | GE-Chat: A Graph Enhanced RAG Framework for Evidential Response
Generation of LLMs | 5 pages, 4 figures, accepted to IJCAI2025 demo track | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models are now key assistants in human decision-making
processes. However, a common note always seems to follow: "LLMs can make
mistakes. Be careful with important info." This points to the reality that not
all outputs from LLMs are dependable, and users must evaluate them manually.
The challenge deepens as hallucinated responses, often presented with seemingly
plausible explanations, create complications and raise trust issues among
users. To tackle such issue, this paper proposes GE-Chat, a knowledge Graph
enhanced retrieval-augmented generation framework to provide Evidence-based
response generation. Specifically, when the user uploads a material document, a
knowledge graph will be created, which helps construct a retrieval-augmented
agent, enhancing the agent's responses with additional knowledge beyond its
training corpus. Then we leverage Chain-of-Thought (CoT) logic generation,
n-hop sub-graph searching, and entailment-based sentence generation to realize
accurate evidence retrieval. We demonstrate that our method improves the
existing models' performance in terms of identifying the exact evidence in a
free-form context, providing a reliable way to examine the resources of LLM's
conclusion and help with the judgment of the trustworthiness.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:17:35 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Da",
"Longchao",
""
],
[
"Shah",
"Parth Mitesh",
""
],
[
"Liou",
"Kuan-Ru",
""
],
[
"Zhang",
"Jiaxing",
""
],
[
"Wei",
"Hua",
""
]
] |
2505.10144 | Lukas Radl | Xuechang Tu, Lukas Radl, Michael Steiner, Markus Steinberger, Bernhard
Kerbl, Fernando de la Torre | VRSplat: Fast and Robust Gaussian Splatting for Virtual Reality | I3D'25 (PACMCGIT); Project Page: https://cekavis.site/VRSplat/ | Proc. ACM Comput. Graph. Interact. Tech., volume 8(1), May 2025 | null | null | cs.GR cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | 3D Gaussian Splatting (3DGS) has rapidly become a leading technique for
novel-view synthesis, providing exceptional performance through efficient
software-based GPU rasterization. Its versatility enables real-time
applications, including on mobile and lower-powered devices. However, 3DGS
faces key challenges in virtual reality (VR): (1) temporal artifacts, such as
popping during head movements, (2) projection-based distortions that result in
disturbing and view-inconsistent floaters, and (3) reduced framerates when
rendering large numbers of Gaussians, falling below the critical threshold for
VR. Compared to desktop environments, these issues are drastically amplified by
large field-of-view, constant head movements, and high resolution of
head-mounted displays (HMDs). In this work, we introduce VRSplat: we combine
and extend several recent advancements in 3DGS to address challenges of VR
holistically. We show how the ideas of Mini-Splatting, StopThePop, and Optimal
Projection can complement each other, by modifying the individual techniques
and core 3DGS rasterizer. Additionally, we propose an efficient foveated
rasterizer that handles focus and peripheral areas in a single GPU launch,
avoiding redundant computations and improving GPU utilization. Our method also
incorporates a fine-tuning step that optimizes Gaussian parameters based on
StopThePop depth evaluations and Optimal Projection. We validate our method
through a controlled user study with 25 participants, showing a strong
preference for VRSplat over other configurations of Mini-Splatting. VRSplat is
the first, systematically evaluated 3DGS approach capable of supporting modern
VR applications, achieving 72+ FPS while eliminating popping and
stereo-disrupting floaters.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:17:48 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Tu",
"Xuechang",
""
],
[
"Radl",
"Lukas",
""
],
[
"Steiner",
"Michael",
""
],
[
"Steinberger",
"Markus",
""
],
[
"Kerbl",
"Bernhard",
""
],
[
"de la Torre",
"Fernando",
""
]
] |
2505.10145 | Ruizhi Qiu | Ruizhi Qiu, Yang Liu | An Integrated UVM-TLM Co-Simulation Framework for RISC-V Functional
Verification and Performance Evaluation | 7 pages, 3 figures, This work is under consideration for conference
publication | null | null | null | cs.AR cs.PF | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The burgeoning RISC-V ecosystem necessitates efficient verification
methodologies for complex processors. Traditional approaches often struggle to
concurrently evaluate functional correctness and performance, or balance
simulation speed with modeling accuracy. This paper introduces an integrated
co-simulation framework leveraging Universal Verification Methodology (UVM) and
Transaction-Level Modeling (TLM) for RISC-V processor validation. We present a
configurable UVM-TLM model (vmodel) of a superscalar, out-of-order RISC-V core,
featuring key microarchitectural modeling techniques such as credit-based
pipeline flow control. This environment facilitates unified functional
verification via co-simulation against the Spike ISA simulator and enables
early-stage performance assessment using benchmarks like CoreMark, orchestrated
within UVM. The methodology prioritizes integration, simulation efficiency, and
acceptable fidelity for architectural exploration over cycle-level precision.
