id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2502.19390 | Minjoo Lim | Minjoo Lim, Bogyeong Kang, and Tae-Eui Kam | Multi-modal Contrastive Learning for Tumor-specific Missing Modality
Synthesis | null | null | null | null | eess.IV cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-modal magnetic resonance imaging (MRI) is essential for providing
complementary information about brain anatomy and pathology, leading to more
accurate diagnoses. However, obtaining high-quality multi-modal MRI in a
clinical setting is difficult due to factors such as time constraints, high
costs, and patient movement artifacts. To overcome this difficulty, there is
increasing interest in developing generative models that can synthesize missing
target modality images from the available source ones. Therefore, our team,
PLAVE, design a generative model for missing MRI that integrates multi-modal
contrastive learning with a focus on critical tumor regions. Specifically, we
integrate multi-modal contrastive learning, tailored for multiple source
modalities, and enhance its effectiveness by selecting features based on
entropy during the contrastive learning process. Additionally, our network not
only generates the missing target modality images but also predicts
segmentation outputs, simultaneously. This approach improves the generator's
capability to precisely generate tumor regions, ultimately improving
performance in downstream segmentation tasks. By leveraging a combination of
contrastive, segmentation, and additional self-representation losses, our model
effectively reflects target-specific information and generate high-quality
target images. Consequently, our results in the Brain MR Image Synthesis
challenge demonstrate that the proposed model excelled in generating the
missing modality.
| [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 18:34:58 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 04:37:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lim",
"Minjoo",
""
],
[
"Kang",
"Bogyeong",
""
],
[
"Kam",
"Tae-Eui",
""
]
] |
2502.19460 | Christian Marius Lillelund | Christian Marius Lillelund, Shi-ang Qi, Russell Greiner | Practical Evaluation of Copula-based Survival Metrics: Beyond the
Independent Censoring Assumption | null | null | null | null | stat.ML cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Conventional survival metrics, such as Harrell's concordance index and the
Brier Score, rely on the independent censoring assumption for valid inference
in the presence of right-censored data. However, when instances are censored
for reasons related to the event of interest, this assumption no longer holds,
as this kind of dependent censoring biases the marginal survival estimates of
popular nonparametric estimators. In this paper, we propose three copula-based
metrics to evaluate survival models in the presence of dependent censoring, and
design a framework to create realistic, semi-synthetic datasets with dependent
censoring to facilitate the evaluation of the metrics. Our empirical analyses
in synthetic and semi-synthetic datasets show that our metrics can give error
estimates that are closer to the true error, mainly in terms of prediction
accuracy.
| [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 10:28:44 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 12:22:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lillelund",
"Christian Marius",
""
],
[
"Qi",
"Shi-ang",
""
],
[
"Greiner",
"Russell",
""
]
] |
2502.19474 | Jose Angel Silva-Guill\'en | Guillermo Parra-Martinez, Alejandro Jimeno-Pozo, Vo Tien Phong, Hector
Sainz-Cruz, Daniel Kaplan, Peleg Emanuel, Yuval Oreg, Pierre A. Pantaleon,
Jose Angel Silva-Guillen, Francisco Guinea | Band Renormalization, Quarter Metals, and Chiral Superconductivity in
Rhombohedral Tetralayer Graphene | 5 pages, 4 figures and supplementary material. Updated version.
Comments are very welcome | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Recently, exotic superconductivity emerging from a spin-and-valley-polarized
metallic phase has been discovered in rhombohedral tetralayer graphene. To
explain this observation, we study the role of electron-electron interactions
in driving flavor symmetry breaking, using the Hartree-Fock (HF) approximation,
and in stabilizing superconductivity mediated by repulsive interactions. Though
mean-field HF correctly predicts the isospin flavors and reproduces the
experimental phase diagram, it overestimates the band renormalization near the
Fermi energy and suppresses superconducting instabilities. To address this, we
introduce a physically motivated scheme that includes internal screening in the
HF calculation. Using this formalism, we find superconductivity arising from
the spin-valley polarized phase for a range of electric fields and electron
dopings. Our findings reproduce the experimental observations and reveal a
p-wave, finite-momentum, time-reversal-symmetry-broken superconducting state,
encouraging further investigation into exotic phases in graphene multilayers.
| [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 19:00:01 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:14:00 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Parra-Martinez",
"Guillermo",
""
],
[
"Jimeno-Pozo",
"Alejandro",
""
],
[
"Phong",
"Vo Tien",
""
],
[
"Sainz-Cruz",
"Hector",
""
],
[
"Kaplan",
"Daniel",
""
],
[
"Emanuel",
"Peleg",
""
],
[
"Oreg",
"Yuval",
""
],
[
"Pantaleon",
"Pierre A.",
""
],
[
"Silva-Guillen",
"Jose Angel",
""
],
[
"Guinea",
"Francisco",
""
]
] |
2502.19836 | Zekai Yu | Ze-Kai Yu, Lei Liu, and Tao Zhu | Slowly rotating black hole in chiral scalar-tensor theory | 13 pages | null | 10.1140/epjc/s10052-025-14096-y | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | The chiral scalar-tensor theory is an extension of the Chern-Simons modified
gravity by introducing couplings between the first and second derivatives of
the scalar field and parity-violating spacetime curvatures. A key feature of
this theory is its explicit breaking of parity symmetry in the gravitational
sector, which is expected to affect the spatial-time component of axisymmetric
spacetime. In this paper, we investigate the effects of the chiral
scalar-tensor theory on slowly rotating black holes by building on known
solutions in the dynamical Chern-Simons modified gravity. Using perturbative
methods with small coupling and slow rotation approximations, we find that the
contributions of the chiral scalar-tensor theory appear at quadratic order in
the spin and cubic order in the coupling constants. Furthermore, we explore the
properties of this solution in the weak field and check its ergosphere and
horizon. In the weak limit, we find that the effects of parity violation are
suppressed in the weak field but could become significant in the strong field
regime. These results provide insights into the behavior of parity-violating
gravity in the presence of rotation and may be used for further investigations
into its observational signatures.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 07:17:41 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 07:53:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yu",
"Ze-Kai",
""
],
[
"Liu",
"Lei",
""
],
[
"Zhu",
"Tao",
""
]
] |
2502.20263 | Rongzhen Zhao | Rongzhen Zhao, Vivienne Wang, Juho Kannala, Joni Pajarinen | Vector-Quantized Vision Foundation Models for Object-Centric Learning | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Perceiving visual scenes as objects and background -- like humans do --
Object-Centric Learning (OCL) aggregates image or video feature maps into
object-level feature vectors, termed \textit{slots}. OCL's self-supervision of
reconstructing the input from these aggregated slots struggles with complex
