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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.08210 | Erica van der Sar | Erica van der Sar, Alessandro Zocca, Sandjai Bhulai | Optimizing Power Grid Topologies with Reinforcement Learning: A Survey
of Methods and Challenges | 60 pages, 26 figures, preprint | null | null | null | eess.SY cs.AI cs.LG cs.SY stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Power grid operation is becoming increasingly complex due to the rising
integration of renewable energy sources and the need for more adaptive control
strategies. Reinforcement Learning (RL) has emerged as a promising approach to
power network control (PNC), offering the potential to enhance decision-making
in dynamic and uncertain environments. The Learning To Run a Power Network
(L2RPN) competitions have played a key role in accelerating research by
providing standardized benchmarks and problem formulations, leading to rapid
advancements in RL-based methods. This survey provides a comprehensive and
structured overview of RL applications for power grid topology optimization,
categorizing existing techniques, highlighting key design choices, and
identifying gaps in current research. Additionally, we present a comparative
numerical study evaluating the impact of commonly applied RL-based methods,
offering insights into their practical effectiveness. By consolidating existing
research and outlining open challenges, this survey aims to provide a
foundation for future advancements in RL-driven power grid optimization.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 02:27:30 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:22:35 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"van der Sar",
"Erica",
""
],
[
"Zocca",
"Alessandro",
""
],
[
"Bhulai",
"Sandjai",
""
]
] |
2504.08353 | Ren Li | Ren Li, Cong Cao, Corentin Dumery, Yingxuan You, Hao Li, Pascal Fua | Single View Garment Reconstruction Using Diffusion Mapping Via Pattern
Coordinates | SIGGRAPH 2025 | null | null | null | cs.GR cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Reconstructing 3D clothed humans from images is fundamental to applications
like virtual try-on, avatar creation, and mixed reality. While recent advances
have enhanced human body recovery, accurate reconstruction of garment geometry
-- especially for loose-fitting clothing -- remains an open challenge. We
present a novel method for high-fidelity 3D garment reconstruction from single
images that bridges 2D and 3D representations. Our approach combines Implicit
Sewing Patterns (ISP) with a generative diffusion model to learn rich garment
shape priors in a 2D UV space. A key innovation is our mapping model that
establishes correspondences between 2D image pixels, UV pattern coordinates,
and 3D geometry, enabling joint optimization of both 3D garment meshes and the
corresponding 2D patterns by aligning learned priors with image observations.
Despite training exclusively on synthetically simulated cloth data, our method
generalizes effectively to real-world images, outperforming existing approaches
on both tight- and loose-fitting garments. The reconstructed garments maintain
physical plausibility while capturing fine geometric details, enabling
downstream applications including garment retargeting and texture manipulation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 08:39:18 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 07:51:50 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Li",
"Ren",
""
],
[
"Cao",
"Cong",
""
],
[
"Dumery",
"Corentin",
""
],
[
"You",
"Yingxuan",
""
],
[
"Li",
"Hao",
""
],
[
"Fua",
"Pascal",
""
]
] |
2504.08706 | GuCheol Jeong | Gu-Cheol Jeong, Stefano Dalla Gasperina, Ashish D. Deshpande, Lillian
Chin, and Roberto Mart\'in-Mart\'in | BiFlex: A Passive Bimodal Stiffness Flexible Wrist for Manipulation in
Unstructured Environments | 8 pages, 10 figures | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robotic manipulation in unstructured, humancentric environments poses a dual
challenge: achieving the precision need for delicate free-space operation while
ensuring safety during unexpected contact events. Traditional wrists struggle
to balance these demands, often relying on complex control schemes or
complicated mechanical designs to mitigate potential damage from force
overload. In response, we present BiFlex, a flexible robotic wrist that uses a
soft buckling honeycomb structure to provides a natural bimodal stiffness
response. The higher stiffness mode enables precise household object
manipulation, while the lower stiffness mode provides the compliance needed to
adapt to external forces. We design BiFlex to maintain a fingertip deflection
of less than 1 cm while supporting loads up to 500g and create a BiFlex wrist
for many grippers, including Panda, Robotiq, and BaRiFlex. We validate BiFlex
under several real-world experimental evaluations, including surface wiping,
precise pick-and-place, and grasping under environmental constraints. We
demonstrate that BiFlex simplifies control while maintaining precise object
manipulation and enhanced safety in real-world applications.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 17:16:13 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 18:26:32 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Jeong",
"Gu-Cheol",
""
],
[
"Gasperina",
"Stefano Dalla",
""
],
[
"Deshpande",
"Ashish D.",
""
],
[
"Chin",
"Lillian",
""
],
[
"Martín-Martín",
"Roberto",
""
]
] |
2504.08968 | Roman Nikolaenko | Roman Nikolaenko | A model for massive quark-gluon string fragmentation | 54 pages, 6 figures. Considers classic relativistic string to model
hadronization only. V2: corrected the mistake in string rotation plane
angles, does not effect main formulae and conclusions | null | null | null | hep-ph | http://creativecommons.org/publicdomain/zero/1.0/ | A highly under-looked problem in string models of hadronization is taking
into account the mass of the parton system and the conservation of the total
angular momentum, as these factors can significantly influence the time-space
picture of the fragmentation process. In this paper, I present a method to
properly define the Nambu-Goto relativistic string with non-zero mass based on
the finite order eigenfunction expansion of the initial conditions of the
Cauchy problem for string motion. The principles of the fragmentation process
are derived for these strings. 4-momentum and total angular momentum
conservation laws are taken into account, along with Virasoro conditions. It is
shown that very non-trivial consequences arise for non-zero mass strings,
including the discrete spectrum of string spin, the discreteness of the
possible string breaking points, the close-to-Regge behavior of the spin-mass
relation for the light string fragments and the natural stopping mechanism for
fragmentation.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 20:47:45 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 09:37:33 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Nikolaenko",
"Roman",
""
]
] |
2504.09192 | Zhiyong Wang | Zhiyong Wang | Towards More Efficient, Robust, Instance-adaptive, and Generalizable
Sequential Decision making | Ph.D. Thesis | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The primary goal of my Ph.D. study is to develop provably efficient and
practical algorithms for data-driven sequential decision-making under
uncertainty. My work focuses on reinforcement learning (RL), multi-armed
bandits, and their applications, including recommendation systems, computer
networks, video analytics, and large language models (LLMs). Sequential
decision-making methods, such as bandits and RL, have demonstrated remarkable
success - ranging from outperforming human players in complex games like Atari
and Go to advancing robotics, recommendation systems, and fine-tuning LLMs.
Despite these successes, many established algorithms rely on idealized models
that can fail under model misspecifications or adversarial perturbations,
particularly in settings where accurate prior knowledge of the underlying model
class is unavailable or where malicious users operate within dynamic systems.
These challenges are pervasive in real-world applications, where robust and
adaptive solutions are critical. Furthermore, while worst-case guarantees
provide theoretical reliability, they often fail to capture instance-dependent
performance, which can lead to more efficient and practical solutions. Another
key challenge lies in generalizing to new, unseen environments, a crucial
requirement for deploying these methods in dynamic and unpredictable settings.
To address these limitations, my research aims to develop more efficient,
robust, instance-adaptive, and generalizable sequential decision-making
algorithms for both reinforcement learning and bandits. Towards this end, I
focus on developing more efficient, robust, instance-adaptive, and
generalizable for both general reinforcement learning (RL) and bandits.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 12:17:20 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 02:02:31 GMT"
},
{
"version": "v3",
"created": "Wed, 14 May 2025 04:22:02 GMT"
},
{
"version": "v4",
"created": "Thu, 15 May 2025 06:21:11 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Zhiyong",
""
]
] |
2504.09315 | Faisal Haque Bappy | Tahrim Hossain, Faisal Haque Bappy, Tarannum Shaila Zaman, Raiful
Hasan, Tariqul Islam | SmartShift: A Secure and Efficient Approach to Smart Contract Migration | null | null | 10.1145/3696630.3728517 | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | Blockchain and smart contracts have emerged as revolutionary technologies
transforming distributed computing. While platform evolution and smart
contracts' inherent immutability necessitate migrations both across and within
chains, migrating the vast amounts of critical data in these contracts while
maintaining data integrity and minimizing operational disruption presents a
significant challenge. To address these challenges, we present SmartShift, a
framework that enables secure and efficient smart contract migrations through
intelligent state partitioning and progressive function activation, preserving
operational continuity during transitions. Our comprehensive evaluation
demonstrates that SmartShift significantly reduces migration downtime while
ensuring robust security, establishing a foundation for efficient and secure
smart contract migration systems.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 19:28:23 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:12:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hossain",
"Tahrim",
""
],
[
"Bappy",
"Faisal Haque",
""
],
[
"Zaman",
"Tarannum Shaila",
""
],
[
"Hasan",
"Raiful",
""
],
[
"Islam",
"Tariqul",
""
]
] |
2504.10823 | Ayoung Lee | Ayoung Lee, Ryan Sungmo Kwon, Peter Railton, Lu Wang | CLASH: Evaluating Language Models on Judging High-Stakes Dilemmas from
Multiple Perspectives | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Navigating high-stakes dilemmas involving conflicting values is challenging
even for humans, let alone for AI. Yet prior work in evaluating the reasoning
capabilities of large language models (LLMs) in such situations has been
limited to everyday scenarios. To close this gap, this work first introduces
CLASH (Character perspective-based LLM Assessments in Situations with
High-stakes), a meticulously curated dataset consisting of 345 high-impact
dilemmas along with 3,795 individual perspectives of diverse values. In
particular, we design CLASH in a way to support the study of critical aspects
of value-based decision-making processes which are missing from prior work,
including understanding decision ambivalence and psychological discomfort as
well as capturing the temporal shifts of values in characters' perspectives. By
benchmarking 10 open and closed frontier models, we uncover several key
findings. (1) Even the strongest models, such as GPT-4o and Claude-Sonnet,
achieve less than 50% accuracy in identifying situations where the decision
should be ambivalent, while they perform significantly better in clear-cut
scenarios. (2) While LLMs reasonably predict psychological discomfort as marked
by human, they inadequately comprehend perspectives involving value shifts,
indicating a need for LLMs to reason over complex values. (3) Our experiments
also reveal a significant correlation between LLMs' value preferences and their
steerability towards a given value. (4) Finally, LLMs exhibit greater
steerability when engaged in value reasoning from a third-party perspective,
compared to a first-person setup, though certain value pairs benefit uniquely
from the first-person framing.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 02:54:16 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 21:15:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Lee",
"Ayoung",
""
],
[
"Kwon",
"Ryan Sungmo",
""
],
[
"Railton",
"Peter",
""
],
[
"Wang",
"Lu",
""
]
] |
2504.10889 | Aiman Farooq | Shripad Pate, Aiman Farooq, Suvrankar Datta, Musadiq Aadil Sheikh,
Atin Kumar, and Deepak Mishra | Fine-Grained Rib Fracture Diagnosis with Hyperbolic Embeddings: A
Detailed Annotation Framework and Multi-Label Classification Model | null | MICCAI 2025 | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Accurate rib fracture identification and classification are essential for
treatment planning. However, existing datasets often lack fine-grained
annotations, particularly regarding rib fracture characterization, type, and
precise anatomical location on individual ribs. To address this, we introduce a
novel rib fracture annotation protocol tailored for fracture classification.
Further, we enhance fracture classification by leveraging cross-modal
embeddings that bridge radiological images and clinical descriptions. Our
approach employs hyperbolic embeddings to capture the hierarchical nature of
fracture, mapping visual features and textual descriptions into a shared
non-Euclidean manifold. This framework enables more nuanced similarity
computations between imaging characteristics and clinical descriptions,
accounting for the inherent hierarchical relationships in fracture taxonomy.
Experimental results demonstrate that our approach outperforms existing methods
across multiple classification tasks, with average recall improvements of 6% on
the AirRib dataset and 17.5% on the public RibFrac dataset.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 05:47:09 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 04:29:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Pate",
"Shripad",
""
],
[
"Farooq",
"Aiman",
""
],
[
"Datta",
"Suvrankar",
""
],
[
"Sheikh",
"Musadiq Aadil",
""
],
[
"Kumar",
"Atin",
""
],
[
"Mishra",
"Deepak",
""
]
] |
2504.11403 | Qiutong Wang | Qiutong Wang | Counting irreducible representations of general linear groups and
unitary groups | null | null | null | null | math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $G$ be a general linear group over $\BR$, $\BC$, or $\BH$, or a real
unitary group. In this paper, we precisely describe the number of isomorphism
classes of irreducible Casselman-Wallach representations of $G$ with a given
infinitesimal character and a given associated variety, expressed in terms of
certain combinatorial data called painted Young diagrams and assigned Young
diagrams.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 17:18:20 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 10:32:56 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 15:59:21 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Qiutong",
""
]
] |
2504.11683 | Prasenjit Das | Sayantan Mondal and Prasenjit Das | Velocity Distribution and Diffusion of an Athermal Inertial
Run-and-Tumble Particle in a Shear-Thinning Medium | 20 Pages, 7 Figures, Accepted to Physics of Fluids | Physics of Fluids 37, 057114 (2025) | 10.1063/5.0261651 | null | cond-mat.stat-mech cond-mat.soft | http://creativecommons.org/publicdomain/zero/1.0/ | We study the dynamics of an athermal inertial active particle moving in a
shear-thinning medium in $d=1$. The viscosity of the medium is modeled using a
Coulomb-tanh function, while the activity is represented by an asymmetric
dichotomous noise with strengths $-\Delta$ and $\mu\Delta$, transitioning
between these states at a rate $\lambda$. Starting from the Fokker-Planck~(FP)
equation for the time-dependent probability distributions $P(v,-\Delta,t)$ and
$P(v,\mu\Delta,t)$ of the particle's velocity $v$ at time $t$, moving under the
influence of active forces $-\Delta$ and $\mu\Delta$ respectively, we
analytically derive the steady-state velocity distribution function $P_s(v)$,
explicitly dependent on $\mu$. Also, we obtain a quadrature expression for the
effective diffusion coefficient $D_e$ for the symmetric active force
case~($\mu=1$). For a given $\Delta$ and $\mu$, we show that $P_s(v)$ exhibits
multiple transitions as $\lambda$ is varied. Subsequently, we numerically
compute $P_s(v)$, the mean-squared velocity $\langle v^2\rangle(t)$, and the
diffusion coefficient $D_e$ by solving the particle's equation of motion, all
of which show excellent agreement with the analytical results in the
steady-state. Finally, we examine the universal nature of the transitions in
$P_s(v)$ by considering an alternative functional form of medium's viscosity
that also capture the shear-thinning behavior.
| [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 00:52:27 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Mondal",
"Sayantan",
""
],
[
"Das",
"Prasenjit",
""
]
] |
2504.12988 | Yannis Montreuil | Yannis Montreuil, Axel Carlier, Lai Xing Ng, Wei Tsang Ooi | Why Ask One When You Can Ask $k$? Two-Stage Learning-to-Defer to the
Top-$k$ Experts | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Although existing Learning-to-Defer (L2D) frameworks support multiple
experts, they allocate each query to a single expert, limiting their ability to
leverage collective expertise in complex decision-making scenarios. To address
this, we introduce the first framework for Top-$k$ Learning-to-Defer, enabling
systems to defer each query to the $k$ most cost-effective experts. Our
formulation strictly generalizes classical two-stage L2D by supporting
multi-expert deferral-a capability absent in prior work. We further propose
Top-$k(x)$ Learning-to-Defer, an adaptive extension that learns the optimal
number of experts per query based on input complexity, expert quality, and
consultation cost. We introduce a novel surrogate loss that is
Bayes-consistent, $(\mathcal{R}, \mathcal{G})$-consistent, and independent of
the cardinality parameter $k$, enabling efficient reuse across different values
of $k$. We show that classical model cascades arise as a special case of our
method, situating our framework as a strict generalization of both selective
deferral and cascaded inference. Experiments on classification and regression
demonstrate that Top-$k$ and Top-$k(x)$ yield improved accuracy--cost
trade-offs, establishing a new direction for multi-expert deferral in
Learning-to-Defer.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 14:50:40 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 07:02:20 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 10:25:18 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Montreuil",
"Yannis",
""
],
[
"Carlier",
"Axel",
""
],
[
"Ng",
"Lai Xing",
""
],
[
"Ooi",
"Wei Tsang",
""
]
] |
2504.13016 | Borja Genoves Guzman | Borja Genoves Guzman, Ma\"it\'e Brandt-Pearce | ORIS allocation to minimize the outage probability in a multi-user VLC
scenario | null | B. G. Guzman and M. Brandt-Pearce, "ORIS allocation to minimize
the outage probability in a multi-user VLC scenario," in IEEE Photonics
Technology Letters, 2025 | 10.1109/LPT.2025.3566979 | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visible Light Communication (VLC) is a promising solution to address the
growing demand for wireless data, leveraging the widespread use of
light-emitting diodes (LEDs) as transmitters. However, its deployment is
challenged by link blockages that cause connectivity outages. Optical
reconfigurable intelligent surfaces (ORISs) have recently emerged as a solution
to mitigate these disruptions. This work considers a multi-user VLC system and
investigates the optimal association of ORISs to LEDs and users to minimize the
outage probability while limiting the number of ORISs used. Numerical results
from our proposed optimization algorithm demonstrate that using ORISs can
reduce the outage probability by up to 85% compared to a no-ORIS scenario.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 15:24:25 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 08:28:23 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guzman",
"Borja Genoves",
""
],
[
"Brandt-Pearce",
"Maïté",
""
]
] |
2504.13231 | Isar Nejadgholi | Braeden Sherritt and Isar Nejadgholi and Marzieh Amini | WildFireCan-MMD: A Multimodal Dataset for Classification of
User-Generated Content During Wildfires in Canada | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Rapid information access is vital during wildfires, yet traditional data
sources are slow and costly. Social media offers real-time updates, but
extracting relevant insights remains a challenge. We present WildFireCan-MMD, a
new multimodal dataset of X posts from recent Canadian wildfires, annotated
across twelve key themes. Evaluating both vision-language models and
custom-trained classifiers, we show that while zero-shot prompting offers quick
deployment, even simple trained models outperform them when labelled data is
available. Our best-performing transformer-based fine-tuned model reaches 83%
f-score, outperforming gpt4 by 23%. As a use case, we demonstrate how this
model can be used to uncover trends during wildfires. Our findings highlight
the enduring importance of tailored datasets and task-specific training.
Importantly, such datasets should be localized, as disaster response
requirements vary across regions and contexts.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 14:43:56 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:47:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sherritt",
"Braeden",
""
],
[
"Nejadgholi",
"Isar",
""
],
[
"Amini",
"Marzieh",
""
]
] |
2504.13670 | Guangyu Zhu | Guangyu Zhu, Xidong Mu, Li Guo, Shibiao Xu, Yuanwei Liu, and Naofal
Al-Dhahir | Pinching-Antenna Systems (PASS)-enabled Secure Wireless Communications | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A novel pinching-antenna systems (PASS)-enabled secure wireless communication
framework is proposed. By dynamically adjusting the positions of dielectric
particles, namely pinching antennas (PAs), along the waveguides, PASS
introduces a novel concept of pinching beamforming to enhance the performance
of physical layer security. A fundamental PASS-enabled secure communication
system is considered with one legitimate user and one eavesdropper. Both
single-waveguide and multiple-waveguide scenarios are studied. 1) For the
single-waveguide scenario, the secrecy rate (SR) maximization is formulated to
optimize the pinching beamforming. A PA-wise successive tuning (PAST) algorithm
is proposed, which ensures constructive signal superposition at the legitimate
user while inducing a destructive legitimate signal at the eavesdropper. 2) For
the multiple-waveguide scenario, artificial noise (AN) is employed to further
improve secrecy performance. A pair of practical transmission architectures are
developed: waveguide division (WD) and waveguide multiplexing (WM). The key
difference lies in whether each waveguide carries a single type of signal or a
mixture of signals with baseband beamforming. For the SR maximization problem
under the WD case, a two-stage algorithm is developed, where the pinching
beamforming is designed with the PAST algorithm and the baseband power
allocation among AN and legitimate signals is solved using successive convex
approximation (SCA). For the WM case, an alternating optimization algorithm is
developed, where the baseband beamforming is optimized with SCA and the
pinching beamforming is designed employing particle swarm optimization.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 13:04:56 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 02:03:34 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zhu",
"Guangyu",
""
],
[
"Mu",
"Xidong",
""
],
[
"Guo",
"Li",
""
],
[
"Xu",
"Shibiao",
""
],
[
"Liu",
"Yuanwei",
""
],
[
"Al-Dhahir",
"Naofal",
""
]
] |
2504.14040 | Van Sy Mai | Van Sy Mai, Abderrahim Amlou, Amar Abane, Abdella Battou | Towards Optimal Orders for Entanglement Swapping in Path Graphs: A
Greedy Approach | 11 pages, 11 figures | null | null | null | quant-ph cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | This paper considers the problem of finding an optimal order for entanglement
swapping in a heterogeneous path of quantum repeaters so as to maximize the
path throughput defined as the delivery rate of end-to-end entanglements. The
primary difficulty in addressing this problem lies in the vast array of
possible swapping orders for large paths and the complexity of the expected
throughput, which depends on the attributes of each node and edge along the
path, as well as the order of swapping. To cope with these issues, we first
propose simple approximations in estimating the swapping outcome between two
entanglement distributions that can run in constant time, thereby providing an
efficient approach for evaluating and comparing different swapping orders,
allowing us to solve the problem exactly for small paths. Second, as the number
of possible orders grows exponentially with the number of repeaters in the
path, we develop an efficient heuristic based on the greedy selection of nodes
to sequentially perform swaps according to their swapping scores, defined as
the expected number of entanglements resulting from their swaps. The scores are
local but dynamic in the sense that they depend not just on the entanglement
distributions available on the path but also on prior swapping decisions.
