id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-21 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.07851 | H{\aa}kan Karlsson Faronius | H{\aa}kan Karlsson Faronius, Pedro Zuidberg Dos Martires | Independence Is Not an Issue in Neurosymbolic AI | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | A popular approach to neurosymbolic AI is to take the output of the last
layer of a neural network, e.g. a softmax activation, and pass it through a
sparse computation graph encoding certain logical constraints one wishes to
enforce. This induces a probability distribution over a set of random
variables, which happen to be conditionally independent of each other in many
commonly used neurosymbolic AI models. Such conditionally independent random
variables have been deemed harmful as their presence has been observed to
co-occur with a phenomenon dubbed deterministic bias, where systems learn to
deterministically prefer one of the valid solutions from the solution space
over the others. We provide evidence contesting this conclusion and show that
the phenomenon of deterministic bias is an artifact of improperly applying
neurosymbolic AI.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 15:28:36 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 10:29:19 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Faronius",
"Håkan Karlsson",
""
],
[
"Martires",
"Pedro Zuidberg Dos",
""
]
] |
2504.07873 | O.A. Veliev | O. A. Veliev | Spectral Periodic Differential Operators of Odd Order | null | null | null | null | math.SP | http://creativecommons.org/licenses/by/4.0/ | In this paper, we establish a condition on the coefficients of the
differential operators L generated by an ordinary differential expression of
odd order with periodic, complex-valued coefficients, under which the operator
L is a spectral operator.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 15:46:53 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 10:01:09 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Veliev",
"O. A.",
""
]
] |
2504.07884 | Ryoki Hamano | Ryoki Hamano, Masahiro Nomura, Shota Saito, Kento Uchida, Shinichi
Shirakawa | CatCMA with Margin: Stochastic Optimization for Continuous, Integer, and
Categorical Variables | Accepted at Genetic and Evolutionary Computation Conference (GECCO
'25). We have corrected typographical errors in equations (16), (39), (40),
(56), and (57) | null | 10.1145/3712256.3726471 | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study focuses on mixed-variable black-box optimization (MV-BBO),
addressing continuous, integer, and categorical variables. Many real-world
MV-BBO problems involve dependencies among these different types of variables,
requiring efficient methods to optimize them simultaneously. Recently,
stochastic optimization methods leveraging the mechanism of the covariance
matrix adaptation evolution strategy have shown promising results in
mixed-integer or mixed-category optimization. However, such methods cannot
handle the three types of variables simultaneously. In this study, we propose
CatCMA with Margin (CatCMAwM), a stochastic optimization method for MV-BBO that
jointly optimizes continuous, integer, and categorical variables. CatCMAwM is
developed by incorporating a novel integer handling into CatCMA, a
mixed-category black-box optimization method employing a joint distribution of
multivariate Gaussian and categorical distributions. The proposed integer
handling is carefully designed by reviewing existing integer handlings and
following the design principles of CatCMA. Even when applied to mixed-integer
problems, it stabilizes the marginal probability and improves the convergence
performance of continuous variables. Numerical experiments show that CatCMAwM
effectively handles the three types of variables, outperforming
state-of-the-art Bayesian optimization methods and baselines that simply
incorporate existing integer handlings into CatCMA.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 15:59:22 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 17:45:42 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 04:08:10 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Hamano",
"Ryoki",
""
],
[
"Nomura",
"Masahiro",
""
],
[
"Saito",
"Shota",
""
],
[
"Uchida",
"Kento",
""
],
[
"Shirakawa",
"Shinichi",
""
]
] |
2504.08012 | Yuseon Kim | Yuseon Kim and Kyongseok Park | SRVP: Strong Recollection Video Prediction Model Using Attention-Based
Spatiotemporal Correlation Fusion | This paper has been accepted to CVPR 2025 Precognition Workshop | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video prediction (VP) generates future frames by leveraging spatial
representations and temporal context from past frames. Traditional recurrent
neural network (RNN)-based models enhance memory cell structures to capture
spatiotemporal states over extended durations but suffer from gradual loss of
object appearance details. To address this issue, we propose the strong
recollection VP (SRVP) model, which integrates standard attention (SA) and
reinforced feature attention (RFA) modules. Both modules employ scaled
dot-product attention to extract temporal context and spatial correlations,
which are then fused to enhance spatiotemporal representations. Experiments on
three benchmark datasets demonstrate that SRVP mitigates image quality
degradation in RNN-based models while achieving predictive performance
comparable to RNN-free architectures.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 07:36:50 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:18:13 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kim",
"Yuseon",
""
],
[
"Park",
"Kyongseok",
""
]
] |
2504.08330 | Kurt Girstmair | Kurt Girstmair | Cotangent power sums and character coordinates | null | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | We show that certain sums studied in two recent papers are basically
character coordinates (as they are called in the literature). These sums
involve values of Dirichlet characters and powers of $\cot(\pi k/n)$, $1\le
k\le n-1$. We also show that a basic tool for the study of these sums was
already given in 1987, in the form of the character coordinates of so-called
cotangent numbers. By means of this tool, we obtain the results of the said
papers in a simple and lucid way. We also show that the coefficients of the
linear combinations used in the said papers are essentially the same.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 07:53:40 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 14:00:12 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Girstmair",
"Kurt",
""
]
] |
2504.08334 | Hongyi Guan | Hongyi Guan, Yichuan Gao, Chenlu Miao, Haoyang Wu, Hang Zhu, Mingfeng
Lin, Huayue Liang | Efficient Architecture for RISC-V Vector Memory Access | null | null | null | null | cs.AR cs.DC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Vector processors frequently suffer from inefficient memory accesses,
particularly for strided and segment patterns. While coalescing strided
accesses is a natural solution, effectively gathering or scattering elements at
fixed strides remains challenging. Naive approaches rely on high-overhead
crossbars that remap any byte between memory and registers, leading to physical
design issues. Segment operations require row-column transpositions, typically
handled using either element-level in-place transposition (degrading
performance) or large buffer-based bulk transposition (incurring high area
overhead). In this paper, we present EARTH, a novel vector memory access
architecture designed to overcome these challenges through shifting-based
optimizations. For strided accesses, EARTH integrates specialized shift
networks for gathering and scattering elements. After coalescing multiple
accesses within the same cache line, data is routed between memory and
registers through the shifting network with minimal overhead. For segment
operations, EARTH employs a shifted register bank enabling direct column-wise
access, eliminating dedicated segment buffers while providing high-performance,
in-place bulk transposition. Implemented on FPGA with Chisel HDL based on an
open-source RISC-V vector unit, EARTH enhances performance for strided memory
accesses, achieving 4x-8x speedups in benchmarks dominated by strided
operations. Compared to conventional designs, EARTH reduces hardware area by 9%
and power consumption by 41%, significantly advancing both performance and
efficiency of vector processors.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 07:59:06 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 14:22:19 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 05:57:08 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Guan",
"Hongyi",
""
],
[
"Gao",
"Yichuan",
""
],
[
"Miao",
"Chenlu",
""
],
[
"Wu",
"Haoyang",
""
],
[
"Zhu",
"Hang",
""
],
[
"Lin",
"Mingfeng",
""
],
[
"Liang",
"Huayue",
""
]
] |
2504.08525 | Ye Ye | Ye Ye | Task Memory Engine (TME): A Structured Memory Framework with Graph-Aware
Extensions for Multi-Step LLM Agent Tasks | 14 pages, 5 figures. Preprint prepared for future submission.
Includes implementation and token-efficiency analysis. Code at
https://github.com/biubiutomato/TME-Agent | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) are increasingly used as autonomous agents for
multi-step tasks. However, most existing frameworks fail to maintain a
structured understanding of the task state, often relying on linear prompt
concatenation or shallow memory buffers. This leads to brittle performance,
frequent hallucinations, and poor long-range coherence. In this work, we
propose the Task Memory Engine (TME), a lightweight and structured memory
module that tracks task execution using a hierarchical Task Memory Tree (TMT).
Each node in the tree corresponds to a task step, storing relevant input,
output, status, and sub-task relationships. We introduce a prompt synthesis
method that dynamically generates LLM prompts based on the active node path,
significantly improving execution consistency and contextual grounding. Through
case studies and comparative experiments on multi-step agent tasks, we
demonstrate that TME leads to better task completion accuracy and more
interpretable behavior with minimal implementation overhead. A reference
implementation of the core TME components is available at
https://github.com/biubiutomato/TME-Agent, including basic examples and
structured memory integration. While the current implementation uses a
tree-based structure, TME is designed to be graph-aware, supporting reusable
substeps, converging task paths, and shared dependencies. This lays the
groundwork for future DAG-based memory architectures.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 13:38:36 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:38:19 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 14:48:13 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Ye",
"Ye",
""
]
] |
2504.08629 | Pablo Villegas G\'ongora | Pablo Villegas | Strange Attractors in Complex Networks | 5 pages, 4 figures, and Supplemental Material. Accepted in Phys. Rev.
E | Phys. Rev. E 111, L042301 (2025) | 10.1103/PhysRevE.111.L042301 | null | cond-mat.stat-mech cond-mat.dis-nn physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Disorder and noise in physical systems often disrupt spatial and temporal
regularity, yet chaotic systems reveal how order can emerge from unpredictable
behavior. Complex networks, spatial analogs of chaos, exhibit disordered,
non-Euclidean architectures with hidden symmetries, hinting at spontaneous
order. Finding low-dimensional embeddings that reveal network patterns and link
them to dimensionality that governs universal behavior remains a fundamental
open challenge, as it needs to bridge the gap between microscopic disorder and
macroscopic regularities. Here, the minimal space revealing key network
properties is introduced, showing that non-integer dimensions produce
chaotic-like attractors.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:31:30 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Villegas",
"Pablo",
""
]
] |
2504.08662 | Tianzhe Zhou | Tianzhe Zhou and Chun Huang | Hidden Twin Star Solutions from an Agnostic Speed-of-Sound Model:
Confronting XTE J1814--338's Extreme Compactness | Submitted to Astrophysical Journal | null | null | null | astro-ph.HE gr-qc nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The twin star configuration, where two neutron stars share the same mass but
exhibit different radii, arises from a strong first-order phase transition
within the stellar interior. In widely used equation of state (EoS)
meta-models, such as the Polytrope (PP) and Speed-of-Sound (CS) models, this
first-order phase transition behavior can be naturally mimicked by tuning some
model parameters. Here, we systematically explore the under-explored parameter
space within one of a widely adopted CS model that leads to twin stars via a
strong first-order phase transition. Within this twin-star subspace, we perform
a comprehensive Bayesian analysis that integrates mass--radius (MR) constraints
from X-ray observations of rotation-powered millisecond pulsars. The resultant
twin star branch, situated within the 1--1.2 $M_{\odot}$ mass range and
approximately 7 km in radius, surprisingly coincides with the MR ranges
proposed for the recent anomaly in the Accreting Millisecond X-ray Pulsars XTE
J1814--338 (J1814), suggesting a hybrid twin star configuration. Moreover,
incorporating the J1814 observation as an additional constraint yields an
extreme phase transition pressure $P_{\text{trans}} = 108.9_{-4.85}^{+6.46}$
MeV/fm$^3$, a transition density of $\varepsilon_{\text{trans}}/\varepsilon_0 =
4.847_{-0.134}^{+0.271}$(where $\varepsilon_0$ is the nuclear saturation energy
density) and an energy density jump $\Delta \varepsilon =
558.7_{-278.7}^{+303.6}$ MeV/fm$^3$, corresponding to $\Delta
\varepsilon/\varepsilon_0 = 3.716_{-1.854}^{+2.020}$. Notably, to satisfy all
astrophysical constraints, the speed of sound inside of the hybrid twin star
core is driven toward the speed of light ($c_s^2/c^2 > 0.9$), indicating the
potential presence of strongly interacting, exotic matter in this core region.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 16:11:19 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 14:34:06 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhou",
"Tianzhe",
""
],
[
"Huang",
"Chun",
""
]
] |
2504.08713 | Sahil Sethi | Sahil Sethi, David Chen, Thomas Statchen, Michael C. Burkhart, Nipun
Bhandari, Bashar Ramadan, Brett Beaulieu-Jones | ProtoECGNet: Case-Based Interpretable Deep Learning for Multi-Label ECG
Classification with Contrastive Learning | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Deep learning-based electrocardiogram (ECG) classification has shown
impressive performance but clinical adoption has been slowed by the lack of
transparent and faithful explanations. Post hoc methods such as saliency maps
may fail to reflect a model's true decision process. Prototype-based reasoning
offers a more transparent alternative by grounding decisions in similarity to
learned representations of real ECG segments, enabling faithful, case-based
explanations. We introduce ProtoECGNet, a prototype-based deep learning model
for interpretable, multi-label ECG classification. ProtoECGNet employs a
structured, multi-branch architecture that reflects clinical interpretation
workflows: it integrates a 1D CNN with global prototypes for rhythm
classification, a 2D CNN with time-localized prototypes for morphology-based
reasoning, and a 2D CNN with global prototypes for diffuse abnormalities. Each
branch is trained with a prototype loss designed for multi-label learning,
combining clustering, separation, diversity, and a novel contrastive loss that
encourages appropriate separation between prototypes of unrelated classes while
allowing clustering for frequently co-occurring diagnoses. We evaluate
ProtoECGNet on all 71 diagnostic labels from the PTB-XL dataset, demonstrating
competitive performance relative to state-of-the-art black-box models while
providing structured, case-based explanations. To assess prototype quality, we
conduct a structured clinician review of the final model's projected
prototypes, finding that they are rated as representative and clear.
ProtoECGNet shows that prototype learning can be effectively scaled to complex,
multi-label time-series classification, offering a practical path toward
transparent and trustworthy deep learning models for clinical decision support.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 17:23:37 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 22:08:45 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Sethi",
"Sahil",
""
],
[
"Chen",
"David",
""
],
[
"Statchen",
"Thomas",
""
],
[
"Burkhart",
"Michael C.",
""
],
[
"Bhandari",
"Nipun",
""
],
[
"Ramadan",
"Bashar",
""
],
[
"Beaulieu-Jones",
"Brett",
""
]
] |
2504.08738 | Chengxuan Xia | Qianye Wu, Chengxuan Xia, Sixuan Tian | AI-Driven Sentiment Analytics: Unlocking Business Value in the
E-Commerce Landscape_v1 | 7 pages | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The rapid growth of e-commerce has led to an overwhelming volume of customer
feedback, from product reviews to service interactions. Extracting meaningful
insights from this data is crucial for businesses aiming to improve customer
satisfaction and optimize decision-making. This paper presents an AI-driven
sentiment analysis system designed specifically for e-commerce applications,
balancing accuracy with interpretability. Our approach integrates traditional
machine learning techniques with modern deep learning models, allowing for a
more nuanced understanding of customer sentiment while ensuring transparency in
decision-making. Experimental results show that our system outperforms standard
sentiment analysis methods, achieving an accuracy of 89.7% on diverse,
large-scale datasets. Beyond technical performance, real-world implementation
across multiple e-commerce platforms demonstrates tangible improvements in
customer engagement and operational efficiency. This study highlights both the
potential and the challenges of applying AI to sentiment analysis in a
commercial setting, offering insights into practical deployment strategies and
areas for future refinement.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 18:56:22 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 05:59:02 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wu",
"Qianye",
""
],
[
"Xia",
"Chengxuan",
""
],
[
"Tian",
"Sixuan",
""
]
] |
2504.08754 | TongYoung Kim | Tongyoung Kim, Jeongeun Lee, Soojin Yoon, Sunghwan Kim, Dongha Lee | Towards Personalized Conversational Sales Agents : Contextual User
Profiling for Strategic Action | null | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Conversational Recommender Systems (CRSs) aim to engage users in dialogue to
provide tailored recommendations. While traditional CRSs focus on eliciting
preferences and retrieving items, real-world e-commerce interactions involve
more complex decision-making, where users consider multiple factors beyond
simple attributes. To bridge this gap, we introduce Conversational Sales
(CSales), a novel task that unifies preference elicitation, recommendation, and
persuasion to better support user decision-making. For a realistic evaluation
of CSales, we present CSUser, an LLM-based user simulator constructed from
real-world data, modeling diverse user profiles with needs and personalities.
Additionally, we propose CSI, a conversational sales agent that proactively
infers contextual profiles through dialogue for personalized action planning.
Extensive experiments demonstrate that CSUser effectively replicates real-world
users and emphasize the importance of contextual profiling for strategic action
selection, ultimately driving successful purchases in e-commerce.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 15:49:52 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 08:21:39 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 07:59:48 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kim",
"Tongyoung",
""
],
[
"Lee",
"Jeongeun",
""
],
[
"Yoon",
"Soojin",
""
],
[
"Kim",
"Sunghwan",
""
],
[
"Lee",
"Dongha",
""
]
] |
2504.08785 | Gaetan De Rassenfosse | Gaetan de Rassenfosse | What is patent quality? | The Patentist Living Literature Review 4: 1-5 | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Patent quality is a critical yet complex concept of interest to inventors,
legal experts, and policymakers alike. This article demystifies patent quality
by examining its various dimensions and definitions. It begins by discussing
invention quality, emphasizing that the notion of 'technological superiority'
is often too abstract to capture practical impact. The work distinguishes
between incremental and radical innovations, highlighting that inventions can
be evaluated by improvements over previous technologies and their economic
potential. Four definitions of patent quality are explored: one based on
invention quality, another linking quality with economic value, a third
focusing on drafting clarity, and a fourth measuring compliance with legal
standards. The article explains that high-quality patents may not necessarily
protect valuable inventions and that weak patents can hold significant economic
value due to legal advantages. This accessible review provides a comprehensive
framework that benefits a broad audience interested in innovation policy and
intellectual property.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 21:11:04 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 22:53:30 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"de Rassenfosse",
"Gaetan",
""
]
] |
2504.09029 | William Cook | William Cook | A Hierarchical Decomposition of Kullback-Leibler Divergence:
Disentangling Marginal Mismatches from Statistical Dependencies | 17 pages, 3 figures | null | null | null | cs.IT math.IT math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | The Kullback-Leibler (KL) divergence is a foundational measure for comparing
probability distributions. Yet in multivariate settings, its single value often
obscures the underlying reasons for divergence, conflating mismatches in
individual variable distributions (marginals) with effects arising from
statistical dependencies. We derive an algebraically exact, additive, and
hierarchical decomposition of the KL divergence between a joint distribution
P(X1,...,Xn) and a standard product reference distribution Q(X1,...,Xn) =
product_i q(Xi), where variables are assumed independent and identically
distributed according to a common reference q. The total divergence precisely
splits into two primary components: (1) the summed divergence of each marginal
distribution Pi(Xi) from the common reference q(Xi), quantifying marginal
deviations; and (2) the total correlation (or multi-information), capturing the
total statistical dependency among variables. Leveraging Mobius inversion on
the subset lattice, we further decompose this total correlation term into a
hierarchy of signed contributions from distinct pairwise, triplet, and
higher-order statistical interactions, expressed using standard Shannon
information quantities. This decomposition provides an algebraically complete
and interpretable breakdown of KL divergence using established information
measures, requiring no approximations or model assumptions. Numerical
validation using hypergeometric sampling confirms the decomposition's exactness
to machine precision across diverse system configurations. This framework
enables a precise diagnosis of divergence origins--distinguishing marginal
effects from interaction effects--with potential applications across machine
learning, econometrics, and complex systems analysis.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 00:59:07 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 21:48:52 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Cook",
"William",
""
]
] |
2504.09109 | Ganxi Xu | Ganxi Xu, Jinyi Long, Jia Zhang | Probability Distribution Alignment and Low-Rank Weight Decomposition for
Source-Free Domain Adaptive Brain Decoding | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Brain decoding currently faces significant challenges in individual
differences, modality alignment, and high-dimensional embeddings. To address
individual differences, researchers often use source subject data, which leads
to issues such as privacy leakage and heavy data storage burdens. In modality
alignment, current works focus on aligning the softmax probability distribution
but neglect the alignment of marginal probability distributions, resulting in
modality misalignment. Additionally, images and text are aligned separately
with fMRI without considering the complex interplay between images and text,
leading to poor image reconstruction. Finally, the enormous dimensionality of
CLIP embeddings causes significant computational costs. Although the
dimensionality of CLIP embeddings can be reduced by ignoring the number of
patches obtained from images and the number of tokens acquired from text, this
comes at the cost of a significant drop in model performance, creating a
dilemma. To overcome these limitations, we propose a source-free domain
adaptation-based brain decoding framework.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 07:28:16 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 22:42:51 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Xu",
"Ganxi",
""
],
[
"Long",
"Jinyi",
""
],
[
"Zhang",
"Jia",
""
]
] |
2504.09158 | Oliver Buchmueller | C. F. A. Baynham, R. Hobson, O. Buchmueller, D. Evans, L. Hawkins, L.