object textures, thus Vision Foundation Model (VFM) representations are used as
the aggregation input and reconstruction target. However, existing methods
leverage VFM representations in diverse ways and often fail to fully exploit
their potential. In response, we propose a clean architecture --
Vector-Quantized VFMs for OCL (VQ-VFM-OCL, or VVO) -- that unifies mainstream
OCL methods. The key to our unification is simple yet effective, just shared
quantizing the same VFM representation as the reconstruction target. Through
mathematical modeling and statistical verification, we further analyze why VFM
representations facilitate OCL aggregation and how their shared quantization as
reconstruction targets strengthens OCL supervision. Experiments show that
across different VFMs, aggregators and decoders, our VVO consistently
outperforms baselines in object discovery and recognition, as well as
downstream visual prediction and reasoning. The source code is available in
supplemental files.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 16:51:13 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 08:41:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhao",
"Rongzhen",
""
],
[
"Wang",
"Vivienne",
""
],
[
"Kannala",
"Juho",
""
],
[
"Pajarinen",
"Joni",
""
]
] |
2502.20816 | Angelo Milfont MsEE | Angelo Milfont and Alvaro Veiga | Structural breaks detection and variable selection in dynamic linear
regression via the Iterative Fused LASSO in high dimension | 13 pages, 6 figures | null | null | null | econ.EM | http://creativecommons.org/licenses/by/4.0/ | We aim to develop a time series modeling methodology tailored to
high-dimensional environments, addressing two critical challenges: variable
selection from a large pool of candidates, and the detection of structural
break points, where the model's parameters shift. This effort centers on
formulating a least squares estimation problem with regularization constraints,
drawing on techniques such as Fused LASSO and AdaLASSO, which are
well-established in machine learning. Our primary achievement is the creation
of an efficient algorithm capable of handling high-dimensional cases within
practical time limits. By addressing these pivotal challenges, our methodology
holds the potential for widespread adoption. To validate its effectiveness, we
detail the iterative algorithm and benchmark its performance against the widely
recognized Path Algorithm for Generalized Lasso. Comprehensive simulations and
performance analyses highlight the algorithm's strengths. Additionally, we
demonstrate the methodology's applicability and robustness through simulated
case studies and a real-world example involving a stock portfolio dataset.
These examples underscore the methodology's practical utility and potential
impact across diverse high-dimensional settings.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 08:01:35 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:00:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Milfont",
"Angelo",
""
],
[
"Veiga",
"Alvaro",
""
]
] |
2503.00199 | Jacob Hiesener | Jacob M. Hiesener, C. Alex Kaylor, Joshua J. Wong, Prankush Agarwal,
Stephen E. Ralph | Seeded Topology Optimization for Commercial Foundry Integrated Photonics | 21 pages, 9 figures, submitted to Optics Express | null | null | null | physics.optics physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | We present a seeded topology optimization methodology for integrated photonic
devices fabricated on foundry platforms that yields improved performance
compared to traditional topology optimization. We employ blurring filters and a
DRC correction algorithm to more readily meet design rule checks resulting in
devices with fewer artifacts and improved correlation between simulation and
measurements. We apply this process to an ultra-compact TE modal multiplexer, a
TE mode converter, a polarization rotator, and a high-contrast grating
reflector. The measured insertion loss of the TE mode converter was reduced
from 1.37 dB to 0.64 dB through this optimization strategy. This approach
enables the use of physics-informed device topologies in inverse design and
maintains compliance with foundry constraints throughout optimization.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 21:38:15 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 00:05:39 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hiesener",
"Jacob M.",
""
],
[
"Kaylor",
"C. Alex",
""
],
[
"Wong",
"Joshua J.",
""
],
[
"Agarwal",
"Prankush",
""
],
[
"Ralph",
"Stephen E.",
""
]
] |
2503.00223 | Pengcheng Jiang | Pengcheng Jiang, Jiacheng Lin, Lang Cao, Runchu Tian, SeongKu Kang,
Zifeng Wang, Jimeng Sun, Jiawei Han | DeepRetrieval: Hacking Real Search Engines and Retrievers with Large
Language Models via Reinforcement Learning | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Information retrieval systems are crucial for enabling effective access to
large document collections. Recent approaches have leveraged Large Language
Models (LLMs) to enhance retrieval performance through query augmentation, but
often rely on expensive supervised learning or distillation techniques that
require significant computational resources and hand-labeled data. We introduce
DeepRetrieval, a reinforcement learning (RL) approach that trains LLMs for
query generation through trial and error without supervised data (reference
query). Using retrieval metrics as rewards, our system generates queries that
maximize retrieval performance. DeepRetrieval outperforms leading methods on
literature search with 65.07% (vs. previous SOTA 24.68%) recall for publication
search and 63.18% (vs. previous SOTA 32.11%) recall for trial search using
real-world search engines. DeepRetrieval also dominates in evidence-seeking
retrieval, classic information retrieval and SQL database search. With only 3B
parameters, it outperforms industry-leading models like GPT-4o and
Claude-3.5-Sonnet on 11/13 datasets. These results demonstrate that our RL
approach offers a more efficient and effective paradigm for information
retrieval. Our data and code are available at:
https://github.com/pat-jj/DeepRetrieval.
| [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 22:16:42 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Mar 2025 18:01:15 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 03:17:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jiang",
"Pengcheng",
""
],
[
"Lin",
"Jiacheng",
""
],
[
"Cao",
"Lang",
""
],
[
"Tian",
"Runchu",
""
],
[
"Kang",
"SeongKu",
""
],
[
"Wang",
"Zifeng",
""
],
[
"Sun",
"Jimeng",
""
],
[
"Han",
"Jiawei",
""
]
] |
2503.00362 | Pin Ju Tsai | Sheng-Hung Wang and Po-Han Chen and Cheng-Yu Yang and Yen-Hung Chen
and Pin-Ju Tsai | Harnessing Hybrid Frequency-Entangled Qudits through Quantum
Interference | 21 pages, 12 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | High-dimensional (HD) quantum entanglement expands the Hilbert space,
offering a robust framework for quantum information processing with enhanced
capacity and error resilience. In this work, we present a novel HD
frequency-domain entangled state, the hybrid frequency-entangled qudit (HFEQ),
generated via Hong-Ou-Mandel (HOM) interference, exhibiting both
discrete-variable (DV) and continuous-variable (CV) characteristics in the
frequency domain. By tuning HOM interference, we generate and control HFEQs
with dimensions $D=5,7,9,11, confirming their DV nature. Franson interferometry
confirms the global frequency correlations with visibility exceeding 98% and
verifies the CV entanglement within individual frequency modes with visibility
greater than 95%. Our findings provide deeper insight into the physical nature
of frequency-entangled qudits generated by quantum interference and introduce a
novel resource for HD time-frequency quantum information processing.