Finally, we illustrate the efficiency and effectiveness of our proposed model
and approach through extensive experimentation conducted using a general
quantum network simulator.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 19:10:13 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:06:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Mai",
"Van Sy",
""
],
[
"Amlou",
"Abderrahim",
""
],
[
"Abane",
"Amar",
""
],
[
"Battou",
"Abdella",
""
]
] |
2504.14756 | Alfred Michel Grundland | {\L}ukasz Chomienia, Alfred Michel Grundland | Lie modules analysis of hydrodynamic-type systems | null | null | null | null | math-ph math.AP math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The objective of this paper is to study nonlinear superpositions of Riemann
wave solutions admitted by quasilinear hyperbolic first-order systems of
partial differential equations. Particular attention is devoted to the analysis
of non-elastic wave superpositions that cannot be decomposed into pairwise
independent interactions of waves (quasi-rectifiability). In the case of the
compressible Euler system, we describe the structure of the
infinite-dimensional Lie algebra of vector fields associated with waves. We
prove that a certain class of Lie modules associated with a hydrodynamic-type
systems can be uniquely transformed into a real Lie algebra in an
angle-preserving manner. For the Euler system, we then demonstrate the
connection between the transformed finite-dimensional Lie algebra and the
infinite-dimensional algebra associated with waves. This enables a detailed
investigation of the geometry of wave superpositions using tools from
differential geometry and Lie group theory. In particular, we study the
geometry of the manifold of wave superpositions in terms of deformations of
submanifolds corresponding to Lie subalgebras associated with waves.
Additionally, we present new methods and criteria for determining the
quasi-rectifiability of Riemann k-waves.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 22:33:57 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 17:38:00 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Chomienia",
"Łukasz",
""
],
[
"Grundland",
"Alfred Michel",
""
]
] |
2504.14791 | Xing Wang | Xing M. Wang | Quantum Measurement Without Collapse or Many Worlds: The Branched
Hilbert Subspace Interpretation | 15 pages, 1 figure, 1 table | null | null | null | quant-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The interpretation of quantum measurements remains contested between
collapse-based frameworks like the Copenhagen Interpretation and no-collapse
approaches like the Many-Worlds Interpretation (MWI). We propose the Branched
Hilbert Subspace Interpretation (BHSI) as a minimalist alternative that
preserves unitarity while avoiding both wavefunction collapse and ontological
excess. BHSI models measurement as the unitary branching of a system's local
Hilbert space into decoherent subspaces, formalized through causal state
updates using branching and disengaging operators. Unlike MWI, BHSI avoids
parallel worlds by maintaining a single-world ontology in which branching is
confined to observable subspaces; unlike Copenhagen, it eliminates collapse
while recovering the Born rule through branch weights. Through physically
meaningful subspace records, the framework resolves quantum paradoxes such as
particle-wave duality, Wigner and his friend, black hole radiation, etc. It
remains consistent with interference patterns, entanglement correlations, and
information preservation. By comparing BHSI with QBism, Relational Quantum
Mechanics, and modal interpretations, as well as analyzing quantum
teleportation (where locally decoherent Hilbert subspaces are observed), we
demonstrate its advantages as a causally sound and empirically grounded
approach that reconciles unitary evolution with definite measurement outcomes
without metaphysical proliferation.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 01:34:57 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Apr 2025 01:36:04 GMT"
},
{
"version": "v3",
"created": "Wed, 14 May 2025 18:55:27 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Xing M.",
""
]
] |
2504.14869 | K.X. Hu | Xin-Ke Hu (BIT), Jia-Xuan Li (BIT), Yu-Wei Yu (BIT), Ji-Shun Lian
(BIT), Wei Deng (GXU), Hai-Ming Zhang (GXU), Jin Zhang (BIT) | X-ray Polarization of the High-Synchrotron-Peaked BL Lac H 1426+428 | 21 pages, 9 figures, 3 tables, Accepted for publication in ApJ | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report the X-ray polarization properties of the high-synchrotron-peaked BL
Lac H 1426+428, based on two-epoch observational data from the Imaging X-ray
Polarimetry Explorer (IXPE). For the first observation, only an upper limit of
polarization degree ($\Pi_{\rm X}$), $\Pi_{\rm X}<19.5\%$, at the $99\%$
confidence level (C.L.) is determined. In contrast, for the second observation,
we derive $\Pi_{\rm X}=20.6\%\pm2.9\%$ with a polarization angle ($\psi_{\rm
X}$) of $\psi_{\rm X}=116.1^{\circ}\pm4.1^{\circ}$ at a C.L. of 7.1 $\sigma$.
The time-resolved and energy-resolved polarization analysis reveals no
significant variation in $\psi_{\rm X}$ and no detectable polarization within
narrower energy bins for the first observation, while the polarization during
the second observation is dominated by low-energy photons. Furthermore, the
X-rays during the second observation are found to be in a higher flux state
with a harder spectrum compared to that observed during the first observation,
consistent with a {\it harder-when-brighter} behavior. We propose that the
enhanced X-ray emission observed during the second observation is produced by
shock-accelerated electrons within an ordered magnetic field region via
synchrotron radiation. Nonetheless, no significant detection of polarization
during the first IXPE observation may be due to the limited number of detected
photons.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 05:38:20 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 02:34:15 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hu",
"Xin-Ke",
"",
"BIT"
],
[
"Li",
"Jia-Xuan",
"",
"BIT"
],
[
"Yu",
"Yu-Wei",
"",
"BIT"
],
[
"Lian",
"Ji-Shun",
"",
"BIT"
],
[
"Deng",
"Wei",
"",
"GXU"
],
[
"Zhang",
"Hai-Ming",
"",
"GXU"
],
[
"Zhang",
"Jin",
"",
"BIT"
]
] |
2504.15023 | Taketo Yoshida | Taketo Yoshida, Tohru Nagao, Yoshiki Toba, Akatoki Noboriguchi, Kohei
Ichikawa, Hendrik Hildebrandt, Naomichi Yutani, Kenneth C. Chambers, Ryo
Iwamoto, Seira Kobayashi, Masamune Oguri, Ken Osato, Kohei Shibata, Yuxing
Zhong | Dust-obscured Galaxies with Broken Power-law Spectral Energy
Distributions Discovered by UNIONS | 13 pages, 10 figures, 2 tables, accepted for publication in ApJ | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We report on the spectral energy distributions (SEDs) of infrared-bright
dust-obscured galaxies (DOGs) with $(i - [22])_{\rm AB} \geq 7.0$. Using
photometry from the deep and wide Ultraviolet Near-Infrared Optical Northern
Survey, combined with near-IR and mid-IR data from the UKIRT Infrared Deep Sky
Survey and the Wide-field Infrared Survey Explorer, we successfully identified
382 DOGs in $\sim$ 170 deg$^2$. Among them, the vast majority (376 DOGs) were
classified into two subclasses: bump DOGs (132/376) and power-law (PL) DOGs
(244/376), which are dominated by star formation and active galactic nucleus
(AGN), respectively. Through the SED analysis, we found that roughly half
(120/244) of the PL DOGs show ``broken'' power-law SEDs. The significant red
slope from optical to near-IR in the SEDs of these ``broken power-law DOGs''
(BPL DOGs) probably reflects their large amount of dust extinction. In other
words, BPL DOGs are more heavily obscured AGNs, compared to PL DOGs with
non-broken power-law SEDs.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 11:11:41 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 04:12:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Yoshida",
"Taketo",
""
],
[
"Nagao",
"Tohru",
""
],
[
"Toba",
"Yoshiki",
""
],
[
"Noboriguchi",
"Akatoki",
""
],
[
"Ichikawa",
"Kohei",
""
],
[
"Hildebrandt",
"Hendrik",
""
],
[
"Yutani",
"Naomichi",
""
],
[
"Chambers",
"Kenneth C.",
""
],
[
"Iwamoto",
"Ryo",
""
],
[
"Kobayashi",
"Seira",
""
],
[
"Oguri",
"Masamune",
""
],
[
"Osato",
"Ken",
""
],
[
"Shibata",
"Kohei",
""
],
[
"Zhong",
"Yuxing",
""
]
] |
2504.15285 | Mario De Lucio | Mario de Lucio, Jacobo Diaz, Alberto de Castro, Luis E. Romera | AneuPy: An open source Python tool for creating simulation-ready
geometries of abdominal aortic aneurysms | 14 pages, 5 figures | null | null | null | physics.med-ph cs.CE cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | Abdominal aortic aneurysms (AAAs) are localized dilatations of the abdominal
aorta that can lead to life-threatening rupture if left untreated. AAAs
primarily affect older individuals, with high mortality rates following
rupture, so early diagnosis and risk assessment are critical. The geometrical
characteristics of an AAA, such as its maximum diameter, asymmetry, and wall
thickness, are extremely significant in biomechanical models for the assessment
of rupture risk. Despite the growing use of computational modeling for AAA
investigation, there is a notable gap in accessible, open-source software
capable of generating simulation-ready geometries for biomechanical and
hemodynamic simulations. To address this gap, we introduce \textbf{AneuPy}, an
open-source Python-based tool designed to create both idealized and
patient-specific AAA geometric models. \textbf{AneuPy} is a fast and automated
approach for generating aneurysm geometries from minimal input data, allowing
for extensive parameter customization. By automating the creation of
simulation-ready geometries for finite element analysis (FEA), computational
fluid dynamics (CFD), or fluid-structure interaction (FSI) models,
\textbf{AneuPy} can facilitate research in AAA and improve patient-specific
risk prediction.
| [
{
"version": "v1",
"created": "Thu, 13 Mar 2025 16:05:32 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 17:59:31 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"de Lucio",
"Mario",
""
],
[
"Diaz",
"Jacobo",
""
],
[
"de Castro",
"Alberto",
""
],
[
"Romera",
"Luis E.",
""
]
] |
2504.16251 | Chia-Che Tsai | Vijay Dhanraj, Harpreet Singh Chwarla, Tao Zhang, Daniel Manila, Eric
Thomas Schneider, Erica Fu, Mona Vij, Chia-Che Tsai, Donald E. Porter | Adaptive and Efficient Dynamic Memory Management for Hardware Enclaves | 12 pages, 10 figures | null | null | null | cs.OS cs.CR | http://creativecommons.org/licenses/by/4.0/ | The second version of Intel Software Guard Extensions (Intel SGX), or SGX2,
adds dynamic management of enclave memory and threads. The first version
required the address space and thread counts to be fixed before execution. The
Enclave Dynamic Memory Management (EDMM) feature of SGX2 has the potential to
lower launch times and overall execution time. Despite reducing the enclave
loading time by 28--93%, straightforward EDMM adoption strategies actually slow
execution time down by as much as 58%. Using the Gramine library OS as a
representative enclave runtime environment, this paper shows how to recover
EDMM performance. The paper explains how implementing mutual distrust between
the OS and enclave increases the cost of modifying page mappings. The paper
then describes and evaluates a series of optimizations on application
benchmarks, showing that these optimizations effectively eliminate the
overheads of EDMM while retaining EDMM's performance and flexibility gains.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 20:28:44 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 19:58:34 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Dhanraj",
"Vijay",
""
],
[
"Chwarla",
"Harpreet Singh",
""
],
[
"Zhang",
"Tao",
""
],
[
"Manila",
"Daniel",
""
],
[
"Schneider",
"Eric Thomas",
""
],
[
"Fu",
"Erica",
""
],
[
"Vij",
"Mona",
""
],
[
"Tsai",
"Chia-Che",
""
],
[
"Porter",
"Donald E.",
""
]
] |
2504.16287 | Anwesh Ray | Eknath Ghate, Anwesh Ray | Deformations of reducible Galois representations with large Selmer
$p$-rank | Version 2: 43 pages; minor corrections incorporated | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | Let $p\geq 5$ be a prime number. In this paper, we construct Galois
representations associated with modular forms for which the dimension of the
$p$-torsion in the Bloch-Kato Selmer group can be made arbitrarily large. Our
result extends similar results known for small primes, such as Matsuno's work
on Tate-Shafarevich groups of elliptic curves. Extending the technique of
Hamblen and Ramakrishna, we lift residually reducible Galois representations to
modular representations for which the associated Greenberg Selmer groups are
minimally generated by a large number of elements over the Iwasawa algebra. We
deduce that there is an isogenous lattice for which the Bloch-Kato Selmer group
has large $p$-rank.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 21:49:49 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:50:15 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ghate",
"Eknath",
""
],
[
"Ray",
"Anwesh",
""
]
] |
2504.17185 | Shuiyin Liu | Shuiyin Liu and Amin Sakzad | Compact Lattice-Coded (Multi-Recipient) Kyber without CLT Independence
Assumption | 8 Tables, 3 Figures | null | null | null | cs.CR cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work presents a joint design of encoding and encryption procedures for
public key encryptions (PKEs) and key encapsulation mechanism (KEMs) such as
Kyber, without relying on the assumption of independent decoding noise
components, achieving reductions in both communication overhead (CER) and
decryption failure rate (DFR). Our design features two techniques: ciphertext
packing and lattice packing. First, we extend the Peikert-Vaikuntanathan-Waters
(PVW) method to Kyber: $\ell$ plaintexts are packed into a single ciphertext.
This scheme is referred to as P$_\ell$-Kyber. We prove that the P$_\ell$-Kyber
is IND-CCA secure under the M-LWE hardness assumption. We show that the
decryption decoding noise entries across the $\ell$ plaintexts (also known as
layers) are mutually independent. Second, we propose a cross-layer lattice
encoding scheme for the P$_\ell$-Kyber, where every $\ell$ cross-layer
information symbols are encoded to a lattice point. This way we obtain a
\emph{coded} P$_\ell$-Kyber, where the decoding noise entries for each lattice
point are mutually independent. Therefore, the DFR analysis does not require
the assumption of independence among the decryption decoding noise entries.
Both DFR and CER are greatly decreased thanks to ciphertext packing and lattice
packing. We demonstrate that with $\ell=24$ and Leech lattice encoder, the
proposed coded P$_\ell$-KYBER1024 achieves DFR $<2^{-281}$ and CER $ = 4.6$,
i.e., a decrease of CER by $90\%$ compared to KYBER1024.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 01:39:36 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 03:11:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Liu",
"Shuiyin",
""
],
[
"Sakzad",
"Amin",
""
]
] |
2504.17212 | Tianxing Ma | Ting Guo, Lufeng Zhang, Tianxing Ma and Hai-Qing Lin | Dual role of stripe phase on superconducting correlation in a bilayer
square lattice | 8 pages and 7 figures | Phys. Rev. B 111, 195127(2025) | 10.1103/PhysRevB.111.195127 | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While the stripe phase has been observed not only in monolayer cuprates but
also in bilayer cuprates, research on its behavior in bilayer cuprates has been
limited. Using constrained path quantum Monte Carlo, we explore the effect of
stripes on the bilayer square lattice. We find the system exhibits short-range
antiferromagnetism, which is enhanced by stripes and is strongest when the
electron density of the interstriped rows reaches half-filling. The hole doping
concentration plays a crucial role in the interaction between stripes and
superconductivity. The $d$-wave pairing is enhanced by stripe potential $V_0$
at the hole doping $\delta_h=1/4$, whereas it is suppressed by stripe potential
$V_0$ at the hole doping $\delta_h=1/8$. We elucidate this phenomenon through
an analysis of the magnetism of the interstriped rows. Furthermore, the
effective $d$-wave pairing is stronger in the bilayer model compared to the
monolayer model when stripes are introduced on the square lattice. Overall, our
unbiased numerical simulations provide a further understanding of the crossed
bilayer square lattice model.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 02:54:31 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guo",
"Ting",
""
],
[
"Zhang",
"Lufeng",
""
],
[
"Ma",
"Tianxing",
""
],
[
"Lin",
"Hai-Qing",
""
]
] |
2504.17671 | YuanChang Ye | Yuanchang Ye and Weiyan Wen | Data-Driven Calibration of Prediction Sets in Large Vision-Language
Models Based on Inductive Conformal Prediction | Accepted by ICIPCA 2025 | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study addresses the critical challenge of hallucination mitigation in
Large Vision-Language Models (LVLMs) for Visual Question Answering (VQA) tasks
through a Split Conformal Prediction (SCP) framework. While LVLMs excel in
multi-modal reasoning, their outputs often exhibit hallucinated content with
high confidence, posing risks in safety-critical applications. We propose a
model-agnostic uncertainty quantification method that integrates dynamic
threshold calibration and cross-modal consistency verification. By partitioning
data into calibration and test sets, the framework computes nonconformity
scores to construct prediction sets with statistical guarantees under
user-defined risk levels ($\alpha$). Key innovations include: (1) rigorous
control of \textbf{marginal coverage} to ensure empirical error rates remain
strictly below $\alpha$; (2) dynamic adjustment of prediction set sizes
inversely with $\alpha$, filtering low-confidence outputs; (3) elimination of
prior distribution assumptions and retraining requirements. Evaluations on
benchmarks (ScienceQA, MMMU) with eight LVLMs demonstrate that SCP enforces
theoretical guarantees across all $\alpha$ values. The framework achieves
stable performance across varying calibration-to-test split ratios,
underscoring its robustness for real-world deployment in healthcare, autonomous
systems, and other safety-sensitive domains. This work bridges the gap between
theoretical reliability and practical applicability in multi-modal AI systems,
offering a scalable solution for hallucination detection and uncertainty-aware
decision-making.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 15:39:46 GMT"
},
{
"version": "v2",
"created": "Fri, 25 Apr 2025 09:34:59 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 16:24:49 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ye",
"Yuanchang",
""
],
[
"Wen",
"Weiyan",
""
]
] |
2504.17752 | Zhihui Gao | Zhihui Gao, Sri Krishna Vadlamani, Kfir Sulimany, Dirk Englund, and
Tingjun Chen | Disaggregated Deep Learning via In-Physics Computing at Radio Frequency | 11 pages, 4 figures. Supplementary Information: 54 pages, 20 figures,
1 table | null | null | null | cs.ET cs.LG eess.SP physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern edge devices, such as cameras, drones, and Internet-of-Things nodes,
rely on deep learning to enable a wide range of intelligent applications,
including object recognition, environment perception, and autonomous
navigation. However, deploying deep learning models directly on the often
resource-constrained edge devices demands significant memory footprints and
computational power for real-time inference using traditional digital computing
architectures. In this paper, we present WISE, a novel computing architecture
for wireless edge networks designed to overcome energy constraints in deep
learning inference. WISE achieves this goal through two key innovations:
disaggregated model access via wireless broadcasting and in-physics computation
of general complex-valued matrix-vector multiplications directly at radio
frequency. Using a software-defined radio platform with wirelessly broadcast
model weights over the air, we demonstrate that WISE achieves 95.7% image
classification accuracy with ultra-low operation power of 6.0 fJ/MAC per
client, corresponding to a computation efficiency of 165.8 TOPS/W. This
approach enables energy-efficient deep learning inference on wirelessly
connected edge devices, achieving more than two orders of magnitude improvement
in efficiency compared to traditional digital computing.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 17:10:18 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Gao",
"Zhihui",
""
],
[
"Vadlamani",
"Sri Krishna",
""
],
[
"Sulimany",
"Kfir",
""
],
[
"Englund",
"Dirk",
""
],
[
"Chen",
"Tingjun",
""
]
] |
2504.17907 | Arunangshu Debnath | Arunangshu Debnath | Coherent nonlinear optical probe for cavity-dressed vibrational mode
mixing: multidimensional double-quantum coherence and photon-echo signal | Submitted 11/04/2025. Update: Part of Eq. C3 (in Appendix) was
missing. It has been updated | null | null | null | quant-ph physics.chem-ph physics.optics | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Cavity modulation of interacting molecular vibrational modes promises an
expanded role of the characteristic vibrations to act as spectroscopic markers.