Experimental results validate functional correctness and significant simulation
speedup over RTL approaches, accelerating design iterations and enhancing
verification coverage.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:18:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Qiu",
"Ruizhi",
""
],
[
"Liu",
"Yang",
""
]
] |
2505.10146 | Jan Hurt | Jan Hurt, Stefan Thurner, Peter Klimek | An Agent-Based Extension to Sector-Wise Input-Output Recovery Models | null | null | null | null | econ.GN q-fin.EC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dynamic input-output models are standard tools for understanding
inter-industry dependencies and how economies respond to shocks like disasters
and pandemics. However, traditional approaches often assume fixed prices,
limiting their ability to capture realistic economic behavior. Here, we
introduce an adaptive extension to dynamic input-output recovery models where
producers respond to shocks through simultaneous price and quantity
adjustments. Our framework preserves the economic constraints of the Leontief
input-output model while converging towards equilibrium configurations based on
sector-specific behavioral parameters. When applied to input-output data, the
model allows us to compute behavioral metrics indicating whether specific
sectors predominantly favor price or quantity adjustments. Using the World
Input-Output Database, we identify strong, consistent regional and
sector-specific behavioral patterns. These findings provide insights into how
different regions employ distinct strategies to manage shocks, thereby
influencing economic resilience and recovery dynamics.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:18:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hurt",
"Jan",
""
],
[
"Thurner",
"Stefan",
""
],
[
"Klimek",
"Peter",
""
]
] |
2505.10147 | Yash Keshwani | Yash, Nikhil Karamchandani, Avishek Ghosh | Near Optimal Best Arm Identification for Clustered Bandits | To be published in ICML 2025 | null | null | null | cs.LG cs.MA | http://creativecommons.org/licenses/by/4.0/ | This work investigates the problem of best arm identification for multi-agent
multi-armed bandits. We consider $N$ agents grouped into $M$ clusters, where
each cluster solves a stochastic bandit problem. The mapping between agents and
bandits is a priori unknown. Each bandit is associated with $K$ arms, and the
goal is to identify the best arm for each agent under a $\delta$-probably
correct ($\delta$-PC) framework, while minimizing sample complexity and
communication overhead.
We propose two novel algorithms: Clustering then Best Arm Identification
(Cl-BAI) and Best Arm Identification then Clustering (BAI-Cl). Cl-BAI uses a
two-phase approach that first clusters agents based on the bandit problems they
are learning, followed by identifying the best arm for each cluster. BAI-Cl
reverses the sequence by identifying the best arms first and then clustering
agents accordingly. Both algorithms leverage the successive elimination
framework to ensure computational efficiency and high accuracy.
We establish $\delta$-PC guarantees for both methods, derive bounds on their
sample complexity, and provide a lower bound for this problem class. Moreover,
when $M$ is small (a constant), we show that the sample complexity of a variant
of BAI-Cl is minimax optimal in an order-wise sense. Experiments on synthetic
and real-world datasets (MovieLens, Yelp) demonstrate the superior performance
of the proposed algorithms in terms of sample and communication efficiency,
particularly in settings where $M \ll N$.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:20:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Yash",
"",
""
],
[
"Karamchandani",
"Nikhil",
""
],
[
"Ghosh",
"Avishek",
""
]
] |
2505.10148 | Yoshihiro Ueda | Yoshihiro Ueda, Makoto Ishihara, Wojciech Roga, and Masahiro Takeoka | Quantum network sensing with efficient multi-partite entanglement
distribution via lossy channels | 10 pages, 6 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum network sensing shows potential to enhance the estimation precision
for functions of spatially distributed parameters beyond the shot noise limit.