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 06:00:54 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 05:52:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Sheng-Hung",
""
],
[
"Chen",
"Po-Han",
""
],
[
"Yang",
"Cheng-Yu",
""
],
[
"Chen",
"Yen-Hung",
""
],
[
"Tsai",
"Pin-Ju",
""
]
] |
2503.00399 | Lijie Yang | Juan Song, Lijie Yang, Mingtao Feng | Extremely low-bitrate Image Compression Semantically Disentangled by
LMMs from a Human Perception Perspective | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It remains a significant challenge to compress images at extremely low
bitrate while achieving both semantic consistency and high perceptual quality.
Inspired by human progressive perception mechanism, we propose a Semantically
Disentangled Image Compression framework (SEDIC) in this paper. Initially, an
extremely compressed reference image is obtained through a learned image
encoder. Then we leverage LMMs to extract essential semantic components,
including overall descriptions, object detailed description, and semantic
segmentation masks. We propose a training-free Object Restoration model with
Attention Guidance (ORAG) built on pre-trained ControlNet to restore object
details conditioned by object-level text descriptions and semantic masks. Based
on the proposed ORAG, we design a multistage semantic image decoder to
progressively restore the details object by object, starting from the extremely
compressed reference image, ultimately generating high-quality and
high-fidelity reconstructions. Experimental results demonstrate that SEDIC
significantly outperforms state-of-the-art approaches, achieving superior
perceptual quality and semantic consistency at extremely low-bitrates ($\le$
0.05 bpp).
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 08:27:11 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Mar 2025 02:03:22 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 11:05:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Song",
"Juan",
""
],
[
"Yang",
"Lijie",
""
],
[
"Feng",
"Mingtao",
""
]
] |
2503.00771 | Yupu Hao | Yupu Hao, Pengfei Cao, Zhuoran Jin, Huanxuan Liao, Yubo Chen, Kang
Liu, Jun Zhao | Evaluating Personalized Tool-Augmented LLMs from the Perspectives of
Personalization and Proactivity | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Personalized tool utilization is essential for aligning large language models
(LLMs) with user preference in interaction scenarios with various tools.
However, most of the current benchmarks primarily focus on either
personalization of text generation or direct tool-utilizing, without
considering both. In this work, we introduce a novel benchmark ETAPP for
evaluating personalized tool invocation, establishing a sandbox environment,
and a comprehensive dataset of 800 testing cases covering diverse user
profiles. To improve the accuracy of our evaluation, we propose a
key-point-based LLM evaluation method, mitigating biases in the LLM-as-a-judge
system by manually annotating key points for each test case and providing them
to LLM as the reference. Additionally, we evaluate the excellent LLMs and
provide an in-depth analysis. Furthermore, we investigate the impact of
different tool-invoking strategies on LLMs' personalization performance and the
effects of fine-tuning in our task. The effectiveness of our preference-setting
and key-point-based evaluation method is also validated. Our findings offer
insights into improving personalized LLM agents. Our Code is available at
https://github.com/hypasd-art/ETAPP.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 07:36:22 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 09:19:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hao",
"Yupu",
""
],
[
"Cao",
"Pengfei",
""
],
[
"Jin",
"Zhuoran",
""
],
[
"Liao",
"Huanxuan",
""
],
[
"Chen",
"Yubo",
""
],
[
"Liu",
"Kang",
""
],
[
"Zhao",
"Jun",
""
]
] |
2503.00882 | Tao Wang | Yu-xin Cheng, De-hao Zhao, Yue-yang Shao, Li Gong, Tao Wang, Xiao-shen
Kang | SU(3) analysis for B(E2) anomaly | Comments are welcome | null | null | null | nucl-th | http://creativecommons.org/licenses/by/4.0/ | B(E2) anomaly is becoming a hot topic in the field of nuclear structure.
Since the B(E2) anomaly was experimentally found, understanding the mechanism
of its production has become an important problem. Theoretical studies have
found that the SU3-IBM and other extended IBM theories can give explanations,
but different mechanisms were found. In this paper, the concept ``SU(3)
analysis'' for B(E2) anomaly is proposed, and some new conclusions are
obtained. The SU(3) third-order interaction $[L\times Q \times L]^{(0)}$ and
level-crossing are both vital for the emergence of the B(E2) anomaly. This
technique can help us to better understand the realistic reason of the B(E2)
anomaly.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 12:53:26 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 15:22:27 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Cheng",
"Yu-xin",
""
],
[
"Zhao",
"De-hao",
""
],
[
"Shao",
"Yue-yang",
""
],
[
"Gong",
"Li",
""
],
[
"Wang",
"Tao",
""
],
[
"Kang",
"Xiao-shen",
""
]
] |
2503.00984 | Stefan Hegselmann | Stefan Hegselmann, Helen Zhou, Elizabeth Healey, Trenton Chang, Caleb
Ellington, Vishwali Mhasawade, Sana Tonekaboni, Peniel Argaw, Haoran Zhang | Machine Learning for Health symposium 2024 -- Findings track | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A collection of the accepted Findings papers that were presented at the 4th
Machine Learning for Health symposium (ML4H 2024), which was held on December
15-16, 2024, in Vancouver, BC, Canada. ML4H 2024 invited high-quality
submissions describing innovative research in a variety of health-related
disciplines including healthcare, biomedicine, and public health. Works could
be submitted to either the archival Proceedings track, or the non-archival
Findings track. The Proceedings track targeted mature, cohesive works with
technical sophistication and high-impact relevance to health. The Findings
track promoted works that would spark new insights, collaborations, and
discussions at ML4H. Both tracks were given the opportunity to share their work
through the in-person poster session. All the manuscripts submitted to ML4H
Symposium underwent a double-blind peer-review process.
| [
{
"version": "v1",
"created": "Sun, 2 Mar 2025 18:47:05 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 17:51:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hegselmann",
"Stefan",
""
],
[
"Zhou",
"Helen",
""
],
[
"Healey",
"Elizabeth",
""
],
[
"Chang",
"Trenton",
""
],
[
"Ellington",
"Caleb",
""
],
[
"Mhasawade",
"Vishwali",
""
],
[
"Tonekaboni",
"Sana",
""
],
[
"Argaw",
"Peniel",
""
],
[
"Zhang",
"Haoran",
""
]
] |
2503.01198 | Xue-Jia Yu | Sheng Yang, Fu Xu, Da-Chuan Lu, Yi-Zhuang You, Hai-Qing Lin, and
Xue-Jia Yu | Deconfined criticality as intrinsically gapless topological state in one
dimension | 4 pages with supplemental materials, 9 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deconfined criticality and gapless topological states have recently attracted
growing attention, as both phenomena go beyond the traditional Landau paradigm.