In addition to the intrinsic, inter-mode and extrinsic, cavity-mediated
inter-mode delocalization, the vibrational modes undergo multiscale relaxation
due to reservoir vibrations. Two complementary coherent multidimensional
spectroscopic techniques are presented, which reveal the dynamical nature of
the mode correlations in the presence relaxation. Simulations are presented for
finite laser pulses. The chosen cavity-vibron mode coupling strengths fall
within the range of inter-vibrational mode coupling, hinting at a parameter
regime where vibrational energy dispersal can be controlled. It is highlighted
that the cavity manipulation of vibrational double excitations, namely,
overtone and combination states, can offer deeper insights into the collective
mode dynamics. These observations set the ground for cavity-controlled
spectroscopic investigation of mode-selective dynamics, where localized
electromagnetic modes function as an effective few-body coupler of vibrations.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 19:45:12 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 13:30:07 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Debnath",
"Arunangshu",
""
]
] |
2504.17975 | T. M. Prosser | T. M. Prosser | A General Theory of Abiogenesis: Thermodynamic Selection and the
Emergence of Life | 38 pages, 3 figures, 1 appendix. This v3 update significantly expands
the original model by introducing the Thermodynamic Abiogenesis Likelihood
Model (TALM), a probabilistic framework incorporating entropy-diffusion
penalties, planetary scenarios, and simulation pathways. All figures updated.
Text reorganized for clarity | null | null | null | physics.bio-ph q-bio.PE | http://creativecommons.org/licenses/by/4.0/ | The origin of life is often approached through the lens of replication,
heredity, or molecular specificity. This paper proposes a thermodynamic
framework in which the emergence of life is driven by the persistence of
reaction pathways that align energetically with fluctuating environmental
inputs. We define a reaction viability inequality based on energy input,
return, resilience, and expenditure, which selects for persistent chemical
configurations without invoking heredity or genetic encoding.
We further incorporate entropic dynamics and spatial constraints into an
augmented persistence function, showing that systems far from equilibrium can
simultaneously increase global entropy while supporting localized chemical
order. These refinements lead to the development of the Thermodynamic
Abiogenesis Likelihood Model (TALM), a probabilistic extension that estimates
the likelihood of persistence-driven selection under various prebiotic and
planetary scenarios.
This framework redefines the conditions under which life-like organization
may emerge and provides a testable, general theory for abiogenesis grounded in
physical law.
| [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 23:04:32 GMT"
},
{
"version": "v2",
"created": "Fri, 9 May 2025 00:07:11 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 13:07:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Prosser",
"T. M.",
""
]
] |
2504.18008 | Nooshin Yousefzadeh Hosseini | Nooshin Yousefzadeh, Rahul Sengupta, Jeremy Dilmore, Sanjay Ranka | TGDT: A Temporal Graph-based Digital Twin for Urban Traffic Corridors | 8 pages, 4 figures, 1 table | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Urban congestion at signalized intersections leads to significant delays,
economic losses, and increased emissions. Existing deep learning models often
lack spatial generalizability, rely on complex architectures, and struggle with
real-time deployment. To address these limitations, we propose the Temporal
Graph-based Digital Twin (TGDT), a scalable framework that integrates Temporal
Convolutional Networks and Attentional Graph Neural Networks for dynamic,
direction-aware traffic modeling and assessment at urban corridors. TGDT
estimates key Measures of Effectiveness (MOEs) for traffic flow optimization at
both the intersection level (e.g., queue length, waiting time) and the corridor
level (e.g., traffic volume, travel time). Its modular architecture and
sequential optimization scheme enable easy extension to any number of
intersections and MOEs. The model outperforms state-of-the-art baselines by
accurately producing high-dimensional, concurrent multi-output estimates. It
also demonstrates high robustness and accuracy across diverse traffic
conditions, including extreme scenarios, while relying on only a minimal set of
traffic features. Fully parallelized, TGDT can simulate over a thousand
scenarios within a matter of seconds, offering a cost-effective, interpretable,
and real-time solution for urban traffic management and optimization.
| [
{
"version": "v1",
"created": "Fri, 25 Apr 2025 01:28:32 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 03:23:36 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Yousefzadeh",
"Nooshin",
""
],
[
"Sengupta",
"Rahul",
""
],
[
"Dilmore",
"Jeremy",
""
],
[
"Ranka",
"Sanjay",
""
]
] |
2504.18336 | Hong-Min Jiang | Hong-Min Jiang, Wen-Qian Dong, Shun-Li Yu, and Z. D. Wang | The possible frustrated superconductivity in the kagome superconductors | 10 pages, 5 figures | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Geometric frustration has long been a subject of enduring interest in
condensed matter physics. While geometric frustration traditionally focuses on
magnetic systems, little attention is paid to the "frustrated
superconductivity" which could arise when the superconducting interaction
conflicts with the crystal symmetry. The recently discovered kagome
superconductors provide a particular opportunity for studying this due to the
fact that the frustrated lattice structure and the interference effect between
the three sublattices can facilitate the frustrated superconducting
interaction. Here, we propose a theory that supports the frustrated
superconducting state, derived from the on-site $s$-wave superconducting
pairing in conjunction with the nearest-neighbor pairings hoping and the unique
geometrical frustrated lattice structure. In this state, whereas the mutual
$2\pi/3$ difference of the superconducting pairing phase causes the six-fold
modulation of the amplitude and breaks the time-reversal symmetry with $4\pi$
phase changes of the superconducting pairing as one following it around the
Fermi surface, it is immune to the impurities without the impurity-induced
in-gap states and produces the pronounced Hebel-Slichter peak of the nuclear
spin-lattice relaxation rate below $T_{c}$. Notably, the theory also reveals a
disorder-induced superconducting pairing transition from the frustrated
superconducting state to an isotropic $s$-wave superconducting state without
traversing the nodal points, recovering and explaining the behavior found in
experiment. This study not only serves as a promising proposal to mediate the
divergent or seemingly contradictory experimental outcomes regarding
superconducting pairing symmetry, but may also pave the way for advancing
investigations into the frustrated superconducting state.
| [
{
"version": "v1",
"created": "Fri, 25 Apr 2025 13:24:11 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 03:28:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Jiang",
"Hong-Min",
""
],
[
"Dong",
"Wen-Qian",
""
],
[
"Yu",
"Shun-Li",
""
],
[
"Wang",
"Z. D.",
""
]
] |
2504.18763 | Antonio Vidiella-Barranco | E.P. Mattos and A. Vidiella-Barranco | Phase-space approach to cavity field dynamics in a squeezed thermal
reservoir | 15 pages, 2 figures, Accepted for publication in Annals of Physics | Ann.Phys. 479, 170055 (2025) | 10.1016/j.aop.2025.170055 | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Quantum systems, such as a single-mode cavity field coupled to a thermal
bath, typically experience destructive effects due to interactions with their
noisy environment. When the bath combines both thermal fluctuations and a
nonclassical feature like quadrature squeezing, forming a squeezed thermal
reservoir, the system's behaviour can change substantially. In this work, we
study the evolution of the cavity field in this generalized environment using
an alternative phase-space approach based on the Glauber-Sudarshan
$P$-function. We derive a compact analytical expression for the time-dependent
$P$-function for arbitrary initial cavity field states and demonstrate its
utility through specific examples. Additionally, we obtain analytical
expressions, as a function of time, for some statistical properties of the
cavity field, as well as for the nonclassical depth, $\tau_m$, a
nonclassicality measure calculated directly from the $P$-function.
| [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 01:48:28 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Mattos",
"E. P.",
""
],
[
"Vidiella-Barranco",
"A.",
""
]
] |
2504.18940 | Jonathan White | J. S. White, V. Ukleev, L. Yu, Y. Tokura, Y. Taguchi, K. Karube | Anisotropy-dependent decay of room temperature metastable skyrmions and
a nascent double-$q$ spin texture in Co$_{8}$Zn$_{9}$Mn$_{3}$ | 22 pages, 6 figures | Adv. Mater. 2501146 (2025) | 10.1002/adma.202501146 | null | cond-mat.mtrl-sci cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Chiral cubic Co-Zn-Mn magnets exhibit diverse topological spin textures,
including room-temperature skyrmion phases and robust far-from-equilibrium
metastable states. Despite recent advances in understanding metastable
skyrmions, the interplay between compositional disorder and varying magnetic
anisotropy on the stability and decay of metastable textures, particularly near
room temperature, remains incompletely understood. In this work, the
equilibrium and metastable skyrmion formation in Co$_{8}$Zn$_{9}$Mn$_{3}$ is
examined, revealing transformations between distinct metastable spin textures
induced by temperature and magnetic field. At room temperature, the decay
dynamics of metastable skyrmions exhibits a strong dependence on magnetic
anisotropy, showcasing a route towards tailoring relaxation behavior.
Furthermore, a nascent double-$q$ spin texture, characterized by two coexisting
magnetic modulation vectors $q$, is identified as a minority phase alongside
the conventional triple-$q$ hexagonal skyrmion lattice. This double-$q$ texture
can be quenched as a metastable state, suggesting both its topological
character, and its role as a potential intermediary of metastable skyrmion
decay. These findings provide new insights into the tunability of equilibrium
and metastable topological spin textures via chemical composition and magnetic
anisotropy, offering strategies for designing materials with customizable and
dynamic skyrmion properties for advanced technological applications.
| [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 14:53:38 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"White",
"J. S.",
""
],
[
"Ukleev",
"V.",
""
],
[
"Yu",
"L.",
""
],
[
"Tokura",
"Y.",
""
],
[
"Taguchi",
"Y.",
""
],
[
"Karube",
"K.",
""
]
] |
2504.18960 | Tetsuya Takaishi | Tetsuya Takaishi | Impact of the COVID-19 pandemic on the financial market efficiency of
price returns, absolute returns, and volatility increment: Evidence from
stock and cryptocurrency markets | 20 pages, 10 figures | null | 10.3390/jrfm18050237 | null | q-fin.ST | http://creativecommons.org/licenses/by/4.0/ | This study examines the impact of the coronavirus disease 2019 (COVID-19)
pandemic on market efficiency by analyzing three time series -- price returns,
absolute returns, and volatility increments -- in stock (Deutscher Aktienindex,
Nikkei 225, Shanghai Stock Exchange (SSE), and Volatility Index) and
cryptocurrency (Bitcoin and Ethereum) markets. The effect is found to vary by
asset class and market. In the stock market, while the pandemic did not
influence the Hurst exponent of volatility increments, it affected that of
returns and absolute returns (except in the SSE, where returns remained
unaffected). In the cryptocurrency market, the pandemic did not alter the Hurst
exponent for any time series but influenced the strength of multifractality in
returns and absolute returns. Some Hurst exponent time series exhibited a
gradual decline over time, complicating the assessment of pandemic-related
effects. Consequently, segmented analyses by pandemic periods may erroneously
suggest an impact, warranting caution in period-based studies.
| [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 16:02:54 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Takaishi",
"Tetsuya",
""
]
] |
2504.19139 | Yun Qu | Yun Qu, Qi Cheems Wang, Yixiu Mao, Yiqin Lv, Xiangyang Ji | Fast and Robust: Task Sampling with Posterior and Diversity Synergies
for Adaptive Decision-Makers in Randomized Environments | ICML 2025 | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Task robust adaptation is a long-standing pursuit in sequential
decision-making. Some risk-averse strategies, e.g., the conditional
value-at-risk principle, are incorporated in domain randomization or meta
reinforcement learning to prioritize difficult tasks in optimization, which
demand costly intensive evaluations. The efficiency issue prompts the
development of robust active task sampling to train adaptive policies, where
risk-predictive models are used to surrogate policy evaluation. This work
characterizes the optimization pipeline of robust active task sampling as a
Markov decision process, posits theoretical and practical insights, and
constitutes robustness concepts in risk-averse scenarios. Importantly, we
propose an easy-to-implement method, referred to as Posterior and Diversity
Synergized Task Sampling (PDTS), to accommodate fast and robust sequential
decision-making. Extensive experiments show that PDTS unlocks the potential of
robust active task sampling, significantly improves the zero-shot and few-shot
adaptation robustness in challenging tasks, and even accelerates the learning
process under certain scenarios. Our project website is at
https://thu-rllab.github.io/PDTS_project_page.
| [
{
"version": "v1",
"created": "Sun, 27 Apr 2025 07:27:17 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 08:16:08 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 01:51:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Qu",
"Yun",
""
],
[
"Wang",
"Qi Cheems",
""
],
[
"Mao",
"Yixiu",
""
],
[
"Lv",
"Yiqin",
""
],
[
"Ji",
"Xiangyang",
""
]
] |
2504.19458 | Taoyu Su | Taoyu Su, Jiawei Sheng, Duohe Ma, Xiaodong Li, Juwei Yue, Mengxiao
Song, Yingkai Tang, Tingwen Liu | Mitigating Modality Bias in Multi-modal Entity Alignment from a Causal
Perspective | Accepted by SIGIR 2025, 11 pages, 10 figures, 4 tables, | null | 10.1145/3726302.3730037 | null | cs.MM cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-Modal Entity Alignment (MMEA) aims to retrieve equivalent entities from
different Multi-Modal Knowledge Graphs (MMKGs), a critical information
retrieval task. Existing studies have explored various fusion paradigms and
consistency constraints to improve the alignment of equivalent entities, while
overlooking that the visual modality may not always contribute positively.
Empirically, entities with low-similarity images usually generate
unsatisfactory performance, highlighting the limitation of overly relying on
visual features. We believe the model can be biased toward the visual modality,
leading to a shortcut image-matching task. To address this, we propose a
counterfactual debiasing framework for MMEA, termed CDMEA, which investigates
visual modality bias from a causal perspective. Our approach aims to leverage
both visual and graph modalities to enhance MMEA while suppressing the direct
causal effect of the visual modality on model predictions. By estimating the
Total Effect (TE) of both modalities and excluding the Natural Direct Effect
(NDE) of the visual modality, we ensure that the model predicts based on the
Total Indirect Effect (TIE), effectively utilizing both modalities and reducing
visual modality bias. Extensive experiments on 9 benchmark datasets show that
CDMEA outperforms 14 state-of-the-art methods, especially in low-similarity,
high-noise, and low-resource data scenarios.
| [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 03:48:23 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Apr 2025 13:58:48 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 09:07:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Su",
"Taoyu",
""
],
[
"Sheng",
"Jiawei",
""
],
[
"Ma",
"Duohe",
""
],
[
"Li",
"Xiaodong",
""
],
[
"Yue",
"Juwei",
""
],
[
"Song",
"Mengxiao",
""
],
[
"Tang",
"Yingkai",
""
],
[
"Liu",
"Tingwen",
""
]
] |
2504.19709 | Davide Ricci | Davide Ricci, Fulvio Laudisio, Alessandro Lorenzetto, Marco De
Pascale, Andrea Baruffolo, Daniele Vassallo, Domenico Barbato, Maria Bergomi,
Florian Briegel, Elena Carolo, Simone Di Filippo, Marco Dima, Valentina
D'Orazi, Tania Sofia Gomes Machado, Davide Greggio, Luca Marafatto, Dino
Mesa, Lars Mohr, Gabriele Rodeghiero, Kalyan Kumar Radhakrishnan
Santhakumari, Gabriele Umbriaco, Valentina Viotto, Jacopo Farinato | SHINS, the SHARK-NIR Instrument Control Software | 26 pages, 20 figures. Accepted for publication in JATIS | null | 10.1117/1.JATIS.11.2.025005 | null | astro-ph.IM astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | SHARK-NIR is a new compact instrument for coronagraphic imaging, direct
imaging, and coronagraphic spectroscopy in the near-infrared wavelengths
mounted at LBT. Taking advantage of the telescope's adaptive optics system, it
provides high contrast imaging with coronagraphic and spectroscopic
capabilities and is focused on the direct imaging of exoplanets and
circumstellar discs. We present SHINS, the SHARK-NIR instrument control
software, mainly realized with the TwiceAsNice framework from MPIA - Heidelberg
and the ICE framework using the C++ programming language. We describe how we
implemented the software components controlling several instrument subsystems,
through the adaptation of already tested libraries from other instruments at
LBT, such as LINC-NIRVANA. The scientific detector comes with its own readout
electronic and control software interfaced with our software through INDI. We
describe the C++ core software Observation Control Software, responsible for
dispatching commands to the subsystems, also implementing a software solution
to avoid a potential collision between motorized components, fully transparent
to final users. It exposes an ICE interface and can be controlled by clients
developed in different languages. Observation, calibration, and maintenance
procedures are implemented by means of template scripts, written in python
language, controlling Observation Control Software through its ICE interface.
These templates and their parameters are configured using "ESO-style", XML
Observation Blocks prepared by observers, or in general SHARK-NIR users. The
high-level control is carried out by REST HTTP APIs implemented in a python
back-end, also acting as a web server for the several browser-based front-end
GUIs. Finally, we present the first scientific results obtained by SHARK-NIR
using coronagraphic mode.
| [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 12:03:49 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ricci",
"Davide",
""
],
[
"Laudisio",
"Fulvio",
""
],
[
"Lorenzetto",
"Alessandro",
""
],
[
"De Pascale",
"Marco",
""
],
[
"Baruffolo",
"Andrea",
""
],
[
"Vassallo",
"Daniele",
""
],
[
"Barbato",
"Domenico",
""
],
[
"Bergomi",
"Maria",
""
],
[
"Briegel",
"Florian",
""
],
[
"Carolo",
"Elena",
""
],
[
"Di Filippo",
"Simone",
""
],
[
"Dima",
"Marco",
""
],
[
"D'Orazi",
"Valentina",
""
],
[
"Machado",
"Tania Sofia Gomes",
""
],
[
"Greggio",
"Davide",
""
],
[
"Marafatto",
"Luca",
""
],
[
"Mesa",
"Dino",
""
],
[
"Mohr",
"Lars",
""
],
[
"Rodeghiero",
"Gabriele",
""
],
[
"Santhakumari",
"Kalyan Kumar Radhakrishnan",
""
],
[
"Umbriaco",
"Gabriele",
""
],
[
"Viotto",
"Valentina",
""
],
[
"Farinato",
"Jacopo",
""
]
] |
2504.20043 | Juan Urrutia | John Ellis, Malcolm Fairbairn, Juan Urrutia and Ville Vaskonen | Starlight from JWST: Implications for star formation and dark matter
models | 15 pages, 11 figures, a PBH constraint plot has been added | null | null | KCL-PH-TH/2025-12, CERN-TH-2025-085, AION-REPORT/2025-03 | astro-ph.CO astro-ph.GA hep-ph | http://creativecommons.org/licenses/by/4.0/ | We confront the star formation rate in different dark matter (DM) models with
UV luminosity data from JWST up to $z\simeq25$ and legacy data from HST. We
find that a transition from a Salpeter population to top-heavy Pop-III stars is
likely at $z\simeq10$ and that beyond $z=10-15$ the feedback from supernovae
and active galactic nuclei is progressively reduced, so that at $z\simeq25$ the
production of stars is almost free from any feedback. We compare fuzzy and warm
DM models that suppress small-scale structures with the CDM paradigm, finding
that the fuzzy DM mass $> 4.5 \times 10^{-22}{\rm eV}$ and the warm DM mass $>
1.5\, {\rm keV}$ at the 95\% CL. The fits of the star formation rate
parametrization do not depend strongly on the DM properties within the allowed
range. We find no preference over CDM for enhanced matter perturbations
associated with axion miniclusters or primordial black holes. The scale of the
enhancement of the power spectrum should be $> 27\,{\rm Mpc}^{-1}$ at the 95\%
CL, excluding axion miniclusters produced for $m_a < 7.5 \times 10^{-17}\,{\rm
eV}$ or heavy primordial black holes that constitute a fraction $f_{\rm PBH} >
{\rm Max}[88 M_{\odot}/m_{\rm PBH}, 10^{-4} (m_{\rm PBH}/10^4
M_{\odot})^{-0.09}]$ of DM.