Iannizzotto-Venezze, A. Josset, D. Lee, E. Pasatembou, B. E. Sauer, M. R.
Tarbutt, T. Walker, O. Ennis, U. Chauhan, A. Brzakalik, S. Dey, S. Hedges, B.
Stray, M. Langlois, K. Bongs, T. Hird, S. Lellouch, M. Holynski, B. Bostwick,
J. Chen, Z. Eyler, V. Gibson, T. L. Harte, C. C. Hsu, M. Karzazi, C. Lu, B.
Millward, J. Mitchell, N. Mouelle, B. Panchumarthi, J. Scheper, U. Schneider,
X. Su, Y. Tang, K. Tkalcec, M. Zeuner, S. Zhang, Y. Zhi, L. Badurina, A.
Beniwal, D. Blas, J. Carlton, J. Ellis, C. McCabe, G. Parish, D. Pathak
Govardhan, V. Vaskonen, T. Bowcock, K. Bridges, A. Carroll, J. Coleman, G.
Elertas, S. Hindley, C. Metelko, H. Throssell, J. N. Tinsley, E. Bentine, M.
Booth, D. Bortoletto, N. Callaghan, C. Foot, C. Gomez-Monedero, K. Hughes, A.
James, T. Leese, A. Lowe, J. March-Russell, J. Sander, J. Schelfhout, I.
Shipsey, D. Weatherill, D. Wood, M.G. Bason, K. Hussain, H. Labiad, A.L.
Marchant, T.C. Thornton, T. Valenzuela, S.N. Balashov, P. Majewski, D.
Newbold, M.G.D. van der Grinten, Z. Pan, Z. Tam, I. Wilmut, K. Clarke, A.
Vick | A Prototype Atom Interferometer to Detect Dark Matter and Gravitational
Waves | V2: minor fixes of author list issues and references | null | null | AION-REPORT/2025-02 | hep-ex astro-ph.IM gr-qc hep-ph physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The AION project has built a tabletop prototype of a single-photon
long-baseline atom interferometer using the 87Sr clock transition - a type of
quantum sensor designed to search for dark matter and gravitational waves. Our
prototype detector operates at the Standard Quantum Limit (SQL), producing a
signal with no unexpected noise beyond atom shot noise. Importantly, the
detector remains at the SQL even when additional laser phase noise is
introduced, emulating conditions in a long-baseline detector such as AION or
AEDGE where significant laser phase deviations will accumulate during long atom
interrogation times. Our results mark a key milestone in extending atom
interferometers to long baselines. Such interferometers can complement
laser-interferometer gravitational wave detectors by accessing the
mid-frequency gravitational wave band around 1 Hz, and can search for physics
beyond the Standard Model.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 09:45:16 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 14:23:14 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Baynham",
"C. F. A.",
""
],
[
"Hobson",
"R.",
""
],
[
"Buchmueller",
"O.",
""
],
[
"Evans",
"D.",
""
],
[
"Hawkins",
"L.",
""
],
[
"Iannizzotto-Venezze",
"L.",
""
],
[
"Josset",
"A.",
""
],
[
"Lee",
"D.",
""
],
[
"Pasatembou",
"E.",
""
],
[
"Sauer",
"B. E.",
""
],
[
"Tarbutt",
"M. R.",
""
],
[
"Walker",
"T.",
""
],
[
"Ennis",
"O.",
""
],
[
"Chauhan",
"U.",
""
],
[
"Brzakalik",
"A.",
""
],
[
"Dey",
"S.",
""
],
[
"Hedges",
"S.",
""
],
[
"Stray",
"B.",
""
],
[
"Langlois",
"M.",
""
],
[
"Bongs",
"K.",
""
],
[
"Hird",
"T.",
""
],
[
"Lellouch",
"S.",
""
],
[
"Holynski",
"M.",
""
],
[
"Bostwick",
"B.",
""
],
[
"Chen",
"J.",
""
],
[
"Eyler",
"Z.",
""
],
[
"Gibson",
"V.",
""
],
[
"Harte",
"T. L.",
""
],
[
"Hsu",
"C. C.",
""
],
[
"Karzazi",
"M.",
""
],
[
"Lu",
"C.",
""
],
[
"Millward",
"B.",
""
],
[
"Mitchell",
"J.",
""
],
[
"Mouelle",
"N.",
""
],
[
"Panchumarthi",
"B.",
""
],
[
"Scheper",
"J.",
""
],
[
"Schneider",
"U.",
""
],
[
"Su",
"X.",
""
],
[
"Tang",
"Y.",
""
],
[
"Tkalcec",
"K.",
""
],
[
"Zeuner",
"M.",
""
],
[
"Zhang",
"S.",
""
],
[
"Zhi",
"Y.",
""
],
[
"Badurina",
"L.",
""
],
[
"Beniwal",
"A.",
""
],
[
"Blas",
"D.",
""
],
[
"Carlton",
"J.",
""
],
[
"Ellis",
"J.",
""
],
[
"McCabe",
"C.",
""
],
[
"Parish",
"G.",
""
],
[
"Govardhan",
"D. Pathak",
""
],
[
"Vaskonen",
"V.",
""
],
[
"Bowcock",
"T.",
""
],
[
"Bridges",
"K.",
""
],
[
"Carroll",
"A.",
""
],
[
"Coleman",
"J.",
""
],
[
"Elertas",
"G.",
""
],
[
"Hindley",
"S.",
""
],
[
"Metelko",
"C.",
""
],
[
"Throssell",
"H.",
""
],
[
"Tinsley",
"J. N.",
""
],
[
"Bentine",
"E.",
""
],
[
"Booth",
"M.",
""
],
[
"Bortoletto",
"D.",
""
],
[
"Callaghan",
"N.",
""
],
[
"Foot",
"C.",
""
],
[
"Gomez-Monedero",
"C.",
""
],
[
"Hughes",
"K.",
""
],
[
"James",
"A.",
""
],
[
"Leese",
"T.",
""
],
[
"Lowe",
"A.",
""
],
[
"March-Russell",
"J.",
""
],
[
"Sander",
"J.",
""
],
[
"Schelfhout",
"J.",
""
],
[
"Shipsey",
"I.",
""
],
[
"Weatherill",
"D.",
""
],
[
"Wood",
"D.",
""
],
[
"Bason",
"M. G.",
""
],
[
"Hussain",
"K.",
""
],
[
"Labiad",
"H.",
""
],
[
"Marchant",
"A. L.",
""
],
[
"Thornton",
"T. C.",
""
],
[
"Valenzuela",
"T.",
""
],
[
"Balashov",
"S. N.",
""
],
[
"Majewski",
"P.",
""
],
[
"Newbold",
"D.",
""
],
[
"van der Grinten",
"M. G. D.",
""
],
[
"Pan",
"Z.",
""
],
[
"Tam",
"Z.",
""
],
[
"Wilmut",
"I.",
""
],
[
"Clarke",
"K.",
""
],
[
"Vick",
"A.",
""
]
] |
2504.09310 | Sangwoo Park | Osvaldo Simeone, Sangwoo Park, Matteo Zecchin | Conformal Calibration: Ensuring the Reliability of Black-Box AI in
Wireless Systems | submitted for a journal publication | null | null | null | cs.IT cs.LG eess.SP math.IT stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | AI is poised to revolutionize telecommunication networks by boosting
efficiency, automation, and decision-making. However, the black-box nature of
most AI models introduces substantial risk, possibly deterring adoption by
network operators. These risks are not addressed by the current prevailing
deployment strategy, which typically follows a best-effort train-and-deploy
paradigm. This paper reviews conformal calibration, a general framework that
moves beyond the state of the art by adopting computationally lightweight,
advanced statistical tools that offer formal reliability guarantees without
requiring further training or fine-tuning. Conformal calibration encompasses
pre-deployment calibration via uncertainty quantification or hyperparameter
selection; online monitoring to detect and mitigate failures in real time; and
counterfactual post-deployment performance analysis to address "what if"
diagnostic questions after deployment. By weaving conformal calibration into
the AI model lifecycle, network operators can establish confidence in black-box
AI models as a dependable enabling technology for wireless systems.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 19:05:00 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 11:25:54 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Simeone",
"Osvaldo",
""
],
[
"Park",
"Sangwoo",
""
],
[
"Zecchin",
"Matteo",
""
]
] |
2504.09347 | Xuran Meng | Xuran Meng and Yi Li | Inferring Outcome Means of Exponential Family Distributions Estimated by
Deep Neural Networks | 44 pages, 6 figures, 5 tables | null | null | null | stat.ML cs.LG math.ST stat.TH | http://creativecommons.org/licenses/by-sa/4.0/ | While deep neural networks (DNNs) are widely used for prediction, inference
on DNN-estimated subject-specific means for categorical or exponential family
outcomes remains underexplored. We address this by proposing a DNN estimator
under generalized nonparametric regression models (GNRMs) and developing a
rigorous inference framework. Unlike existing approaches that assume
independence between prediction errors and inputs to establish the error bound,
a condition often violated in GNRMs, we allow for dependence and our
theoretical analysis demonstrates the feasibility of drawing inference under
GNRMs. To implement inference, we consider an Ensemble Subsampling Method (ESM)
that leverages U-statistics and the Hoeffding decomposition to construct
reliable confidence intervals for DNN estimates. We show that, under GNRM
settings, ESM enables model-free variance estimation and accounts for
heterogeneity among individuals in the population. Through simulations under
nonparametric logistic, Poisson, and binomial regression models, we demonstrate
the effectiveness and efficiency of our method. We further apply the method to
the electronic Intensive Care Unit (eICU) dataset, a large-scale collection of
anonymized health records from ICU patients, to predict ICU readmission risk
and offer patient-centric insights for clinical decision-making.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 21:32:42 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 15:55:26 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Meng",
"Xuran",
""
],
[
"Li",
"Yi",
""
]
] |
2504.09394 | Joseph Liu | Joseph Liu, Yoonsoo Nam, Xinyue Cui, Swabha Swayamdipta | Evaluation Under Imperfect Benchmarks and Ratings: A Case Study in Text
Simplification | 9 pages, 6 figures | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Despite the successes of language models, their evaluation remains a daunting
challenge for new and existing tasks. We consider the task of text
simplification, commonly used to improve information accessibility, where
evaluation faces two major challenges. First, the data in existing benchmarks
might not reflect the capabilities of current language models on the task,
often containing disfluent, incoherent, or simplistic examples. Second,
existing human ratings associated with the benchmarks often contain a high
degree of disagreement, resulting in inconsistent ratings; nevertheless,
existing metrics still have to show higher correlations with these imperfect
ratings. As a result, evaluation for the task is not reliable and does not
reflect expected trends (e.g., more powerful models being assigned higher
scores). We address these challenges for the task of text simplification
through three contributions. First, we introduce SynthSimpliEval, a synthetic
benchmark for text simplification featuring simplified sentences generated by
models of varying sizes. Through a pilot study, we show that human ratings on
our benchmark exhibit high inter-annotator agreement and reflect the expected
trend: larger models produce higher-quality simplifications. Second, we show
that auto-evaluation with a panel of LLM judges (LLMs-as-a-jury) often suffices
to obtain consistent ratings for the evaluation of text simplification. Third,
we demonstrate that existing learnable metrics for text simplification benefit
from training on our LLMs-as-a-jury-rated synthetic data, closing the gap with
pure LLMs-as-a-jury for evaluation. Overall, through our case study on text
simplification, we show that a reliable evaluation requires higher quality test
data, which could be obtained through synthetic data and LLMs-as-a-jury
ratings.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 01:36:47 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 23:18:26 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Liu",
"Joseph",
""
],
[
"Nam",
"Yoonsoo",
""
],
[
"Cui",
"Xinyue",
""
],
[
"Swayamdipta",
"Swabha",
""
]
] |
2504.09446 | Zhu Yimin | Lincoln Linlin Xu, Yimin Zhu, Zack Dewis, Zhengsen Xu, Motasem
Alkayid, Mabel Heffring, Saeid Taleghanidoozdoozan | Sparse Deformable Mamba for Hyperspectral Image Classification | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Although Mamba models significantly improve hyperspectral image (HSI)
classification, one critical challenge is the difficulty in building the
sequence of Mamba tokens efficiently. This paper presents a Sparse Deformable
Mamba (SDMamba) approach for enhanced HSI classification, with the following
contributions. First, to enhance Mamba sequence, an efficient Sparse Deformable
Sequencing (SDS) approach is designed to adaptively learn the ''optimal"
sequence, leading to sparse and deformable Mamba sequence with increased detail
preservation and decreased computations. Second, to boost spatial-spectral
feature learning, based on SDS, a Sparse Deformable Spatial Mamba Module
(SDSpaM) and a Sparse Deformable Spectral Mamba Module (SDSpeM) are designed
for tailored modeling of the spatial information spectral information. Last, to
improve the fusion of SDSpaM and SDSpeM, an attention based feature fusion
approach is designed to integrate the outputs of the SDSpaM and SDSpeM. The
proposed method is tested on several benchmark datasets with many
state-of-the-art approaches, demonstrating that the proposed approach can
achieve higher accuracy with less computation, and better detail small-class
preservation capability.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 06:08:19 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 22:17:09 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Xu",
"Lincoln Linlin",
""
],
[
"Zhu",
"Yimin",
""
],
[
"Dewis",
"Zack",
""
],
[
"Xu",
"Zhengsen",
""
],
[
"Alkayid",
"Motasem",
""
],
[
"Heffring",
"Mabel",
""
],
[
"Taleghanidoozdoozan",
"Saeid",
""
]
] |
2504.09530 | Shuchao Duan | Shuchao Duan, Amirhossein Dadashzadeh, Alan Whone and Majid Mirmehdi | Trajectory-guided Motion Perception for Facial Expression Quality
Assessment in Neurological Disorders | Accepted to IEEE FG 2025 (preprint) | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Automated facial expression quality assessment (FEQA) in neurological
disorders is critical for enhancing diagnostic accuracy and improving patient
care, yet effectively capturing the subtle motions and nuances of facial muscle
movements remains a challenge. We propose to analyse facial landmark
trajectories, a compact yet informative representation, that encodes these
subtle motions from a high-level structural perspective. Hence, we introduce
Trajectory-guided Motion Perception Transformer (TraMP-Former), a novel FEQA
framework that fuses landmark trajectory features for fine-grained motion
capture with visual semantic cues from RGB frames, ultimately regressing the
combined features into a quality score. Extensive experiments demonstrate that
TraMP-Former achieves new state-of-the-art performance on benchmark datasets
with neurological disorders, including PFED5 (up by 6.51%) and an augmented
Toronto NeuroFace (up by 7.62%). Our ablation studies further validate the
efficiency and effectiveness of landmark trajectories in FEQA. Our code is
available at https://github.com/shuchaoduan/TraMP-Former.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 11:30:57 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:59:54 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Duan",
"Shuchao",
""
],
[
"Dadashzadeh",
"Amirhossein",
""
],
[
"Whone",
"Alan",
""
],
[
"Mirmehdi",
"Majid",
""
]
] |
2504.09555 | Jinhao Li | Jinhao Li, Zijian Chen, Runze Jiang, Tingzhu Chen, Changbo Wang,
Guangtao Zhai | Mitigating Long-tail Distribution in Oracle Bone Inscriptions: Dataset,
Model, and Benchmark | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The oracle bone inscription (OBI) recognition plays a significant role in
understanding the history and culture of ancient China. However, the existing
OBI datasets suffer from a long-tail distribution problem, leading to biased
performance of OBI recognition models across majority and minority classes.
With recent advancements in generative models, OBI synthesis-based data
augmentation has become a promising avenue to expand the sample size of
minority classes. Unfortunately, current OBI datasets lack large-scale
structure-aligned image pairs for generative model training. To address these
problems, we first present the Oracle-P15K, a structure-aligned OBI dataset for
OBI generation and denoising, consisting of 14,542 images infused with domain
knowledge from OBI experts. Second, we propose a diffusion model-based pseudo
OBI generator, called OBIDiff, to achieve realistic and controllable OBI
generation. Given a clean glyph image and a target rubbing-style image, it can
effectively transfer the noise style of the original rubbing to the glyph
image. Extensive experiments on OBI downstream tasks and user preference
studies show the effectiveness of the proposed Oracle-P15K dataset and
demonstrate that OBIDiff can accurately preserve inherent glyph structures
while transferring authentic rubbing styles effectively.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 13:03:25 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 09:29:01 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Li",
"Jinhao",
""
],
[
"Chen",
"Zijian",
""
],
[
"Jiang",
"Runze",
""
],
[
"Chen",
"Tingzhu",
""
],
[
"Wang",
"Changbo",
""
],
[
"Zhai",
"Guangtao",
""
]
] |
2504.09566 | Jiaquan Zhang | Chenghao Li, Chaoning Zhang, Yi Lu, Jiaquan Zhang, Qigan Sun, Xudong
Wang, Jiwei Wei, Guoqing Wang, Yang Yang, Heng Tao Shen | Syzygy of Thoughts: Improving LLM CoT with the Minimal Free Resolution | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Chain-of-Thought (CoT) prompting enhances the reasoning of large language
models (LLMs) by decomposing problems into sequential steps, mimicking human
logic and reducing errors. However, complex tasks with vast solution spaces and
vague constraints often exceed the capacity of a single reasoning chain.
Inspired by Minimal Free Resolution (MFR) in commutative algebra and algebraic
geometry, we propose Syzygy of Thoughts (SoT)-a novel framework that extends
CoT by introducing auxiliary, interrelated reasoning paths. SoT captures deeper
logical dependencies, enabling more robust and structured problem-solving. MFR
decomposes a module into a sequence of free modules with minimal rank,
providing a structured analytical approach to complex systems. This method
introduces the concepts of "Module", "Betti numbers","Freeness", "Mapping",
"Exactness" and "Minimality", enabling the systematic decomposition of the
original complex problem into logically complete minimal subproblems while
preserving key problem features and reducing reasoning length. We tested SoT
across diverse datasets (e.g., GSM8K, MATH) and models (e.g., GPT-4o-mini,
Qwen2.5), achieving inference accuracy that matches or surpasses mainstream
CoTs standards. Additionally, by aligning the sampling process with algebraic
constraints, our approach enhances the scalability of inference time in LLMs,
ensuring both transparent reasoning and high performance. Our code will be
publicly available at https://github.com/dlMARiA/Syzygy-of-thoughts.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 13:35:41 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 05:02:01 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Li",
"Chenghao",
""
],
[
"Zhang",
"Chaoning",
""
],
[
"Lu",
"Yi",
""
],
[
"Zhang",
"Jiaquan",
""
],
[
"Sun",
"Qigan",
""
],
[
"Wang",
"Xudong",
""
],
[
"Wei",
"Jiwei",
""
],
[
"Wang",
"Guoqing",
""
],
[
"Yang",
"Yang",
""
],
[
"Shen",
"Heng Tao",
""
]
] |
2504.09634 | Saeid Jamshidi | Saeid Jamshidi, Kawser Wazed Nafi, Amin Nikanjam, Foutse Khomh | Evaluating Machine Learning-Driven Intrusion Detection Systems in IoT:
Performance and Energy Consumption | Computers & Industrial Engineering (2025) | null | 10.1016/j.cie.2025.111103 | null | cs.NI | http://creativecommons.org/licenses/by/4.0/ | In the evolving landscape of the Internet of Things (IoT), Machine Learning
(ML)-based Intrusion Detection Systems (IDS) represent a significant
advancement, especially when integrated with Software-Defined Networking (SDN).
These systems play a critical role in enhancing security infrastructure within
resource-constrained IoT systems. Despite their growing adoption, limited
research has explored the impact of ML-based IDS on key performance metrics,
such as CPU load, CPU usage, and energy consumption, particularly under
real-time cyber threats. This study bridges that gap through an empirical
evaluation of cutting-edge ML-based IDSs deployed at the edge of IoT networks
under both benign and attack scenarios. Additionally, we investigate how SDN's
centralized control and dynamic resource management influence IDS performance.