However, the deep connection between these two critical states, particularly in
lattice realization, remains insufficiently explored. In this Letter, we reveal
that certain deconfined criticality can be regarded as an intrinsically gapless
topological state without gapped counterparts in a one dimensional lattice
model. Using a combination of field-theoretic arguments and large-scale
numerical simulations, we establish the global phase diagram of the model,
which features deconfined critical lines separating two distinct spontaneous
symmetry breaking ordered phases. More importantly, we unambiguously
demonstrate that the mixed anomaly inherent to deconfined criticality enforces
topologically robust edge modes near the boundary, providing a general
mechanism by which deconfined criticality manifests as a gapless topological
state. Our findings not only offer a new perspective on deconfined criticality
but also deepen our understanding of gapless topological phases of matter.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 05:49:45 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 00:09:48 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Sheng",
""
],
[
"Xu",
"Fu",
""
],
[
"Lu",
"Da-Chuan",
""
],
[
"You",
"Yi-Zhuang",
""
],
[
"Lin",
"Hai-Qing",
""
],
[
"Yu",
"Xue-Jia",
""
]
] |
2503.01426 | Lina Zhao | Shubin Fu and Lina Zhao | A novel multipoint stress control volume method for linear elasticity on
quadrilateral grids | 26 pages, 6 figures | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we develop a novel control volume method that is locally
conservative and locking-free for linear elasticity problem on quadrilateral
grids. The symmetry of stress is weakly imposed through the introduction of a
Lagrange multiplier. As such, the method involves three unknowns: stress,
displacement and rotation. To ensure the well-posedness of the scheme, a pair
of carefully defined finite element spaces is used for the stress, displacement
and rotation such that the inf-sup condition holds. An appealing feature of the
method is that piecewise constant functions are used for the approximations of
stress, displacement and rotation, which greatly simplifies the implementation.
In particular, the stress space is defined delicately such that the stress
bilinear form is localized around each vertex, which allows for the local
elimination of the stress, resulting in a cell-centered system. By choosing
different definitions of the space for rotation, we develop two variants of the
method. In particular, the first method uses a constant function for rotation
over the interaction region, which allows for further elimination and results
in a cell-centered system involving displacement only. A rigorous error
analysis is performed for the proposed scheme. We show the optimal convergence
for $L^2$-error of the stress and rotation. Moreover, we can also prove the
superconvergence for $L^2$-error of displacement. Extensive numerical
simulations indicate that our method is efficient and accurate, and can handle
problems with discontinuous coefficients.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 11:28:48 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 09:33:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fu",
"Shubin",
""
],
[
"Zhao",
"Lina",
""
]
] |
2503.01911 | Masashi Sakiyama | Akira Iino, Masashi Sakiyama | Planar lattices and equilateral odd-gons | 7 pages, to appear in Yokohama Mathematical Journal | Yokohama Math. J. 70 (2024) 235-242 | 10.18880/0002001755 | null | math.CO math.MG math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For a planar integral lattice $L$, let $\nu(L)$ denote the square-free part
of the integer $D(L)^2$, where $D(L)$ stands for the area of a fundamental
parallelogram of $L$. For each odd integer $n$ with $3 \leq n<29$, a planar
lattice $L$ contains an equilateral $n$-gon if and only if $L$ is similar to an
integral lattice $L'$ such that $\nu(L')\equiv 3 \pmod 4$ and the largest prime
factor $p$ of $\nu(L')$ satisfies $p \leq n$. Moreover, such $L$ contains a
convex equilateral $n$-gon, which answers a problem posed by Maehara.
| [
{
"version": "v1",
"created": "Sat, 1 Mar 2025 04:59:41 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Iino",
"Akira",
""
],
[
"Sakiyama",
"Masashi",
""
]
] |
2503.02028 | Cristoforo Iossa | Augustus Brown, Francesco Galvagno, Alba Grassi, Cristoforo Iossa,
Congkao Wen | Large charge meets semiclassics in $\mathcal{N}=4$ super Yang-Mills | 49 pages, 4 figures | null | null | null | hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the large-charge sector of $\mathcal{N}=4$ super Yang-Mills theory
(SYM) with $SU(N)$ gauge group by constructing a special class of half-BPS
heavy operators, termed "canonical operators". Such operators exhibit
remarkable simplicity in the large-charge 't Hooft limit, where the dimension
of the operators $\Delta \to \infty$ with $\Delta\, g_{\text{YM}}^2$ held
finite. Canonical operator insertions in this regime map $\mathcal{N}=4$ SYM
onto the Coulomb branch, by assigning a classical profile to the scalar fields
with non-vanishing values along the diagonals given by the roots of unity. We
follow a semiclassical approach to study two-point, three-point and
Heavy-Heavy-Light-Light (HHLL) correlators. In particular we show that HHLL
correlators in the large-charge 't Hooft limit are computed as two-point
functions in a background determined by the classical profiles. We provide
consistent evidence of our findings by computing the same observables via
supersymmetric localization.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 20:02:17 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:14:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Brown",
"Augustus",
""
],
[
"Galvagno",
"Francesco",
""
],
[
"Grassi",
"Alba",
""
],
[
"Iossa",
"Cristoforo",
""
],
[
"Wen",
"Congkao",
""
]
] |
2503.02055 | Nikolaos Dimakis | N. Dimakis | Time operator from parametrization invariance and implications for
cosmology | 9 pages, no figures, Latex2e source file, to appear in PRD | null | null | null | gr-qc hep-th quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated by the parametrization invariance of cosmological Lagrangians and
their equivalence to systems describing the motion of particles in curved
backgrounds, we identify the phase space analogue of the notion of proper time.