| [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 17:59:58 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 12:59:27 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ellis",
"John",
""
],
[
"Fairbairn",
"Malcolm",
""
],
[
"Urrutia",
"Juan",
""
],
[
"Vaskonen",
"Ville",
""
]
] |
2504.20262 | Yaroslav Ivanashev | Yaroslav Ivanashev | On closure properties of TotP | 11 pages | null | null | null | cs.CC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The class TotP consists of functions that count the number of all paths of a
nondeterministic polynomial-time Turing machine. In this paper, we give a
second definition of the class TotP in terms of certificates and verifiers, and
present a few natural closure properties of this class. We also prove that the
closure of TotP under left composition with the class FP+ is equivalent to TotP
= FP+ and P = PP, and give examples of FP+-functions such that if TotP is
closed under composition with them, then it is closed under composition with
FP+.
| [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 21:07:13 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 04:14:11 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ivanashev",
"Yaroslav",
""
]
] |
2504.20538 | Gianluca Castignani | G. Castignani, F. Combes, P. Salom\'e, A. Edge, and P. Jablonka | Molecular gas in cool-core brightest cluster galaxies at $z\simeq0.4$ | 18 pages, 10 figures, 4 tables, accepted for publication in Astronomy
& Astrophysics, revised version after grammar corrections | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Brightest cluster galaxies (BCG) are today passive and very massive galaxies
at the center of their clusters, still accreting mass through swallowing
companions, and flows of cold gas, regulated by radio-mode active galactic
nucleus (AGN) feedback. However, their formation history is still a matter of
debate. We report new results based on millimeter observations performed with
the Northern Extended Millimeter Array (NOEMA) interferometer, mapping the cold
molecular gas (CO) that feeds the star formation of distant BCGs. We selected
three among the strongest cool-core BCGs at intermediate redshifts
($z\simeq0.4$), namely RX 1532, MACS 1447, and CHIPS 1911. Previous unresolved
millimeter observations and multi-wavelength analysis showed that they are
among the most star forming (${\rm SFR}\simeq100~ M_\odot/{\rm yr}$) and gas
rich ($M_{H_2}\simeq10^{11}~M_\odot$) BCGs at intermediate redshifts. The
selected BCGs are thus caught in a phase of rapid mass assembly, which makes
them ideal targets for high-resolution observations of their molecular gas. By
combining NOEMA intensity and velocity maps with archival images from the
Hubble Space Telescope, we detect in-situ star formation, filaments of
accreting cold gas likely regulated by AGN feedback, disturbed morphology
associated with tidal tails of molecular gas, as well as gas compression and
tails originated from stripping of gas. While effective condensation of the
intra-cluster medium is required to explain the large molecular gas reservoirs,
the BCGs exhibit a broad variety of environment-driven mechanisms responsible
for the processing of their cold gas: flows of cooling gas (RX 1532), ram
pressure or sloshing of the intra-cluster medium (MACS 1447), and galactic
tides (CHIPS 1911). This study thus sheds new insights on the physical
mechanisms responsible for the mass assembly of galaxies hosting AGN at the
center of clusters.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 08:31:45 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 09:36:36 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Castignani",
"G.",
""
],
[
"Combes",
"F.",
""
],
[
"Salomé",
"P.",
""
],
[
"Edge",
"A.",
""
],
[
"Jablonka",
"P.",
""
]
] |
2504.20817 | Pascal Thomas | Quang Dieu Nguyen, Pascal J. Thomas | Removable sets for pseudoconvexity for weakly smooth boundaries | We corrected inaccuracies in our presentation (and proof) of the
previous result by Grauert & Remmert; 16 pages | null | null | null | math.CV math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that for bounded domains in $\mathbb C^n$ with $\mathcal C^{1,1}$
smooth boundary, if there is a closed set $F$ of $2n-1$-Lebesgue measure $0$
such that $\partial \Omega \setminus F$ is $\mathcal C^{2}$-smooth and locally
pseudoconvex at every point, then $\Omega$ is globally pseudoconvex. Unlike in
the globally $\mathcal C^{2}$-smooth case, the condition ``$F$ of (relative)
empty interior'' is not enough to obtain such a result. We also give some
results under peak-set type hypotheses, which in particular provide a new proof
of an old result of Grauert and Remmert about removable sets for
pseudoconvexity under minimal hypotheses of boundary regularity.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 14:36:57 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:18:49 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Nguyen",
"Quang Dieu",
""
],
[
"Thomas",
"Pascal J.",
""
]
] |
2504.20916 | Priya Priya | Monu Kadyan, Priya, and Sanjay Kumar Singh | A generalization of Ramanujan's sum over finite groups | null | null | null | null | math.NT math.CO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Let $G$ be a finite group, and let $x \in G$. Define $[x^G] := \{ y \in G :
\langle x^G \rangle = \langle y^G \rangle \}$, where $\langle x^G \rangle$
denotes the normal subgroup of $G$ generated by the conjugacy class of $x$. In
this paper, we determine an explicit formula for the eigenvalues of the normal
Cayley graph $\text{Cay}(G, [x^G])$. These eigenvalues can be viewed as a
generalization of classical Ramanujan's sum in the setting of finite groups.
Surprisingly, the formula we derive for the eigenvalues of $\text{Cay}(G,
[x^G])$ extends the known formula of classical Ramanujan's sum to the context
of finite groups. This generalization not only enrich the theory of Ramanujan's
sum but also provide new tools in spectral graph theory, representation theory,
and algebraic number theory.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 16:35:01 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 06:05:47 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Kadyan",
"Monu",
""
],
[
"Priya",
"",
""
],
[
"Singh",
"Sanjay Kumar",
""
]
] |
2504.20931 | Rolando Ramos | Rolando Ramos, David Whiting | Generalized cluster algebras are subquotients of cluster algebras | 27 pages | null | null | null | math.RA math.AC | http://creativecommons.org/licenses/by/4.0/ | Generalized Cluster Algebras (GCA) are generalizations of Cluster Algebras
(CA) with higher-order exchange relations. Previously, Chekhov-Shapiro
conjectured that every GCA can be embedded into a CA. In this paper, we prove a
modified version of this conjecture by providing a construction that realizes a
given GCA as subquotient of some CA, as an algebra over the ground ring of the
GCA via restriction of scalars.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 16:48:33 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Apr 2025 19:28:14 GMT"
},
{
"version": "v3",
"created": "Wed, 14 May 2025 18:21:37 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ramos",
"Rolando",
""
],
[
"Whiting",
"David",
""
]
] |
2504.21149 | Suho Oh | SuHo Oh | Permutation ensembles and acyclicity on product of simplices | 20 pages, 14 tables, 3 figures | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose the study of $S_n$-ensembles: $n \times n$ arrays of permutations
of $[n]$ that encode the boundary data of $n\Delta_{n-1}$. We characterize
precisely when an $S_4$-ensemble contains a permutation appearing exactly four
times, leading to a complete description of the types of boundary of
$4\Delta_3$ that can be extended to fine mixed subdivisions.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 20:05:48 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:42:23 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Oh",
"SuHo",
""
]
] |
2504.21843 | Adarsha H A | H. A. Adarsha (MCNS, India), Chandrachur Chakraborty (MCNS, India),
Sudip Bhattacharyya (TIFR, India) | Accretion inside astrophysical objects : Effects of rotation and
viscosity | 16 pages, 15 figures, 1 table, accepted for publication in Physical
Review D | Phys. Rev. D 111, 103033 (2025) | 10.1103/PhysRevD.111.103033 | null | astro-ph.HE astro-ph.CO gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sub-solar mass black holes could show up in gravitational observations in
future and near-solar mass black holes might have been involved in the events
GW190425 and GW190814. Since they cannot form from the stellar evolution, their
creation requires exotic mechanisms. One such mechanism involves the capture of
dark matter particles by stellar objects and their thermalization. When the
criterion for the collapse of these dark matter particles is satisfied, a tiny
endoparasitic black hole (EBH) forms and then it accretes matter from the host.
The EBH may transmute the host into a black hole of nearly the same mass as the
host or lesser, depending on the type of accretion. We examine this complex and
poorly explored accretion mechanism, considering the effects of rotation and
viscosity but ignoring some other effects, such as those of pressure and
magnetic field, as the first step. Using a general framework to assess the
effects of rotation and viscosity on accretion, we show that the accretion
could be stalled in some white dwarfs, but not in neutron stars. The stalled
accretion should cause an opening in the host's polar regions, the extent of
which depends on the mass and spin of the host.
| [
{
"version": "v1",
"created": "Wed, 30 Apr 2025 17:52:25 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Adarsha",
"H. A.",
"",
"MCNS, India"
],
[
"Chakraborty",
"Chandrachur",
"",
"MCNS, India"
],
[
"Bhattacharyya",
"Sudip",
"",
"TIFR, India"
]
] |
2505.00551 | Chong Zhang | Chong Zhang, Yue Deng, Xiang Lin, Bin Wang, Dianwen Ng, Hai Ye,
Xingxuan Li, Yao Xiao, Zhanfeng Mo, Qi Zhang, Lidong Bing | 100 Days After DeepSeek-R1: A Survey on Replication Studies and More
Directions for Reasoning Language Models | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recent development of reasoning language models (RLMs) represents a novel
evolution in large language models. In particular, the recent release of
DeepSeek-R1 has generated widespread social impact and sparked enthusiasm in
the research community for exploring the explicit reasoning paradigm of
language models. However, the implementation details of the released models
have not been fully open-sourced by DeepSeek, including DeepSeek-R1-Zero,
DeepSeek-R1, and the distilled small models. As a result, many replication
studies have emerged aiming to reproduce the strong performance achieved by
DeepSeek-R1, reaching comparable performance through similar training
procedures and fully open-source data resources. These works have investigated
feasible strategies for supervised fine-tuning (SFT) and reinforcement learning
from verifiable rewards (RLVR), focusing on data preparation and method design,
yielding various valuable insights. In this report, we provide a summary of
recent replication studies to inspire future research. We primarily focus on
SFT and RLVR as two main directions, introducing the details for data
construction, method design and training procedure of current replication
studies. Moreover, we conclude key findings from the implementation details and
experimental results reported by these studies, anticipating to inspire future
research. We also discuss additional techniques of enhancing RLMs, highlighting
the potential of expanding the application scope of these models, and
discussing the challenges in development. By this survey, we aim to help
researchers and developers of RLMs stay updated with the latest advancements,
and seek to inspire new ideas to further enhance RLMs.
| [
{
"version": "v1",
"created": "Thu, 1 May 2025 14:28:35 GMT"
},
{
"version": "v2",
"created": "Fri, 2 May 2025 07:44:19 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 14:16:03 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zhang",
"Chong",
""
],
[
"Deng",
"Yue",
""
],
[
"Lin",
"Xiang",
""
],
[
"Wang",
"Bin",
""
],
[
"Ng",
"Dianwen",
""
],
[
"Ye",
"Hai",
""
],
[
"Li",
"Xingxuan",
""
],
[
"Xiao",
"Yao",
""
],
[
"Mo",
"Zhanfeng",
""
],
[
"Zhang",
"Qi",
""
],
[
"Bing",
"Lidong",
""
]
] |
2505.00885 | Dongho Chae | Dongho Chae | Liouville type theorem for double Beltrami solutions of the Hall-MHD
system in $\Bbb R^3$ | There is a calculation mistake, and the result is not yet valid | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we prove Liouville type theorem for the double Beltrami
solutions to the stationary Hall-MHD equations in $\Bbb R^3$. Let $(u, B)$ be a
smooth double Beltrami solution to the stationary Hall-MHD equations in $\Bbb
R^3$, satisfying $\int_{\Bbb R^3} (|u|^q + |B|^q )dx <+\infty$ for some $q\in
[2, 3)$, then $u=B=0$. In the case of $B=0$ the theorem reduces the previously
known Liouville type result for the Beltrami solutions to the Euler equations.
| [
{
"version": "v1",
"created": "Thu, 1 May 2025 22:01:19 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 01:09:41 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Chae",
"Dongho",
""
]
] |
2505.01047 | Letian Yi | Letian Yi, Siyuan Yang, Ying Cui, Zhilu Lai | Transforming physics-informed machine learning to convex optimization | 33 pages,14 figures | null | null | null | cs.CE math.OC physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Physics-Informed Machine Learning (PIML) offers a powerful paradigm of
integrating data with physical laws to address important scientific problems,
such as parameter estimation, inferring hidden physics, equation discovery, and
state prediction, etc. However, PIML still faces many serious optimization
challenges that significantly restrict its applications. In this study, we
propose a comprehensive framework that transforms PIML to convex optimization
to overcome all these limitations, referred to as Convex-PIML. The linear
combination of B-splines is utilized to approximate the data, promoting the
convexity of the loss function. By replacing the non-convex components of the
loss function with convex approximations, the problem is further converted into
a sequence of successively refined approximated convex optimization problems.
This conversion allows the use of well-established convex optimization
algorithms, obtaining solutions effectively and efficiently. Furthermore, an
adaptive knot optimization method based on error estimate is introduced to
mitigate the spectral bias issue of PIML, further improving the performance.
The proposed theoretically guaranteed framework is tested in scenarios with
distinct types of physical prior. The results indicate that optimization
problems are effectively solved in these scenarios, highlighting the potential
of the framework for broad applications.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 06:43:09 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:37:58 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Yi",
"Letian",
""
],
[
"Yang",
"Siyuan",
""
],
[
"Cui",
"Ying",
""
],
[
"Lai",
"Zhilu",
""
]
] |
2505.01256 | Andre Opris | Andre Opris | A First Runtime Analysis of NSGA-III on a Many-Objective Multimodal
Problem: Provable Exponential Speedup via Stochastic Population Update | To appear at IJCAI 2025 | null | null | null | cs.NE | http://creativecommons.org/publicdomain/zero/1.0/ | The NSGA-III is a prominent algorithm in evolutionary many-objective
optimization. It is well-suited for optimizing functions with more than three
objectives, setting it apart from the classic NSGA-II. However, theoretical
insights about NSGA-III of when and why it performs well are still in its early
development. This paper addresses this point and conducts a rigorous runtime
analysis of NSGA-III on the many-objective \textsc{OneJumpZeroJump} benchmark
(\textsc{OjZj} for short), providing the first runtime bounds where the number
of objectives is constant. We show that NSGA-III finds the Pareto front of
\textsc{OjZj} in time $O(n^{k+d/2}+ \mu n \ln(n))$ where $n$ is the problem
size, $d$ is the number of objectives, $k$ is the gap size, a problem specific
parameter, if its population size $\mu \in 2^{O(n)}$ is at least
$(2n/d+1)^{d/2}$. Notably, NSGA-III is faster than NSGA-II by a factor of
$\mu/n^{d/2}$ for some $\mu \in \omega(n^{d/2})$. We also show that a
stochastic population update, proposed by Bian et al., provably guarantees a
speedup of order $\Theta((k/b)^{k-1})$ in the runtime where $b>0$ is a
constant. To our knowledge, this is the first rigorous runtime analysis of
NSGA-III on \textsc{OjZj}. Proving these bounds requires a much deeper
understanding of the population dynamics of NSGA-III than previous papers
achieved.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 13:26:05 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:20:06 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Opris",
"Andre",
""
]
] |
2505.01360 | Andrea Tommasi | Andr\'ea Tommasi, Felipe S\'aez-Leiva, Michel Peyret, Riad Hassani,
Maurine Montagnat | Modelling spontaneous ductile (viscous) strain localisation on Earth | 16 pages, 12 figures | null | null | null | physics.geo-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A characteristic feature of the Earth is that diffuse thermal convection in
the mantle produces localized deformation at the surface: Plate Tectonics.
However, modelling this phenomenon remains a challenge, due to inability to
simulate spontaneous strain localisation in the deep sections of the plates,
where rocks deform by ductile processes and behave as highly viscous fluids.
Analysis of shear zones - the main expression of strain localisation in nature
- reveals that mechanical heterogeneity is key. Here we posit that the
bottleneck for self-consistent generation of strain localisation in
geodynamical models is poor representation of this heterogeneity and its
evolution during deformation, in particular at small scales. To bypass this
obstacle, we introduce a stochastic description of the mechanical properties of
the rocks, which evolves as a function of the local work rate. This approach
not only successfully generates steady-state shear zones from random
rheological heterogeneity fields, but reproduces the full range of observations
in nature and experiments. It allows defining the necessary conditions and a
regime diagram for viscous strain localisation and quantifying the resulting
bulk softening, paving the way for self-consistent modelling of the development
of plate boundaries and initiation of Plate Tectonics on Earth.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 15:57:06 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:42:30 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Tommasi",
"Andréa",
""
],
[
"Sáez-Leiva",
"Felipe",
""
],
[
"Peyret",
"Michel",
""
],
[
"Hassani",
"Riad",
""
],
[
"Montagnat",
"Maurine",
""
]
] |
2505.01719 | Yibo Yang | Zhi-Cheng Hu, Ji-Hao Wang, Xiangyu Jiang, Liuming Liu, Shi-Hao Su,
Peng Sun, and Yi-Bo Yang | Realization of all-to-all fermion propagator for the first principle
high accuracy strong interaction prediction | 20 pages, 15 figures, some of statements updated | null | null | null | hep-lat hep-ph nucl-th | http://creativecommons.org/licenses/by/4.0/ | We propose a ``blending" algorithm that projects the all-to-all fermion
propagator onto spatial low-frequency modes (LFM) combined with a stochastic
estimate of spatial high-frequency modes (SHFM) at each time slice. This
approach enables the calculation of arbitrary-point correlation functions for
arbitrary hadron states in strongly interacting quantum field theories (QFT)
with fermions, such as quantum chromodynamics (QCD). Specifically, LFM allows
the construction of spatially extended hadron states below a certain energy
threshold by diagonalizing multi-fermion interpolation fields. Meanwhile, the
local interactions required for N-point correlation functions in QFT can be
approximated in an unbiased manner through a reweighted summation of both LFM
and SHFM contributions. To demonstrate the efficiency of this algorithm, we
obtained {\color{black} $g_A^u=0.895(15)$, $g_A^d=-0.338(15)$,
$g_A^s=-0.0245(72)$, $g_A^{u+d+s}=0.533(28)$ and $g_A^{u-d}=1.2339(43)$ } for
nucleon at $m_{\pi}=300$ MeV and $a=0.077$ fm using 40 configurations. The
consistency check of the pion electric form factor and charge radius derived
from 3-point and 4-point correlation functions is also provided.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 07:11:50 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 13:48:56 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hu",
"Zhi-Cheng",
""
],
[
"Wang",
"Ji-Hao",
""
],
[
"Jiang",
"Xiangyu",
""
],
[
"Liu",
"Liuming",
""
],
[
"Su",
"Shi-Hao",
""
],
[
"Sun",
"Peng",
""
],
[
"Yang",
"Yi-Bo",
""
]
] |
2505.01848 | Takahisa Igata | Takahisa Igata | Deflection Angle in the Strong Deflection Limit and Quasinormal Modes in
Stationary Axisymmetric Spacetimes | 5 pages, no figures; v2: references added | null | null | null | gr-qc astro-ph.HE hep-th | http://creativecommons.org/licenses/by/4.0/ | We derive a coordinate-invariant expression for the photon deflection angle
in the strong deflection limit (SDL) of stationary axisymmetric spacetimes. The
key logarithmic-divergence coefficient is shown to depend only on quantities
locally measurable by a zero-angular-momentum observer -- curvature scalars,
the circumferential radius, and the proper angular velocity. The same
coefficient governs the damping rate of quasinormal modes (QNMs) in the eikonal
limit, establishing a curvature-based, model-independent connection between
QNMs and lensing in the SDL near rotating compact objects.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 15:44:12 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 06:50:36 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Igata",
"Takahisa",
""
]
] |
2505.01964 | WenTi Guo | Wen-Ti Guo, Junqi Xu, Yurong Yang, Huaiqiang Wang, Haijun Zhang | Altermagnetic type-II Multiferroics with N\'{e}el-order-locked Electric
Polarization | 7 pages, 4 figures | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Altermagnetism, an emergent magnetic phase featuring compensated collinear
magnetic moments and momentum-dependent spin splittings, has recently garnered
widespread interest. A critical issue concerns whether the unconventional spin
structures can generate spontaneous electric polarization in altermagnets,
thereby achieving type-II multiferroicity. Here, with the combination of
symmetry analysis and microscopic theory, we explicitly demonstrate the
generation of electric polarization by altermagnetic N\'eel order, and
establish a microscopic mechanism of N\'eel-order-locked electric polarization
in altermagnetic multiferroics.~We further reveal these pronounced
magnetoelectric coupling behaviors and classify them into eight distinct
categories for two-dimensional altermagnets governed by layer group symmetries.