The key resource required for this task is possibly multi-partite quantum
entanglement. The photonic entanglement is the most natural for this task;
however, distributing it over long distances presents significant difficulties,
mainly because of unavoidable loss in communication channels. In this research,
we analyze a quantum network sensing protocol based on a recently proposed,
efficient GHZ state distribution scheme. In comparison to conventional methods
based on entanglement distribution , our protocol shows the decreasing
loss-induced estimation error of certain functions of distributed parameters
including their arbitrary linear combinations.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:22:30 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ueda",
"Yoshihiro",
""
],
[
"Ishihara",
"Makoto",
""
],
[
"Roga",
"Wojciech",
""
],
[
"Takeoka",
"Masahiro",
""
]
] |
2505.10149 | Mirai Ikebuchi | Mirai Ikebuchi | Homological Invariants of Higher-Order Equational Theories | to appear at LICS 2025 | null | null | null | cs.LO math.CT math.LO | http://creativecommons.org/licenses/by/4.0/ | Many first-order equational theories, such as the theory of groups or boolean
algebras, can be presented by a smaller set of axioms than the original one.
Recent studies showed that a homological approach to equational theories gives
us inequalities to obtain lower bounds on the number of axioms. In this paper,
we extend this result to higher-order equational theories. More precisely, we
consider simply typed lambda calculus with product and unit types and study
sets of equations between lambda terms. Then, we define homology groups of the
given equational theory and show that a lower bound on the number of equations
can be computed from the homology groups.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:22:59 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ikebuchi",
"Mirai",
""
]
] |
2505.10150 | Qiushi Liang | Qiushi Liang, Yeyue Cai, Jianhua Mo, and Meixia Tao | CFARNet: Learning-Based High-Resolution Multi-Target Detection for
Rainbow Beam Radar | 6 pages, 6 figures | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Millimeter-wave (mmWave) OFDM radar equipped with rainbow beamforming,
enabled by joint phase-time arrays (JPTAs), provides wide-angle coverage and is
well-suited for fast real-time target detection and tracking. However, accurate
detection of multiple closely spaced targets remains a key challenge for
conventional signal processing pipelines, particularly those relying on
constant false alarm rate (CFAR) detectors. This paper presents CFARNet, a
learning-based processing framework that replaces CFAR with a convolutional
neural network (CNN) for peak detection in the angle-Doppler domain. The
network predicts target subcarrier indices, which guide angle estimation via a
known frequency-angle mapping and enable high-resolution range and velocity
estimation using the MUSIC algorithm. Extensive simulations demonstrate that
CFARNet significantly outperforms a CFAR+MUSIC baseline, especially under low
transmit power and dense multi-target conditions. The proposed method offers
superior angular resolution, enhanced robustness in low-SNR scenarios, and
improved computational efficiency, highlighting the potential of data-driven
approaches for high-resolution mmWave radar sensing.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:23:09 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Liang",
"Qiushi",
""
],
[
"Cai",
"Yeyue",
""
],
[
"Mo",
"Jianhua",
""
],
[
"Tao",
"Meixia",
""
]
] |
2505.10151 | Endong Sun | Endong Sun, Yuqing Zhu and Matthew Howard | Training People to Reward Robots | 6 pages | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Learning from demonstration (LfD) is a technique that allows expert teachers
to teach task-oriented skills to robotic systems. However, the most effective
way of guiding novice teachers to approach expert-level demonstrations
quantitatively for specific teaching tasks remains an open question. To this
end, this paper investigates the use of machine teaching (MT) to guide novice
teachers to improve their teaching skills based on reinforcement learning from
demonstration (RLfD). The paper reports an experiment in which novices receive
MT-derived guidance to train their ability to teach a given motor skill with
only 8 demonstrations and generalise this to previously unseen ones. Results