We define the corresponding quantum operator, which results in being
canonically conjugate to that of the vanishing Hamiltonian. In the context of
particle dynamics, this leads to an uncertainty relation of the form $\Delta
E_0 \,\Delta T \geq \hbar$, where $E_0$ is the rest energy of the particle. By
studying the non-relativistic limit, we show that the action of the operator
reduces to multiplication by the classical time coordinate. Finally, we derive
the generic expression for the introduced time variable in the cosmological
setting.
| [
{
"version": "v1",
"created": "Mon, 3 Mar 2025 21:11:43 GMT"
},
{
"version": "v2",
"created": "Wed, 2 Apr 2025 00:11:21 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 15:48:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dimakis",
"N.",
""
]
] |
2503.02634 | Haiwen Wu | Haiwen Wu, Bayu Jayawardhana, Dabo Xu | Velocity-free task-space regulator for robot manipulators with external
disturbances | null | null | null | null | eess.SY cs.RO cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses the problem of task-space robust regulation of robot
manipulators subject to external disturbances. A velocity-free control law is
proposed by combining the internal model principle and the passivity-based
output-feedback control approach. The resulting controller not only ensures
asymptotic convergence of the regulation error but also rejects unwanted
external sinusoidal disturbances. The potential of the proposed method lies in
its simplicity, intuitiveness, and straightforward gain selection criteria for
the synthesis of multi-joint robot manipulator control systems.
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 13:59:51 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:13:04 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wu",
"Haiwen",
""
],
[
"Jayawardhana",
"Bayu",
""
],
[
"Xu",
"Dabo",
""
]
] |
2503.02973 | Zhuoyue Lyu | Zhuoyue Lyu, Per Ola Kristensson | Objestures: Everyday Objects Meet Mid-Air Gestures for Expressive
Interaction | Under Review | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Everyday objects and mid-air gestures have been explored as input modalities,
but each has its strengths and limitations - for example, objects offer
tangibility but rely on their physical presence; gestures are convenient but
lack haptic feedback. We introduce Objestures ("Obj" + "Gestures"), five
interaction types that utilize both modalities for a design space of expressive
and playful interaction. To evaluate its usefulness, we conducted a user study
(N = 12) assessing whether it can effectively support basic 3D tasks such as
rotation and scaling and found it has performance comparable to or better than
the headset's native freehand manipulation. To understand its user experience,
we conducted case studies on three example applications - Sound, Draw, and
Shadow - with the same participants, who found it intuitive, engaging, and
expressive, and were interested in its everyday use. We further illustrate 30
examples to showcase how Objestures can enrich everyday interactions and
discuss its limitations and implications. https://www.zhuoyuelyu.com/objestures
| [
{
"version": "v1",
"created": "Tue, 4 Mar 2025 19:57:54 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 06:32:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lyu",
"Zhuoyue",
""
],
[
"Kristensson",
"Per Ola",
""
]
] |
2503.03096 | Xia Zhang | Xia Zhang, Leilei Wei, Ming Liu | $C$-existence families, $C$-semigroups and their associated abstract
Cauchy problems in complete random normed modules | 21 pages | null | null | null | math.FA math.PR | http://creativecommons.org/licenses/by/4.0/ | In this paper, we first introduce the notion of a (mild) $C$-existence family
in complete random normed modules, then we prove that a (mild) $C$-existence
family can guarantee the existence of the (mild) solutions of the associated
abstract Cauchy problem in the random setting. Second, we investigate several
important properties peculiar to locally almost surely bounded $C$-semigroups
in complete random normed modules, which are not involved in the classical
theory of $C$-semigroups. Finally, based on the above work, some relations
among $C$-existence families, $C$-semigroups and their associated abstract
Cauchy problems in complete random normed modules are established, which extend
and improve some known results. Besides, an application to a type of stochastic
differential equations is also given.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 01:26:45 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 04:09:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Xia",
""
],
[
"Wei",
"Leilei",
""
],
[
"Liu",
"Ming",
""
]
] |
2503.03378 | Shunxuan He | Shunxuan He, Yang Huang, XinYi Li, Huawei Zhang, Gaochao Liu, Timothy
C. Beers, Hong Wu, Zhou Fan | Photometric-Metallicity and Distance Estimates for $\sim$70,000 RR Lyrae
Stars from the Zwicky Transient Facility | 29 pages, 31 figures and 7 tables, accepted by ApJS, the RRL
parameter catalogs are available at https://zenodo.org/records/14561442 | null | 10.3847/1538-4365/adbcad | null | astro-ph.SR astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Utilizing Zwicky Transient Facility (ZTF) data and existing RR Lyrae stars
(RRLs) catalogs, this study achieves the first calibration of the $P -
\phi_{31} - R_{21} - \text{[Fe/H]}$ and $P-\phi_{31}-A_{2}-A_{1}-\text{[Fe/H]}$
relations in the ZTF photometric system for RRab and RRc stars. We also
re-calibrate the period-absolute magnitude-metallicity (PMZ) and
period-Wesenheit-metallicity (PWZ) relations in the ZTF $gri$-bands for RRab
and RRc stars. Based on nearly 4100 stars with precise measurements of $P$,
$\phi_{31}$, $A_{2}$, and $A_{1}$, and available spectroscopic-metallicity
estimates, the photometric-metallicity relations exhibit strong internal
consistency across different bands, supporting the use of a weighted averaging
method for the final estimates. The photometric-metallicity estimates of
globular clusters based on RR Lyrae members also show excellent agreement with
high-resolution spectroscopic measurements, with typical scatter of 0.15 dex
for RRab stars and 0.14 dex for RRc stars, respectively. Using hundreds of
local RRLs with newly derived photometric metallicities and precise Gaia Data
Release 3 parallaxes, we establish the PMZ and PWZ relations in multiple bands.