Our experimental framework compares traditional ML-based IDS with deep learning
(DL)-based counterparts, both with and without SDN integration. Results reveal
that edge-deployed ML-based IDSs significantly impact system performance during
cyber threats, with marked increases in resource consumption. SDN integration
further influences these outcomes, emphasizing the need for optimized
architectural design. Statistical analysis using ANOVA confirms the
significance of our findings. This research provides critical insights into the
performance and trade-offs of deploying ML-based IDSs in edge-based IoT
systems.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 16:11:10 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Jamshidi",
"Saeid",
""
],
[
"Nafi",
"Kawser Wazed",
""
],
[
"Nikanjam",
"Amin",
""
],
[
"Khomh",
"Foutse",
""
]
] |
2504.09722 | Lucy Oswald | L. S. Oswald, P. Weltevrede, B. Posselt, S. Johnston, A. Karastergiou,
M. E. Lower | The Thousand-Pulsar-Array programme on MeerKAT -- XVI. Mapping the
Galactic magnetic field with pulsar observations | 20 pages, 10 figures, 2 tables, accepted for publication in MNRAS | null | null | null | astro-ph.GA astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Measuring the magnetic field of the Milky Way reveals the structure and
evolution of the galaxy. Pulsar rotation measures (RMs) provide a means to
probe this Galactic magnetic field (GMF) in three dimensions. We use the
largest single-origin data set of pulsar measurements, from the MeerKAT
Thousand-Pulsar-Array, to map out GMF components parallel to pulsar lines of
sight. We also present these measurements for easy integration into the
consolidated RM catalogue, RMTable. Focusing on the Galactic disk, we
investigate competing theories of how the GMF relates to the spiral arms,
comparing our observational map with five analytic models of magnetic field
structure. We also analyse RMs to extragalactic radio sources, to help build up
a three-dimensional picture of the magnetic structure of the galaxy. In
particular, our large number of measurements allows us to investigate differing
magnetic field behaviour in the upper and lower halves of the Galactic plane.
We find that the GMF is best explained as following the spiral arms in a
roughly bisymmetric structure, with antisymmetric parity with respect to the
Galactic plane. This picture is complicated by variations in parity on
different spiral arms, and the parity change location appears to be shifted by
a distance of 0.15 kpc perpendicular to the Galactic plane. This indicates a
complex relationship between the large-scale distributions of matter and
magnetic fields in our galaxy. Future pulsar discoveries will help reveal the
origins of this relationship with greater precision, as well as probing the
locations of local magnetic field inhomogenities.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 21:08:16 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 17:28:34 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Oswald",
"L. S.",
""
],
[
"Weltevrede",
"P.",
""
],
[
"Posselt",
"B.",
""
],
[
"Johnston",
"S.",
""
],
[
"Karastergiou",
"A.",
""
],
[
"Lower",
"M. E.",
""
]
] |
2504.09744 | Mst Shamima Khanom | Mst Shamima Khanom, Benjamin W. Keller and Javier Ignacio Saavedra
Moreno | Unveiling the drivers of the Baryon Cycles with Interpretable Multi-step
Machine Learning and Simulations | 28 pages, 12 figures, submitted to The Open Journal of Astrophysics | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We present a new approach for understanding how galaxies lose or retain
baryons by utilizing a pipeline of two machine learning methods applied the
IllustrisTNG100 simulation. We employed a Random Forest Regressor and
Explainable Boosting Machine (EBM) model to connect the retained baryon
fraction of approximately 10^5 simulated galaxies to their properties. We
employed Random Forest models to filter and used the five most significant
properties to train an EBM. Interaction functions identified by the EBM
highlight the relationship between baryon fraction and three different galactic
mass measurements, the location of the rotation curve peak, and the velocity
dispersion. This interpretable machine learning-based approach provides a
promising pathway for understanding the baryon cycle in galaxies.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 22:43:19 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:25:18 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Khanom",
"Mst Shamima",
""
],
[
"Keller",
"Benjamin W.",
""
],
[
"Moreno",
"Javier Ignacio Saavedra",
""
]
] |
2504.09752 | Christopher Glein | Christopher R. Glein, Xinting Yu, Cindy N. Luu | Deciphering Sub-Neptune Atmospheres: New Insights from Geochemical
Models of TOI-270 d | Accepted by ApJ, 49 pages, 11 figures, 5 tables | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | The nature of sub-Neptunes is one of the hottest topics in exoplanetary
science. Temperate sub-Neptunes are of special interest because some could be
habitable. Here, we consider whether these planets might instead be rocky
worlds with thick, hot atmospheres. Can recent JWST observations of TOI-270 d
be understood in terms of such a model? We perform thermochemical equilibrium
calculations to infer conditions of quenching of C-H-O-N species. Our results
indicate apparent CO$_2$-CH$_4$ equilibrium between ~900 and ~1100 K. The CO
abundance should be quenched higher in the atmosphere where the equilibrium
CO/CO$_2$ ratio is lower, potentially explaining a lack of CO. N$_2$ is
predicted to dominate the nitrogen budget. We confirm that the atmosphere of
TOI-270 d is strongly enriched in both C and O$_g$$_a$$_s$ relative to
protosolar H, whereas N is likely to be less enriched or even depleted. We
attempt to reproduce these enrichments by modeling the atmosphere as nebular
gas that extracted heavy elements from accreted solids. This type of model can
explain the C/H and O$_g$$_a$$_s$/H ratios, but despite supersolar C/N ratios
provided by solids, the NH$_3$ abundance will probably be too high unless there
is a nitrogen sink in addition to N$_2$. A magma ocean may be implied, and
indeed the oxygen fugacity of the deep atmosphere seems sufficiently low to
support the sequestration of reduced N in silicate melt. The evaluation
presented here demonstrates that exoplanetary geochemistry now approaches a
level of sophistication comparable to that achieved within our own solar
system.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 23:07:20 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 15:30:14 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Glein",
"Christopher R.",
""
],
[
"Yu",
"Xinting",
""
],
[
"Luu",
"Cindy N.",
""
]
] |
2504.09775 | Abhimanyu Rajeshkumar Bambhaniya | Abhimanyu Rajeshkumar Bambhaniya, Hanjiang Wu, Suvinay Subramanian,
Sudarshan Srinivasan, Souvik Kundu, Amir Yazdanbakhsh, Midhilesh Elavazhagan,
Madhu Kumar, Tushar Krishna | Understanding and Optimizing Multi-Stage AI Inference Pipelines | Inference System Design for Multi-Stage AI Inference Pipelines. 13
Pages, 15 Figues, 3 Tables | null | null | null | cs.AR cs.AI cs.DC cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The rapid evolution of Large Language Models (LLMs) has driven the need for
increasingly sophisticated inference pipelines and hardware platforms. Modern
LLM serving extends beyond traditional prefill-decode workflows, incorporating
multi-stage processes such as Retrieval Augmented Generation (RAG), key-value
(KV) cache retrieval, dynamic model routing, and multi step reasoning. These
stages exhibit diverse computational demands, requiring distributed systems
that integrate GPUs, ASICs, CPUs, and memory-centric architectures. However,
existing simulators lack the fidelity to model these heterogeneous,
multi-engine workflows, limiting their ability to inform architectural
decisions.
To address this gap, we introduce HERMES, a Heterogeneous Multi-stage LLM
inference Execution Simulator. HERMES models diverse request stages; including
RAG, KV retrieval, reasoning, prefill, and decode across complex hardware
hierarchies. HERMES supports heterogeneous clients executing multiple models
concurrently unlike prior frameworks while incorporating advanced batching
strategies and multi-level memory hierarchies. By integrating real hardware
traces with analytical modeling, HERMES captures critical trade-offs such as
memory bandwidth contention, inter-cluster communication latency, and batching
efficiency in hybrid CPU-accelerator deployments. Through case studies, we
explore the impact of reasoning stages on end-to-end latency, optimal batching
strategies for hybrid pipelines, and the architectural implications of remote
KV cache retrieval. HERMES empowers system designers to navigate the evolving
landscape of LLM inference, providing actionable insights into optimizing
hardware-software co-design for next-generation AI workloads.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 00:29:49 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 17:34:04 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Bambhaniya",
"Abhimanyu Rajeshkumar",
""
],
[
"Wu",
"Hanjiang",
""
],
[
"Subramanian",
"Suvinay",
""
],
[
"Srinivasan",
"Sudarshan",
""
],
[
"Kundu",
"Souvik",
""
],
[
"Yazdanbakhsh",
"Amir",
""
],
[
"Elavazhagan",
"Midhilesh",
""
],
[
"Kumar",
"Madhu",
""
],
[
"Krishna",
"Tushar",
""
]
] |
2504.09798 | Sandya Wijaya | Sandya Wijaya, Jacob Bolano, Alejandro Gomez Soteres, Shriyanshu Kode,
Yue Huang, Anant Sahai | ReadMe.LLM: A Framework to Help LLMs Understand Your Library | 10 pages, 15 figures | null | null | null | cs.SE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Large Language Models (LLMs) often struggle with code generation tasks
involving niche software libraries. Existing code generation techniques with
only human-oriented documentation can fail -- even when the LLM has access to
web search and the library is documented online. To address this challenge, we
propose ReadMe$.$LLM, LLM-oriented documentation for software libraries. By
attaching the contents of ReadMe$.$LLM to a query, performance consistently
improves to near-perfect accuracy, with one case study demonstrating up to 100%
success across all tested models. We propose a software development lifecycle
where LLM-specific documentation is maintained alongside traditional software
updates. In this study, we present two practical applications of the
ReadMe$.$LLM idea with diverse software libraries, highlighting that our
proposed approach could generalize across programming domains.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 01:57:43 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 21:40:48 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wijaya",
"Sandya",
""
],
[
"Bolano",
"Jacob",
""
],
[
"Soteres",
"Alejandro Gomez",
""
],
[
"Kode",
"Shriyanshu",
""
],
[
"Huang",
"Yue",
""
],
[
"Sahai",
"Anant",
""
]
] |
2504.09891 | Ken Hayami | Kouta Sugihara and Ken Hayami | NR-SSOR right preconditioned RRGMRES for arbitrary singular systems and
least squares problems | arXiv admin note: text overlap with arXiv:2310.16442 | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | GMRES is known to determine a least squares solution of $ A x = b $ where $ A
\in R^{n \times n} $ without breakdown for arbitrary $ b \in R^n $, and initial
iterate $ x_0 \in R^n $ if and only if $ A $ is range-symmetric, i.e. $ R(A^T)
= R(A) $, where $ A $ may be singular and $ b $ may not be in the range space $
R(A) $ of $ A $.
In this paper, we propose applying the Range Restricted GMRES (RRGMRES) to $
A C A^T z = b $, where $ C \in R^{n \times n} $ is symmetric positive definite.
This determines a least squares solution $ x = C A^T z $ of $ A x = b $ without
breakdown for arbitrary (singular) matrix $ A \in R^{n \times n} $ and $ b, x_0
\in R^n $, and is much more stable and accurate compared to GMRES, RRGMRES and
MINRES-QLP applied to $ A x = b $ for inconsistent problems when $ b \notin
R(A) $. In particular, we propose applying the NR-SSOR as the inner iteration
right preconditioner, which also works efficiently for least squares problems $
\min_{x \in R^n} \| b - A x\|_2 $ for $ A \in R^{m \times n} $ and arbitrary $
b \in R^m $.
Numerical experiments demonstrate the validity of the proposed method.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 05:34:02 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 05:44:33 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Sugihara",
"Kouta",
""
],
[
"Hayami",
"Ken",
""
]
] |
2504.09971 | Ilan Komargodski | Ilan Komargodski, Itamar Schen, Omri Weinstein | Proofs of Useful Work from Arbitrary Matrix Multiplication | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | We revisit the longstanding open problem of implementing Nakamoto's
proof-of-work (PoW) consensus based on a real-world computational task $T(x)$
(as opposed to artificial random hashing), in a truly permissionless setting
where the miner itself chooses the input $x$. The challenge in designing such a
Proof-of-Useful-Work (PoUW) protocol, is using the native computation of $T(x)$
to produce a PoW certificate with prescribed hardness and with negligible
computational overhead over the worst-case complexity of $T(\cdot)$ -- This
ensures malicious miners cannot ``game the system" by fooling the verifier to
accept with higher probability compared to honest miners (while using similar
computational resources). Indeed, obtaining a PoUW with $O(1)$-factor overhead
is trivial for any task $T$, but also useless.
Our main result is a PoUW for the task of Matrix Multiplication $MatMul(A,B)$
of arbitrary matrices with $1+o(1)$ multiplicative overhead compared to naive
$MatMul$ (even in the presence of Fast Matrix Multiplication-style algorithms,
which are currently impractical). We conjecture that our protocol has optimal
security in the sense that a malicious prover cannot obtain any significant
advantage over an honest prover. This conjecture is based on reducing hardness
of our protocol to the task of solving a batch of low-rank random linear
equations which is of independent interest.
Since $MatMul$s are the bottleneck of AI compute as well as countless
industry-scale applications, this primitive suggests a concrete design of a new
L1 base-layer protocol, which nearly eliminates the energy-waste of Bitcoin
mining -- allowing GPU consumers to reduce their AI training and inference
costs by ``re-using" it for blockchain consensus, in exchange for block rewards
(2-for-1). This blockchain is currently under construction.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 08:22:40 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 18:46:12 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Komargodski",
"Ilan",
""
],
[
"Schen",
"Itamar",
""
],
[
"Weinstein",
"Omri",
""
]
] |
2504.10001 | Junlin Hao | Junlin Hao, Peiheng Wang, Haoyang Wang, Xinggong Zhang, Zongming Guo | GaussVideoDreamer: 3D Scene Generation with Video Diffusion and
Inconsistency-Aware Gaussian Splatting | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Single-image 3D scene reconstruction presents significant challenges due to
its inherently ill-posed nature and limited input constraints. Recent advances
have explored two promising directions: multiview generative models that train
on 3D consistent datasets but struggle with out-of-distribution generalization,
and 3D scene inpainting and completion frameworks that suffer from cross-view
inconsistency and suboptimal error handling, as they depend exclusively on
depth data or 3D smoothness, which ultimately degrades output quality and
computational performance. Building upon these approaches, we present
GaussVideoDreamer, which advances generative multimedia approaches by bridging
the gap between image, video, and 3D generation, integrating their strengths
through two key innovations: (1) A progressive video inpainting strategy that
harnesses temporal coherence for improved multiview consistency and faster
convergence. (2) A 3D Gaussian Splatting consistency mask to guide the video
diffusion with 3D consistent multiview evidence. Our pipeline combines three
core components: a geometry-aware initialization protocol, Inconsistency-Aware
Gaussian Splatting, and a progressive video inpainting strategy. Experimental
results demonstrate that our approach achieves 32% higher LLaVA-IQA scores and
at least 2x speedup compared to existing methods while maintaining robust
performance across diverse scenes.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 09:04:01 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 03:01:45 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Apr 2025 07:43:01 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Hao",
"Junlin",
""
],
[
"Wang",
"Peiheng",
""
],
[
"Wang",
"Haoyang",
""
],
[
"Zhang",
"Xinggong",
""
],
[
"Guo",
"Zongming",
""
]
] |
2504.10070 | Kiana Hooshanfar | Kiana Hooshanfar, Alireza Hosseini, Ahmad Kalhor, Babak Nadjar Araabi | DTFSal: Audio-Visual Dynamic Token Fusion for Video Saliency Prediction | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Audio-visual saliency prediction aims to mimic human visual attention by
identifying salient regions in videos through the integration of both visual
and auditory information. Although visual-only approaches have significantly
advanced, effectively incorporating auditory cues remains challenging due to
complex spatio-temporal interactions and high computational demands. To address
these challenges, we propose Dynamic Token Fusion Saliency (DFTSal), a novel
audio-visual saliency prediction framework designed to balance accuracy with
computational efficiency. Our approach features a multi-scale visual encoder
equipped with two novel modules: the Learnable Token Enhancement Block (LTEB),
which adaptively weights tokens to emphasize crucial saliency cues, and the
Dynamic Learnable Token Fusion Block (DLTFB), which employs a shifting
operation to reorganize and merge features, effectively capturing long-range
dependencies and detailed spatial information. In parallel, an audio branch
processes raw audio signals to extract meaningful auditory features. Both
visual and audio features are integrated using our Adaptive Multimodal Fusion
Block (AMFB), which employs local, global, and adaptive fusion streams for
precise cross-modal fusion. The resulting fused features are processed by a
hierarchical multi-decoder structure, producing accurate saliency maps.
Extensive evaluations on six audio-visual benchmarks demonstrate that DFTSal
achieves SOTA performance while maintaining computational efficiency.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 10:17:25 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 06:02:02 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Hooshanfar",
"Kiana",
""
],
[
"Hosseini",
"Alireza",
""
],
[
"Kalhor",
"Ahmad",
""
],
[
"Araabi",
"Babak Nadjar",
""
]
] |
2504.10072 | Lukasz Swiderski | M. Grodzicka-Kobylka, T. Szczesniak, L. Adamowski, L. Swiderski, K.
Brylew, A. Syntfeld-Ka\.zuch, W. K. Warburton, J. S. Carlson, J. J.
Valiente-Dob\'on | Comparison of an OGS/Polystyrene scintillator (BSO-406) with pure OGS
(BSO-100), EJ-276, EJ-309, and M600 scintillators | null | null | null | null | physics.ins-det | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we present an investigation into the scintillation properties
and pulse shape discrimination (PSD) performance of the new BSO-406, which is a
blend of 40% organic glass scintillator and 60% polystyrene. We tested a
cylindrical sample with dimensions of 2x2 inches. The study includes
measurements of neutron-gamma discrimination capability, emission spectra,
photoelectron yield, and the analysis of light pulse shapes originating from
events related to gamma-rays and fast neutrons. The results were compared to
data previously recorded using a pure Organic Glass Scintillator (BSO-100), an
EJ-309 liquid scintillator, and EJ-276 and M600 polyurethane-based plastic
scintillators.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 10:19:05 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 07:43:16 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Grodzicka-Kobylka",
"M.",
""
],
[
"Szczesniak",
"T.",
""
],
[
"Adamowski",
"L.",
""
],
[
"Swiderski",
"L.",
""
],
[
"Brylew",
"K.",
""
],
[
"Syntfeld-Każuch",
"A.",
""
],
[
"Warburton",
"W. K.",
""
],
[
"Carlson",
"J. S.",
""
],
[
"Valiente-Dobón",
"J. J.",
""
]
] |
2504.10143 | Yichao Cai | Yichao Cai, Yuhang Liu, Erdun Gao, Tianjiao Jiang, Zhen Zhang, Anton
van den Hengel, Javen Qinfeng Shi | Negate or Embrace: On How Misalignment Shapes Multimodal Representation
Learning | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | Multimodal representation learning, exemplified by multimodal contrastive
learning (MMCL) using image-text pairs, aims to learn powerful representations
by aligning cues across modalities. This approach relies on the core assumption
that the exemplar image-text pairs constitute two representations of an
identical concept. However, recent research has revealed that real-world
datasets often exhibit misalignment. There are two distinct viewpoints on how
to address this issue: one suggests mitigating the misalignment, and the other
leveraging it. We seek here to reconcile these seemingly opposing perspectives,
and to provide a practical guide for practitioners. Using latent variable
models we thus formalize misalignment by introducing two specific mechanisms:
selection bias, where some semantic variables are missing, and perturbation
bias, where semantic variables are distorted -- both affecting latent variables
shared across modalities. Our theoretical analysis demonstrates that, under
mild assumptions, the representations learned by MMCL capture exactly the
information related to the subset of the semantic variables invariant to
selection and perturbation biases. This provides a unified perspective for
understanding misalignment. Based on this, we further offer actionable insights
into how misalignment should inform the design of real-world ML systems. We
validate our theoretical findings through extensive empirical studies on both
synthetic data and real image-text datasets, shedding light on the nuanced
impact of misalignment on multimodal representation learning.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 11:54:19 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 05:22:32 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Cai",
"Yichao",
""
],
[
"Liu",
"Yuhang",
""
],
[
"Gao",
"Erdun",
""
],
[
"Jiang",
"Tianjiao",
""
],
[
"Zhang",
"Zhen",
""
],
[
"Hengel",
"Anton van den",
""
],
[
"Shi",
"Javen Qinfeng",
""
]
] |
2504.10149 | Michal Danilowski | Michal Danilowski, Soumyajit Chatterjee, Abhirup Ghosh | BoTTA: Benchmarking on-device Test Time Adaptation | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The performance of deep learning models depends heavily on test samples at
runtime, and shifts from the training data distribution can significantly
reduce accuracy. Test-time adaptation (TTA) addresses this by adapting models
during inference without requiring labeled test data or access to the original
training set. While research has explored TTA from various perspectives like
algorithmic complexity, data and class distribution shifts, model
architectures, and offline versus continuous learning, constraints specific to
mobile and edge devices remain underexplored. We propose BoTTA, a benchmark
designed to evaluate TTA methods under practical constraints on mobile and edge
devices. Our evaluation targets four key challenges caused by limited resources
and usage conditions: (i) limited test samples, (ii) limited exposure to
categories, (iii) diverse distribution shifts, and (iv) overlapping shifts
within a sample. We assess state-of-the-art TTA methods under these scenarios
using benchmark datasets and report system-level metrics on a real testbed.