Then we take monolayer MgFe$_2$N$_2$ as a prototypical example of altermagnetic
type-II multiferroics by first-principles calculations. We also propose to
identify the N\'eel order and accompanying electric polarization in
altermagnetic multiferroics by magneto-optical microscopy. Bridging type-II
multiferroics and altermagnets, our work could pave the way for altermagnetic
multifunctional spintronics.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 02:27:05 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 05:11:23 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Guo",
"Wen-Ti",
""
],
[
"Xu",
"Junqi",
""
],
[
"Yang",
"Yurong",
""
],
[
"Wang",
"Huaiqiang",
""
],
[
"Zhang",
"Haijun",
""
]
] |
2505.02145 | Ameth Ndiaye | Abdou Bousso and Ameth Ndiaye | Ricci Solitons on the Poincar\'e upper half plane | null | null | null | null | math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we characterize the Ricci soliton equations on the Poincar\'e
upper half plane .
First we classify all Ricci soliton and Ricci Bourguignon soliton in the half
plane of Poincar\'e
and after we generalize those equations in $\mathbb{H}^n$. We obtain some
nice properties of the soliton
about their geodesic flows.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 15:13:40 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 12:08:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bousso",
"Abdou",
""
],
[
"Ndiaye",
"Ameth",
""
]
] |
2505.02387 | Xiusi Chen | Xiusi Chen, Gaotang Li, Ziqi Wang, Bowen Jin, Cheng Qian, Yu Wang,
Hongru Wang, Yu Zhang, Denghui Zhang, Tong Zhang, Hanghang Tong, Heng Ji | RM-R1: Reward Modeling as Reasoning | 24 pages, 8 figures | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Reward modeling is essential for aligning large language models (LLMs) with
human preferences through reinforcement learning (RL). To provide accurate
reward signals, a reward model (RM) should stimulate deep thinking and conduct
interpretable reasoning before assigning a score or a judgment. Inspired by
recent advances of long chain-of-thought (CoT) on reasoning-intensive tasks, we
hypothesize and validate that integrating reasoning capabilities into reward
modeling significantly enhances RM's interpretability and performance. To this
end, we introduce a new class of generative reward models -- Reasoning Reward
Models (ReasRMs) -- which formulate reward modeling as a reasoning task. We
propose a reasoning-oriented training pipeline and train a family of ReasRMs,
RM-R1. RM-R1 features a chain-of-rubrics (CoR) mechanism -- self-generating
sample-level chat rubrics or math/code solutions, and evaluating candidate
responses against them. The training of M-R1 consists of two key stages: (1)
distillation of high-quality reasoning chains and (2) reinforcement learning
with verifiable rewards. Empirically, our models achieve state-of-the-art
performance across three reward model benchmarks on average, outperforming much
larger open-weight models (e.g., INF-ORM-Llama3.1-70B) and proprietary ones
(e.g., GPT-4o) by up to 4.9%. Beyond final performance, we perform thorough
empirical analysis to understand the key ingredients of successful ReasRM
training. To facilitate future research, we release six ReasRM models along
with code and data at https://github.com/RM-R1-UIUC/RM-R1.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 06:11:12 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 04:14:49 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Chen",
"Xiusi",
""
],
[
"Li",
"Gaotang",
""
],
[
"Wang",
"Ziqi",
""
],
[
"Jin",
"Bowen",
""
],
[
"Qian",
"Cheng",
""
],
[
"Wang",
"Yu",
""
],
[
"Wang",
"Hongru",
""
],
[
"Zhang",
"Yu",
""
],
[
"Zhang",
"Denghui",
""
],
[
"Zhang",
"Tong",
""
],
[
"Tong",
"Hanghang",
""
],
[
"Ji",
"Heng",
""
]
] |
2505.02581 | Hector Zenil | Alberto Hern\'andez-Espinosa, Felipe S. Abrah\~ao, Olaf Witkowski,
Hector Zenil | Neurodivergent Influenceability as a Contingent Solution to the AI
Alignment Problem | 44 pages | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The AI alignment problem, which focusses on ensuring that artificial
intelligence (AI), including AGI and ASI, systems act according to human
values, presents profound challenges. With the progression from narrow AI to
Artificial General Intelligence (AGI) and Superintelligence, fears about
control and existential risk have escalated. Here, we investigate whether
embracing inevitable AI misalignment can be a contingent strategy to foster a
dynamic ecosystem of competing agents as a viable path to steer them in more
human-aligned trends and mitigate risks. We explore how misalignment may serve
and should be promoted as a counterbalancing mechanism to team up with
whichever agents are most aligned to human interests, ensuring that no single
system dominates destructively. The main premise of our contribution is that
misalignment is inevitable because full AI-human alignment is a mathematical
impossibility from Turing-complete systems, which we also offer as a proof in
this contribution, a feature then inherited to AGI and ASI systems. We
introduce a change-of-opinion attack test based on perturbation and
intervention analysis to study how humans and agents may change or neutralise
friendly and unfriendly AIs through cooperation and competition. We show that
open models are more diverse and that most likely guardrails implemented in
proprietary models are successful at controlling some of the agents' range of
behaviour with positive and negative consequences while closed systems are more
steerable and can also be used against proprietary AI systems. We also show
that human and AI intervention has different effects hence suggesting multiple
strategies.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 11:33:18 GMT"
},
{
"version": "v2",
"created": "Thu, 8 May 2025 03:02:00 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 01:23:57 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hernández-Espinosa",
"Alberto",
""
],
[
"Abrahão",
"Felipe S.",
""
],
[
"Witkowski",
"Olaf",
""
],
[
"Zenil",
"Hector",
""
]
] |
2505.02658 | George P Efstathiou | George Efstathiou | Baryon Acoustic Oscillations from a Different Angle | Types fixed | null | null | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents an alternative way of analysing Baryon Acoustic
Oscillation (BAO) distance measurements via rotations to define new quantities
Dperp and Dpar. These quantities allow simple tests of consistency with the
Planck LCDM cosmology. The parameter Dperp is determined with negligible
uncertainty from Planck under the assumption of LCDM. Comparing with
measurements from the Dark Energy Spectroscopic Instrument (DESI), we find that
the measurements of Dperp from Data Release 2 (DR2) move into significantly
better agreement with the Planck LCDM cosmology compared to DESI Data Release 1
(DR1). The quantity in the orthogonal direction Dpar provides a measure of the
physical matter density omega_m in the LCDM cosmology. The DR2 measurements of
Dpar remain consistent with Planck LCDM despite the substantial improvement in
their accuracy compared to the earlier DR1 results. From the comparison of
Planck and DESI BAO measurements, we find no significant evidence in support of
evolving dark energy. We also investigate a rotation in the theory space of the
w_0 and w_a parameterization of the dark energy equation-of-state w(z). We show
that the combination of DESI BAO measurements and the CMB constrain w(z=0.5) =
-0.996 pm 0.046, i.e. very close to the value expected for a cosmological
constant. We present a critique of the statistical methodology employed by the
DESI collaboration and argue that it gives a misleading impression of the
evidence in favour of evolving dark energy. An Appendix shows that the
cosmological parameters determined from the Dark Energy Survey 5 Year supernova
sample are in tension with those from
DESI DR2 and parameters determined by Planck.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 14:03:37 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 19:17:47 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Efstathiou",
"George",
""
]
] |
2505.02767 | Zhi-Wei Sun | Zhi-Wei Sun | A family of polynomials and related congruences and series | 25 pages. Add the current Theorem 1.2 and some new conjectures | null | null | null | math.NT math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we study a family of polynomials
$$S_n^{(m)}(x):=\sum_{i,j=0}^n\binom ni^m\binom nj^m\binom{i+j}ix^{i+j}\ \
(m,n=0,1,2,\ldots).$$ For example, we show that
$$\sum_{k=0}^{p-1}S_k^{(0)}(x)\equiv\frac
x{2x-1}\left(1+\left(\frac{1-4x^2}p\right)\right)\pmod p $$ for any odd prime
$p$ and integer $x\not\equiv1/2\pmod p$, where $(\frac{\cdot}p)$ denotes the
Legendre symbol. We also formulate some open conjectures on related congruences
and series for $1/\pi$. For example, we conjecture that
$$\sum_{k=0}^\infty(7k+1)\frac{S_k^{(2)}(1/11)}{9^k}=\frac{5445}{104\sqrt{39}\,\pi}$$
and
$$\sum_{k=0}^\infty(1365k+181)\frac{S_k^{(2)}(1/18)}{16^k}=\frac{1377}{\sqrt2\,\pi}.$$
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 16:22:10 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:55:05 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Sun",
"Zhi-Wei",
""
]
] |
2505.03084 | Shashank Kapoor | Shashank Kapoor, Sanjay Surendranath Girija, Lakshit Arora, Dipen
Pradhan, Ankit Shetgaonkar, Aman Raj | Adversarial Attacks in Multimodal Systems: A Practitioner's Survey | Accepted in IEEE COMPSAC 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The introduction of multimodal models is a huge step forward in Artificial
Intelligence. A single model is trained to understand multiple modalities:
text, image, video, and audio. Open-source multimodal models have made these
breakthroughs more accessible. However, considering the vast landscape of
adversarial attacks across these modalities, these models also inherit
vulnerabilities of all the modalities, and ultimately, the adversarial threat
amplifies. While broad research is available on possible attacks within or
across these modalities, a practitioner-focused view that outlines attack types
remains absent in the multimodal world. As more Machine Learning Practitioners
adopt, fine-tune, and deploy open-source models in real-world applications,
it's crucial that they can view the threat landscape and take the preventive
actions necessary. This paper addresses the gap by surveying adversarial
attacks targeting all four modalities: text, image, video, and audio. This
survey provides a view of the adversarial attack landscape and presents how
multimodal adversarial threats have evolved. To the best of our knowledge, this
survey is the first comprehensive summarization of the threat landscape in the
multimodal world.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 00:41:16 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Kapoor",
"Shashank",
""
],
[
"Girija",
"Sanjay Surendranath",
""
],
[
"Arora",
"Lakshit",
""
],
[
"Pradhan",
"Dipen",
""
],
[
"Shetgaonkar",
"Ankit",
""
],
[
"Raj",
"Aman",
""
]
] |
2505.03093 | Siming He | Siming He, Zachary Osman, Fernando Cladera, Dexter Ong, Nitant Rai,
Patrick Corey Green, Vijay Kumar, Pratik Chaudhari | Estimating the Diameter at Breast Height of Trees in a Forest With a
Single 360 Camera | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Forest inventories rely on accurate measurements of the diameter at breast
height (DBH) for ecological monitoring, resource management, and carbon
accounting. While LiDAR-based techniques can achieve centimeter-level
precision, they are cost-prohibitive and operationally complex. We present a
low-cost alternative that only needs a consumer-grade 360 video camera. Our
semi-automated pipeline comprises of (i) a dense point cloud reconstruction
using Structure from Motion (SfM) photogrammetry software called Agisoft
Metashape, (ii) semantic trunk segmentation by projecting Grounded Segment
Anything (SAM) masks onto the 3D cloud, and (iii) a robust RANSAC-based
technique to estimate cross section shape and DBH. We introduce an interactive
visualization tool for inspecting segmented trees and their estimated DBH. On
61 acquisitions of 43 trees under a variety of conditions, our method attains
median absolute relative errors of 5-9% with respect to "ground-truth" manual
measurements. This is only 2-4% higher than LiDAR-based estimates, while
employing a single 360 camera that costs orders of magnitude less, requires
minimal setup, and is widely available.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 01:09:07 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:24:44 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"He",
"Siming",
""
],
[
"Osman",
"Zachary",
""
],
[
"Cladera",
"Fernando",
""
],
[
"Ong",
"Dexter",
""
],
[
"Rai",
"Nitant",
""
],
[
"Green",
"Patrick Corey",
""
],
[
"Kumar",
"Vijay",
""
],
[
"Chaudhari",
"Pratik",
""
]
] |
2505.03186 | Detao Bai | Detao Bai, Zhiheng Ma, Xihan Wei, Liefeng Bo | CoGenAV: Versatile Audio-Visual Representation Learning via
Contrastive-Generative Synchronization | null | null | null | null | cs.SD cs.CV eess.AS | http://creativecommons.org/licenses/by/4.0/ | The inherent synchronization between a speaker's lip movements, voice, and
the underlying linguistic content offers a rich source of information for
improving speech processing tasks, especially in challenging conditions where
traditional audio-only systems falter. We introduce CoGenAV, a powerful and
data-efficient model designed to learn versatile audio-visual representations
applicable across a wide range of speech and audio-visual tasks. CoGenAV is
trained by optimizing a dual objective derived from natural audio-visual
synchrony, contrastive feature alignment and generative text prediction, using
only 223 hours of labeled data from the LRS2 dataset. This
contrastive-generative synchronization strategy effectively captures
fundamental cross-modal correlations. We showcase the effectiveness and
versatility of the learned CoGenAV representations on multiple benchmarks. When
utilized for Audio-Visual Speech Recognition (AVSR) on LRS2, these
representations contribute to achieving a state-of-the-art Word Error Rate
(WER) of 1.27. They also enable strong performance in Visual Speech Recognition
(VSR) with a WER of 20.5 on LRS2, and significantly improve performance in
noisy environments by over 70%. Furthermore, CoGenAV representations benefit
speech reconstruction tasks, boosting performance in Speech Enhancement and
Separation, and achieve competitive results in audio-visual synchronization
tasks like Active Speaker Detection (ASD). Our model will be open-sourced to
facilitate further development and collaboration within both academia and
industry.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 05:07:11 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 07:21:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bai",
"Detao",
""
],
[
"Ma",
"Zhiheng",
""
],
[
"Wei",
"Xihan",
""
],
[
"Bo",
"Liefeng",
""
]
] |
2505.03292 | Viktor Iv\'ady | Andr\'as T\'ark\'anyi and Viktor Iv\'ady | Understanding Decoherence of the Boron Vacancy Center in Hexagonal Boron
Nitride | null | null | null | null | quant-ph cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Hexagonal boron nitride (hBN) has emerged as a significant material for
quantum sensing, particularly due to its ability to host spin active defects,
such as the negatively charged boron vacancy (V$_\mathrm{B}^-$ center). The
optical addressability of the V$_\mathrm{B}^-$ center and hBN's 2D structure
enable high spatial resolution and integration into various platforms. However,
decoherence due to the strong magnetic noise in hBN imposes fundamental
limitations on the sensitivity of V$_\mathrm{B}^-$ center-based applications.
Understanding the phenomena behind decoherence and identifying parameter
settings that provide the highest performance are essential for advancing
V$_\mathrm{B}^-$ sensors. This study employs state-of-the-art computational
methods to investigate the decoherence of the V$_\mathrm{B}^-$ center in
hexagonal boron nitride across a wide range of magnetic field values from 0 T
up to 3 T. The provided in-depth numerical and analytical analysis reveals an
intricate interplay of various decoherence mechanisms. This study identifies
five distinct magnetic field regions governed by different types of magnetic
interactions with and within the abundant nuclear spin bath. In addition to
magnetic field, the effects of zero-field splitting, nuclear polarization, and
different hyperfine coupling terms are studied, representing an important step
forward in utilizing V$_\mathrm{B}^-$ ensembles in sensing. In particular, this
study proposes operation in the moderate $180-350$ mT magnetic field range in
chemically pure h$^{11}$B$^{15}$N samples, where the coherence time can reach
$1-20$ $\mu$s, significantly exceeding the $\mathcal{O}( 100~\text{ns})$
low-field $T_2$ values.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 08:20:40 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:59:11 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Tárkányi",
"András",
""
],
[
"Ivády",
"Viktor",
""
]
] |
2505.03332 | Evgeny Markhasin | Evgeny Markhasin | AI-Driven Scholarly Peer Review via Persistent Workflow Prompting,
Meta-Prompting, and Meta-Reasoning | 22 pages, 36 pages (references and appendixes) | null | null | null | cs.AI physics.chem-ph | http://creativecommons.org/licenses/by-sa/4.0/ | Critical peer review of scientific manuscripts presents a significant
challenge for Large Language Models (LLMs), partly due to data limitations and
the complexity of expert reasoning. This report introduces Persistent Workflow
Prompting (PWP), a potentially broadly applicable prompt engineering
methodology designed to bridge this gap using standard LLM chat interfaces
(zero-code, no APIs). We present a proof-of-concept PWP prompt for the critical
analysis of experimental chemistry manuscripts, featuring a hierarchical,
modular architecture (structured via Markdown) that defines detailed analysis
workflows. We develop this PWP prompt through iterative application of
meta-prompting techniques and meta-reasoning aimed at systematically codifying
expert review workflows, including tacit knowledge. Submitted once at the start
of a session, this PWP prompt equips the LLM with persistent workflows
triggered by subsequent queries, guiding modern reasoning LLMs through
systematic, multimodal evaluations. Demonstrations show the PWP-guided LLM
identifying major methodological flaws in a test case while mitigating LLM
input bias and performing complex tasks, including distinguishing claims from
evidence, integrating text/photo/figure analysis to infer parameters, executing
quantitative feasibility checks, comparing estimates against claims, and
assessing a priori plausibility. To ensure transparency and facilitate
replication, we provide full prompts, detailed demonstration analyses, and logs
of interactive chats as supplementary resources. Beyond the specific
application, this work offers insights into the meta-development process
itself, highlighting the potential of PWP, informed by detailed workflow
formalization, to enable sophisticated analysis using readily available LLMs
for complex scientific tasks.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 09:06:18 GMT"
},
{
"version": "v2",
"created": "Thu, 8 May 2025 18:27:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Markhasin",
"Evgeny",
""
]
] |
2505.03449 | Dimitrios Tsimpis | Paul Marconnet and Dimitrios Tsimpis | Universal Cosmologies | 39 pages, 12 figures, 9 tables. v2: comments and references added | null | null | null | hep-th astro-ph.CO gr-qc | http://creativecommons.org/licenses/by/4.0/ | Universal cosmologies are exact solutions of 10d type IIA supergravity
containing a 4d Friedmann-Lema\^{i}tre-Robertson-Walker factor, that can also
be repackaged as solutions of 4d models, i.e. as 4d consistent truncations. We
extend the dynamical system analysis of universal cosmologies, beyond the case
of a single exponential potential. For an open universe (negative 3d spatial
curvature), these models generally possess many desirable features: parametric
control of e-folds, late-time acceleration from potentials with steep
exponentials (i.e. in accordance with swampland bounds), small string-loop and
$\alpha'$-corrections, scale separation and/or absence of decompactification.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 11:43:55 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:00:18 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Marconnet",
"Paul",
""
],
[
"Tsimpis",
"Dimitrios",
""
]
] |
2505.03551 | Mark Everitt | Mark J. Everitt | Emergence of a Classical Dirac-Type Spin from On-Shell Factorisation and
Liouvillian Evolution | Draft - improved explanations and text and a discussion on Leibniz
rules added - comments and feedback most welcome | null | null | null | quant-ph physics.class-ph | http://creativecommons.org/licenses/by/4.0/ | While Liouville's theorem is first order in time for the phase-space
distribution itself, the relativistic mass-shell constraint $p^\mu p_\mu = m^2$
is naively second order in energy. We argue it is reasonable to unify both
energy branches within a single Hamiltonian by factorising $(p^2 - m^2)$ in
analogy with Dirac's approach in relativistic quantum mechanics. We show the
resulting matrix-based Liouville equation remains first order and naturally
yields a $4\times4$ matrix-valued probability density function in phase space
as a classical analogue of a relativistic spin-half Wigner function.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 14:02:52 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 07:44:36 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Everitt",
"Mark J.",
""
]
] |
2505.03780 | Burkhard Ringlein | Burkhard Ringlein, Thomas Parnell, Radu Stoica | GPU Performance Portability needs Autotuning | typos, fix grammatical mistakes | null | null | null | cs.AR cs.AI cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As LLMs grow in complexity, achieving state-of-the-art performance requires
tight co-design across algorithms, software, and hardware. Today's reliance on
a single dominant platform limits portability, creates vendor lock-in, and
raises barriers for new AI hardware. In this work, we make the case for
combining just-in-time (JIT) compilation with kernel parameter autotuning to
enable portable LLM inference with state-of-the-art performance without code
changes. Focusing on flash attention -- a widespread performance critical LLM
kernel -- we demonstrate that this approach explores up to 15x more kernel
parameter configurations, produces significantly more diverse code across
multiple dimensions, and even outperforms vendor-optimized implementations by
up to 230%, all while reducing kernel code size by 70x and eliminating manual
code optimizations. Our results highlight autotuning as a promising path to
unlocking model portability across GPU vendors.
| [
{
"version": "v1",
"created": "Wed, 30 Apr 2025 12:57:21 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:26:40 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ringlein",
"Burkhard",
""
],
[
"Parnell",
"Thomas",
""
],
[
"Stoica",
"Radu",
""
]
] |
2505.03795 | Jacob Crandall | Jonathan Skaggs and Jacob W. Crandall | Modeling Human Behavior in a Strategic Network Game with Complex Group
Dynamics | null | null | null | null | cs.SI cs.AI physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | Human networks greatly impact important societal outcomes, including wealth
and health inequality, poverty, and bullying. As such, understanding human
networks is critical to learning how to promote favorable societal outcomes. As
a step toward better understanding human networks, we compare and contrast
several methods for learning, from a small data set, models of human behavior
in a strategic network game called the Junior High Game (JHG). These modeling
methods differ with respect to the assumptions they use to parameterize human
behavior (behavior vs. community-aware behavior) and the moments they model
(mean vs. distribution). Results show that the highest-performing method,
called hCAB, models the distribution of human behavior rather than the mean and
assumes humans use community-aware behavior rather than behavior matching. When
applied to small societies (6-11 individuals), the hCAB model closely mirrors
the population dynamics of human groups (with notable differences).