Validation with globular cluster RR Lyrae members reveals typical distance
errors of 3.1% and 3.0% for the PMZ relations, and 3.1% and 2.6% for the PWZ
relations for RRab and RRc stars, respectively. Compared to PMZ relations, the
PWZ relations are tighter and almost unbiased, making them the recommended
choice for distance calculations. We present a catalog of 73,795 RRLs with
precise photometric metallicities; over 95% of them have accurate distance
measurements. Compared to Gaia DR3, approximately 25,000 RRLs have precise
photometric metallicities and distances derived for the first time.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 10:55:14 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Shunxuan",
""
],
[
"Huang",
"Yang",
""
],
[
"Li",
"XinYi",
""
],
[
"Zhang",
"Huawei",
""
],
[
"Liu",
"Gaochao",
""
],
[
"Beers",
"Timothy C.",
""
],
[
"Wu",
"Hong",
""
],
[
"Fan",
"Zhou",
""
]
] |
2503.03441 | Yu He | Yu He, Zifeng Wang, Jiaxu Li, Fenglin Zhong, Haozhe Yang, Kewen Shi,
Le Wang, Guang Yang and Weisheng Zhao | Geometric Asymmetry-Enhanced Nonreciprocal Supercurrent Transport
Revealed by Second-Harmonic Response | 17 pages, 4 figures, v2 | null | null | null | cond-mat.supr-con | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nonreciprocal transport in superconducting systems serves as a powerful probe
of symmetry-breaking mechanisms, with the superconducting diode effect emerging
as a key manifestation enabling cryogenic rectification. While theoretical
models have extensively explored superconducting nonreciprocity, experimental
verification remains challenging, as conventional transport measurements
struggle to disentangle intrinsic and extrinsic contributions. Nonlinear
transport analysis, particularly second-harmonic response, offers an
alternative approach by providing a sensitive probe for detecting spatial
inversion symmetry breaking in the presence of time-reversal symmetry
violation. Here, we systematically investigate the influence of geometric
symmetry on nonreciprocal transport by comparing two triangular-extended Hall
bar configurations with a symmetric Hall bar control. Second-harmonic nonlinear
transport measurements reveal that the triangular extension significantly
enhances nonreciprocal response, exhibiting a clear dependence on the base
angle of the extension. These findings establish a direct connection between
mesoscopic geometry and macroscopic nonreciprocity, demonstrating how spatial
symmetry and vortex dynamics govern nonlinear transport. This insight offers a
guiding principle for designing superconducting rectification architectures.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 12:17:44 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 02:37:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Yu",
""
],
[
"Wang",
"Zifeng",
""
],
[
"Li",
"Jiaxu",
""
],
[
"Zhong",
"Fenglin",
""
],
[
"Yang",
"Haozhe",
""
],
[
"Shi",
"Kewen",
""
],
[
"Wang",
"Le",
""
],
[
"Yang",
"Guang",
""
],
[
"Zhao",
"Weisheng",
""
]
] |
2503.03550 | Rajesh Selukar | Rajesh Selukar | Semiparametric Growth-Curve Modeling in Hierarchical, Longitudinal
Studies | null | null | null | null | stat.ME stat.CO | http://creativecommons.org/licenses/by/4.0/ | Modeling of growth (or decay) curves arises in many fields such as
microbiology, epidemiology, marketing, and econometrics. Parametric forms like
Logistic and Gompertz are often used for modeling such monotonic patterns.
While useful for compact description, the real-life growth curves rarely follow
these parametric forms perfectly. Therefore, the curve estimation methods that
strike a balance between prior information in the parametric form and fidelity
with the observed data are preferred. In hierarchical, longitudinal studies the
interest lies in comparing the growth curves of different groups while
accounting for the differences between the within-group subjects. This article
describes a flexible state space modeling framework that enables semiparametric
growth curve modeling for the data generated from hierarchical, longitudinal
studies. The methodology, a type of functional mixed effects modeling, is
illustrated with a real-life example of bacterial growth in different settings.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 14:36:28 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Mar 2025 13:52:08 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 17:54:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Selukar",
"Rajesh",
""
]
] |
2503.03736 | Sourajit Das | Sourajit Das, Kirtan Gopal Panda, Navid NaderiAlizadeh | Opportunistic Routing in Wireless Communications via Learnable
State-Augmented Policies | null | null | null | null | eess.SP cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper addresses the challenge of packet-based information routing in
large-scale wireless communication networks. The problem is framed as a
constrained statistical learning task, where each network node operates using
only local information. Opportunistic routing exploits the broadcast nature of
wireless communication to dynamically select optimal forwarding nodes, enabling
the information to reach the destination through multiple relay nodes
simultaneously. To solve this, we propose a State-Augmentation (SA) based
distributed optimization approach aimed at maximizing the total information
handled by the source nodes in the network. The problem formulation leverages
Graph Neural Networks (GNNs), which perform graph convolutions based on the
topological connections between network nodes. Using an unsupervised learning
paradigm, we extract routing policies from the GNN architecture, enabling
optimal decisions for source nodes across various flows. Numerical experiments
demonstrate that the proposed method achieves superior performance when
training a GNN-parameterized model, particularly when compared to baseline
algorithms. Additionally, applying the method to real-world network topologies
and wireless ad-hoc network test beds validates its effectiveness, highlighting
the robustness and transferability of GNNs.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 18:44:56 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 19:46:51 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Das",
"Sourajit",
""
],
[
"Panda",
"Kirtan Gopal",
""
],
[
"NaderiAlizadeh",
"Navid",
""
]
] |
2503.03953 | Morteza Karimzadeh | Aidan Marler, Yannik Roell, Steffen Knoblauch, Jane P. Messina, Thomas
Jaenisch, Morteza Karimzadeh | GeoDEN: A Visual Exploration Tool for Analysing the Geographic Spread of
Dengue Serotypes | null | Computer Graphics Forum (2025) | 10.1111/cgf.70087 | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Static maps and animations remain popular in spatial epidemiology of dengue,
limiting the analytical depth and scope of visualisations. Over half of the
global population live in dengue endemic regions. Understanding the
spatiotemporal dynamics of the four closely related dengue serotypes, and their
immunological interactions, remains a challenge at a global scale. To
facilitate this understanding, we worked with dengue epidemiologists in a
user-centered design framework to create GeoDEN, an exploratory visualisation
tool that empowers experts to investigate spatiotemporal patterns in dengue
serotype reports. The tool has several linked visualisations and filtering
mechanisms, enabling analysis at a range of spatial and temporal scales. To
identify successes and failures, we present both insight-based and value-driven
evaluations. Our domain experts found GeoDEN valuable, verifying existing
hypotheses and uncovering novel insights that warrant further investigation by
the epidemiology community. The developed visual exploration approach can be
adapted for exploring other epidemiology and disease incident datasets.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 22:54:38 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 18:57:32 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Marler",
"Aidan",
""
],
[
"Roell",
"Yannik",
""
],
[
"Knoblauch",
"Steffen",
""
],
[
"Messina",
"Jane P.",
""
],
[
"Jaenisch",
"Thomas",
""
],
[
"Karimzadeh",
"Morteza",
""
]
] |
2503.04104 | Zichong Li | Zichong Li, Xinyu Feng, Yuheng Cai, Zixuan Zhang, Tianyi Liu, Chen
Liang, Weizhu Chen, Haoyu Wang and Tuo Zhao | LLMs Can Generate a Better Answer by Aggregating Their Own Responses | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have shown remarkable capabilities across tasks,
yet they often require additional prompting techniques when facing complex
problems. While approaches like self-correction and response selection have
emerged as popular solutions, recent studies have shown these methods perform
poorly when relying on the LLM itself to provide feedback or selection
criteria. We argue this limitation stems from the fact that common LLM
post-training procedures lack explicit supervision for discriminative judgment
tasks. In this paper, we propose Generative Self-Aggregation (GSA), a novel
prompting method that improves answer quality without requiring the model's
discriminative capabilities. GSA first samples multiple diverse responses from
the LLM, then aggregates them to obtain an improved solution. Unlike previous
approaches, our method does not require the LLM to correct errors or compare
response quality; instead, it leverages the model's generative abilities to
synthesize a new response based on the context of multiple samples. While GSA
shares similarities with the self-consistency (SC) approach for response
aggregation, SC requires specific verifiable tokens to enable majority voting.