Furthermore, unlike prior work, we align with on-device requirements by
advocating periodic adaptation instead of continuous inference-time adaptation.
Experiments reveal key insights: many recent TTA algorithms struggle with small
datasets, fail to generalize to unseen categories, and depend on the diversity
and complexity of distribution shifts. BoTTA also reports device-specific
resource use. For example, while SHOT improves accuracy by $2.25\times$ with
$512$ adaptation samples, it uses $1.08\times$ peak memory on Raspberry Pi
versus the base model. BoTTA offers actionable guidance for TTA in real-world,
resource-constrained deployments.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 12:00:00 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:16:19 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Danilowski",
"Michal",
""
],
[
"Chatterjee",
"Soumyajit",
""
],
[
"Ghosh",
"Abhirup",
""
]
] |
2504.10156 | Ben David Golan | Ben David Golan | The missing elements in the telegraph equations | null | null | null | null | physics.gen-ph | http://creativecommons.org/publicdomain/zero/1.0/ | The conventional modeling of transmission lines relies on the classical
telegraph equations, originally formulated over 150 years ago. These equations
are typically derived by representing the line as an assembly of infinitesimal
inductive, capacitive, and resistive elements. However, this formulation is
fundamentally flawed, as a transmission line cannot be accurately described
through a discretized model of infinitesimal lumped components. Instead, a more
rigorous approach should derive the governing equations directly from Maxwells
equations in conjunction with Ohms law. This paper presents such a derivation
and introduces a corrected formulation, herein referred to as the Trump
Equations.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 12:12:43 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 03:54:22 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Golan",
"Ben David",
""
]
] |
2504.10185 | Soumyadeep Pal | Soumyadeep Pal, Changsheng Wang, James Diffenderfer, Bhavya Kailkhura,
Sijia Liu | LLM Unlearning Reveals a Stronger-Than-Expected Coreset Effect in
Current Benchmarks | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language model unlearning has become a critical challenge in ensuring
safety and controlled model behavior by removing undesired data-model
influences from the pretrained model while preserving general utility.
Significant recent efforts have been dedicated to developing LLM unlearning
benchmarks such as WMDP (Weapons of Mass Destruction Proxy) and MUSE (Machine
Unlearning Six-way Evaluation), facilitating standardized unlearning
performance assessment and method comparison. Despite their usefulness, we
uncover for the first time a novel coreset effect within these benchmarks.
Specifically, we find that LLM unlearning achieved with the original (full)
forget set can be effectively maintained using a significantly smaller subset
(functioning as a "coreset"), e.g., as little as 5% of the forget set, even
when selected at random. This suggests that LLM unlearning in these benchmarks
can be performed surprisingly easily, even in an extremely low-data regime. We
demonstrate that this coreset effect remains strong, regardless of the LLM
unlearning method used, such as NPO (Negative Preference Optimization) and RMU
(Representation Misdirection Unlearning), the popular ones in these benchmarks.
The surprisingly strong coreset effect is also robust across various data
selection methods, ranging from random selection to more sophisticated
heuristic approaches. We explain the coreset effect in LLM unlearning through a
keyword-based perspective, showing that keywords extracted from the forget set
alone contribute significantly to unlearning effectiveness and indicating that
current unlearning is driven by a compact set of high-impact tokens rather than
the entire dataset. We further justify the faithfulness of coreset-unlearned
models along additional dimensions, such as mode connectivity and robustness to
jailbreaking attacks. Codes are available at
https://github.com/OPTML-Group/MU-Coreset.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 12:38:37 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 14:45:55 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Pal",
"Soumyadeep",
""
],
[
"Wang",
"Changsheng",
""
],
[
"Diffenderfer",
"James",
""
],
[
"Kailkhura",
"Bhavya",
""
],
[
"Liu",
"Sijia",
""
]
] |
2504.10215 | Konstantin Kunze | Konstantin Kunze | Public Health Insurance of Children and Maternal Labor Market Outcomes | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper exploits variation resulting from a series of federal and state
Medicaid expansions between 1977 and 2017 to estimate the effects of children's
increased access to public health insurance on the labor market outcomes of
their mothers. The results imply that the extended Medicaid eligibility of
children leads to positive labor supply responses of single mothers and to
negative labor supply responses of married mothers. The analysis of mechanisms
suggests that extended children's Medicaid eligibility positively affects
take-up of Medicaid and health of children.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 13:32:57 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:25:04 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kunze",
"Konstantin",
""
]
] |
2504.10337 | Wenlei Shi | Wenlei Shi, Xing Jin | Heimdall: test-time scaling on the generative verification | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An AI system can create and maintain knowledge only to the extent that it can
verify that knowledge itself. Recent work on long Chain-of-Thought reasoning
has demonstrated great potential of LLMs on solving competitive problems, but
their verification ability remains to be weak and not sufficiently
investigated. In this paper, we propose Heimdall, the long CoT verification LLM
that can accurately judge the correctness of solutions. With pure reinforcement
learning, we boost the verification accuracy from 62.5% to 94.5% on competitive
math problems. By scaling with repeated sampling, the accuracy further
increases to 97.5%. Through human evaluation, Heimdall demonstrates impressive
generalization capabilities, successfully detecting most issues in challenging
math proofs, the type of which is not included during training. Furthermore, we
propose Pessimistic Verification to extend the functionality of Heimdall to
scaling up the problem solving. It calls Heimdall to judge the solutions from a
solver model and based on the pessimistic principle, selects the most likely
correct solution with the least uncertainty. Taking
DeepSeek-R1-Distill-Qwen-32B as the solver model, Pessimistic Verification
improves the solution accuracy on AIME2025 from 54.2% to 70.0% with 16x compute
budget and to 83.3% with more compute budget. With the stronger solver Gemini
2.5 Pro, the score reaches 93.0%. Finally, we prototype an automatic knowledge
discovery system, a ternary system where one poses questions, another provides
solutions, and the third verifies the solutions. Using the data synthesis work
NuminaMath for the first two components, Heimdall effectively identifies
problematic records within the dataset and reveals that nearly half of the data
is flawed, which interestingly aligns with the recent ablation studies from
NuminaMath.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 15:46:33 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 14:58:26 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Shi",
"Wenlei",
""
],
[
"Jin",
"Xing",
""
]
] |
2504.10524 | Reza Bozorgpour | Reza Bozorgpour, Jacob R. Rammer | Hemodynamic Markers: CFD-Based Prediction of Cerebral Aneurysm Rupture
Risk | 16 figures, 24 pages | null | null | null | q-bio.QM physics.bio-ph physics.med-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This study investigates the influence of aneurysm evolution on hemodynamic
characteristics within the sac region. Using computational fluid dynamics
(CFD), blood flow through the parent vessel and aneurysm sac was analyzed to
assess the impact on wall shear stress (WSS), time-averaged wall shear stress
(TAWSS), and the oscillatory shear index (OSI), key indicators of rupture risk.
Additionally, Relative Residence Time (RRT) and Endothelial Cell Activation
Potential (ECAP) were examined to provide a broader understanding of the
aneurysm's hemodynamic environment. Six distinct cerebral aneurysm (CA) models,
all from individuals of the same gender, were selected to minimize
gender-related variability.
Results showed that unruptured cases exhibited higher WSS and TAWSS, along
with lower OSI and RRT values patterns consistent with stable flow conditions
supporting vascular integrity. In contrast, ruptured cases had lower WSS and
TAWSS, coupled with elevated OSI and RRT, suggesting disturbed and oscillatory
flow commonly linked to aneurysm wall weakening. ECAP was also higher in
ruptured cases, indicating increased endothelial activation under unstable
flow. Notably, areas with the highest OSI and RRT often aligned with vortex
centers, reinforcing the association between disturbed flow and aneurysm
instability.
These findings highlight the value of combining multiple hemodynamic
parameters for rupture risk assessment. Including RRT and ECAP provides deeper
insight into flow endothelium-interactions, offering a stronger basis for
evaluating aneurysm stability and guiding treatment decisions.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 22:59:31 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 02:10:45 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Bozorgpour",
"Reza",
""
],
[
"Rammer",
"Jacob R.",
""
]
] |
2504.10530 | Sarat Babu Moka Dr | Sarat Moka, Christian Hirsch, Volker Schmidt, and Dirk Kroese | Efficient Rare-Event Simulation for Random Geometric Graphs via
Importance Sampling | 29 Pages, 2 figures | null | null | null | math.PR stat.CO | http://creativecommons.org/licenses/by/4.0/ | Random geometric graphs defined on Euclidean subspaces, also called Gilbert
graphs, are widely used to model spatially embedded networks across various
domains. In such graphs, nodes are located at random in Euclidean space, and
any two nodes are connected by an edge if they lie within a certain distance
threshold. Accurately estimating rare-event probabilities related to key
properties of these graphs, such as the number of edges and the size of the
largest connected component, is important in the assessment of risk associated
with catastrophic incidents, for example. However, this task is computationally
challenging, especially for large networks. Importance sampling offers a viable
solution by concentrating computational efforts on significant regions of the
graph. This paper explores the application of an importance sampling method to
estimate rare-event probabilities, highlighting its advantages in reducing
variance and enhancing accuracy. Through asymptotic analysis and experiments,
we demonstrate the effectiveness of our methodology, contributing to improved
analysis of Gilbert graphs and showcasing the broader applicability of
importance sampling in complex network analysis.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 14:48:46 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 02:11:35 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Moka",
"Sarat",
""
],
[
"Hirsch",
"Christian",
""
],
[
"Schmidt",
"Volker",
""
],
[
"Kroese",
"Dirk",
""
]
] |
2504.10561 | Runqing Wu | Runqing Wu, Kaihui Huang, Hanyi Zhang, Fei Ye | Self-Controlled Dynamic Expansion Model for Continual Learning | 10 pages, 3 figures, 6 tables, Continual Learning, Cross-Domain
Continual Learning, Mixture Model | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Continual Learning (CL) epitomizes an advanced training paradigm wherein
prior data samples remain inaccessible during the acquisition of new tasks.
Numerous investigations have delved into leveraging a pre-trained Vision
Transformer (ViT) to enhance model efficacy in continual learning. Nonetheless,
these approaches typically utilize a singular, static backbone, which
inadequately adapts to novel tasks, particularly when engaging with diverse
data domains, due to a substantial number of inactive parameters. This paper
addresses this limitation by introducing an innovative Self-Controlled Dynamic
Expansion Model (SCDEM), which orchestrates multiple distinct trainable
pre-trained ViT backbones to furnish diverse and semantically enriched
representations. Specifically, by employing the multi-backbone architecture as
a shared module, the proposed SCDEM dynamically generates a new expert with
minimal parameters to accommodate a new task. A novel Collaborative
Optimization Mechanism (COM) is introduced to synergistically optimize multiple
backbones by harnessing prediction signals from historical experts, thereby
facilitating new task learning without erasing previously acquired knowledge.
Additionally, a novel Feature Distribution Consistency (FDC) approach is
proposed to align semantic similarity between previously and currently learned
representations through an optimal transport distance-based mechanism,
effectively mitigating negative knowledge transfer effects. Furthermore, to
alleviate over-regularization challenges, this paper presents a novel Dynamic
Layer-Wise Feature Attention Mechanism (DLWFAM) to autonomously determine the
penalization intensity on each trainable representation layer. An extensive
series of experiments have been conducted to evaluate the proposed
methodology's efficacy, with empirical results corroborating that the approach
attains state-of-the-art performance.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 15:22:51 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:13:45 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wu",
"Runqing",
""
],
[
"Huang",
"Kaihui",
""
],
[
"Zhang",
"Hanyi",
""
],
[
"Ye",
"Fei",
""
]
] |
2504.10741 | Julio Jaramillo | Julio C\'esar Jaramillo Quiceno | $q$-Heisenberg Algebra in $\otimes^{2}-$Tensor Space | null | null | null | null | math-ph math.MP quant-ph | http://creativecommons.org/licenses/by/4.0/ | In this paper, we introduce the $q$-Heisenberg algebra in the tensor product
space $\otimes^2$. We establish its algebraic properties and provide
applications to the theory of non-monogenic functions. Our results extend known
constructions in $q$-deformed algebras and offer new insights into functional
analysis in non-commutative settings.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 22:26:09 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 02:13:32 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Quiceno",
"Julio César Jaramillo",
""
]
] |
2504.10758 | Marcos Agnoletto | Marcos Agnoletto, Julio C. Correa Hoyos, M\'arcio Fabiano da Silva,
and Stefano Nardulli | Allard's interior $\varepsilon$-Regularity Theorem in Alexandrov spaces | 88 pages, 5 figures, any comments or suggestions are welcome. Please
send them to Marcos Agnoletto at [email protected]. In the second
version, we corrected a typo that appeared in Theorem A, Theorem B, and
Corollary A. Specifically, we replaced $p \in \mathbb{N}$ with $p \in
\mathbb{R}$ and added new references | null | null | null | math.DG math.AP math.MG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we prove Allard's Interior $\varepsilon$-Regularity Theorem
for $m$-dimensional varifolds with generalized mean curvature in $L^p_{loc}$,
for $p \in \mathbb{R}$ such that $p>m$, in Alexandrov spaces of dimension $n$
with double-sided bounded intrinsic sectional curvature. We first give an
intrinsic proof of this theorem in the case of varifolds in Riemannian
manifolds of dimension $n$ whose metric tensor is at least of class
$\mathcal{C}^2$, without using Nash's Isometric Embedding Theorem. This
approach provides explicitly computable constants that depend only on $n$, $m$,
the injectivity radius and bounds on the sectional curvature, which is
essential for proving our main theorem, as we establish it through a density
argument in the topological space of Riemannian manifolds with positive lower
bounds on the injectivity radius and double-sided bounds on sectional
curvature, equipped with the $\mathcal{C}^{1,\alpha}$ topology, for every
$\alpha \in ]0,1[$ (in fact, it is enough with the $W^{2,q}$ topology for some
suitable $q$ large enough).
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 23:18:09 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 15:14:56 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Agnoletto",
"Marcos",
""
],
[
"Hoyos",
"Julio C. Correa",
""
],
[
"da Silva",
"Márcio Fabiano",
""
],
[
"Nardulli",
"Stefano",
""
]
] |
2504.10786 | Katherine Storrs | Gene Tangtartharakul and Katherine R. Storrs | Visual Language Models show widespread visual deficits on
neuropsychological tests | 31 pages, 3 figures, 1 supplementary document with 1 figure and 51
sample images; corrected typo in Fig 1 | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Visual Language Models (VLMs) show remarkable performance in visual reasoning
tasks, successfully tackling college-level challenges that require high-level
understanding of images. However, some recent reports of VLMs struggling to
reason about elemental visual concepts like orientation, position, continuity,
and occlusion suggest a potential gulf between human and VLM vision. Here we
use the toolkit of neuropsychology to systematically assess the capabilities of
three state-of-the-art VLMs across visual domains. Using 51 tests drawn from
six clinical and experimental batteries, we characterise the visual abilities
of leading VLMs relative to normative performance in healthy adults. While the
models excel in straightforward object recognition tasks, we find widespread
deficits in low- and mid-level visual abilities that would be considered
clinically significant in humans. These selective deficits, profiled through
validated test batteries, suggest that an artificial system can achieve complex
object recognition without developing foundational visual concepts that in
humans require no explicit training.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 01:04:56 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:27:42 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Tangtartharakul",
"Gene",
""
],
[
"Storrs",
"Katherine R.",
""
]
] |
2504.10794 | Xiangxiang Wang | Xiangxiang Wang and Jumpei Takata | Cataclysmic Variable Candidates Identified in eROSITA-DE DR1,XMM-Newton,
Swift, and ROSAT Catalogs | 14 pages,11 figures, accepted for publication in A&A | null | null | null | astro-ph.IM astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cataclysmic variables (CVs) are binary systems with a white dwarf accreting
matter from a low-mass star, making them significant sources of X-ray emission
in the Galaxy. We present a systematic search for X-ray emitting CV candidates
by cross-matching four X-ray catalogs (eROSITA, XMM-Newton, Swift, and ROSAT)
with Gaia sources located in the bridge region between the main sequence and
white dwarf cooling sequence in the Hertzsprung-Russell diagram. From 444
candidates (267 confirmed CVs and 177 new candidates), we detect orbital
modulation in 56 sources using ZTF/TESS light curves. The eROSITA catalog
contributes 51% of candidates, outperforming other surveys due to its wider sky
coverage and higher sensitivity (~10^{-14} erg cm^{-2} s^{-1}). Our method
demonstrates the efficiency of combining X-ray data with time-domain analysis
for CV identification, with future eROSITA observations expected to expand the
population of X-ray emitting CVs.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 01:33:42 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 06:19:57 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wang",
"Xiangxiang",
""
],
[
"Takata",
"Jumpei",
""
]
] |
2504.10796 | Lukas-Benedikt Fiechtner | Lukas-Benedikt Fiechtner and Jose Blanchet | Wasserstein Distributionally Robust Regret Optimization | null | null | null | null | math.OC cs.LG | http://creativecommons.org/licenses/by/4.0/ | Distributionally Robust Optimization (DRO) is a popular framework for
decision-making under uncertainty, but its adversarial nature can lead to
overly conservative solutions. To address this, we study ex-ante
Distributionally Robust Regret Optimization (DRRO), focusing on
Wasserstein-based ambiguity sets which are popular due to their links to
regularization and machine learning. We provide a systematic analysis of
Wasserstein DRRO, paralleling known results for Wasserstein DRO. Under
smoothness and regularity conditions, we show that Wasserstein DRRO coincides
with Empirical Risk Minimization (ERM) up to first-order terms, and exactly so
in convex quadratic settings. We revisit the Wasserstein DRRO newsvendor
problem, where the loss is the maximum of two linear functions of demand and
decision. Extending [25], we show that the regret can be computed by maximizing
two one-dimensional concave functions. For more general loss functions
involving the maximum of multiple linear terms in multivariate random variables
and decision vectors, we prove that computing the regret and thus also the DRRO
policy is NP-hard. We then propose a convex relaxation for these more general
Wasserstein DRRO problems and demonstrate its strong empirical performance.