indicate that the MT-guidance not only enhances robot learning performance by
89% on the training skill but also causes a 70% improvement in robot learning
performance on skills not seen by subjects during training. These findings
highlight the effectiveness of MT-guidance in upskilling human teaching
behaviours, ultimately improving demonstration quality in RLfD.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:23:56 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sun",
"Endong",
""
],
[
"Zhu",
"Yuqing",
""
],
[
"Howard",
"Matthew",
""
]
] |
2505.10152 | Yikang Wei | Yikang Wei | Multi-Source Collaborative Style Augmentation and Domain-Invariant
Learning for Federated Domain Generalization | IJCAI 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated domain generalization aims to learn a generalizable model from
multiple decentralized source domains for deploying on the unseen target
domain. The style augmentation methods have achieved great progress on domain
generalization. However, the existing style augmentation methods either explore
the data styles within isolated source domain or interpolate the style
information across existing source domains under the data decentralization
scenario, which leads to limited style space. To address this issue, we propose
a Multi-source Collaborative Style Augmentation and Domain-invariant learning
method (MCSAD) for federated domain generalization. Specifically, we propose a
multi-source collaborative style augmentation module to generate data in the
broader style space. Furthermore, we conduct domain-invariant learning between
the original data and augmented data by cross-domain feature alignment within
the same class and classes relation ensemble distillation between different
classes to learn a domain-invariant model. By alternatively conducting
collaborative style augmentation and domain-invariant learning, the model can
generalize well on unseen target domain. Extensive experiments on multiple
domain generalization datasets indicate that our method significantly
outperforms the state-of-the-art federated domain generalization methods.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:26:17 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wei",
"Yikang",
""
]
] |
2505.10153 | Paul Mulholland | P. Mulholland, A. Zocco, M. C. L. Morren, K. Aleynikova, M. J.
Pueschel, J. H. E. Proll, P. W. Terry | Resonant theory of kinetic ballooning modes in general toroidal geometry | null | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The linear theory of the kinetic-ballooning-mode (KBM) instability is
extended to capture a weakly-driven regime in general toroidal geometry where
the destabilization is caused by the magnetic-drift resonance of the ions. Such
resonantly-destabilized KBMs are characterized by broad eigenfunctions along
the magnetic field line and near-marginal positive growth rates, even well
below the beta threshold of their non-resonant counterparts. This
unconventional (or sub-threshold) KBM, when destabilized, has been shown to
catalyze an enhancement of turbulent transport in the Wendelstein 7-X (W7-X)
stellarator [1, 2]. Simplifying the energy dependence of key resonant
quantities allows for an analytical treatment of this KBM using the
physics-based ordering from the more general equations of Tang, Connor, and
Hastie [3]. Results are then compared with high-fidelity gyrokinetic
simulations for the (st)KBM in W7-X and the conventional KBM in a circular
tokamak at both high and low magnetic shear, where good agreement is obtained
in all cases. This reduced KBM model provides deeper insight into
(sub-threshold) KBMs and their relationship with geometry, and shows promise
for aiding in transport model development and geometry-based turbulence
optimization efforts going forward.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:26:30 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Mulholland",
"P.",
""
],
[
"Zocco",
"A.",
""
],
[
"Morren",
"M. C. L.",
""
],
[
"Aleynikova",
"K.",
""
],
[
"Pueschel",
"M. J.",
""
],
[
"Proll",
"J. H. E.",
""
],
[
"Terry",
"P. W.",
""
]
] |
2505.10154 | Mahdi Kohan Sefidi | Mahdi Kohan Sefidi | The Impact of Node Addition and Deletion on Network Production
Fluctuations | null | null | null | null | econ.TH | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Production networks, dynamic systems of firms linked through input-output
relationships, transmit microeconomic shocks into macroeconomic fluctuations.