In contrast, our approach is more general and can be applied to open-ended
tasks. Empirical evaluation demonstrates that GSA effectively improves response
quality across various tasks, including mathematical reasoning, knowledge-based
problems, and open-ended generation tasks such as code synthesis and
conversational responses.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 05:25:43 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 21:53:51 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Zichong",
""
],
[
"Feng",
"Xinyu",
""
],
[
"Cai",
"Yuheng",
""
],
[
"Zhang",
"Zixuan",
""
],
[
"Liu",
"Tianyi",
""
],
[
"Liang",
"Chen",
""
],
[
"Chen",
"Weizhu",
""
],
[
"Wang",
"Haoyu",
""
],
[
"Zhao",
"Tuo",
""
]
] |
2503.04117 | Soumya Sahu | Soumya Sahu, Thomas Mathew, Dulal K. Bhaumik | Fiducial Confidence Intervals for Agreement Measures Among Raters Under
a Generalized Linear Mixed Effects Model | null | null | null | null | stat.ME stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A generalization of the classical concordance correlation coefficient (CCC)
is considered under a three-level design where multiple raters rate every
subject over time, and each rater is rating every subject multiple times at
each measuring time point. The ratings can be discrete or continuous. A
methodology is developed for the interval estimation of the CCC based on a
suitable linearization of the model along with an adaptation of the fiducial
inference approach. The resulting confidence intervals have satisfactory
coverage probabilities and shorter expected widths compared to the interval
based on Fisher Z-transformation, even under moderate sample sizes. Two real
applications available in the literature are discussed. The first application
is based on a clinical trial to determine if various treatments are more
effective than a placebo for treating knee pain associated with osteoarthritis.
The CCC was used to assess agreement among the manual measurements of the joint
space widths on plain radiographs by two raters, and the computer-generated
measurements of digitalized radiographs. The second example is on a
corticospinal tractography, and the CCC was once again applied in order to
evaluate the agreement between a well-trained technologist and a
neuroradiologist regarding the measurements of fiber number in both the right
and left corticospinal tracts. Other relevant applications of our general
approach are highlighted in many areas including artificial intelligence.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 05:55:01 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 18:26:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sahu",
"Soumya",
""
],
[
"Mathew",
"Thomas",
""
],
[
"Bhaumik",
"Dulal K.",
""
]
] |
2503.04350 | Joana Sim\~oes | Joana Sim\~oes and Jo\~ao Correia | EDCA - An Evolutionary Data-Centric AutoML Framework for Efficient
Pipelines | preprint for EvoApplications 2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated Machine Learning (AutoML) gained popularity due to the increased
demand for Machine Learning (ML) specialists, allowing them to apply ML
techniques effortlessly and quickly. AutoML implementations use optimisation
methods to identify the most effective ML solution for a given dataset, aiming
to improve one or more predefined metrics. However, most implementations focus
on model selection and hyperparameter tuning. Despite being an important factor
in obtaining high-performance ML systems, data quality is usually an overlooked
part of AutoML and continues to be a manual and time-consuming task. This work
presents EDCA, an Evolutionary Data Centric AutoML framework. In addition to
the traditional tasks such as selecting the best models and hyperparameters,
EDCA enhances the given data by optimising data processing tasks such as data
reduction and cleaning according to the problems' needs. All these steps create
an ML pipeline that is optimised by an evolutionary algorithm. To assess its
effectiveness, EDCA was compared to FLAML and TPOT, two frameworks at the top
of the AutoML benchmarks. The frameworks were evaluated in the same conditions
using datasets from AMLB classification benchmarks. EDCA achieved statistically
similar results in performance to FLAML and TPOT but used significantly less
data to train the final solutions. Moreover, EDCA experimental results reveal
that a good performance can be achieved using less data and efficient ML
algorithm aspects that align with Green AutoML guidelines
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 11:46:07 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:57:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Simões",
"Joana",
""
],
[
"Correia",
"João",
""
]
] |
2503.04707 | Mengdi Wang | Mengdi Wang, Efe Bozkir, Enkelejda Kasneci | Iris Style Transfer: Enhancing Iris Recognition with Style Features and
Privacy Preservation through Neural Style Transfer | The 2025 ACM Symposium on Eye Tracking Research & Applications (ETRA) | ACM 2577-6193/2025/6-ART21 | 10.1145/3729413 | null | cs.CV cs.HC | http://creativecommons.org/licenses/by/4.0/ | Iris texture is widely regarded as a gold standard biometric modality for
authentication and identification. The demand for robust iris recognition
methods, coupled with growing security and privacy concerns regarding iris
attacks, has escalated recently. Inspired by neural style transfer, an advanced
technique that leverages neural networks to separate content and style
features, we hypothesize that iris texture's style features provide a reliable
foundation for recognition and are more resilient to variations like rotation
and perspective shifts than traditional approaches. Our experimental results
support this hypothesis, showing a significantly higher classification accuracy
compared to conventional features. Further, we propose using neural style
transfer to obfuscate the identifiable iris style features, ensuring the
protection of sensitive biometric information while maintaining the utility of
eye images for tasks like eye segmentation and gaze estimation. This work opens
new avenues for iris-oriented, secure, and privacy-aware biometric systems.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 18:55:21 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:29:04 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Mengdi",
""
],
[
"Bozkir",
"Efe",
""
],
[
"Kasneci",
"Enkelejda",
""
]
] |
2503.04844 | Hank Gerba | Hank Gerba | Narrative Context Protocol: an Author-centric Storytelling Framework for
Generative AI | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative AI promises to finally realize dynamic, personalized storytelling
technologies across a range of media. To date, experimentation with generative
AI in the field of procedural narrative generation has been quite promising
from a technical perspective. However, fundamental narrative dilemmas remain,
such as the balance between player agency and narrative coherence, and no
rigorous narrative standard has been proposed to specifically leverage the
strengths of generative AI. In this paper, we propose the Narrative Context
Protocol (NCP), an open and extensible standard designed to place writers at
the center of future narrative design workflows and enable interoperability
across authoring platforms. By encoding an author's intent according to an
objective narrative model, the NCP enables narrative portability as well as
intent-based constraints for generative systems.