Finally, we provide an upper bound on the optimality gap of our relaxation and
show it improves over recent alternatives.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 01:47:11 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:02:18 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Fiechtner",
"Lukas-Benedikt",
""
],
[
"Blanchet",
"Jose",
""
]
] |
2504.10803 | Haining Pan | Haining Pan, Thomas Iadecola, E. M. Stoudenmire, J. H. Pixley | Control-driven critical fluctuations across quantum trajectories | 21 Pages, 19 figures | null | null | null | quant-ph cond-mat.dis-nn cond-mat.stat-mech cond-mat.str-el physics.atm-clus | http://creativecommons.org/publicdomain/zero/1.0/ | Monitored quantum circuits in which entangling unitary dynamics compete with
projective local measurements can host measurement-induced phase transitions
witnessed by entanglement measures at late times. Adding feedback conditioned
on the measurement outcomes gives rise to another type of phase transition
witnessed by local order parameters and correlation functions. These
transitions, known as control or absorbing-state transitions, generically occur
within the area-law entanglement phase and are thought to be governed by
classical physics in that their critical exponents match those of the classical
limit of the model. In this work, we examine quantum features of these
transitions, focusing on a Bernoulli circuit model with a well-defined
classical limit. First we demonstrate that, in the local basis defined by the
absorbing state, the steady-state quantum coherence undergoes a phase
transition at the control transition, where its logarithm changes
discontinuously from volume- to area-law scaling. Second, we analyze the
control transition from the perspective of fluctuations in observables, which
carry two contributions: classical fluctuations over circuit realizations
(present in the classical limit), and quantum fluctuations over trajectories
and states (both absent in the classical limit). Both contributions can be
estimated in experiments without post-selection. The circuit-to-circuit
fluctuations, the dominant contribution, carry the critical behavior of the
classical limit. However, the subleading quantum fluctuations that represent
fluctuations between different quantum "worlds" also go critical at the control
transition. These critical quantum fluctuations at the control transition also
occur in other models, and we discuss how they can be measured experimentally
without post-selection.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 01:59:09 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Pan",
"Haining",
""
],
[
"Iadecola",
"Thomas",
""
],
[
"Stoudenmire",
"E. M.",
""
],
[
"Pixley",
"J. H.",
""
]
] |
2504.10872 | Nitish Singh | Nitish Singh, S. Sriram, Ramya Sethuram, Bharat Kumar Yerra, Rahuldeb
Burman, G. Nataraj, C Chethan, P. Madan Mohan Kemkar, K Sagayanathan, Saikat
Das, Francis Xavier Rozario J | Design and Fabrication of a lightweight three-lens corrector system for
the 2.34-m Vainu Bappu Telescope | 25 pages, 11 figures, Accepted in Experimental Astronomy | null | null | null | astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Vainu Bappu Telescope (VBT) is a 2.34-m reflector, primarily supported
on-axis field of view, offering high-resolution and low-to-medium resolution
spectroscopic observations in its prime and Cassegrain configurations. This
study presents the design and fabrication of a compact, lightweight,
three-element wide-field corrector (WFC) utilizing three spherical lenses to
cover a polychromatic wavelength range over a 30$'$ FoV at prime focus. The WFC
design was optimized using ZEMAX, ensuring precision in aberrations,
tolerances, and atmospheric dispersion. The fabricated lenses met stringent
tolerances, with a $\pm$1 mm deviation in radius of curvature and $\pm$2 mm
deviation in center thickness. A mechanical mount was developed to integrate
all the WFC lenses, and wavefront error testing for the WFC system was
performed using ZYGO interferometry, yielding a Wavefront Error of 0.05
$\lambda$. Laboratory performance tests were designed and conducted using a
dedicated setup with achromatic lenses and 100 $\mu m$ fiber-coupled
polychromatic light source showed a deviation of 0.1 pixel on-axis and 0.5
pixel at the extreme off-axis field compared to the ZEMAX design, demonstrating
that the optical performance of WFC is with minimal aberrations across the
entire FoV. The successful integration of the WFC at the VBT prime focus will
increase the FoV, enabling the multi-fiber, multi-spectrograph setup in 30$'$
field that will facilitate both OMR and Echelle spectrograph to be used on the
same night along with the addition of new multi-object spectrograph and an
integral field unit instrument. This will mark a significant upgrade for the
VBT, broadening its research potential, and expanding its observational
versatility.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 05:03:00 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:08:12 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Singh",
"Nitish",
""
],
[
"Sriram",
"S.",
""
],
[
"Sethuram",
"Ramya",
""
],
[
"Yerra",
"Bharat Kumar",
""
],
[
"Burman",
"Rahuldeb",
""
],
[
"Nataraj",
"G.",
""
],
[
"Chethan",
"C",
""
],
[
"Kemkar",
"P. Madan Mohan",
""
],
[
"Sagayanathan",
"K",
""
],
[
"Das",
"Saikat",
""
],
[
"J",
"Francis Xavier Rozario",
""
]
] |
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 | null | 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"
}
] | 2025-04-17T00:00:00 | [
[
"Pate",
"Shripad",
""
],
[
"Farooq",
"Aiman",
""
],
[
"Datta",
"Suvrankar",
""
],
[
"Sheikh",
"Musadiq Aadil",
""
],
[
"Kumar",
"Atin",
""
],
[
"Mishra",
"Deepak",
""
]
] |
2504.10955 | Lei Yu | Lei Yu, Qian Ge, Ke Han, Wen Ji, Yueqi Liu | Designing optimal subsidy schemes and recycling plans for sustainable
treatment of construction and demolition waste | null | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | More than 10 billion tons of construction and demolition waste (CW) are
generated globally each year, exerting a significant impact on the environment.
In the CW recycling process, the government and the carrier are the two primary
stakeholders. The carrier is responsible for transporting CW from production
sites to backfill sites or processing facilities, with a primary focus on
transport efficiency and revenue. Meanwhile, the government aims to minimize
pollution from the recycling system, which is influenced by transport modes,
shipment distances, and the processing methods used for CW. This paper develops
a bi-objective, bi-level optimization model to address these challenges. The
upper-level model is a linear programming model that optimizes the government's
subsidy scheme, while the lower-level model is a minimum-cost flow model that
optimizes the carrier's recycling plan. A hybrid heuristic solution method is
proposed to tackle the problem's complexity. A case study in Chengdu, China,
demonstrates the computational efficiency of the model and its small solution
gap. With an optimized subsidy scheme and recycling plan, pollution can be
reduced by over 29.29% through a relatively small investment in subsidies.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 08:00:46 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 03:01:16 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Yu",
"Lei",
""
],
[
"Ge",
"Qian",
""
],
[
"Han",
"Ke",
""
],
[
"Ji",
"Wen",
""
],
[
"Liu",
"Yueqi",
""
]
] |
2504.10956 | JianMing Yan | Jian-Ming Yan, Tao Zhu, Qiang Wu | The effects of asymptotically flat $R^2$ spacetime on black hole image
of Sagittarius A* | 11 pages, 5 figures | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new class of analytically expressible vacuum solutions has recently been
discovered for pure ${R}^2$ gravity, building upon Buchdahl's seminal work from
1962. These solutions, inspired by Buchdahl's framework, offer a promising
avenue for testing ${R}^2$ gravity against astrophysical observations. Within a
subset of asymptotically flat Buchdahl-inspired vacuum spacetimes, we introduce
a free parameter $\epsilon$ to characterize deviations from the Schwarzschild
metric, which is recovered in the limit $\epsilon = 0$. In this study, we
employ the publicly available code \textit{ipole} to simulate black hole images
under the Buchdahl-inspired metric, with a focus on the black hole at the
center of the Milky Way, Sagittarius A* (Sgr A*). Our simulations show that
both the shadow size and photon ring diameter decrease monotonically with
increasing $\epsilon$. By exploring a range of observational inclination
angles, we find that the photon ring diameter being a direct observable is only
weakly sensitive to the inclination angle. We further constrain the parameter
$\epsilon$ by comparing our simulation results with the Event Horizon Telescope
(EHT) observations of Sgr A*. The obtained bounds are consistent with those
previously derived from the orbital motion of the S2 star, but provide tighter
constraints. In addition, we analyze the influence of the Buchdahl-inspired
spacetime on the polarization patterns near the black hole and find its impact
to be minimal. In contrast, the observational inclination angle has a
substantial effect on the observed polarization structure, highlighting the
dominant role of viewing geometry in shaping polarization features.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 08:02:35 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 15:53:48 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Yan",
"Jian-Ming",
""
],
[
"Zhu",
"Tao",
""
],
[
"Wu",
"Qiang",
""
]
] |
2504.10962 | Amith Manoharan Ph. D | Edvin Martin Andrejev, Amith Manoharan, Karl-Eerik Unt, and Arun Kumar
Singh | $\pi$-MPPI: A Projection-based Model Predictive Path Integral Scheme for
Smooth Optimal Control of Fixed-Wing Aerial Vehicles | 8 pages, 4 figures, submitted to IEEE RA-L | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | Model Predictive Path Integral (MPPI) is a popular sampling-based Model
Predictive Control (MPC) algorithm for nonlinear systems. It optimizes
trajectories by sampling control sequences and averaging them. However, a key
issue with MPPI is the non-smoothness of the optimal control sequence, leading
to oscillations in systems like fixed-wing aerial vehicles (FWVs). Existing
solutions use post-hoc smoothing, which fails to bound control derivatives.
This paper introduces a new approach: we add a projection filter $\pi$ to
minimally correct control samples, ensuring bounds on control magnitude and
higher-order derivatives. The filtered samples are then averaged using MPPI,
leading to our $\pi$-MPPI approach. We minimize computational overhead by using
a neural accelerated custom optimizer for the projection filter. $\pi$-MPPI
offers a simple way to achieve arbitrary smoothness in control sequences. While
we focus on FWVs, this projection filter can be integrated into any MPPI
pipeline. Applied to FWVs, $\pi$-MPPI is easier to tune than the baseline,
resulting in smoother, more robust performance.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 08:11:02 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 08:00:07 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Andrejev",
"Edvin Martin",
""
],
[
"Manoharan",
"Amith",
""
],
[
"Unt",
"Karl-Eerik",
""
],
[
"Singh",
"Arun Kumar",
""
]
] |
2504.10964 | Evagoras Makridis Mr | Evagoras Makridis, Gabriele Oliva, Kasagatta Ramesh Narahari,
Mohammadreza Doostmohammadian, Usman A. Khan, and Themistoklis Charalambous | Distributed Optimization with Gradient Tracking over Heterogeneous
Delay-Prone Directed Networks | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | In this paper, we address the distributed optimization problem over
unidirectional networks with possibly time-invariant heterogeneous bounded
transmission delays. In particular, we propose a modified version of the
Accelerated Distributed Directed OPTimization (ADD-OPT) algorithm, herein
called Robustified ADD-OPT (R-ADD-OPT), which is able to solve the distributed
optimization problem, even when the communication links suffer from
heterogeneous but bounded transmission delays. We show that if the gradient
step-size of the R-ADD-OPT algorithm is within a certain range, which also
depends on the maximum time delay in the network, then the nodes are guaranteed
to converge to the optimal solution of the distributed optimization problem.
The range of the gradient step-size that guarantees convergence can be computed
a priori based on the maximum time delay in the network.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 08:14:30 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 09:33:54 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Makridis",
"Evagoras",
""
],
[
"Oliva",
"Gabriele",
""
],
[
"Narahari",
"Kasagatta Ramesh",
""
],
[
"Doostmohammadian",
"Mohammadreza",
""
],
[
"Khan",
"Usman A.",
""
],
[
"Charalambous",
"Themistoklis",
""
]
] |
2504.10974 | Zhisheng Zhang | Zhisheng Zhang, Peng Zhang, Fengxiang Wang, Liangli Ma, Fuchun Sun | Self-Supervised Enhancement of Forward-Looking Sonar Images: Bridging
Cross-Modal Degradation Gaps through Feature Space Transformation and
Multi-Frame Fusion | null | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Enhancing forward-looking sonar images is critical for accurate underwater
target detection. Current deep learning methods mainly rely on supervised
training with simulated data, but the difficulty in obtaining high-quality
real-world paired data limits their practical use and generalization. Although
self-supervised approaches from remote sensing partially alleviate data
shortages, they neglect the cross-modal degradation gap between sonar and
remote sensing images. Directly transferring pretrained weights often leads to
overly smooth sonar images, detail loss, and insufficient brightness. To
address this, we propose a feature-space transformation that maps sonar images
from the pixel domain to a robust feature domain, effectively bridging the
degradation gap. Additionally, our self-supervised multi-frame fusion strategy
leverages complementary inter-frame information to naturally remove speckle
noise and enhance target-region brightness. Experiments on three self-collected
real-world forward-looking sonar datasets show that our method significantly
outperforms existing approaches, effectively suppressing noise, preserving
detailed edges, and substantially improving brightness, demonstrating strong
potential for underwater target detection applications.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 08:34:56 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 15:58:55 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Zhisheng",
""
],
[
"Zhang",
"Peng",
""
],
[
"Wang",
"Fengxiang",
""
],
[
"Ma",
"Liangli",
""
],
[
"Sun",
"Fuchun",
""
]
] |
2504.11014 | Eunsoo Im | Eunsoo Im, Jung Kwon Lee, Changhyun Jee | GATE3D: Generalized Attention-based Task-synergized Estimation in 3D* | 9pages, 1 supple | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | The emerging trend in computer vision emphasizes developing universal models
capable of simultaneously addressing multiple diverse tasks. Such universality
typically requires joint training across multi-domain datasets to ensure
effective generalization. However, monocular 3D object detection presents
unique challenges in multi-domain training due to the scarcity of datasets
annotated with accurate 3D ground-truth labels, especially beyond typical
road-based autonomous driving contexts. To address this challenge, we introduce
a novel weakly supervised framework leveraging pseudo-labels. Current
pretrained models often struggle to accurately detect pedestrians in non-road
environments due to inherent dataset biases. Unlike generalized image-based 2D
object detection models, achieving similar generalization in monocular 3D
detection remains largely unexplored. In this paper, we propose GATE3D, a novel
framework designed specifically for generalized monocular 3D object detection
via weak supervision. GATE3D effectively bridges domain gaps by employing
consistency losses between 2D and 3D predictions. Remarkably, our model
achieves competitive performance on the KITTI benchmark as well as on an
indoor-office dataset collected by us to evaluate the generalization
capabilities of our framework. Our results demonstrate that GATE3D
significantly accelerates learning from limited annotated data through
effective pre-training strategies, highlighting substantial potential for
broader impacts in robotics, augmented reality, and virtual reality
applications. Project page: https://ies0411.github.io/GATE3D/
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 09:37:54 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:38:27 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Im",
"Eunsoo",
""
],
[
"Lee",
"Jung Kwon",
""
],
[
"Jee",
"Changhyun",
""
]
] |
2504.11074 | Gianmarco Mengaldo Dr | Zhou Fang, Gianmarco Mengaldo | Dynamical errors in machine learning forecasts | null | null | null | null | cs.LG cs.AI physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | In machine learning forecasting, standard error metrics such as mean absolute
error (MAE) and mean squared error (MSE) quantify discrepancies between
predictions and target values. However, these metrics do not directly evaluate
the physical and/or dynamical consistency of forecasts, an increasingly
critical concern in scientific and engineering applications.
Indeed, a fundamental yet often overlooked question is whether machine
learning forecasts preserve the dynamical behavior of the underlying system.
Addressing this issue is essential for assessing the fidelity of machine
learning models and identifying potential failure modes, particularly in
applications where maintaining correct dynamical behavior is crucial.
In this work, we investigate the relationship between standard forecasting
error metrics, such as MAE and MSE, and the dynamical properties of the
underlying system. To achieve this goal, we use two recently developed
dynamical indices: the instantaneous dimension ($d$), and the inverse
persistence ($\theta$). Our results indicate that larger forecast errors --
e.g., higher MSE -- tend to occur in states with higher $d$ (higher complexity)
and higher $\theta$ (lower persistence). To further assess dynamical
consistency, we propose error metrics based on the dynamical indices that
measure the discrepancy of the forecasted $d$ and $\theta$ versus their correct
values. Leveraging these dynamical indices-based metrics, we analyze direct and
recursive forecasting strategies for three canonical datasets -- Lorenz,
Kuramoto-Sivashinsky equation, and Kolmogorov flow -- as well as a real-world
weather forecasting task. Our findings reveal substantial distortions in
dynamical properties in ML forecasts, especially for long forecast lead times
or long recursive simulations, providing complementary information on ML
forecast fidelity that can be used to improve ML models.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 11:16:13 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 08:48:34 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Fang",
"Zhou",
""
],
[
"Mengaldo",
"Gianmarco",
""
]
] |
2504.11101 | Yulong Zhang | Yulong Zhang, Tianyi Liang, Xinyue Huang, Erfei Cui, Xu Guo, Pei Chu,
Chenhui Li, Ru Zhang, Wenhai Wang, Gongshen Liu | Consensus Entropy: Harnessing Multi-VLM Agreement for Self-Verifying and
Self-Improving OCR | null | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Optical Character Recognition (OCR) task is important for evaluating
Vision-Language Models (VLMs) and providing high-quality data sources for LLM
training data. While state-of-the-art VLMs show improved average OCR accuracy,
they still struggle with sample-level quality degradation and lack reliable
automatic detection of low-quality outputs. We introduce Consensus Entropy
(CE), a training-free post-inference method that quantifies OCR uncertainty by
aggregating outputs from multiple VLMs. Our approach exploits a key insight:
correct VLM OCR predictions converge in output space while errors diverge. We
develop a lightweight multi-model framework that effectively identifies
problematic samples, selects the best outputs and combines model strengths.
Experiments across multiple OCR benchmarks and VLMs demonstrate that CE
outperforms VLM-as-judge approaches and single-model baselines at the same cost
and achieves state-of-the-art results across multiple metrics. For instance,
our solution demonstrates: achieving 15.2% higher F1 scores than VLM-as-judge
methods in quality verification, delivering 6.0% accuracy gains on mathematical
calculation tasks, and requiring rephrasing only 7.3% of inputs while
maintaining overall performance. Notably, the entire process requires neither
training nor supervision while maintaining plug-and-play functionality
throughout.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 11:51:18 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 03:22:14 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Yulong",
""
],
[
"Liang",
"Tianyi",
""
],
[
"Huang",
"Xinyue",
""
],
[
"Cui",
"Erfei",
""
],
[
"Guo",
"Xu",
""
],
[
"Chu",
"Pei",
""
],
[
"Li",
"Chenhui",
""
],
[
"Zhang",
"Ru",
""
],
[
"Wang",
"Wenhai",
""
],
[
"Liu",
"Gongshen",
""
]
] |
2504.11129 | Paul Vaandrager | P. Vaandrager, M.L. Lekala and S.A. Rakityansky | $R$-matrix type parametrization of the Jost function for extracting the
resonance parameters from scattering data | null | null | 10.1140/epja/s10050-025-01549-x | null | quant-ph nucl-th | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A new method is proposed for fitting non-relativistic binary-scattering data
and for extracting the parameters of possible quantum resonances in the
compound system that is formed during the collision. The method combines the
well-known $R$-matrix approach with the analysis based on the semi-analytic
representation of the Jost functions. It is shown that such a combination has
the advantages of both these approaches, namely, the number of the fitting
parameters remains relatively small (as for the $R$-matrix approach) and the
proper analytic structure of the $S$-matrix is preserved (as for the Jost
function method). It is also shown that the new formalism, although closely
related to the $R$-matrix method, has the benefit of no dependence on an
arbitrary channel radius. The efficiency and accuracy of the proposed method
are tested using a model single-channel potential. Artificial ``experimental''
data generated with this potential are fitted, and its known resonances are
successfully recovered as zeros of the Jost function on the appropriate sheet
of the Riemann surface of the energy.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 12:29:57 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Vaandrager",
"P.",
""
],
[
"Lekala",
"M. L.",
""
],
[
"Rakityansky",
"S. A.",
""
]
] |
2504.11168 | William Hackett | William Hackett, Lewis Birch, Stefan Trawicki, Neeraj Suri, Peter
Garraghan | Bypassing Prompt Injection and Jailbreak Detection in LLM Guardrails | 12 pages, 5 figures, 6 tables | null | null | null | cs.CR cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLMs) guardrail systems are designed to protect
against prompt injection and jailbreak attacks. However, they remain vulnerable
to evasion techniques. We demonstrate two approaches for bypassing LLM prompt
injection and jailbreak detection systems via traditional character injection
methods and algorithmic Adversarial Machine Learning (AML) evasion techniques.
Through testing against six prominent protection systems, including Microsoft's
Azure Prompt Shield and Meta's Prompt Guard, we show that both methods can be
used to evade detection while maintaining adversarial utility achieving in some
instances up to 100% evasion success. Furthermore, we demonstrate that
adversaries can enhance Attack Success Rates (ASR) against black-box targets by
leveraging word importance ranking computed by offline white-box models. Our
findings reveal vulnerabilities within current LLM protection mechanisms and
highlight the need for more robust guardrail systems.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 13:16:02 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 15:33:06 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Hackett",
"William",
""
],
[
"Birch",
"Lewis",
""
],
[
"Trawicki",
"Stefan",
""
],
[
"Suri",
"Neeraj",
""
],
[
"Garraghan",
"Peter",
""
]
] |
2504.11170 | Taewook Kang | Taewook Kang, Bum-Jae You, Juyoun Park, Yisoo Lee | A Real-time Anomaly Detection Method for Robots based on a Flexible and
Sparse Latent Space | 20 pages, 11 figures | null | null | null | cs.RO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The growing demand for robots to operate effectively in diverse environments
necessitates the need for robust real-time anomaly detection techniques during
robotic operations. However, deep learning-based models in robotics face
significant challenges due to limited training data and highly noisy signal
features. In this paper, we present Sparse Masked Autoregressive Flow-based
Adversarial AutoEncoders model to address these problems. This approach
integrates Masked Autoregressive Flow model into Adversarial AutoEncoders to
construct a flexible latent space and utilize Sparse autoencoder to efficiently
focus on important features, even in scenarios with limited feature space. Our
experiments demonstrate that the proposed model achieves a 4.96% to 9.75%
higher area under the receiver operating characteristic curve for
pick-and-place robotic operations with randomly placed cans, compared to
existing state-of-the-art methods. Notably, it showed up to 19.67% better
performance in scenarios involving collisions with lightweight objects.