While prior studies often assume static networks, real-world economies feature
continuous firm entry (node addition) and exit (node deletion). We develop a
probabilistic model to analyze how these dynamics affect production volatility
and network resilience. Integrating Leontief input-output frameworks with
controllability theory. By quantifying fluctuations as expected values under
probabilistic node dynamics, we identify trade-offs between adaptability and
stability. Methodologically, we unify Kalman rank criteria and minimum input
theory, offering policymakers insights to balance innovation-driven entry with
safeguards against destabilizing exits.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:26:55 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sefidi",
"Mahdi Kohan",
""
]
] |
2505.10155 | Davide Emilio Quadrellaro | Tapani Hyttinen, Gianluca Paolini, Davide Emilio Quadrellaro | A New Construction Principle | null | null | null | null | math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We use the framework of Abstract Elementary Classes ($\mathrm{AEC}$s) to
introduce a new Construction Principle $\mathrm{CP}(\mathbf{K},\ast)$, which
strictly generalises the Construction Principle of Eklof, Mekler and Shelah and
allows for many novel applications beyond the setting of universal algebra. In
particular, we show that $\mathrm{CP}(\mathbf{K},\ast)$ holds in the classes of
free products of cyclic groups of fixed order, direct sums of a fixed
torsion-free abelian group of rank 1 which is not $\mathbb{Q}$, (infinite) free
$(k,n)$-Steiner systems, and (infinite) free generalised $n$-gons. From this we
derive, in ZFC, that these classes of structures are not axiomatisable in the
logic $\mathfrak{L}_{\infty,\omega_1}$, and, under $V=L$, that they are not
axiomatisable in $\mathfrak{L}_{\infty,\infty}$.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:28:12 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hyttinen",
"Tapani",
""
],
[
"Paolini",
"Gianluca",
""
],
[
"Quadrellaro",
"Davide Emilio",
""
]
] |
2505.10156 | Rintaro Niimi | Rintaro Niimi, Chikara Furusawa and Yusuke Himeoka | Population dynamics of generalist/specialist strategies in the
feast-famine cycle | null | null | null | null | q-bio.PE | http://creativecommons.org/licenses/by/4.0/ | Microbial populations exhibit a broad spectrum of nutrient utilization
strategies, ranging from strategies utilizing diverse nutrients, called
"generalist," to ones being highly adapted to specific nutrients, called
"specialists." The mathematical conditions for the diversification of nutrient
utilization strategies are central questions in theoretical ecology. Previous
studies have shown that trade-offs among different resource utilization
functions that cells cannot utilize broad types of substrates at near-maximum
speed are crucial for the emergence of diverse strategies. However, in natural
settings, nutrient availability often fluctuates over time, imposing additional
trade-offs on cells. Cells that grow rapidly under nutrient-rich conditions
will suffer a higher death rate under nutrient-poor conditions, creating a
growth-death trade-off that intersects with the classical resource-use
trade-off. Here, we introduce a unified mathematical model that simultaneously
incorporates the resource-use trade-off and the growth-death trade-off. The
nutrient supply was modeled as discrete stochastic events, capturing realistic
temporal fluctuations. We show that the relative balance between growth and
death rates critically influences the dominance of either generalist or
specialist strategies. Specifically, under conditions of high average growth
rates among different environments and a weak trade-off between growth and
death rates, generalists prevail. In contrast, when the growth-death trade-off
is intense, specialists emerge as the dominant strategy. Our findings reveal
that accounting for the growth-death trade-off is crucial for understanding how
microbial communities adapt and evolve in temporally varying environments.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:28:40 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Niimi",
"Rintaro",
""
],
[
"Furusawa",
"Chikara",
""
],
[
"Himeoka",
"Yusuke",
""
]
] |
2505.10157 | Shuma Sugiura | Shuma Sugiura and Masahito Ueda | Quantum feedback cooling of a trapped nanoparticle by using a low-pass
filter | 12 pages, 6 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a low-pass-filter (LPF) feedback control for cooling a trapped
particle with a low-pass filter, and show that it can achieve the minimum
phonon occupation number lower than cold damping and linear-quadratic-Gaussian
(LQG) control, which are the standard methods of ground-state cooling of a
levitated nanoparticle. For the detection efficiency of $90\%$, the achievable
phonon occupation number with our LPF control is about one third and a half of
that of cold damping and LQG control, respectively, and thus our method has a
decisive advantage to reach the absolute ground state.
| [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:30:43 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sugiura",
"Shuma",
""
],
[
"Ueda",
"Masahito",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.