Additionally, in a user study, human participants were unable to distinguish
individual hCAB agents from other humans, thus illustrating that the hCAB model
also produces plausible (individual) human behavior in this strategic network
game.
| [
{
"version": "v1",
"created": "Thu, 1 May 2025 18:13:20 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 17:57:28 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Skaggs",
"Jonathan",
""
],
[
"Crandall",
"Jacob W.",
""
]
] |
2505.03825 | Anushiya Arunan | Anushiya Arunan, Yan Qin, Xiaoli Li, Yuen Chau | Intelligently Augmented Contrastive Tensor Factorization: Empowering
Multi-dimensional Time Series Classification in Low-Data Environments | Accepted in Expert Systems with Applications
(DOI:https://doi.org/10.1016/j.eswa.2025.127889) | null | 10.1016/j.eswa.2025.127889 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Classification of multi-dimensional time series from real-world systems
require fine-grained learning of complex features such as cross-dimensional
dependencies and intra-class variations-all under the practical challenge of
low training data availability. However, standard deep learning (DL) struggles
to learn generalizable features in low-data environments due to model
overfitting. We propose a versatile yet data-efficient framework, Intelligently
Augmented Contrastive Tensor Factorization (ITA-CTF), to learn effective
representations from multi-dimensional time series. The CTF module learns core
explanatory components of the time series (e.g., sensor factors, temporal
factors), and importantly, their joint dependencies. Notably, unlike standard
tensor factorization (TF), the CTF module incorporates a new contrastive loss
optimization to induce similarity learning and class-awareness into the learnt
representations for better classification performance. To strengthen this
contrastive learning, the preceding ITA module generates targeted but
informative augmentations that highlight realistic intra-class patterns in the
original data, while preserving class-wise properties. This is achieved by
dynamically sampling a "soft" class prototype to guide the warping of each
query data sample, which results in an augmentation that is intelligently
pattern-mixed between the "soft" class prototype and the query sample. These
augmentations enable the CTF module to recognize complex intra-class variations
despite the limited original training data, and seek out invariant class-wise
properties for accurate classification performance. The proposed method is
comprehensively evaluated on five different classification tasks. Compared to
standard TF and several DL benchmarks, notable performance improvements up to
18.7% were achieved.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 11:28:13 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 14:07:22 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Arunan",
"Anushiya",
""
],
[
"Qin",
"Yan",
""
],
[
"Li",
"Xiaoli",
""
],
[
"Chau",
"Yuen",
""
]
] |
2505.03883 | Takuya Takahashi | Takuya Takahashi, Hayato Motohashi, Kazufumi Takahashi | Resonance of black hole quasinormal modes in coupled systems | 21 pages, 8 figures | null | null | RESCEU-10/25 | gr-qc hep-th | http://creativecommons.org/licenses/by/4.0/ | Black hole quasinormal modes (QNMs) can exhibit resonant excitations
associated with avoided crossings in their complex frequency spectrum. Such
resonance phenomena can serve as novel signatures for probing new physics,
where additional degrees of freedom are commonly introduced. Motivated by this
possibility, we investigate QNMs in systems where multiple degrees of freedom
are coupled with each other, and introduce a definition of excitation factors
suitable for such systems. To demonstrate our formulation, we apply it to a
black hole in the Einstein-Maxwell-axion theory, where we find that avoided
crossings can appear even between longest-lived modes originating from the
fundamental modes of different degrees of freedom, in contrast to the Kerr case
in General Relativity. We show that the excitation factors are indeed amplified
as a manifestation of resonance at parameter values corresponding to the
avoided crossings.
| [
{
"version": "v1",
"created": "Tue, 6 May 2025 18:00:04 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 13:16:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Takahashi",
"Takuya",
""
],
[
"Motohashi",
"Hayato",
""
],
[
"Takahashi",
"Kazufumi",
""
]
] |
2505.04553 | Shanyu Han | Shanyu Han, Yang Liu, Xiang Yu | Risk-sensitive Reinforcement Learning Based on Convex Scoring Functions | 35 pages | null | null | null | q-fin.MF cs.AI q-fin.RM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a reinforcement learning (RL) framework under a broad class of
risk objectives, characterized by convex scoring functions. This class covers
many common risk measures, such as variance, Expected Shortfall, entropic
Value-at-Risk, and mean-risk utility. To resolve the time-inconsistency issue,
we consider an augmented state space and an auxiliary variable and recast the
problem as a two-state optimization problem. We propose a customized
Actor-Critic algorithm and establish some theoretical approximation guarantees.
A key theoretical contribution is that our results do not require the Markov
decision process to be continuous. Additionally, we propose an auxiliary
variable sampling method inspired by the alternating minimization algorithm,
which is convergent under certain conditions. We validate our approach in
simulation experiments with a financial application in statistical arbitrage
trading, demonstrating the effectiveness of the algorithm.
| [
{
"version": "v1",
"created": "Wed, 7 May 2025 16:31:42 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:40:05 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Han",
"Shanyu",
""
],
[
"Liu",
"Yang",
""
],
[
"Yu",
"Xiang",
""
]
] |
2505.04560 | Guanghui Wang | Guanghui Wang, Zhiyong Yang, Zitai Wang, Shi Wang, Qianqian Xu,
Qingming Huang | ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge
Distillation via $\alpha$-$\beta$-Divergence | ICML 2025 Spotlight | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Knowledge Distillation (KD) transfers knowledge from a large teacher model to
a smaller student model by minimizing the divergence between their output
distributions, typically using forward Kullback-Leibler divergence (FKLD) or
reverse KLD (RKLD). It has become an effective training paradigm due to the
broader supervision information provided by the teacher distribution compared
to one-hot labels. We identify that the core challenge in KD lies in balancing
two mode-concentration effects: the \textbf{\textit{Hardness-Concentration}}
effect, which refers to focusing on modes with large errors, and the
\textbf{\textit{Confidence-Concentration}} effect, which refers to focusing on
modes with high student confidence. Through an analysis of how probabilities
are reassigned during gradient updates, we observe that these two effects are
entangled in FKLD and RKLD, but in extreme forms. Specifically, both are too
weak in FKLD, causing the student to fail to concentrate on the target class.
In contrast, both are too strong in RKLD, causing the student to overly
emphasize the target class while ignoring the broader distributional
information from the teacher. To address this imbalance, we propose ABKD, a
generic framework with $\alpha$-$\beta$-divergence. Our theoretical results
show that ABKD offers a smooth interpolation between FKLD and RKLD, achieving
an effective trade-off between these effects. Extensive experiments on 17
language/vision datasets with 12 teacher-student settings confirm its efficacy.
The code is available at https://github.com/ghwang-s/abkd.
| [
{
"version": "v1",
"created": "Wed, 7 May 2025 16:48:49 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:13:43 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Guanghui",
""
],
[
"Yang",
"Zhiyong",
""
],
[
"Wang",
"Zitai",
""
],
[
"Wang",
"Shi",
""
],
[
"Xu",
"Qianqian",
""
],
[
"Huang",
"Qingming",
""
]
] |
2505.04609 | Anthony Taylor | Anthony J. Taylor, Vasily Kokorev, Dale D. Kocevski, Hollis B. Akins,
Fergus Cullen, Mark Dickinson, Steven L. Finkelstein, Pablo Arrabal Haro,
Volker Bromm, Mauro Giavalisco, Kohei Inayoshi, Stephanie Juneau, Gene C. K.
Leung, Pablo G. Perez-Gonzalez, Rachel S. Somerville, Jonathan R. Trump,
Ricardo O. Amorin, Guillermo Barro, Denis Burgarella, Madisyn Brooks, Adam
Carnall, Caitlin M. Casey, Yingjie Cheng, John Chisholm, Katherine
Chworowsky, Kelcey Davis, Callum T. Donnan, James S. Dunlop, Richard S.
Ellis, Vital Fernandez, Seiji Fujimoto, Norman A. Grogin, Ansh R. Gupta,
Nimish P. Hathi, Intae Jung, Michaela Hirschmann, Jeyhan S. Kartaltepe, Anton
M. Koekemoer, Rebecca L. Larson, Ho-Hin Leung, Mario Llerena, Ray A. Lucas,
Derek J. McLeod, Ross McLure, Lorenzo Napolitano, Casey Papovich, Thomas M.
Stanton, Roberta Tripodi, Xin Wang, Stephen M. Wilkins, L. Y. Aaron Yung, and
Jorge A. Zavala | CAPERS-LRD-z9: A Gas Enshrouded Little Red Dot Hosting a Broad-line AGN
at z=9.288 | 22 pages, 7 figures, 2 tables. Submitted to ApJL. V2: Updated
acknowledgements. V3: Fixed plotting error in Figure 7 and updated name of
"MoM-150135" to "MoM-BH*-1" | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We present CAPERS-LRD-z9, a little red dot (LRD) which we confirm to be a
$z=9.288$ broad-line AGN (BLAGN). First identified as a high-redshift LRD
candidate from PRIMER NIRCam photometry, follow-up NIRSpec/PRISM spectroscopy
of CAPERS-LRD-z9 from the CANDELS-Area Prism Epoch of Reionization Survey
(CAPERS) has revealed a broad $3500$ km s$^{-1}$ H$\beta$ emission line and
narrow [O III]$\lambda\lambda4959,5007$ lines, indicative of a BLAGN. Based on
the broad H$\beta$ line, we compute a canonical black-hole mass of
$\log(M_{\textrm{BH}}/M_{\odot})=7.58\pm0.15$, although full consideration of
systematic uncertainties yields a conservative range of
$6.65<\log(M_{\textrm{BH}}/M_{\odot})<8.50$. These observations suggest that
either a massive black hole seed, or a lighter stellar remnant seed undergoing
periods of super-Eddington accretion, is necessary to grow such a massive black
hole in $\lesssim500$ Myr of cosmic time. CAPERS-LRD-z9 exhibits a strong
Balmer break, consistent with a central AGN surrounded by dense ($\sim
10^{10}\textrm{ cm}^{-3}$) neutral gas. We model CAPERS-LRD-z9 using CLOUDY to
fit the emission red-ward of the Balmer break with a dense gas-enshrouded AGN,
and bagpipes to fit the rest-ultraviolet emission as a host-galaxy stellar
population. This upper limit on the stellar mass of the host galaxy
($<10^9\,{\rm M_\odot}$) implies that the black-hole to stellar mass ratio may
be extremely large, possibly $>5\%$ (although systematic uncertainties on the
black-hole mass prevent strong conclusions). However, the shape of the UV
continuum differs from typical high-redshift star-forming galaxies, indicating
that this UV emission may also be of AGN origin, and hence the true stellar
mass of the host may be still lower.
| [
{
"version": "v1",
"created": "Wed, 7 May 2025 17:53:48 GMT"
},
{
"version": "v2",
"created": "Fri, 9 May 2025 22:00:46 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 11:02:02 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Taylor",
"Anthony J.",
""
],
[
"Kokorev",
"Vasily",
""
],
[
"Kocevski",
"Dale D.",
""
],
[
"Akins",
"Hollis B.",
""
],
[
"Cullen",
"Fergus",
""
],
[
"Dickinson",
"Mark",
""
],
[
"Finkelstein",
"Steven L.",
""
],
[
"Haro",
"Pablo Arrabal",
""
],
[
"Bromm",
"Volker",
""
],
[
"Giavalisco",
"Mauro",
""
],
[
"Inayoshi",
"Kohei",
""
],
[
"Juneau",
"Stephanie",
""
],
[
"Leung",
"Gene C. K.",
""
],
[
"Perez-Gonzalez",
"Pablo G.",
""
],
[
"Somerville",
"Rachel S.",
""
],
[
"Trump",
"Jonathan R.",
""
],
[
"Amorin",
"Ricardo O.",
""
],
[
"Barro",
"Guillermo",
""
],
[
"Burgarella",
"Denis",
""
],
[
"Brooks",
"Madisyn",
""
],
[
"Carnall",
"Adam",
""
],
[
"Casey",
"Caitlin M.",
""
],
[
"Cheng",
"Yingjie",
""
],
[
"Chisholm",
"John",
""
],
[
"Chworowsky",
"Katherine",
""
],
[
"Davis",
"Kelcey",
""
],
[
"Donnan",
"Callum T.",
""
],
[
"Dunlop",
"James S.",
""
],
[
"Ellis",
"Richard S.",
""
],
[
"Fernandez",
"Vital",
""
],
[
"Fujimoto",
"Seiji",
""
],
[
"Grogin",
"Norman A.",
""
],
[
"Gupta",
"Ansh R.",
""
],
[
"Hathi",
"Nimish P.",
""
],
[
"Jung",
"Intae",
""
],
[
"Hirschmann",
"Michaela",
""
],
[
"Kartaltepe",
"Jeyhan S.",
""
],
[
"Koekemoer",
"Anton M.",
""
],
[
"Larson",
"Rebecca L.",
""
],
[
"Leung",
"Ho-Hin",
""
],
[
"Llerena",
"Mario",
""
],
[
"Lucas",
"Ray A.",
""
],
[
"McLeod",
"Derek J.",
""
],
[
"McLure",
"Ross",
""
],
[
"Napolitano",
"Lorenzo",
""
],
[
"Papovich",
"Casey",
""
],
[
"Stanton",
"Thomas M.",
""
],
[
"Tripodi",
"Roberta",
""
],
[
"Wang",
"Xin",
""
],
[
"Wilkins",
"Stephen M.",
""
],
[
"Yung",
"L. Y. Aaron",
""
],
[
"Zavala",
"Jorge A.",
""
]
] |
2505.04890 | Sora Kang | Sora Kang and Joonhwan Lee | Theatrical Language Processing: Exploring AI-Augmented Improvisational
Acting and Scriptwriting with LLMs | ISEA2025 (30th International Symposium on Electronic/Emerging Art) | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The increasing convergence of artificial intelligence has opened new avenues,
including its emerging role in enhancing creativity. It is reshaping
traditional creative practices such as actor improvisation, which often
struggles with predictable patterns, limited interaction, and a lack of
engaging stimuli. In this paper, we introduce a new concept, Theatrical
Language Processing (TLP), and an AI-driven creativity support tool,
Scribble.ai, designed to augment actors' creative expression and spontaneity
through interactive practice. We conducted a user study involving tests and
interviews with fourteen participants. Our findings indicate that: (1) Actors
expanded their creativity when faced with AI-produced irregular scenarios; (2)
The AI's unpredictability heightened their problem-solving skills, specifically
in interpreting unfamiliar situations; (3) However, AI often generated
excessively detailed scripts, which limited interpretive freedom and hindered
subtext exploration. Based on these findings, we discuss the new potential in
enhancing creative expressions in film and theater studies through an AI-driven
tool.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 01:50:32 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 03:20:29 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Kang",
"Sora",
""
],
[
"Lee",
"Joonhwan",
""
]
] |
2505.05113 | Martin Tassy | Alex Nezlobin and Martin Tassy | Loss-Versus-Rebalancing under Deterministic and Generalized block-times | 16 pages, 2 figures | null | null | null | q-fin.MF math.PR q-fin.PM q-fin.PR q-fin.TR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although modern blockchains almost universally produce blocks at fixed
intervals, existing models still lack an analytical formula for the
loss-versus-rebalancing (LVR) incurred by Automated Market Makers (AMMs)
liquidity providers in this setting. Leveraging tools from random walk theory,
we derive the following closed-form approximation for the per block per unit of
liquidity expected LVR under constant block time:
\[ \overline{\mathrm{ARB}}= \frac{\,\sigma_b^{2}}
{\,2+\sqrt{2\pi}\,\gamma/(|\zeta(1/2)|\,\sigma_b)\,}+O\!\bigl(e^{-\mathrm{const}\tfrac{\gamma}{\sigma_b}}\bigr)\;\approx\;
\frac{\sigma_b^{2}}{\,2 + 1.7164\,\gamma/\sigma_b}, \] where $\sigma_b$ is the
intra-block asset volatility, $\gamma$ the AMM spread and $\zeta$ the Riemann
Zeta function. Our large Monte Carlo simulations show that this formula is in
fact quasi-exact across practical parameter ranges.
Extending our analysis to arbitrary block-time distributions as well, we
demonstrate both that--under every admissible inter-block law--the probability
that a block carries an arbitrage trade converges to a universal limit, and
that only constant block spacing attains the asymptotically minimal LVR. This
shows that constant block intervals provide the best possible protection
against arbitrage for liquidity providers.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 10:30:24 GMT"
},
{
"version": "v2",
"created": "Fri, 9 May 2025 03:08:16 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 10:51:35 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Nezlobin",
"Alex",
""
],
[
"Tassy",
"Martin",
""
]
] |
2505.05145 | Xinyan Hu | Xinyan Hu, Kayo Yin, Michael I. Jordan, Jacob Steinhardt, Lijie Chen | Understanding In-context Learning of Addition via Activation Subspaces | 20 pages | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | To perform in-context learning, language models must extract signals from
individual few-shot examples, aggregate these into a learned prediction rule,
and then apply this rule to new examples. How is this implemented in the
forward pass of modern transformer models? To study this, we consider a
structured family of few-shot learning tasks for which the true prediction rule
is to add an integer $k$ to the input. We find that Llama-3-8B attains high
accuracy on this task for a range of $k$, and localize its few-shot ability to
just three attention heads via a novel optimization approach. We further show
the extracted signals lie in a six-dimensional subspace, where four of the
dimensions track the unit digit and the other two dimensions track overall
magnitude. We finally examine how these heads extract information from
individual few-shot examples, identifying a self-correction mechanism in which
mistakes from earlier examples are suppressed by later examples. Our results
demonstrate how tracking low-dimensional subspaces across a forward pass can
provide insight into fine-grained computational structures.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 11:32:46 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 07:19:33 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hu",
"Xinyan",
""
],
[
"Yin",
"Kayo",
""
],
[
"Jordan",
"Michael I.",
""
],
[
"Steinhardt",
"Jacob",
""
],
[
"Chen",
"Lijie",
""
]
] |
2505.05159 | Linhan Ma | Linhan Ma, Dake Guo, He Wang, Jin Xu, Lei Xie | FlexSpeech: Towards Stable, Controllable and Expressive Text-to-Speech | 10 pages, 5 figures | null | null | null | eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current speech generation research can be categorized into two primary
classes: non-autoregressive and autoregressive. The fundamental distinction
between these approaches lies in the duration prediction strategy employed for
predictable-length sequences. The NAR methods ensure stability in speech
generation by explicitly and independently modeling the duration of each
phonetic unit. Conversely, AR methods employ an autoregressive paradigm to
predict the compressed speech token by implicitly modeling duration with Markov
properties. Although this approach improves prosody, it does not provide the
structural guarantees necessary for stability. To simultaneously address the
issues of stability and naturalness in speech generation, we propose
FlexSpeech, a stable, controllable, and expressive TTS model. The motivation
behind FlexSpeech is to incorporate Markov dependencies and preference
optimization directly on the duration predictor to boost its naturalness while
maintaining explicit modeling of the phonetic units to ensure stability.