Additionally, unlike the existing state-of-the-art model, our model performs
inferences within 1 millisecond, ensuring real-time anomaly detection. These
capabilities make our model highly applicable to machine learning-based robotic
safety systems in dynamic environments. The code will be made publicly
available after acceptance.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 13:17:14 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 08:50:55 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kang",
"Taewook",
""
],
[
"You",
"Bum-Jae",
""
],
[
"Park",
"Juyoun",
""
],
[
"Lee",
"Yisoo",
""
]
] |
2504.11177 | Bingbing Wang | Bingbing Wang, Junxi Yu, Prakash Sharma, Cheng-Cheng Liu | Pressure-Tunable Generalized Wigner Crystal and Fractional Chern
Insulator in twisted MoTe$_2$ | null | null | null | null | cond-mat.mes-hall cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the forming of low-energy flat bands, the moir\'e superlattices of the
transition metal dichalcogenides are fascinating platforms for studying novel
correlated states when such flat bands are fractionally filled, with the
Coulomb interaction dominating. Here, we demonstrate that pressure can
efficiently tune the flatness and quantum geometry of the single-particle bands
in twisted bilayer MoTe$_2$ ($\textit{t}$MoTe$_2$). By fractionally filling the
topmost valence band, we find that pressure can act as a flexible means to
modulate the fractional Chern insulator (FCI) and the generalized Wigner
crystal (GWC) and control their many-body topological phase transitions.
Moreover, our results indicate a remarkable correspondence between the
single-particle band geometry and the formation of FCI and GWC. As the recent
experiments report the presence of FCI phases in $\textit{t}$MoTe$_2$, our
predictions could be readily implemented experimentally.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 13:32:42 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wang",
"Bingbing",
""
],
[
"Yu",
"Junxi",
""
],
[
"Sharma",
"Prakash",
""
],
[
"Liu",
"Cheng-Cheng",
""
]
] |
2504.11184 | Josiah Chong | Shau Poh Chong and Peter T\"or\"ok | A simple, robust and cost-effective method to achieve dispersion
matching in swept source OCT | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | Optical path length and dispersion matching in both measurement and reference
arms of an OCT system is critical for achieving bandwidth-limited axial
resolution. To minimize or eliminate dispersion mismatch, most, if not all,
fiber-based OCT realisations employ a reference arm configuration that is as
closely identical to the measurement arm as possible. This typically includes a
collimator, dispersion compensating material (or sometimes a set of lenses), as
well as a mirror (or retro-reflector) mounted on a translation stage. However,
this solution makes the total instrument cost higher and the setup bulkier than
necessary and it also renders the reference arm mechanically unstable. Here, a
simple yet robust, low-cost reference arm setup is presented and its ability to
compensate for measurement arm dispersion is demonstrated. We use a single-mode
fiber cleaved and polished perpendicular to the fiber axis to construct the
reference arm. The length and material of the fibre is determined by
considering the optical path length and dispersion of the measurement arm.
Experimental images demonstrate the operation of the novel reference arm in our
Swept-source Optical Coherence Tomography.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 13:40:50 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 01:48:16 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Chong",
"Shau Poh",
""
],
[
"Török",
"Peter",
""
]
] |
2504.11197 | Shangyu Liu | Shangyu Liu, Zhenzhe Zheng, Xiaoyao Huang, Fan Wu, Guihai Chen, Jie Wu | Efficient Distributed Retrieval-Augmented Generation for Enhancing
Language Model Performance | null | null | null | null | cs.LG cs.AI cs.DC cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Small language models (SLMs) support efficient deployments on
resource-constrained edge devices, but their limited capacity compromises
inference performance. Retrieval-augmented generation (RAG) is a promising
solution to enhance model performance by integrating external databases,
without requiring intensive on-device model retraining. However, large-scale
public databases and user-specific private contextual documents are typically
located on the cloud and the device separately, while existing RAG
implementations are primarily centralized. To bridge this gap, we propose
DRAGON, a distributed RAG framework to enhance on-device SLMs through both
general and personal knowledge without the risk of leaking document privacy.
Specifically, DRAGON decomposes multi-document RAG into multiple parallel token
generation processes performed independently and locally on the cloud and the
device, and employs a newly designed Speculative Aggregation, a dual-side
speculative algorithm to avoid frequent output synchronization between the
cloud and device. A new scheduling algorithm is further introduced to identify
the optimal aggregation side based on real-time network conditions. Evaluations
on real-world hardware testbed demonstrate a significant performance
improvement of DRAGON-up to 1.9x greater gains over standalone SLM compared to
the centralized RAG, substantial reduction in per-token latency, and negligible
Time to First Token (TTFT) overhead.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 13:53:08 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 03:32:23 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Liu",
"Shangyu",
""
],
[
"Zheng",
"Zhenzhe",
""
],
[
"Huang",
"Xiaoyao",
""
],
[
"Wu",
"Fan",
""
],
[
"Chen",
"Guihai",
""
],
[
"Wu",
"Jie",
""
]
] |
2504.11212 | Martin Rumpf | Samuel Weidemaier, Florine Hartwig, Josua Sassen, Sergio Conti, Mirela
Ben-Chen, Martin Rumpf | SDFs from Unoriented Point Clouds using Neural Variational Heat
Distances | 14 pages, 16 figures, 4 tables | null | null | null | math.NA cs.GR cs.LG cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a novel variational approach for computing neural Signed Distance
Fields (SDF) from unoriented point clouds. To this end, we replace the commonly
used eikonal equation with the heat method, carrying over to the neural domain
what has long been standard practice for computing distances on discrete
surfaces. This yields two convex optimization problems for whose solution we
employ neural networks: We first compute a neural approximation of the
gradients of the unsigned distance field through a small time step of heat flow
with weighted point cloud densities as initial data. Then we use it to compute
a neural approximation of the SDF. We prove that the underlying variational
problems are well-posed. Through numerical experiments, we demonstrate that our
method provides state-of-the-art surface reconstruction and consistent SDF
gradients. Furthermore, we show in a proof-of-concept that it is accurate
enough for solving a PDE on the zero-level set.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 14:13:54 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Weidemaier",
"Samuel",
""
],
[
"Hartwig",
"Florine",
""
],
[
"Sassen",
"Josua",
""
],
[
"Conti",
"Sergio",
""
],
[
"Ben-Chen",
"Mirela",
""
],
[
"Rumpf",
"Martin",
""
]
] |
2504.11218 | Yang Liu | Zeming Wei, Junyi Lin, Yang Liu, Weixing Chen, Jingzhou Luo, Guanbin
Li, Liang Lin | 3DAffordSplat: Efficient Affordance Reasoning with 3D Gaussians | The first large-scale 3D Gaussians Affordance Reasoning Benchmark | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D affordance reasoning is essential in associating human instructions with
the functional regions of 3D objects, facilitating precise, task-oriented
manipulations in embodied AI. However, current methods, which predominantly
depend on sparse 3D point clouds, exhibit limited generalizability and
robustness due to their sensitivity to coordinate variations and the inherent
sparsity of the data. By contrast, 3D Gaussian Splatting (3DGS) delivers
high-fidelity, real-time rendering with minimal computational overhead by
representing scenes as dense, continuous distributions. This positions 3DGS as
a highly effective approach for capturing fine-grained affordance details and
improving recognition accuracy. Nevertheless, its full potential remains
largely untapped due to the absence of large-scale, 3DGS-specific affordance
datasets. To overcome these limitations, we present 3DAffordSplat, the first
large-scale, multi-modal dataset tailored for 3DGS-based affordance reasoning.
This dataset includes 23,677 Gaussian instances, 8,354 point cloud instances,
and 6,631 manually annotated affordance labels, encompassing 21 object
categories and 18 affordance types. Building upon this dataset, we introduce
AffordSplatNet, a novel model specifically designed for affordance reasoning
using 3DGS representations. AffordSplatNet features an innovative cross-modal
structure alignment module that exploits structural consistency priors to align
3D point cloud and 3DGS representations, resulting in enhanced affordance
recognition accuracy. Extensive experiments demonstrate that the 3DAffordSplat
dataset significantly advances affordance learning within the 3DGS domain,
while AffordSplatNet consistently outperforms existing methods across both seen
and unseen settings, highlighting its robust generalization capabilities.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 14:21:47 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 05:03:44 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wei",
"Zeming",
""
],
[
"Lin",
"Junyi",
""
],
[
"Liu",
"Yang",
""
],
[
"Chen",
"Weixing",
""
],
[
"Luo",
"Jingzhou",
""
],
[
"Li",
"Guanbin",
""
],
[
"Lin",
"Liang",
""
]
] |
2504.11224 | Burak Gurlek | Burak Gurlek, Shubham Sharma, Paolo Lazzaroni, Angel Rubio, Mariana
Rossi | Accurate Machine Learning Interatomic Potentials for Polyacene Molecular
Crystals: Application to Single Molecule Host-Guest Systems | null | null | null | null | cond-mat.mtrl-sci physics.chem-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Emerging machine learning interatomic potentials (MLIPs) offer a promising
solution for large-scale accurate material simulations, but stringent tests
related to the description of vibrational dynamics in molecular crystals remain
scarce. Here, we develop a general MLIP by leveraging the graph neural
network-based MACE architecture and active-learning strategies to accurately
capture vibrational dynamics across a range of polyacene-based molecular
crystals, namely naphthalene, anthracene, tetracene and pentacene. Through
careful error propagation, we show that these potentials are accurate and
enable the study of anharmonic vibrational features, vibrational lifetimes, and
vibrational coupling. In particular, we investigate large-scale host-guest
systems based on these molecular crystals, showing the capacity of
molecular-dynamics-based techniques to explain and quantify vibrational
coupling between host and guest nuclear motion. Our results establish a
framework for understanding vibrational signatures in large-scale complex
molecular systems and thus represent an important step for engineering
vibrational interactions in molecular environments.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 14:27:21 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Gurlek",
"Burak",
""
],
[
"Sharma",
"Shubham",
""
],
[
"Lazzaroni",
"Paolo",
""
],
[
"Rubio",
"Angel",
""
],
[
"Rossi",
"Mariana",
""
]
] |
2504.11261 | Hannes Petrenz | Hannes Petrenz, Johannes K\"ohler and Francesco Borrelli | Robust MPC for Uncertain Linear Systems -- Combining Model Adaptation
and Iterative Learning | Github link to the example:
https://github.com/HannesPetrenz/RALMPC_Linear_Uncertain_Systems | null | null | null | eess.SY cs.SY math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a robust adaptive learning Model Predictive Control (MPC)
framework for linear systems with parametric uncertainties and additive
disturbances performing iterative tasks. The approach iteratively refines the
parameter estimates using set membership estimation. Performance enhancement
over iterations is achieved by learning the terminal cost from data. Safety is
enforced using a terminal set, which is also learned iteratively. The proposed
method guarantees recursive feasibility, constraint satisfaction, and a robust
bound on the closed-loop cost. Numerical simulations on a mass-spring-damper
system demonstrate improved computational efficiency and control performance
compared to an existing robust adaptive MPC approach.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 15:00:34 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 06:16:36 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Petrenz",
"Hannes",
""
],
[
"Köhler",
"Johannes",
""
],
[
"Borrelli",
"Francesco",
""
]
] |
2504.11290 | Joshua Otten | Joshua Otten, Antonios Anastasopoulos, Kevin Moran | Automated Python Translation | 15 pages, 4 figures, 17 tables | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Python is one of the most commonly used programming languages in industry and
education. Its English keywords and built-in functions/modules allow it to come
close to pseudo-code in terms of its readability and ease of writing. However,
those who do not speak English may not experience these advantages. In fact,
they may even be hindered in their ability to understand Python code, as the
English nature of its terms creates an additional layer of overhead. To that
end, we introduce the task of automatically translating Python's natural
modality (keywords, error types, identifiers, etc.) into other human languages.
This presents a unique challenge, considering the abbreviated nature of these
forms, as well as potential untranslatability of advanced
mathematical/programming concepts across languages. We therefore create an
automated pipeline to translate Python into other human languages, comparing
strategies using machine translation and large language models. We then use
this pipeline to acquire translations from five common Python libraries
(pytorch, pandas, tensorflow, numpy, and random) in seven languages, and do a
quality test on a subset of these terms in French, Greek, and Bengali. We hope
this will provide a clearer path forward towards creating a universal Python,
accessible to anyone regardless of nationality or language background.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 15:30:22 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:49:21 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Otten",
"Joshua",
""
],
[
"Anastasopoulos",
"Antonios",
""
],
[
"Moran",
"Kevin",
""
]
] |
2504.11346 | Jie Wu | Yu Gao, Lixue Gong, Qiushan Guo, Xiaoxia Hou, Zhichao Lai, Fanshi Li,
Liang Li, Xiaochen Lian, Chao Liao, Liyang Liu, Wei Liu, Yichun Shi, Shiqi
Sun, Yu Tian, Zhi Tian, Peng Wang, Rui Wang, Xuanda Wang, Xun Wang, Ye Wang,
Guofeng Wu, Jie Wu, Xin Xia, Xuefeng Xiao, Zhonghua Zhai, Xinyu Zhang, Qi
Zhang, Yuwei Zhang, Shijia Zhao, Jianchao Yang, Weilin Huang | Seedream 3.0 Technical Report | Seedream 3.0 Technical Report | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present Seedream 3.0, a high-performance Chinese-English bilingual image
generation foundation model. We develop several technical improvements to
address existing challenges in Seedream 2.0, including alignment with
complicated prompts, fine-grained typography generation, suboptimal visual
aesthetics and fidelity, and limited image resolutions. Specifically, the
advancements of Seedream 3.0 stem from improvements across the entire pipeline,
from data construction to model deployment. At the data stratum, we double the
dataset using a defect-aware training paradigm and a dual-axis collaborative
data-sampling framework. Furthermore, we adopt several effective techniques
such as mixed-resolution training, cross-modality RoPE, representation
alignment loss, and resolution-aware timestep sampling in the pre-training
phase. During the post-training stage, we utilize diversified aesthetic
captions in SFT, and a VLM-based reward model with scaling, thereby achieving
outputs that well align with human preferences. Furthermore, Seedream 3.0
pioneers a novel acceleration paradigm. By employing consistent noise
expectation and importance-aware timestep sampling, we achieve a 4 to 8 times
speedup while maintaining image quality. Seedream 3.0 demonstrates significant
improvements over Seedream 2.0: it enhances overall capabilities, in particular
for text-rendering in complicated Chinese characters which is important to
professional typography generation. In addition, it provides native
high-resolution output (up to 2K), allowing it to generate images with high
visual quality.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 16:19:07 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 16:23:31 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Gao",
"Yu",
""
],
[
"Gong",
"Lixue",
""
],
[
"Guo",
"Qiushan",
""
],
[
"Hou",
"Xiaoxia",
""
],
[
"Lai",
"Zhichao",
""
],
[
"Li",
"Fanshi",
""
],
[
"Li",
"Liang",
""
],
[
"Lian",
"Xiaochen",
""
],
[
"Liao",
"Chao",
""
],
[
"Liu",
"Liyang",
""
],
[
"Liu",
"Wei",
""
],
[
"Shi",
"Yichun",
""
],
[
"Sun",
"Shiqi",
""
],
[
"Tian",
"Yu",
""
],
[
"Tian",
"Zhi",
""
],
[
"Wang",
"Peng",
""
],
[
"Wang",
"Rui",
""
],
[
"Wang",
"Xuanda",
""
],
[
"Wang",
"Xun",
""
],
[
"Wang",
"Ye",
""
],
[
"Wu",
"Guofeng",
""
],
[
"Wu",
"Jie",
""
],
[
"Xia",
"Xin",
""
],
[
"Xiao",
"Xuefeng",
""
],
[
"Zhai",
"Zhonghua",
""
],
[
"Zhang",
"Xinyu",
""
],
[
"Zhang",
"Qi",
""
],
[
"Zhang",
"Yuwei",
""
],
[
"Zhao",
"Shijia",
""
],
[
"Yang",
"Jianchao",
""
],
[
"Huang",
"Weilin",
""
]
] |
2504.11347 | Soyoung Yoo | Soyoung Yoo and Namwoo Kang | DeepWheel: Generating a 3D Synthetic Wheel Dataset for Design and
Performance Evaluation | 28 pages, 18 figures. Not yet submitted to a journal or conference | null | null | null | cs.CV physics.app-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Data-driven design is emerging as a powerful strategy to accelerate
engineering innovation. However, its application to vehicle wheel design
remains limited due to the lack of large-scale, high-quality datasets that
include 3D geometry and physical performance metrics. To address this gap, this
study proposes a synthetic design-performance dataset generation framework
using generative AI. The proposed framework first generates 2D rendered images
using Stable Diffusion, and then reconstructs the 3D geometry through 2.5D
depth estimation. Structural simulations are subsequently performed to extract
engineering performance data. To further expand the design and performance
space, topology optimization is applied, enabling the generation of a more
diverse set of wheel designs. The final dataset, named DeepWheel, consists of
over 6,000 photo-realistic images and 900 structurally analyzed 3D models. This
multi-modal dataset serves as a valuable resource for surrogate model training,
data-driven inverse design, and design space exploration. The proposed
methodology is also applicable to other complex design domains. The dataset is
released under the Creative Commons Attribution-NonCommercial 4.0
International(CC BY-NC 4.0) and is available on the
https://www.smartdesignlab.org/datasets
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 16:20:00 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 04:26:29 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Yoo",
"Soyoung",
""
],
[
"Kang",
"Namwoo",
""
]
] |
2504.11383 | Maryam Hakimzadeh | Wei Wang, Maryam Hakimzadeh, Haihui Ruan, Somdatta Goswami | Accelerating Multiscale Modeling with Hybrid Solvers: Coupling FEM and
Neural Operators with Domain Decomposition | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Numerical solvers for partial differential equations (PDEs) face challenges
balancing computational cost and accuracy, especially in multiscale and dynamic
systems. Neural operators can significantly speed up simulations; however, they
often face challenges such as error accumulation and limited generalization in
multiphysics problems. This work introduces a novel hybrid framework that
integrates physics-informed DeepONet with FEM through domain decomposition. The
core innovation lies in adaptively coupling FEM and DeepONet subdomains via a
Schwarz alternating method. This methodology strategically allocates
computationally demanding regions to a pre-trained Deep Operator Network, while
the remaining computational domain is solved through FEM. To address dynamic
systems, we integrate the Newmark time-stepping scheme directly into the
DeepONet, significantly mitigating error accumulation in long-term simulations.
Furthermore, an adaptive subdomain evolution enables the ML-resolved region to
expand dynamically, capturing emerging fine-scale features without remeshing.
The framework's efficacy has been validated across a range of solid mechanics
problems, including static, quasi-static, and dynamic regimes, demonstrating
accelerated convergence rates (up to 20% improvement compared to FE-FE
approaches), while preserving solution fidelity with error < 1%. Our case
studies show that our proposed hybrid solver: (1) maintains solution continuity
across subdomain interfaces, (2) reduces computational costs by eliminating
fine mesh requirements, (3) mitigates error accumulation in time-dependent
simulations, and (4) enables automatic adaptation to evolving physical
phenomena. This work bridges the gap between numerical methods and AI-driven
surrogates, offering a scalable pathway for high-fidelity simulations in
engineering and scientific applications.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 16:54:04 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 12:26:48 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Wang",
"Wei",
""
],
[
"Hakimzadeh",
"Maryam",
""
],
[
"Ruan",
"Haihui",
""
],
[
"Goswami",
"Somdatta",
""
]
] |
2504.11396 | Longxiu Huang | HanQin Cai and Longxiu Huang | Property Inheritance for Subtensors in Tensor Train Decompositions | 2025 IEEE International Symposium on Information Theory (ISIT 2025) | IEEE International Symposium on Information Theory, 2025 | null | null | cs.IT math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tensor dimensionality reduction is one of the fundamental tools for modern
data science. To address the high computational overhead, fiber-wise sampled
subtensors that preserve the original tensor rank are often used in designing
efficient and scalable tensor dimensionality reduction. However, the theory of
property inheritance for subtensors is still underdevelopment, that is, how the
essential properties of the original tensor will be passed to its subtensors.
This paper theoretically studies the property inheritance of the two key tensor
properties, namely incoherence and condition number, under the tensor train
setting. We also show how tensor train rank is preserved through fiber-wise
sampling. The key parameters introduced in theorems are numerically evaluated
under various settings. The results show that the properties of interest can be
well preserved to the subtensors formed via fiber-wise sampling. Overall, this
paper provides several handy analytic tools for developing efficient tensor
analysis
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 17:10:38 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Cai",
"HanQin",
""
],
[
"Huang",
"Longxiu",
""
]
] |
2504.11445 | Hengkai Ding | Hengkai Ding, Yiping Shu, Yun Chen, Nan Li, Xiaoyue Cao, James
Nightingale, Qiuhan He, Lei Wang and Hui Li | Pixel-level modelling of group-scale strong lens CASSOWARY 19 | 25 pages, 9 figures, Submitted to the Research in Astronomy and
Astrophysics (RAA) | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present the first high-precision model for the group-scale strong lensing
system CASSOWARY 19 (CSWA19), utilising images from the Hubble Space Telescope
(HST). Sixteen member galaxies identified via the red-sequence method, and the
main halo, all modelled as the dual Pseudo Isothermal Elliptical profile
(dPIE), are incorporated into a parametric lens model alongside an external
shear field. To model the system, we adopt the PyAutoLens software package,
employing a progressive search chain strategy for realizing the transition of
source model from multiple S\'ersic profiles to a brightness-adaptive
pixelization, which uses 1000 pixels in the source plane to reconstruct the
background source corresponding to 177,144 image pixels in the image plane. Our
results indicate that the total mass within the Einstein radius is
$M_{\theta_\mathrm{E}}$ $\approx 1.41\times10^{13}$M$_{\odot}$ and the average
slope of the total mass density $\rho (r)\propto r^{-\gamma}$ is
$\tilde{\gamma}=1.33$ within the effective radius. This slope is shallower than
those measured in galaxies and groups but is closer to those of galaxy
clusters. In addition, our approach successfully resolves the two merging
galaxies in the background source and yields a total magnification of
$\mu=103.18^{+0.23}_{-0.19}$, which is significantly higher than the outcomes
from previous studies of CSWA19. In summary, our research demonstrates the
effectiveness of the brightness-adaptive pixelization source reconstruction
technique for modelling group-scale strong lensing systems. It can serve as a
technical reference for future investigations into pixel-level modelling of the
group- and cluster-scale strong lensing systems.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 17:56:11 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 12:26:12 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Ding",
"Hengkai",
""
],
[
"Shu",
"Yiping",
""
],
[
"Chen",
"Yun",
""
],
[
"Li",
"Nan",
""
],
[
"Cao",
"Xiaoyue",
""
],
[
"Nightingale",
"James",
""
],
[
"He",
"Qiuhan",
""
],
[
"Wang",
"Lei",
""
],
[
"Li",
"Hui",
""
]
] |
2504.11447 | Shengyuan Zhang | An Zhao, Shengyuan Zhang, Ling Yang, Zejian Li, Jiale Wu, Haoran Xu,
AnYang Wei, Perry Pengyun GU, Lingyun Sun | Diffusion Distillation With Direct Preference Optimization For Efficient
3D LiDAR Scene Completion | Our code is public available on
https://github.com/happyw1nd/DistillationDPO | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The application of diffusion models in 3D LiDAR scene completion is limited
due to diffusion's slow sampling speed. Score distillation accelerates
diffusion sampling but with performance degradation, while post-training with
direct policy optimization (DPO) boosts performance using preference data. This
paper proposes Distillation-DPO, a novel diffusion distillation framework for
LiDAR scene completion with preference aligment. First, the student model
generates paired completion scenes with different initial noises. Second, using
LiDAR scene evaluation metrics as preference, we construct winning and losing
sample pairs. Such construction is reasonable, since most LiDAR scene metrics
are informative but non-differentiable to be optimized directly. Third,
Distillation-DPO optimizes the student model by exploiting the difference in
score functions between the teacher and student models on the paired completion
scenes. Such procedure is repeated until convergence. Extensive experiments
demonstrate that, compared to state-of-the-art LiDAR scene completion diffusion
models, Distillation-DPO achieves higher-quality scene completion while
accelerating the completion speed by more than 5-fold. Our method is the first
to explore adopting preference learning in distillation to the best of our
knowledge and provide insights into preference-aligned distillation. Our code
is public available on https://github.com/happyw1nd/DistillationDPO.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 17:57:13 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 02:02:05 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhao",
"An",
""
],
[
"Zhang",
"Shengyuan",
""
],
[
"Yang",
"Ling",
""
],
[
"Li",
"Zejian",
""
],
[
"Wu",
"Jiale",
""
],
[
"Xu",
"Haoran",
""
],
[
"Wei",
"AnYang",
""
],
[
"GU",
"Perry Pengyun",
""
],
[
"Sun",
"Lingyun",
""
]
] |
2504.11454 | ChengYen Hsieh | Cheng-Yen Hsieh, Xinyou Wang, Daiheng Zhang, Dongyu Xue, Fei Ye,
Shujian Huang, Zaixiang Zheng, Quanquan Gu | Elucidating the Design Space of Multimodal Protein Language Models | Project Page: https://bytedance.github.io/dplm/dplm-2.1/ | null | null | null | cs.LG cs.AI q-bio.QM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Multimodal protein language models (PLMs) integrate sequence and token-based
structural information, serving as a powerful foundation for protein modeling,
generation, and design. However, the reliance on tokenizing 3D structures into
discrete tokens causes substantial loss of fidelity about fine-grained
structural details and correlations. In this paper, we systematically elucidate
the design space of multimodal PLMs to overcome their limitations. We identify
tokenization loss and inaccurate structure token predictions by the PLMs as
major bottlenecks. To address these, our proposed design space covers improved
generative modeling, structure-aware architectures and representation learning,
and data exploration. Our advancements approach finer-grained supervision,
demonstrating that token-based multimodal PLMs can achieve robust structural
modeling. The effective design methods dramatically improve the structure
generation diversity, and notably, folding abilities of our 650M model by
reducing the RMSD from 5.52 to 2.36 on PDB testset, even outperforming 3B
baselines and on par with the specialized folding models.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 17:59:43 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 02:35:11 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Hsieh",
"Cheng-Yen",
""
],
[
"Wang",
"Xinyou",
""
],
[
"Zhang",
"Daiheng",
""
],
[
"Xue",
"Dongyu",
""
],
[
"Ye",
"Fei",
""
],
[
"Huang",
"Shujian",
""
],
[
"Zheng",
"Zaixiang",
""
],
[
"Gu",
"Quanquan",
""
]
] |
2504.11459 | Peter Stockinger | Peter Stockinger (PLIDAM, ESCOM) | From Conceptual Data Models to Multimodal Representation | in French language | null | null | null | cs.AI cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 1) Introduction and Conceptual Framework: This document explores the concept
of information design by dividing it into two major practices: defining the
meaning of a corpus of textual data and its visual or multimodal
representation. It draws on expertise in enriching textual corpora,
particularly audiovisual ones, and transforming them into multiple narrative
formats. The text highlights a crucial distinction between the semantic content
of a domain and the modalities of its graphic expression, illustrating this
approach with concepts rooted in structural semiotics and linguistics
traditions.
2) Modeling and Conceptual Design: The article emphasizes the importance of
semantic modeling, often achieved through conceptual networks or graphs. These
tools enable the structuring of knowledge within a domain by accounting for
relationships between concepts, contexts of use, and specific objectives.
Stockinger also highlights the constraints and challenges involved in creating
dynamic and adaptable models, integrating elements such as thesauri or
interoperable ontologies to facilitate the analysis and publication of complex
corpora.
3) Applications and Multimodal Visualization: The text concludes by examining
the practical application of these models in work environments like OKAPI,
developed to analyze, publish, and reuse audiovisual data. It also discusses
innovative approaches such as visual storytelling and document reengineering,
which involve transforming existing content into new resources tailored to
various contexts. These methods emphasize interoperability, flexibility, and
the intelligence of communication systems, paving the way for richer and more
collaborative use of digital data. The content of this document was presented
during the "Semiotics of Information Design" Day organized by Anne
Beyaert-Geslin of the University of Bordeaux Montaigne (MICA laboratory) on
June 21, 2018, in Bordeaux.
| [
{
"version": "v1",
"created": "Wed, 8 Jan 2025 09:15:01 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Stockinger",
"Peter",
"",
"PLIDAM, ESCOM"
]
] |
2504.11461 | Stefan Forcey | Stefan Forcey | Counting plane arrangements via oriented matroids | 22 pages, 13 figures | null | null | null | math.HO math.CO | http://creativecommons.org/licenses/by/4.0/ | Planes are familiar mathematical objects which lie at the subtle boundary
between continuous geometry and discrete combinatorics. A plane is geometrical,
certainly, but the ways that two planes can interact break cleanly into
discrete sets: the planes can intersect or not. Here we review how oriented
matroids can be used to try to capture the combinatorial aspect, giving a way
to encode with finite sets all the ways that $n$ planes can interact. We
mention how the one-to-one correspondence breaks down in 2 dimensions for 9
lines, and in 3D for 8 planes. We include illustrations of all the types of
plane arrangements using $n=4$ and 5.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 00:09:37 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Forcey",
"Stefan",
""
]
] |
2504.11462 | Salomon Ofman | Luc Brisson, Salomon Ofman (IMJ-PRG) | Timaeus' Puzzle of the Innumerable Interstices in a Universe Without
Void | Ancient Philosophy, In press | null | null | null | math.HO physics.hist-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Some of the most challenging problems in Timaeus' cosmology arise from the
geometry of a universe without any void. On the one hand, the universe is
spherical in shape; on the other hand, it must be entirely filled with the four
basic particles that make up all bodies in the universe, each shaped like one
of four regular polyhedra (cubes, tetrahedra, octahedra and icosahedra). The
faces of all these particles are composed of right triangles. However, this
leads to two mathematical impossibilities. 1. Obtaining a spherical surface
from linear surfaces, as it is impossible to create a circle from straight
lines. 2. Obtaining a complete tiling of a sphere using regular polyhedra,
without any voids or intersections between these polyhedra. The first problem
is addressed in another article slated for publication. In the present one, our
focus will be on the second problem, for which we will present a solution
within the framework of Timaeus' cosmology. The crux of this solution lies in a
feature of Timaeus' universe that sets it apart from almost all ancient
cosmologies. Instead of being composed of rigid parts, it is a dynamic living
body in which all basic components are in constant motion, continuously
undergoing both destruction and reconstruction. In the first part, we examine
the main features of Timaeus' cosmology relevant to our issue. In the second
part, we analyze the paradox in details and its consequences for Timaeus'
cosmology. We then discuss the common 'solutions' and highlight their
shortcomings. Finally, we propose a solution that we believe is consistent with
Plato's text and independent of the choice of the major schools of
interpretation of the Timaeus.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 13:57:49 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Brisson",
"Luc",
"",
"IMJ-PRG"
],
[
"Ofman",
"Salomon",
"",
"IMJ-PRG"
]
] |
2504.11463 | Jan-Christoph Schlage-Puchta | Jan-Christoph Schlage-Puchta | On a Problem by Erd\H{o}s and Mirsky on the ratio of the number of
divisors of consecutive integers | null | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $\mathcal{L}$ be the closure of the set of all real numbers $\alpha$,
such that there exist infinitely many integers $n$, such that
$\alpha=\log\frac{d(n+1)}{d(n)}$, where $d$ is the number of divisors of $n$.
We give improved lower bounds for the density of $\mathcal{L}$.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 19:03:39 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Schlage-Puchta",
"Jan-Christoph",
""
]
] |
2504.11464 | Victor Guo | Li Lu, Lingyu Guo, Victor Z. Guo | Improvements on exponential sums related to Piatetski-Shapiro primes | null | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | We prove a new bound to the exponential sum of the form $$ \sum_{h \sim
H}\delta_h \mathop{\sum_{m\sim M}\sum_{n\sim N}}_{mn\sim
x}a_{m}b_{n}\e\big(\alpha mn + h(mn + u)^{\gamma}\big), $$ by a new approach to
the Type I sum. The sum can be applied to many problems related to
Piatetski-Shapiro primes, which are primes of the form $\lfloor n^c \rfloor$.
In this paper, we improve the admissible range of the Balog-Friedlander
condition, which leads to an improvement to the ternary Goldbach problem with
Piatetski-Shapiro primes. We also investigate the distribution of
Piatetski-Shapiro primes in arithmetic progressions, Piatetski-Shapiro primes
in the intersection of multiple Beatty sequences and so on.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 02:56:41 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Lu",
"Li",
""
],
[
"Guo",
"Lingyu",
""
],
[
"Guo",
"Victor Z.",
""
]
] |
2504.11465 | Rajesh Dachiraju | Rajesh Dachiraju | Fourier Analysis: A New Result | null | null | null | null | math.GM | http://creativecommons.org/licenses/by/4.0/ | This article contains a new result in Fourier analysis concerning jump type
discontinuities.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 09:27:53 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Dachiraju",
"Rajesh",
""
]
] |
2504.11466 | Anna Nelson | Anna C. Nelson, Scott A. McKinley, Melissa M. Rolls, Maria-Veronica
Ciocanel | Emergent microtubule properties in a model of filament turnover and
nucleation | 20 pages, 5 figures | null | null | null | physics.bio-ph q-bio.SC | http://creativecommons.org/licenses/by/4.0/ | Microtubules (MTs) are dynamic protein filaments essential for intracellular
organization and transport, particularly in long-lived cells such as neurons.
The plus and minus ends of neuronal MTs switch between growth and shrinking
phases, and the nucleation of new filaments is believed to be regulated in both
healthy and injury conditions. We propose stochastic and deterministic
mathematical models to investigate the impact of filament nucleation and
length-regulation mechanisms on emergent properties such as MT lengths and
numbers in living cells. We expand our stochastic continuous-time Markov chain
model of filament dynamics to incorporate MT nucleation and capture realistic
stochastic fluctuations in MT numbers and tubulin availability. We also propose
a simplified partial differential equation (PDE) model, which allows for
tractable analytical investigation into steady-state MT distributions under
different nucleation and length-regulating mechanisms. We find that the
stochastic and PDE modeling approaches show good agreement in predicted MT
length distributions, and that both MT nucleation and the catastrophe of
large-length MTs regulate MT length distributions. In both frameworks, multiple
mechanistic combinations achieve the same average MT length. The models
proposed can predict parameter regimes where the system is scarce in tubulin,
the building block of MTs, and suggest that low filament nucleation regimes are
characterized by high variation in MT lengths, while high nucleation regimes
drive high variation in MT numbers. These mathematical frameworks have the
potential to improve our understanding of MT regulation in both healthy and
injured neurons.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 16:00:02 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Nelson",
"Anna C.",
""
],
[
"McKinley",
"Scott A.",
""
],
[
"Rolls",
"Melissa M.",
""
],
[
"Ciocanel",
"Maria-Veronica",
""
]
] |
2504.11467 | Qianxue Zhang | Qianxue Zhang and Eiman Kanjo | MultiCore+TPU Accelerated Multi-Modal TinyML for Livestock Behaviour
Recognition | 11 pages, 10 figures | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The advancement of technology has revolutionised the agricultural industry,
transitioning it from labour-intensive farming practices to automated,
AI-powered management systems. In recent years, more intelligent livestock
monitoring solutions have been proposed to enhance farming efficiency and
productivity. This work presents a novel approach to animal activity
recognition and movement tracking, leveraging tiny machine learning (TinyML)
techniques, wireless communication framework, and microcontroller platforms to
develop an efficient, cost-effective livestock sensing system. It collects and
fuses accelerometer data and vision inputs to build a multi-modal network for
three tasks: image classification, object detection, and behaviour recognition.
The system is deployed and evaluated on commercial microcontrollers for
real-time inference using embedded applications, demonstrating up to
270$\times$ model size reduction, less than 80ms response latency, and on-par
performance comparable to existing methods. The incorporation of the TinyML
technique allows for seamless data transmission between devices, benefiting use
cases in remote locations with poor Internet connectivity. This work delivers a
robust, scalable IoT-edge livestock monitoring solution adaptable to diverse
farming needs, offering flexibility for future extensions.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 12:55:08 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Qianxue",
""
],
[
"Kanjo",
"Eiman",
""
]
] |
2504.11468 | Guiming Hardy Chen | Hardy Chen, Haoqin Tu, Fali Wang, Hui Liu, Xianfeng Tang, Xinya Du,
Yuyin Zhou, Cihang Xie | SFT or RL? An Early Investigation into Training R1-Like Reasoning Large
Vision-Language Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | This work revisits the dominant supervised fine-tuning (SFT) then
reinforcement learning (RL) paradigm for training Large Vision-Language Models
(LVLMs), and reveals a key finding: SFT can significantly undermine subsequent
RL by inducing ``pseudo reasoning paths'' imitated from expert models. While
these paths may resemble the native reasoning paths of RL models, they often
involve prolonged, hesitant, less informative steps, and incorrect reasoning.
To systematically study this effect, we introduce VLAA-Thinking, a new
multimodal dataset designed to support reasoning in LVLMs. Constructed via a
six-step pipeline involving captioning, reasoning distillation, answer rewrite
and verification, VLAA-Thinking comprises high-quality, step-by-step visual
reasoning traces for SFT, along with a more challenging RL split from the same
data source. Using this dataset, we conduct extensive experiments comparing
SFT, RL and their combinations. Results show that while SFT helps models learn
reasoning formats, it often locks aligned models into imitative, rigid
reasoning modes that impede further learning. In contrast, building on the
Group Relative Policy Optimization (GRPO) with a novel mixed reward module
integrating both perception and cognition signals, our RL approach fosters more
genuine, adaptive reasoning behavior. Notably, our model VLAA-Thinker, based on
Qwen2.5VL 3B, achieves top-1 performance on Open LMM Reasoning Leaderboard
(https://huggingface.co/spaces/opencompass/Open_LMM_Reasoning_Leaderboard)
among 4B scale LVLMs, surpassing the previous state-of-the-art by 1.8%. We hope
our findings provide valuable insights in developing reasoning-capable LVLMs
and can inform future research in this area.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 16:54:05 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Chen",
"Hardy",
""
],
[
"Tu",
"Haoqin",
""
],
[
"Wang",
"Fali",
""
],
[
"Liu",
"Hui",
""
],
[
"Tang",
"Xianfeng",
""
],
[
"Du",
"Xinya",
""
],
[
"Zhou",
"Yuyin",
""
],
[
"Xie",
"Cihang",
""
]
] |
2504.11469 | Guillaume Garret | Guillaume Garret, Antoine Vacavant, Carole Frindel | Do Segmentation Models Understand Vascular Structure? A Blob-Based XAI
Framework | Open access version of an article submitted to Medical Image
Understanding and Analysis (MIUA) 2025 | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | Deep learning models have achieved impressive performance in medical image
segmentation, yet their black-box nature limits clinical adoption. In vascular
applications, trustworthy segmentation should rely on both local image cues and
global anatomical structures, such as vessel connectivity or branching.