Specifically, we decompose the speech generation task into two components: an
AR duration predictor and a NAR acoustic model. The acoustic model is trained
on a substantial amount of data to learn to render audio more stably, given
reference audio prosody and phone durations. The duration predictor is
optimized in a lightweight manner for different stylistic variations, thereby
enabling rapid style transfer while maintaining a decoupled relationship with
the specified speaker timbre. Experimental results demonstrate that our
approach achieves SOTA stability and naturalness in zero-shot TTS. More
importantly, when transferring to a specific stylistic domain, we can
accomplish lightweight optimization of the duration module solely with about
100 data samples, without the need to adjust the acoustic model, thereby
enabling rapid and stable style transfer.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 11:55:19 GMT"
},
{
"version": "v2",
"created": "Fri, 9 May 2025 07:25:32 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 08:28:37 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Ma",
"Linhan",
""
],
[
"Guo",
"Dake",
""
],
[
"Wang",
"He",
""
],
[
"Xu",
"Jin",
""
],
[
"Xie",
"Lei",
""
]
] |
2505.05476 | Anthony Lin | Anthony Lin | Attachment: a predictive coding approach | 27 pages; major revision to model, emphasizing the role of
domain-general processes in the attachment system | null | null | null | q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | We introduce a novel predictive coding framework for studying attachment
theory. Building off an established model of attachment, the
dynamic-maturational model (DMM), as well as the neuroanatomical Embodied
Predictive Interoception Coding (EPIC) model of interoception and emotion, we
not only elucidate how neural processes can shape attachment strategies, but
also explore how early attachment experiences can shape those processes in the
first place.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 16:32:42 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 21:57:42 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Lin",
"Anthony",
""
]
] |
2505.05513 | Muhammad Junaid Asif | Muhammad Junaid Asif, Hamza Khan, Rabia Tehseen, Syed Tahir Hussain
Rizvi, Mujtaba Asad, Shazia Saqib, Rana Fayyaz Ahmad | Exploring Convolutional Neural Networks for Rice Grain Classification:
An Explainable AI Approach | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Rice is an essential staple food worldwide that is important in promoting
international trade, economic growth, and nutrition. Asian countries such as
China, India, Pakistan, Thailand, Vietnam, and Indonesia are notable for their
significant contribution to the cultivation and utilization of rice. These
nations are also known for cultivating different rice grains, including short
and long grains. These sizes are further classified as basmati, jasmine, kainat
saila, ipsala, arborio, etc., catering to diverse culinary preferences and
cultural traditions. For both local and international trade, inspecting and
maintaining the quality of rice grains to satisfy customers and preserve a
country's reputation is necessary. Manual quality check and classification is
quite a laborious and time-consuming process. It is also highly prone to
mistakes. Therefore, an automatic solution must be proposed for the effective
and efficient classification of different varieties of rice grains. This
research paper presents an automatic framework based on a convolutional neural
network (CNN) for classifying different varieties of rice grains. We evaluated
the proposed model based on performance metrics such as accuracy, recall,
precision, and F1-Score. The CNN model underwent rigorous training and
validation, achieving a remarkable accuracy rate and a perfect area under each
class's Receiver Operating Characteristic (ROC) curve. The confusion matrix
analysis confirmed the model's effectiveness in distinguishing between the
different rice varieties, indicating minimal misclassifications. Additionally,
the integration of explainability techniques such as LIME (Local Interpretable
Model-agnostic Explanations) and SHAP (SHapley Additive exPlanations) provided
valuable insights into the model's decision-making process, revealing how
specific features of the rice grains influenced classification outcomes.
| [
{
"version": "v1",
"created": "Wed, 7 May 2025 10:33:45 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 06:56:31 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Asif",
"Muhammad Junaid",
""
],
[
"Khan",
"Hamza",
""
],
[
"Tehseen",
"Rabia",
""
],
[
"Rizvi",
"Syed Tahir Hussain",
""
],
[
"Asad",
"Mujtaba",
""
],
[
"Saqib",
"Shazia",
""
],
[
"Ahmad",
"Rana Fayyaz",
""
]
] |
2505.05521 | Peiyan Hu | Peiyan Hu, Haodong Feng, Yue Wang, Zhiming Ma | Model-Based Closed-Loop Control Algorithm for Stochastic Partial
Differential Equation Control | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Neural operators have demonstrated promise in modeling and controlling
systems governed by Partial Differential Equations (PDEs). Beyond PDEs,
Stochastic Partial Differential Equations (SPDEs) play a critical role in
modeling systems influenced by randomness, with applications in finance,
physics, and beyond. However, controlling SPDE-governed systems remains a
significant challenge. On the one hand, the regularity of the system's state
(which can be intuitively understood as smoothness) deteriorates, making
modeling and generalization more challenging. On the other hand, this
stochasticity also renders control more unstable and thus less accurate. To
address this gap, we propose the Model-Based Closed-Loop Control Algorithm
(MB-CC), the first model-based closed-loop control method for SPDEs. MB-CC
introduces two key innovations to enhance control robustness and efficiency: a
Regularity Feature (RF) block and a closed-loop strategy with an
operator-encoded policy network. The RF block, inspired by the regularity
structure theory of SPDEs, addresses noise-induced irregularities by
transforming the network's input, including the system state and
noise-perturbed external forces, into a refined feature space for improved
forward prediction. Compared to previous works using regularity features, we
introduce a new parameterization, data augmentation, and extend the RF block as
a plug-and-play component. Additionally, to achieve closed-loop control, we
introduce an operator-encoded policy network to map the current state to
optimal control, which integrates physical priors and swiftly makes decisions
based on states returned by the environment. We conduct a systematic evaluation
of MB-CC on two notable SPDEs, showcasing its effectiveness and efficiency. The
ablation studies show its ability to handle stochasticity more effectively.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 05:23:43 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 12:58:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Hu",
"Peiyan",
""
],
[
"Feng",
"Haodong",
""
],
[
"Wang",
"Yue",
""
],
[
"Ma",
"Zhiming",
""
]
] |
2505.05676 | Roc\'io D\'iaz Mart\'in | Akram Aldroubi, Roc\'io D\'iaz Mart\'in, Ivan Medri, Kristofor E. Pas,
Gustavo K. Rohde, Abu Hasnat Mohammad Rubaiyat | An Efficient Transport-Based Dissimilarity Measure for Time Series
Classification under Warping Distortions | null | null | null | null | cs.IT math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time Series Classification (TSC) is an important problem with numerous
applications in science and technology. Dissimilarity-based approaches, such as
Dynamic Time Warping (DTW), are classical methods for distinguishing time
series when time deformations are confounding information. In this paper,
starting from a deformation-based model for signal classes we define a problem
statement for time series classification problem. We show that, under
theoretically ideal conditions, a continuous version of classic 1NN-DTW method
can solve the stated problem, even when only one training sample is available.
In addition, we propose an alternative dissimilarity measure based on Optimal
Transport and show that it can also solve the aforementioned problem statement
at a significantly reduced computational cost. Finally, we demonstrate the
application of the newly proposed approach in simulated and real time series
classification data, showing the efficacy of the method.
| [
{
"version": "v1",
"created": "Thu, 8 May 2025 22:26:37 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 19:26:04 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Aldroubi",
"Akram",
""
],
[
"Martín",
"Rocío Díaz",
""
],
[
"Medri",
"Ivan",
""
],
[
"Pas",
"Kristofor E.",
""
],
[
"Rohde",
"Gustavo K.",
""
],
[
"Rubaiyat",
"Abu Hasnat Mohammad",
""
]
] |
2505.05901 | Hanzhe Liang | Hanzhe Liang, Aoran Wang, Jie Zhou, Xin Jin, Can Gao, Jinbao Wang | Examining the Source of Defects from a Mechanical Perspective for 3D
Anomaly Detection | 26 pages | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we explore a novel approach to 3D anomaly detection (AD) that
goes beyond merely identifying anomalies based on structural characteristics.
Our primary perspective is that most anomalies arise from unpredictable
defective forces originating from both internal and external sources. To
address these anomalies, we seek out opposing forces that can help correct
them. Therefore, we introduce the Mechanics Complementary Model-based Framework
for the 3D-AD task (MC4AD), which generates internal and external corrective
forces for each point. We first propose a Diverse Anomaly-Generation (DA-Gen)
module designed to simulate various types of anomalies. Next, we present the
Corrective Force Prediction Network (CFP-Net), which uses complementary
representations for point-level analysis to simulate the different
contributions from internal and external corrective forces. To ensure the
corrective forces are constrained effectively, we have developed a combined
loss function that includes a new symmetric loss and an overall loss. Notably,
we implement a Hierarchical Quality Control (HQC) strategy based on a three-way
decision process and contribute a dataset titled Anomaly-IntraVariance, which
incorporates intraclass variance to evaluate our model. As a result, the
proposed MC4AD has been proven effective through theory and experimentation.
The experimental results demonstrate that our approach yields nine
state-of-the-art performances, achieving optimal results with minimal
parameters and the fastest inference speed across five existing datasets, in
addition to the proposed Anomaly-IntraVariance dataset. The source is available
at https://github.com/hzzzzzhappy/MC4AD
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 09:09:08 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:46:43 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Liang",
"Hanzhe",
""
],
[
"Wang",
"Aoran",
""
],
[
"Zhou",
"Jie",
""
],
[
"Jin",
"Xin",
""
],
[
"Gao",
"Can",
""
],
[
"Wang",
"Jinbao",
""
]
] |
2505.05922 | Haoqi Wu | Haoqi Wu, Wei Dai, Li Wang, Qiang Yan | Cape: Context-Aware Prompt Perturbation Mechanism with Differential
Privacy | to be published in ICML 2025 | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLMs) have gained significant popularity due to their
remarkable capabilities in text understanding and generation. However, despite
their widespread deployment in inference services such as ChatGPT, concerns
about the potential leakage of sensitive user data have arisen. Existing
solutions primarily rely on privacy-enhancing technologies to mitigate such
risks, facing the trade-off among efficiency, privacy, and utility. To narrow
this gap, we propose Cape, a context-aware prompt perturbation mechanism based
on differential privacy, to enable efficient inference with an improved
privacy-utility trade-off. Concretely, we introduce a hybrid utility function
that better captures the token similarity. Additionally, we propose a
bucketized sampling mechanism to handle large sampling space, which might lead
to long-tail phenomenons. Extensive experiments across multiple datasets, along
with ablation studies, demonstrate that Cape achieves a better privacy-utility
trade-off compared to prior state-of-the-art works.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 09:54:07 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 09:31:11 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wu",
"Haoqi",
""
],
[
"Dai",
"Wei",
""
],
[
"Wang",
"Li",
""
],
[
"Yan",
"Qiang",
""
]
] |
2505.05926 | Milad Khademi Nori | Milad Khademi Nori, Il-Min Kim, Guanghui Wang | Autoencoder-Based Hybrid Replay for Class-Incremental Learning | Accepted ICML 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In class-incremental learning (CIL), effective incremental learning
strategies are essential to mitigate task confusion and catastrophic
forgetting, especially as the number of tasks $t$ increases. Current exemplar
replay strategies impose $\mathcal{O}(t)$ memory/compute complexities. We
propose an autoencoder-based hybrid replay (AHR) strategy that leverages our
new hybrid autoencoder (HAE) to function as a compressor to alleviate the
requirement for large memory, achieving $\mathcal{O}(0.1 t)$ at the worst case
with the computing complexity of $\mathcal{O}(t)$ while accomplishing
state-of-the-art performance. The decoder later recovers the exemplar data
stored in the latent space, rather than in raw format. Additionally, HAE is
designed for both discriminative and generative modeling, enabling
classification and replay capabilities, respectively. HAE adopts the charged
particle system energy minimization equations and repulsive force algorithm for
the incremental embedding and distribution of new class centroids in its latent
space. Our results demonstrate that AHR consistently outperforms recent
baselines across multiple benchmarks while operating with the same
memory/compute budgets. The source code is included in the supplementary
material and will be open-sourced upon publication.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 09:59:12 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 02:46:39 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Nori",
"Milad Khademi",
""
],
[
"Kim",
"Il-Min",
""
],
[
"Wang",
"Guanghui",
""
]
] |
2505.05957 | Peter R\"oseler | Peter R\"oseler, Oliver Schaudt, Helmut Berg, Christian Bauckhage,
Matthias Koch | Efficient Quantum Convolutional Neural Networks for Image
Classification: Overcoming Hardware Constraints | null | null | null | null | quant-ph cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | While classical convolutional neural networks (CNNs) have revolutionized
image classification, the emergence of quantum computing presents new
opportunities for enhancing neural network architectures. Quantum CNNs (QCNNs)
leverage quantum mechanical properties and hold potential to outperform
classical approaches. However, their implementation on current noisy
intermediate-scale quantum (NISQ) devices remains challenging due to hardware
limitations. In our research, we address this challenge by introducing an
encoding scheme that significantly reduces the input dimensionality. We
demonstrate that a primitive QCNN architecture with 49 qubits is sufficient to
directly process $28\times 28$ pixel MNIST images, eliminating the need for
classical dimensionality reduction pre-processing. Additionally, we propose an
automated framework based on expressibility, entanglement, and complexity
characteristics to identify the building blocks of QCNNs, parameterized quantum
circuits (PQCs). Our approach demonstrates advantages in accuracy and
convergence speed with a similar parameter count compared to both hybrid QCNNs
and classical CNNs. We validated our experiments on IBM's Heron r2 quantum
processor, achieving $96.08\%$ classification accuracy, surpassing the
$71.74\%$ benchmark of traditional approaches under identical training
conditions. These results represent one of the first implementations of image
classifications on real quantum hardware and validate the potential of quantum
computing in this area.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 11:09:52 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:43:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Röseler",
"Peter",
""
],
[
"Schaudt",
"Oliver",
""
],
[
"Berg",
"Helmut",
""
],
[
"Bauckhage",
"Christian",
""
],
[
"Koch",
"Matthias",
""
]
] |
2505.06046 | Joshua Harris | Joshua Harris, Fan Grayson, Felix Feldman, Timothy Laurence, Toby
Nonnenmacher, Oliver Higgins, Leo Loman, Selina Patel, Thomas Finnie, Samuel
Collins and Michael Borowitz | Healthy LLMs? Benchmarking LLM Knowledge of UK Government Public Health
Information | 24 pages, 10 pages main text | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | As Large Language Models (LLMs) become widely accessible, a detailed
understanding of their knowledge within specific domains becomes necessary for
successful real world use. This is particularly critical in public health,
where failure to retrieve relevant, accurate, and current information could
significantly impact UK residents. However, currently little is known about LLM
knowledge of UK Government public health information. To address this issue,
this paper introduces a new benchmark, PubHealthBench, with over 8000 questions
for evaluating LLMs' Multiple Choice Question Answering (MCQA) and free form
responses to public health queries. To create PubHealthBench we extract free
text from 687 current UK government guidance documents and implement an
automated pipeline for generating MCQA samples. Assessing 24 LLMs on
PubHealthBench we find the latest private LLMs (GPT-4.5, GPT-4.1 and o1) have a
high degree of knowledge, achieving >90% accuracy in the MCQA setup, and
outperform humans with cursory search engine use. However, in the free form
setup we see lower performance with no model scoring >75%. Importantly we find
in both setups LLMs have higher accuracy on guidance intended for the general
public. Therefore, there are promising signs that state of the art (SOTA) LLMs
are an increasingly accurate source of public health information, but
additional safeguards or tools may still be needed when providing free form
responses on public health topics.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 13:42:59 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:14:47 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Harris",
"Joshua",
""
],
[
"Grayson",
"Fan",
""
],
[
"Feldman",
"Felix",
""
],
[
"Laurence",
"Timothy",
""
],
[
"Nonnenmacher",
"Toby",
""
],
[
"Higgins",
"Oliver",
""
],
[
"Loman",
"Leo",
""
],
[
"Patel",
"Selina",
""
],
[
"Finnie",
"Thomas",
""
],
[
"Collins",
"Samuel",
""
],
[
"Borowitz",
"Michael",
""
]
] |
2505.06085 | Hiari Pizzini Cavagna | Hiari Pizzini Cavagna, Daniele Cesarini, Andrea Bartolini | Assessing Tenstorrent's RISC-V MatMul Acceleration Capabilities | Accepted to the Computational Aspects of Deep Learning Workshop at
ISC High Performance 2025. To appear in the ISC High Performance 2025
Workshop Proceedings | null | null | null | cs.PF cs.AI cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The increasing demand for generative AI as Large Language Models (LLMs)
services has driven the need for specialized hardware architectures that
optimize computational efficiency and energy consumption. This paper evaluates
the performance of the Tenstorrent Grayskull e75 RISC-V accelerator for basic
linear algebra kernels at reduced numerical precision, a fundamental operation
in LLM computations. We present a detailed characterization of Grayskull's
execution model, gridsize, matrix dimensions, data formats, and numerical
precision impact computational efficiency. Furthermore, we compare Grayskull's
performance against state-of-the-art architectures with tensor acceleration,
including Intel Sapphire Rapids processors and two NVIDIA GPUs (V100 and A100).
Whilst NVIDIA GPUs dominate raw performance, Grayskull demonstrates a
competitive trade-off between power consumption and computational throughput,
reaching a peak of 1.55 TFLOPs/Watt with BF16.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 14:29:37 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 13:07:31 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Cavagna",
"Hiari Pizzini",
""
],
[
"Cesarini",
"Daniele",
""
],
[
"Bartolini",
"Andrea",
""
]
] |
2505.06111 | Qingwen Bu | Qingwen Bu, Yanting Yang, Jisong Cai, Shenyuan Gao, Guanghui Ren,
Maoqing Yao, Ping Luo, Hongyang Li | UniVLA: Learning to Act Anywhere with Task-centric Latent Actions | Accepted to RSS 2025. Code is available at
https://github.com/OpenDriveLab/UniVLA | null | null | null | cs.RO cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A generalist robot should perform effectively across various environments.
However, most existing approaches heavily rely on scaling action-annotated data
to enhance their capabilities. Consequently, they are often limited to single
physical specification and struggle to learn transferable knowledge across
different embodiments and environments. To confront these limitations, we
propose UniVLA, a new framework for learning cross-embodiment
vision-language-action (VLA) policies. Our key innovation is to derive
task-centric action representations from videos with a latent action model.
This enables us to exploit extensive data across a wide spectrum of embodiments
and perspectives. To mitigate the effect of task-irrelevant dynamics, we
incorporate language instructions and establish a latent action model within
the DINO feature space. Learned from internet-scale videos, the generalist
policy can be deployed to various robots through efficient latent action
decoding. We obtain state-of-the-art results across multiple manipulation and
navigation benchmarks, as well as real-robot deployments. UniVLA achieves
superior performance over OpenVLA with less than 1/20 of pretraining compute
and 1/10 of downstream data. Continuous performance improvements are observed
as heterogeneous data, even including human videos, are incorporated into the
training pipeline. The results underscore UniVLA's potential to facilitate
scalable and efficient robot policy learning.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 15:11:13 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:31:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Bu",
"Qingwen",
""
],
[
"Yang",
"Yanting",
""
],
[
"Cai",
"Jisong",
""
],
[
"Gao",
"Shenyuan",
""
],
[
"Ren",
"Guanghui",
""
],
[
"Yao",
"Maoqing",
""
],
[
"Luo",
"Ping",
""
],
[
"Li",
"Hongyang",
""
]
] |
2505.06221 | Niana Mohammed | Niana N. Mohammed, Jessie C. Runnoe, Michael Eracleous, Tamara
Bogdanovi\'c, Daniel Stern, Joseph Simon, Maria Charisi, T. Joseph W. Lazio,
Kaitlyn Szekerczes, Steinn Sigurdsson, Collin Dabbieri | Testing the Hypothesis that the Quasar J0950+5128 Harbors a Supermassive
Black Hole Binary | null | null | null | null | astro-ph.GA astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Spectroscopic observations of the quasar J0950+5128 spanning 22 years show
that its broad H$\beta$ emission line displays monotonic radial velocity
variations consistent with those expected from a supermassive binary. At the
same time, the profile of the broad H$\beta$ line is changing, necessitating
careful measurements. We present robust measurements of the broad H$\beta$
velocity shifts. We fit the spectra with a multi-component parametric model to
isolate the broad H$\beta$ by removing the stellar and non-stellar continua,
the Fe ii line blends, and various narrow and broad emission lines in its
vicinity. We measure the H$\beta$ velocity shifts via cross-correlation, both
before and after decomposing the spectra, and we demonstrate that they can be
reliably measured in spite of variations in the line profiles. We incorporate
radial velocity "jitter" into the radial velocity curve to emulate typical
quasar variability, and fit the curve with a Keplerian orbit model. The fit
yields a period of 33 years and an eccentricity of 0.65, with lower limits on
the semi-major axis and black hole mass of $10^{-2}$ pc and $10^7\ M_\odot$,
respectively. Thus, this quasar is a compelling supermassive binary candidate.