However, the extent to which models leverage such global context remains
unclear. We present a novel explainability pipeline for 3D vessel segmentation,
combining gradient-based attribution with graph-guided point selection and a
blob-based analysis of Saliency maps. Using vascular graphs extracted from
ground truth, we define anatomically meaningful points of interest (POIs) and
assess the contribution of input voxels via Saliency maps. These are analyzed
at both global and local scales using a custom blob detector. Applied to IRCAD
and Bullitt datasets, our analysis shows that model decisions are dominated by
highly localized attribution blobs centered near POIs. Attribution features
show little correlation with vessel-level properties such as thickness,
tubularity, or connectivity -- suggesting limited use of global anatomical
reasoning. Our results underline the importance of structured explainability
tools and highlight the current limitations of segmentation models in capturing
global vascular context.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 12:42:52 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Garret",
"Guillaume",
""
],
[
"Vacavant",
"Antoine",
""
],
[
"Frindel",
"Carole",
""
]
] |
2504.11470 | Huaxiang Zhang | Huaxiang Zhang, Hao Zhang, Aoran Mei, Zhongxue Gan, Guo-Niu Zhu | SO-DETR: Leveraging Dual-Domain Features and Knowledge Distillation for
Small Object Detection | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Detection Transformer-based methods have achieved significant advancements in
general object detection. However, challenges remain in effectively detecting
small objects. One key difficulty is that existing encoders struggle to
efficiently fuse low-level features. Additionally, the query selection
strategies are not effectively tailored for small objects. To address these
challenges, this paper proposes an efficient model, Small Object Detection
Transformer (SO-DETR). The model comprises three key components: a dual-domain
hybrid encoder, an enhanced query selection mechanism, and a knowledge
distillation strategy. The dual-domain hybrid encoder integrates spatial and
frequency domains to fuse multi-scale features effectively. This approach
enhances the representation of high-resolution features while maintaining
relatively low computational overhead. The enhanced query selection mechanism
optimizes query initialization by dynamically selecting high-scoring anchor
boxes using expanded IoU, thereby improving the allocation of query resources.
Furthermore, by incorporating a lightweight backbone network and implementing a
knowledge distillation strategy, we develop an efficient detector for small
objects. Experimental results on the VisDrone-2019-DET and UAVVaste datasets
demonstrate that SO-DETR outperforms existing methods with similar
computational demands. The project page is available at
https://github.com/ValiantDiligent/SO_DETR.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 13:47:37 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Huaxiang",
""
],
[
"Zhang",
"Hao",
""
],
[
"Mei",
"Aoran",
""
],
[
"Gan",
"Zhongxue",
""
],
[
"Zhu",
"Guo-Niu",
""
]
] |
2504.11471 | Qingyao Zhang | Qingyao Zhang | Vaidya-Reissner-Nordstr\"om Extension On the White-hole Region and the
White Hole Evaporation Dynamic | 30 pages, 2 figures | null | null | null | gr-qc math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop an analytic model that extends classical white hole geometry by
incorporating both radiative dynamics and electric charge. Starting from a
maximal analytic extension of the Schwarzschild white hole via Kruskal Szekeres
coordinates, we introduce a time dependent mass function, representative of
outgoing null dust to model evaporation. Building on this foundation, the study
then integrates the Reissner-Nordstr\"om framework to obtain a dynamic, charged
white hole solution in double null coordinates. In the resulting Vaidya
Reissner Nordstr\"om metric, both the Bondi mass and the associated charge
decrease monotonically with retarded time, capturing the interplay of radiation
and electromagnetic effects. Detailed analysis of horizon behavior reveals how
mass loss and charge shedding modify the causal structure, ensuring that energy
conditions are preserved and cosmic censorship is maintained.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 18:10:05 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Qingyao",
""
]
] |
2504.11472 | Kebin Contreras MSc | Kebin Contreras, Brayan Monroy, and Jorge Bacca | High Dynamic Range Modulo Imaging for Robust Object Detection in
Autonomous Driving | null | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Object detection precision is crucial for ensuring the safety and efficacy of
autonomous driving systems. The quality of acquired images directly influences
the ability of autonomous driving systems to correctly recognize and respond to
other vehicles, pedestrians, and obstacles in real-time. However, real
environments present extreme variations in lighting, causing saturation
problems and resulting in the loss of crucial details for detection.
Traditionally, High Dynamic Range (HDR) images have been preferred for their
ability to capture a broad spectrum of light intensities, but the need for
multiple captures to construct HDR images is inefficient for real-time
applications in autonomous vehicles. To address these issues, this work
introduces the use of modulo sensors for robust object detection. The modulo
sensor allows pixels to `reset/wrap' upon reaching saturation level by
acquiring an irradiance encoding image which can then be recovered using
unwrapping algorithms. The applied reconstruction techniques enable HDR
recovery of color intensity and image details, ensuring better visual quality
even under extreme lighting conditions at the cost of extra time. Experiments
with the YOLOv10 model demonstrate that images processed using modulo images
achieve performance comparable to HDR images and significantly surpass
saturated images in terms of object detection accuracy. Moreover, the proposed
modulo imaging step combined with HDR image reconstruction is shorter than the
time required for conventional HDR image acquisition.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 21:35:26 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Contreras",
"Kebin",
""
],
[
"Monroy",
"Brayan",
""
],
[
"Bacca",
"Jorge",
""
]
] |
2504.11473 | Aida Ramezani | Warren Zhu, Aida Ramezani, Yang Xu | Visual moral inference and communication | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Humans can make moral inferences from multiple sources of input. In contrast,
automated moral inference in artificial intelligence typically relies on
language models with textual input. However, morality is conveyed through
modalities beyond language. We present a computational framework that supports
moral inference from natural images, demonstrated in two related tasks: 1)
inferring human moral judgment toward visual images and 2) analyzing patterns
in moral content communicated via images from public news. We find that models
based on text alone cannot capture the fine-grained human moral judgment toward
visual stimuli, but language-vision fusion models offer better precision in
visual moral inference. Furthermore, applications of our framework to news data
reveal implicit biases in news categories and geopolitical discussions. Our
work creates avenues for automating visual moral inference and discovering
patterns of visual moral communication in public media.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 00:46:27 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhu",
"Warren",
""
],
[
"Ramezani",
"Aida",
""
],
[
"Xu",
"Yang",
""
]
] |
2504.11474 | Younghun Kwon | Byunggun Kim and Younghun Kwon | Local Temporal Feature Enhanced Transformer with ROI-rank Based Masking
for Diagnosis of ADHD | null | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | In modern society, Attention-Deficit/Hyperactivity Disorder (ADHD) is one of
the common mental diseases discovered not only in children but also in adults.
In this context, we propose a ADHD diagnosis transformer model that can
effectively simultaneously find important brain spatiotemporal biomarkers from
resting-state functional magnetic resonance (rs-fMRI). This model not only
learns spatiotemporal individual features but also learns the correlation with
full attention structures specialized in ADHD diagnosis. In particular, it
focuses on learning local blood oxygenation level dependent (BOLD) signals and
distinguishing important regions of interest (ROI) in the brain. Specifically,
the three proposed methods for ADHD diagnosis transformer are as follows.
First, we design a CNN-based embedding block to obtain more expressive
embedding features in brain region attention. It is reconstructed based on the
previously CNN-based ADHD diagnosis models for the transformer. Next, for
individual spatiotemporal feature attention, we change the attention method to
local temporal attention and ROI-rank based masking. For the temporal features
of fMRI, the local temporal attention enables to learn local BOLD signal
features with only simple window masking. For the spatial feature of fMRI,
ROI-rank based masking can distinguish ROIs with high correlation in ROI
relationships based on attention scores, thereby providing a more specific
biomarker for ADHD diagnosis. The experiment was conducted with various types
of transformer models. To evaluate these models, we collected the data from 939
individuals from all sites provided by the ADHD-200 competition. Through this,
the spatiotemporal enhanced transformer for ADHD diagnosis outperforms the
performance of other different types of transformer variants. (77.78ACC
76.60SPE 79.22SEN 79.30AUC)
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 02:03:35 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kim",
"Byunggun",
""
],
[
"Kwon",
"Younghun",
""
]
] |
2504.11475 | Punit Kumar | Priya Mishra and Punit Kumar | A relativistic model for Quantum Plasmas | 9 pages , 0 figures | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new model to study the dynamics of relativistic quantum plasmas using the
quantum electrodynamical (QED) approach has been constructed to analyze the
quantum effects, relativistic corrections, and electromagnetic interactions.
Considering the covariant Lagrangian function and EulerLagrange equation, the
equations of motion have been established describing the interaction of strong
electromagnetic waves in plasma. These equations of motion constitute a model
for the propagation of relativistic laser pulse through high density quantum
plasma. Our model specifically takes the effects of four spin and four velocity
into account during the interaction process. This model is applicable to high
density plasmas in all ranges of electromagnetic fields which includes
astrophysical environments, high power laser plasma interactions, etc.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 04:51:46 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Mishra",
"Priya",
""
],
[
"Kumar",
"Punit",
""
]
] |
2504.11476 | M Tanveer PhD | Ritik Mishra, Mushir Akhtar, M. Tanveer | CI-RKM: A Class-Informed Approach to Robust Restricted Kernel Machines | Accepted in International Joint Conference on Neural Networks (IJCNN)
2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Restricted kernel machines (RKMs) represent a versatile and powerful
framework within the kernel machine family, leveraging conjugate feature
duality to address a wide range of machine learning tasks, including
classification, regression, and feature learning. However, their performance
can degrade significantly in the presence of noise and outliers, which
compromises robustness and predictive accuracy. In this paper, we propose a
novel enhancement to the RKM framework by integrating a class-informed weighted
function. This weighting mechanism dynamically adjusts the contribution of
individual training points based on their proximity to class centers and
class-specific characteristics, thereby mitigating the adverse effects of noisy
and outlier data. By incorporating weighted conjugate feature duality and
leveraging the Schur complement theorem, we introduce the class-informed
restricted kernel machine (CI-RKM), a robust extension of the RKM designed to
improve generalization and resilience to data imperfections. Experimental
evaluations on benchmark datasets demonstrate that the proposed CI-RKM
consistently outperforms existing baselines, achieving superior classification
accuracy and enhanced robustness against noise and outliers. Our proposed
method establishes a significant advancement in the development of kernel-based
learning models, addressing a core challenge in the field.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 11:12:30 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Mishra",
"Ritik",
""
],
[
"Akhtar",
"Mushir",
""
],
[
"Tanveer",
"M.",
""
]
] |
2504.11477 | Zhang Yunkai | Yunkai Zhang, Shiyin Wei, Yong Huang, Yawu Su, Shanshan Lu, Hui Li | SDIGLM: Leveraging Large Language Models and Multi-Modal Chain of
Thought for Structural Damage Identification | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing computer vision(CV)-based structural damage identification models
demonstrate notable accuracy in categorizing and localizing damage. However,
these models present several critical limitations that hinder their practical
application in civil engineering(CE). Primarily, their ability to recognize
damage types remains constrained, preventing comprehensive analysis of the
highly varied and complex conditions encountered in real-world CE structures.
Second, these models lack linguistic capabilities, rendering them unable to
articulate structural damage characteristics through natural language
descriptions. With the continuous advancement of artificial intelligence(AI),
large multi-modal models(LMMs) have emerged as a transformative solution,
enabling the unified encoding and alignment of textual and visual data. These
models can autonomously generate detailed descriptive narratives of structural
damage while demonstrating robust generalization across diverse scenarios and
tasks. This study introduces SDIGLM, an innovative LMM for structural damage
identification, developed based on the open-source VisualGLM-6B architecture.
To address the challenge of adapting LMMs to the intricate and varied operating
conditions in CE, this work integrates a U-Net-based semantic segmentation
module to generate defect segmentation maps as visual Chain of Thought(CoT).
Additionally, a multi-round dialogue fine-tuning dataset is constructed to
enhance logical reasoning, complemented by a language CoT formed through prompt
engineering. By leveraging this multi-modal CoT, SDIGLM surpasses
general-purpose LMMs in structural damage identification, achieving an accuracy
of 95.24% across various infrastructure types. Moreover, the model effectively
describes damage characteristics such as hole size, crack direction, and
corrosion severity.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 11:37:10 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Zhang",
"Yunkai",
""
],
[
"Wei",
"Shiyin",
""
],
[
"Huang",
"Yong",
""
],
[
"Su",
"Yawu",
""
],
[
"Lu",
"Shanshan",
""
],
[
"Li",
"Hui",
""
]
] |
2504.11478 | Hao Kang | Hao Kang, Stathi Fotiadis, Liming Jiang, Qing Yan, Yumin Jia, Zichuan
Liu, Min Jin Chong, Xin Lu | Flux Already Knows - Activating Subject-Driven Image Generation without
Training | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | We propose a simple yet effective zero-shot framework for subject-driven
image generation using a vanilla Flux model. By framing the task as grid-based
image completion and simply replicating the subject image(s) in a mosaic
layout, we activate strong identity-preserving capabilities without any
additional data, training, or inference-time fine-tuning. This "free lunch"
approach is further strengthened by a novel cascade attention design and meta
prompting technique, boosting fidelity and versatility. Experimental results
show that our method outperforms baselines across multiple key metrics in
benchmarks and human preference studies, with trade-offs in certain aspects.
Additionally, it supports diverse edits, including logo insertion, virtual
try-on, and subject replacement or insertion. These results demonstrate that a
pre-trained foundational text-to-image model can enable high-quality,
resource-efficient subject-driven generation, opening new possibilities for
lightweight customization in downstream applications.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 20:41:53 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Kang",
"Hao",
""
],
[
"Fotiadis",
"Stathi",
""
],
[
"Jiang",
"Liming",
""
],
[
"Yan",
"Qing",
""
],
[
"Jia",
"Yumin",
""
],
[
"Liu",
"Zichuan",
""
],
[
"Chong",
"Min Jin",
""
],
[
"Lu",
"Xin",
""
]
] |
2504.11479 | Pengfei Li | Li Pengfei (1 and 2), Liu Wanqi (1), Ha Shuai (2), Pan Yuzhou (2), Fan
Xuhong (2), Du Zhanhui (2), Wan Chengliang (2), Cui Ying (2), Yao Ke (3 and
4), Ma Yue (5), Yang Zhihu (6), Shao Caojie (6), Reinhold Schuch (7), Lu Di
(8), Song Yushou (1), Zhang Hongqiang (2), Chen Ximeng (2), ((1) College of
Nuclear Science and Technology, Harbin Engineering University, Harbin, China,
(2) School of Nuclear Science and Technology, Lanzhou University, Lanzhou,
China, (3) Institute of Modern Physics, Fudan University, Shanghai, China,
(4) Key Laboratory of Nuclear Physics and Ion-Beam Application (MOE), Fudan
University, Shanghai, China, (5) RIKEN Nishina Center, RIKEN, Wako, Japan,
(6) Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou, China,
(7) Physics Department, Stockholm University, Stockholm, Sweden, (8)
Department of Physics, University of Gothenburg, Gothenburg, Sweden) | Transmission of low energy electrons through a polyethylene
terephthalate 800-nm diameter nanocapillary | null | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | The transmission of 2-keV electrons through a polyethylene terephthalate
(PET) nanocapillary with a diameter of 800 nm and a length of 10 {\mu}m is
studied. The transmitted electrons are detected using a microchannel plate
(MCP) with a phosphor screen. It is found that the transmission rate for the
transmitted electrons at the incident energy can reach up to 10% for an aligned
capillary in the beam direction, but drops to less than 1% when the tilt angle
exceeds the geometrically allowable angle. The transmitted electrons with the
incident energy do not move with changes in the tilt angle, so the incident
electrons are not guided in the insulating capillary, which is different from
the behavior of positive ions. In the final stage of transmission, the angular
distribution of the transmitted electrons within the geometrically allowable
angle splits into two peaks along the observation angle perpendicular to the
tilt angle. The time evolution of the transmitted complete angular distribution
shows that when the beam is turned on, the transmission profile forms a single
peak. As the incident charge and time accumulate, the transmission profile
begins to stretch in the plane perpendicular to the tilt angle and gradually
splits into two peaks. When the tilt angle of the nanocapillary exceeds the
geometrically allowable angle, this splitting tends to disappear. Simulation of
charge deposition in the capillary directly exposed to the beam indicates the
formation of positive charge patches, which are not conducive to guiding,
unlike positive ions. Based on the simulation results, we can explain our data
and discuss the possible reasons for the splitting of the transmission angular
profiles.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 07:59:10 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Pengfei",
"Li",
"",
"1 and 2"
],
[
"Wanqi",
"Liu",
"",
"3 and\n 4"
],
[
"Shuai",
"Ha",
"",
"3 and\n 4"
],
[
"Yuzhou",
"Pan",
"",
"3 and\n 4"
],
[
"Xuhong",
"Fan",
"",
"3 and\n 4"
],
[
"Zhanhui",
"Du",
"",
"3 and\n 4"
],
[
"Chengliang",
"Wan",
"",
"3 and\n 4"
],
[
"Ying",
"Cui",
"",
"3 and\n 4"
],
[
"Ke",
"Yao",
"",
"3 and\n 4"
],
[
"Yue",
"Ma",
""
],
[
"Zhihu",
"Yang",
""
],
[
"Caojie",
"Shao",
""
],
[
"Schuch",
"Reinhold",
""
],
[
"Di",
"Lu",
""
],
[
"Yushou",
"Song",
""
],
[
"Hongqiang",
"Zhang",
""
],
[
"Ximeng",
"Chen",
""
]
] |
2504.11480 | Andrea Lucchini | Andrea Lucchini | Finite groups with regular subgroup graph | null | null | null | null | math.GR | http://creativecommons.org/licenses/by/4.0/ | We prove that the subgroup graph of a finite group $G$ is regular if and only
if $G$ is cyclic with square-free order.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 08:10:22 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Lucchini",
"Andrea",
""
]
] |
2504.11481 | Yu Hxiang Chen | Yu-Hxiang Chen, Ju-Shen Huang, Jia-Yu Hung, Chia-Kai Chang | Leveraging Knowledge Graphs and Large Language Models to Track and
Analyze Learning Trajectories | null | null | null | null | cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study addresses the challenges of tracking and analyzing students'
learning trajectories, particularly the issue of inadequate knowledge coverage
in course assessments. Traditional assessment tools often fail to fully cover
course content, leading to imprecise evaluations of student mastery. To tackle
this problem, the study proposes a knowledge graph construction method based on
large language models (LLMs), which transforms learning materials into
structured data and generates personalized learning trajectory graphs by
analyzing students' test data. Experimental results demonstrate that the model
effectively alerts teachers to potential biases in their exam questions and
tracks individual student progress. This system not only enhances the accuracy
of learning assessments but also helps teachers provide timely guidance to
students who are falling behind, thereby improving overall teaching strategies.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 16:27:15 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Chen",
"Yu-Hxiang",
""
],
[
"Huang",
"Ju-Shen",
""
],
[
"Hung",
"Jia-Yu",
""
],
[
"Chang",
"Chia-Kai",
""
]
] |
2504.11482 | Vidya Sudevan | Vidya Sudevan, Fakhreddine Zayer, Rizwana Kausar, Sajid Javed, Hamad
Karki, Giulia De Masi, Jorge Dias | snnTrans-DHZ: A Lightweight Spiking Neural Network Architecture for
Underwater Image Dehazing | null | null | null | null | cs.CV cs.AI cs.PF cs.RO eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Underwater image dehazing is critical for vision-based marine operations
because light scattering and absorption can severely reduce visibility. This
paper introduces snnTrans-DHZ, a lightweight Spiking Neural Network (SNN)
specifically designed for underwater dehazing. By leveraging the temporal
dynamics of SNNs, snnTrans-DHZ efficiently processes time-dependent raw image
sequences while maintaining low power consumption. Static underwater images are
first converted into time-dependent sequences by repeatedly inputting the same
image over user-defined timesteps. These RGB sequences are then transformed
into LAB color space representations and processed concurrently. The
architecture features three key modules: (i) a K estimator that extracts
features from multiple color space representations; (ii) a Background Light
Estimator that jointly infers the background light component from the RGB-LAB
images; and (iii) a soft image reconstruction module that produces haze-free,
visibility-enhanced outputs. The snnTrans-DHZ model is directly trained using a
surrogate gradient-based backpropagation through time (BPTT) strategy alongside
a novel combined loss function. Evaluated on the UIEB benchmark, snnTrans-DHZ
achieves a PSNR of 21.68 dB and an SSIM of 0.8795, and on the EUVP dataset, it
yields a PSNR of 23.46 dB and an SSIM of 0.8439. With only 0.5670 million
network parameters, and requiring just 7.42 GSOPs and 0.0151 J of energy, the
algorithm significantly outperforms existing state-of-the-art methods in terms
of efficiency. These features make snnTrans-DHZ highly suitable for deployment
in underwater robotics, marine exploration, and environmental monitoring.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 19:03:33 GMT"
}
] | 2025-04-17T00:00:00 | [
[
"Sudevan",
"Vidya",
""
],
[
"Zayer",
"Fakhreddine",
""
],
[
"Kausar",
"Rizwana",
""
],
[
"Javed",
"Sajid",
""
],
[
"Karki",
"Hamad",
""
],
[
"De Masi",
"Giulia",
""
],
[
"Dias",
"Jorge",
""
]
] |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.