Continued monitoring is essential to further evaluate its credentials.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 17:56:50 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Mohammed",
"Niana N.",
""
],
[
"Runnoe",
"Jessie C.",
""
],
[
"Eracleous",
"Michael",
""
],
[
"Bogdanović",
"Tamara",
""
],
[
"Stern",
"Daniel",
""
],
[
"Simon",
"Joseph",
""
],
[
"Charisi",
"Maria",
""
],
[
"Lazio",
"T. Joseph W.",
""
],
[
"Szekerczes",
"Kaitlyn",
""
],
[
"Sigurdsson",
"Steinn",
""
],
[
"Dabbieri",
"Collin",
""
]
] |
2505.06341 | Pavel Fileviez Perez | Jon Butterworth, Hridoy Debnath, Joseph Egan, Pavel Fileviez Perez | Local Baryon Number at the LHC | null | null | null | null | hep-ph hep-ex hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The minimal theory in which baryon number is spontaneously broken at the low
scale predicts new fermions, one of which is a dark matter candidate, from
gauge anomaly cancellation. We discuss the production mechanisms and decays of
these new fermions, which include channels with multi-leptons, and channels
with long-lived charged fermions that can give rise to exotic signatures with
'kinked' tracks at the Large Hadron Collider. We evaluate the contraints on the
theory from current LHC searches and measurements, and briefly comment on the
excess in top pair production at threshold recently reported by CMS. We also
discuss predictions for the $h \to \gamma Z_B$ decay, where $h$ is the SM-like
Higgs and $Z_B$ is the new gauge boson associated with baryon number.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 18:00:01 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Butterworth",
"Jon",
""
],
[
"Debnath",
"Hridoy",
""
],
[
"Egan",
"Joseph",
""
],
[
"Perez",
"Pavel Fileviez",
""
]
] |
2505.06466 | Giancarlo Franzese | Alberto Martinez-Serra, Gionni Marchetti, Francesco D'Amico, Ivana
Fenoglio, Barbara Rossi, Marco P. Monopoli, Giancarlo Franzese | Machine Learning Tool to Analyse Spectroscopic Changes in
High-Dimensional Data | 30 pages, 9 figures, added references, revised fig.8 and related
discussion, results unchanged | null | null | null | cond-mat.stat-mech q-bio.QM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | When nanoparticles (NPs) are introduced into a biological solution, layers of
biomolecules form on their surface, creating a corona. Understanding how the
structure of the protein evolves into the corona is essential for evaluating
the safety and toxicity of nanotechnology. However, the influence of NP
properties on protein conformation is not well understood. In this study, we
propose a new method that addresses this issue by analyzing multi-component
spectral data using Machine Learning (ML). We apply the method to fibrinogen, a
crucial protein in human blood plasma, at physiological concentrations while
interacting with hydrophobic carbon or hydrophilic silicon dioxide NPs,
revealing striking differences in the temperature dependence of the protein
structure between the two cases. Our unsupervised ML method a) does not suffer
from the challenges associated with the curse of dimensionality, and b)
simultaneously handles spectral data from various sources. The method offers a
quantitative analysis of protein structural changes upon adsorption and
enhances the understanding of the correlation between protein structure and NP
interactions, which could support the development of nanomedical tools to treat
various conditions.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 23:22:24 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 23:04:00 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Martinez-Serra",
"Alberto",
""
],
[
"Marchetti",
"Gionni",
""
],
[
"D'Amico",
"Francesco",
""
],
[
"Fenoglio",
"Ivana",
""
],
[
"Rossi",
"Barbara",
""
],
[
"Monopoli",
"Marco P.",
""
],
[
"Franzese",
"Giancarlo",
""
]
] |
2505.06470 | Calvin Deutschbein | Calvin Deutschbein, Jimmy Ostler, Hriday Raj | "vcd2df" -- Leveraging Data Science Insights for Hardware Security
Research | 6 pages, no figures, under submission at ACDSA 2025. Added co-author
Hriday Raj during v2 as Hriday joined us for some Spark characterization as a
domain specialist between initial drafts and reaching camera-readiness | null | null | null | cs.CR cs.AR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this work, we hope to expand the universe of security practitioners of
open-source hardware by creating a bridge from hardware design languages (HDLs)
to data science languages like Python and R through libraries that convert VCD
(value change dump) files into data frames, the expected input type of the
modern data science tools. We show how insights can be derived in high-level
languages from register transfer level (RTL) trace data. Additional, we show a
promising future direction in hardware security research leveraging the
parallelism of the Spark DataFrame.
| [
{
"version": "v1",
"created": "Fri, 9 May 2025 23:53:34 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:50:14 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Deutschbein",
"Calvin",
""
],
[
"Ostler",
"Jimmy",
""
],
[
"Raj",
"Hriday",
""
]
] |
2505.06512 | Hang Wang | Hang Wang, Zhi-Qi Cheng, Chenhao Lin, Chao Shen and Lei Zhang | HCMA: Hierarchical Cross-model Alignment for Grounded Text-to-Image
Generation | 10 pages, 4 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-to-image synthesis has progressed to the point where models can generate
visually compelling images from natural language prompts. Yet, existing methods
often fail to reconcile high-level semantic fidelity with explicit spatial
control, particularly in scenes involving multiple objects, nuanced relations,
or complex layouts. To bridge this gap, we propose a Hierarchical Cross-Modal
Alignment (HCMA) framework for grounded text-to-image generation. HCMA
integrates two alignment modules into each diffusion sampling step: a global
module that continuously aligns latent representations with textual
descriptions to ensure scene-level coherence, and a local module that employs
bounding-box layouts to anchor objects at specified locations, enabling
fine-grained spatial control. Extensive experiments on the MS-COCO 2014
validation set show that HCMA surpasses state-of-the-art baselines, achieving a
0.69 improvement in Frechet Inception Distance (FID) and a 0.0295 gain in CLIP
Score. These results demonstrate HCMA's effectiveness in faithfully capturing
intricate textual semantics while adhering to user-defined spatial constraints,
offering a robust solution for semantically grounded image generation. Our code
is available at https://github.com/hwang-cs-ime/HCMA.
| [
{
"version": "v1",
"created": "Sat, 10 May 2025 05:02:58 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 05:23:54 GMT"
},
{
"version": "v3",
"created": "Thu, 15 May 2025 01:04:26 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Wang",
"Hang",
""
],
[
"Cheng",
"Zhi-Qi",
""
],
[
"Lin",
"Chenhao",
""
],
[
"Shen",
"Chao",
""
],
[
"Zhang",
"Lei",
""
]
] |
2505.06711 | Junfan Xia | Junfan Xia and Bin Jiang | Efficient Parallelization of Message Passing Neural Networks | 34 pages, 8 figures | null | null | null | physics.chem-ph cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning potentials have achieved great success in accelerating
atomistic simulations. Many of them rely on local descriptors that readily
allow parallelization. More recent message passing neural network (MPNN) models
have demonstrated their superior accuracy and become increasingly popular.
However, parallelizing MPNN models for large-scale simulations across compute
nodes remains a challenge, as the previously argued poor scalability with the
number of MP layers and the necessity of data communication. Here, we propose
an efficient parallel algorithm for MPNN models, in which additional data
communication is minimized among local atoms only in each MP layer without
redundant computation, thus scaling linearly with the layer number. Integrated
with our recursively embedded atom neural network model, this algorithm
demonstrates excellent strong scaling and weak scaling behaviors in several
benchmark systems. This approach enables massive molecular dynamics simulations
on MPNN models for hundreds of millions of atoms as fast as on strictly local
models, vastly extending the applicability of the MPNN potential to an
unprecedented scale. This general parallelization framework can empower various
MPNN models to efficiently simulate very large and complex systems.
| [
{
"version": "v1",
"created": "Sat, 10 May 2025 17:30:22 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 02:22:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Xia",
"Junfan",
""
],
[
"Jiang",
"Bin",
""
]
] |
2505.06761 | Jan Dubi\'nski | Youcef Djenouri, Nassim Belmecheri, Tomasz Michalak, Jan Dubi\'nski,
Ahmed Nabil Belbachir, Anis Yazidi | Learning Graph Representation of Agent Diffusers | Accepted at AAMAS2025 International Conference on Autonomous Agents
and Multiagent Systems | null | null | null | cs.LG cs.MA | http://creativecommons.org/licenses/by/4.0/ | Diffusion-based generative models have significantly advanced text-to-image
synthesis, demonstrating impressive text comprehension and zero-shot
generalization. These models refine images from random noise based on textual
prompts, with initial reliance on text input shifting towards enhanced visual
fidelity over time. This transition suggests that static model parameters might
not optimally address the distinct phases of generation. We introduce LGR-AD
(Learning Graph Representation of Agent Diffusers), a novel multi-agent system
designed to improve adaptability in dynamic computer vision tasks. LGR-AD
models the generation process as a distributed system of interacting agents,
each representing an expert sub-model. These agents dynamically adapt to
varying conditions and collaborate through a graph neural network that encodes
their relationships and performance metrics. Our approach employs a
coordination mechanism based on top-$k$ maximum spanning trees, optimizing the
generation process. Each agent's decision-making is guided by a meta-model that
minimizes a novel loss function, balancing accuracy and diversity. Theoretical
analysis and extensive empirical evaluations show that LGR-AD outperforms
traditional diffusion models across various benchmarks, highlighting its
potential for scalable and flexible solutions in complex image generation
tasks. Code is available at: https://github.com/YousIA/LGR_AD
| [
{
"version": "v1",
"created": "Sat, 10 May 2025 21:42:24 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:32:55 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Djenouri",
"Youcef",
""
],
[
"Belmecheri",
"Nassim",
""
],
[
"Michalak",
"Tomasz",
""
],
[
"Dubiński",
"Jan",
""
],
[
"Belbachir",
"Ahmed Nabil",
""
],
[
"Yazidi",
"Anis",
""
]
] |
2505.06945 | Maryam Farhadizadeh | Maryam Farhadizadeh, Maria Weymann, Michael Bla{\ss}, Johann Kraus,
Christopher Gundler, Sebastian Walter, Noah Hempen, Harald Binder, Nadine
Binder | A systematic review of challenges and proposed solutions in modeling
multimodal data | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Multimodal data modeling has emerged as a powerful approach in clinical
research, enabling the integration of diverse data types such as imaging,
genomics, wearable sensors, and electronic health records. Despite its
potential to improve diagnostic accuracy and support personalized care,
modeling such heterogeneous data presents significant technical challenges.
This systematic review synthesizes findings from 69 studies to identify common
obstacles, including missing modalities, limited sample sizes, dimensionality
imbalance, interpretability issues, and finding the optimal fusion techniques.
We highlight recent methodological advances, such as transfer learning,
generative models, attention mechanisms, and neural architecture search that
offer promising solutions. By mapping current trends and innovations, this
review provides a comprehensive overview of the field and offers practical
insights to guide future research and development in multimodal modeling for
medical applications.
| [
{
"version": "v1",
"created": "Sun, 11 May 2025 11:23:51 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 11:38:48 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Farhadizadeh",
"Maryam",
""
],
[
"Weymann",
"Maria",
""
],
[
"Blaß",
"Michael",
""
],
[
"Kraus",
"Johann",
""
],
[
"Gundler",
"Christopher",
""
],
[
"Walter",
"Sebastian",
""
],
[
"Hempen",
"Noah",
""
],
[
"Binder",
"Harald",
""
],
[
"Binder",
"Nadine",
""
]
] |
2505.07096 | Prithwish Dan | Prithwish Dan and Kushal Kedia and Angela Chao and Edward Weiyi Duan
and Maximus Adrian Pace and Wei-Chiu Ma and Sanjiban Choudhury | X-Sim: Cross-Embodiment Learning via Real-to-Sim-to-Real | null | null | null | null | cs.RO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Human videos offer a scalable way to train robot manipulation policies, but
lack the action labels needed by standard imitation learning algorithms.
Existing cross-embodiment approaches try to map human motion to robot actions,
but often fail when the embodiments differ significantly. We propose X-Sim, a
real-to-sim-to-real framework that uses object motion as a dense and
transferable signal for learning robot policies. X-Sim starts by reconstructing
a photorealistic simulation from an RGBD human video and tracking object
trajectories to define object-centric rewards. These rewards are used to train
a reinforcement learning (RL) policy in simulation. The learned policy is then
distilled into an image-conditioned diffusion policy using synthetic rollouts
rendered with varied viewpoints and lighting. To transfer to the real world,
X-Sim introduces an online domain adaptation technique that aligns real and
simulated observations during deployment. Importantly, X-Sim does not require
any robot teleoperation data. We evaluate it across 5 manipulation tasks in 2
environments and show that it: (1) improves task progress by 30% on average
over hand-tracking and sim-to-real baselines, (2) matches behavior cloning with
10x less data collection time, and (3) generalizes to new camera viewpoints and
test-time changes. Code and videos are available at
https://portal-cornell.github.io/X-Sim/.
| [
{
"version": "v1",
"created": "Sun, 11 May 2025 19:04:00 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 00:43:19 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Dan",
"Prithwish",
""
],
[
"Kedia",
"Kushal",
""
],
[
"Chao",
"Angela",
""
],
[
"Duan",
"Edward Weiyi",
""
],
[
"Pace",
"Maximus Adrian",
""
],
[
"Ma",
"Wei-Chiu",
""
],
[
"Choudhury",
"Sanjiban",
""
]
] |
2505.07119 | Francesco Borsatti | Arianna Stropeni, Francesco Borsatti, Manuel Barusco, Davide Dalle
Pezze, Marco Fabris, Gian Antonio Susto | Towards Scalable IoT Deployment for Visual Anomaly Detection via
Efficient Compression | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Visual Anomaly Detection (VAD) is a key task in industrial settings, where
minimizing operational costs is essential. Deploying deep learning models
within Internet of Things (IoT) environments introduces specific challenges due
to limited computational power and bandwidth of edge devices. This study
investigates how to perform VAD effectively under such constraints by
leveraging compact, efficient processing strategies. We evaluate several data
compression techniques, examining the tradeoff between system latency and
detection accuracy. Experiments on the MVTec AD benchmark demonstrate that
significant compression can be achieved with minimal loss in anomaly detection
performance compared to uncompressed data. Current results show up to 80%
reduction in end-to-end inference time, including edge processing,
transmission, and server computation.
| [
{
"version": "v1",
"created": "Sun, 11 May 2025 21:05:33 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 15:05:10 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Stropeni",
"Arianna",
""
],
[
"Borsatti",
"Francesco",
""
],
[
"Barusco",
"Manuel",
""
],
[
"Pezze",
"Davide Dalle",
""
],
[
"Fabris",
"Marco",
""
],
[
"Susto",
"Gian Antonio",
""
]
] |
2505.07214 | Pascal Spiegler | Pascal Spiegler, Arash Harirpoush, Yiming Xiao | Towards user-centered interactive medical image segmentation in VR with
an assistive AI agent | null | null | null | null | cs.HC cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Crucial in disease analysis and surgical planning, manual segmentation of
volumetric medical scans (e.g. MRI, CT) is laborious, error-prone, and
challenging to master, while fully automatic algorithms can benefit from user
feedback. Therefore, with the complementary power of the latest radiological AI
foundation models and virtual reality (VR)'s intuitive data interaction, we
propose SAMIRA, a novel conversational AI agent that assists users with
localizing, segmenting, and visualizing 3D medical concepts in VR. Through
speech-based interaction, the agent helps users understand radiological
features, locate clinical targets, and generate segmentation masks that can be
refined with just a few point prompts. The system also supports true-to-scale
3D visualization of segmented pathology to enhance patient-specific anatomical
understanding. Furthermore, to determine the optimal interaction paradigm under
near-far attention-switching for refining segmentation masks in an immersive,
human-in-the-loop workflow, we compare VR controller pointing, head pointing,
and eye tracking as input modes. With a user study, evaluations demonstrated a
high usability score (SUS=90.0 $\pm$ 9.0), low overall task load, as well as
strong support for the proposed VR system's guidance, training potential, and
integration of AI in radiological segmentation tasks.
| [
{
"version": "v1",
"created": "Mon, 12 May 2025 03:47:05 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 05:47:33 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Spiegler",
"Pascal",
""
],
[
"Harirpoush",
"Arash",
""
],
[
"Xiao",
"Yiming",
""
]
] |
2505.07247 | Peichao Lai | Peichao Lai, Kexuan Zhang, Yi Lin, Linyihan Zhang, Feiyang Ye, Jinhao
Yan, Yanwei Xu, Conghui He, Yilei Wang, Wentao Zhang, Bin Cui | SAS-Bench: A Fine-Grained Benchmark for Evaluating Short Answer Scoring
with Large Language Models | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | Subjective Answer Grading (SAG) plays a crucial role in education,
standardized testing, and automated assessment systems, particularly for
evaluating short-form responses in Short Answer Scoring (SAS). However,
existing approaches often produce coarse-grained scores and lack detailed
reasoning. Although large language models (LLMs) have demonstrated potential as
zero-shot evaluators, they remain susceptible to bias, inconsistencies with
human judgment, and limited transparency in scoring decisions. To overcome
these limitations, we introduce SAS-Bench, a benchmark specifically designed
for LLM-based SAS tasks. SAS-Bench provides fine-grained, step-wise scoring,
expert-annotated error categories, and a diverse range of question types
derived from real-world subject-specific exams. This benchmark facilitates
detailed evaluation of model reasoning processes and explainability. We also
release an open-source dataset containing 1,030 questions and 4,109 student
responses, each annotated by domain experts. Furthermore, we conduct
comprehensive experiments with various LLMs, identifying major challenges in
scoring science-related questions and highlighting the effectiveness of
few-shot prompting in improving scoring accuracy. Our work offers valuable
insights into the development of more robust, fair, and educationally
meaningful LLM-based evaluation systems.
| [
{
"version": "v1",
"created": "Mon, 12 May 2025 05:43:21 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 11:01:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Lai",
"Peichao",
""
],
[
"Zhang",
"Kexuan",
""
],
[
"Lin",
"Yi",
""
],
[
"Zhang",
"Linyihan",
""
],
[
"Ye",
"Feiyang",
""
],
[
"Yan",
"Jinhao",
""
],
[
"Xu",
"Yanwei",
""
],
[
"He",
"Conghui",
""
],
[
"Wang",
"Yilei",
""
],
[
"Zhang",
"Wentao",
""
],
[
"Cui",
"Bin",
""
]
] |
2505.07344 | Jiacheng Jiang | Yuan Zhang, Jiacheng Jiang, Guoqing Ma, Zhiying Lu, Haoyang Huang,
Jianlong Yuan, Nan Duan | Generative Pre-trained Autoregressive Diffusion Transformer | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present GPDiT, a Generative Pre-trained Autoregressive
Diffusion Transformer that unifies the strengths of diffusion and
autoregressive modeling for long-range video synthesis, within a continuous
latent space. Instead of predicting discrete tokens, GPDiT autoregressively
predicts future latent frames using a diffusion loss, enabling natural modeling
of motion dynamics and semantic consistency across frames. This continuous
autoregressive framework not only enhances generation quality but also endows
the model with representation capabilities. Additionally, we introduce a
lightweight causal attention variant and a parameter-free rotation-based
time-conditioning mechanism, improving both the training and inference
efficiency. Extensive experiments demonstrate that GPDiT achieves strong
performance in video generation quality, video representation ability, and
few-shot learning tasks, highlighting its potential as an effective framework
for video modeling in continuous space.
| [
{
"version": "v1",
"created": "Mon, 12 May 2025 08:32:39 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 10:24:45 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Zhang",
"Yuan",
""
],
[
"Jiang",
"Jiacheng",
""
],
[
"Ma",
"Guoqing",
""
],
[
"Lu",
"Zhiying",
""
],
[
"Huang",
"Haoyang",
""
],
[
"Yuan",
"Jianlong",
""
],
[
"Duan",
"Nan",
""
]
] |
2505.07575 | Samuel Erickson Andersson | Samuel Erickson and Mikael Johansson | Personalized Federated Learning under Model Dissimilarity Constraints | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | One of the defining challenges in federated learning is that of statistical
heterogeneity among clients. We address this problem with KARULA, a regularized
strategy for personalized federated learning, which constrains the pairwise
model dissimilarities between clients based on the difference in their
distributions, as measured by a surrogate for the 1-Wasserstein distance
adapted for the federated setting. This allows the strategy to adapt to highly
complex interrelations between clients, that e.g., clustered approaches fail to
capture. We propose an inexact projected stochastic gradient algorithm to solve
the constrained problem that the strategy defines, and show theoretically that
it converges with smooth, possibly non-convex losses to a neighborhood of a
stationary point with rate O(1/K). We demonstrate the effectiveness of KARULA
on synthetic and real federated data sets.
| [
{
"version": "v1",
"created": "Mon, 12 May 2025 13:54:55 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 16:50:52 GMT"
}
] | 2025-05-16T00:00:00 | [
[
"Erickson",
"Samuel",
""
],
[
"Johansson",
"Mikael",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.