| [
{
"version": "v1",
"created": "Wed, 5 Mar 2025 18:29:15 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Mar 2025 18:26:37 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Mar 2025 21:53:10 GMT"
},
{
"version": "v4",
"created": "Sat, 12 Apr 2025 18:17:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gerba",
"Hank",
""
]
] |
2503.04964 | Agnieszka Hejna | Jacek Dziuba\'nski, Agnieszka Hejna | Characterizations of $H^1$ and Fefferman-Stein decompositions of ${\rm
BMO}$ functions by systems of singular integrals in the Dunkl setting | null | null | null | null | math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend the classical theorem of Uchiyama about constructive
Fefferman-Stein decompositions of ${\rm BMO}$ functions by systems of singular
integrals to the rational Dunkl setting. On $\mathbb{R}^N$ equipped with a root
system $R$ and a multiplicity function $k \geq 0$, let \[ dw(\mathbf{x}) =
\prod_{\alpha \in R} |\langle \alpha, \mathbf{x} \rangle|^{k(\alpha)} \,
d\mathbf{x} \] denote the associated measure, and let $\mathcal{F}$ stand for
the Dunkl transform. Consider a system $(\theta_0, \theta_1, \theta_2, \dots,
\theta_d)$ of functions on $\mathbb{R}^N$ that are smooth away from the origin
and homogeneous of degree zero, with $\theta_0(\xi) \equiv 1$. We prove that if
\[ \text{rank} \left( \begin{array}{ccccc} 1 & \theta_1(\xi) & \theta_2(\xi) &
\ldots & \theta_d(\xi) \\ 1 & \theta_1(-\xi) & \theta_2(-\xi) & \ldots &
\theta_d(-\xi) \end{array} \right) = 2 \quad \text{for all } \xi \in
\mathbb{R}^N \text{ with } \|\xi\| = 1, \] then any compactly supported ${\rm
BMO}(\mathbb{R}^N, \|\mathbf{x} - \mathbf{y}\|, dw)$ function $f$ can be
decomposed into \[ f = g_0 + \sum_{j=1}^d \mathbf{S}^{\{j\}} g_j, \quad \left\|
\sum_{j=0}^d g_j \right\|_{L^\infty} \leq C \|f\|_{\rm BMO}, \] where
$\mathbf{S}^{\{j\}} g = \mathcal{F}^{-1}(\theta_j \mathcal{F}g)$. As a
corollary, we obtain characterizations of the Hardy space $H^1_{\rm Dunkl}$ by
the system of singular integral operators $({\rm Id}, \mathbf{S}^{\{1\}},
\mathbf{S}^{\{2\}}, \dots, \mathbf{S}^{\{d\}})$.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2025 20:51:37 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 15:02:56 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dziubański",
"Jacek",
""
],
[
"Hejna",
"Agnieszka",
""
]
] |
2503.05136 | Ronny Ko | Ronny Ko | The Beginner's Textbook for Fully Homomorphic Encryption | null | null | null | null | cs.CR cs.DM | http://creativecommons.org/licenses/by/4.0/ | Fully Homomorphic Encryption (FHE) is a cryptographic scheme that enables
computations to be performed directly on encrypted data, as if the data were in
plaintext. After all computations are performed on the encrypted data, it can
be decrypted to reveal the result. The decrypted value matches the result that
would have been obtained if the same computations were applied to the plaintext
data.
FHE supports basic operations such as addition and multiplication on
encrypted numbers. Using these fundamental operations, more complex
computations can be constructed, including subtraction, division, logic gates
(e.g., AND, OR, XOR, NAND, MUX), and even advanced mathematical functions such
as ReLU, sigmoid, and trigonometric functions (e.g., sin, cos). These functions
can be implemented either as exact formulas or as approximations, depending on
the trade-off between computational efficiency and accuracy.
Fully Homomorphic Encryption (FHE) enables privacy-preserving machine
learning by allowing a server to process the client's data in its encrypted
form through an ML model. With FHE, the server learns neither the plaintext
version of the input features nor the inference results. Only the client, using
their secret key, can decrypt and access the results at the end of the service
protocol.FHE can also be applied to confidential blockchain services, ensuring
that sensitive data in smart contracts remains encrypted and confidential while
maintaining the transparency and integrity of the execution process. Other
applications of FHE include secure outsourcing of data analytics, encrypted
database queries, privacy-preserving searches, efficient multi-party
computation for digital signatures, and more.
This article is designed to help the reader understand how FHE works from the
mathematical level.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 04:29:11 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 15:18:50 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 03:22:13 GMT"
},
{
"version": "v4",
"created": "Sun, 13 Apr 2025 13:14:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ko",
"Ronny",
""
]
] |
2503.05172 | Louis Hanotel | Louis Hanotel and Elena R. Loubenets | Nonviolation of the CHSH inequality under local spin-1 measurements on
two spin qutrits | 22 pages, 3 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the present paper, based on the general analytical expression
[arXiv:2412.03470] for the maximum of the CHSH expectation under local Alice
and Bob spin-$s$ measurements in a two-qudit state of dimension $d=2s+1$,
$s\geq 1/2$, we analyze whether or not, under spin-$1$ measurements in an
arbitrary two-qutrit state, the CHSH inequality is violated. We find
analytically for a variety of pure nonseparable two-qutrit states and also,
numerically for $1,000,000$ randomly generated pure nonseparable two-qutrit
states, that, under local Alice and Bob spin-$1$ measurements in each of these
nonseparable states, including maximally entangled, the CHSH inequality is not
violated. These results together with the spectral decomposition of a mixed
state lead us to the Conjecture that, under local Alice and Bob spin-$1$
measurements, every nonseparable two-qutrit state, pure or mixed, does not
violate the CHSH inequality. For a variety of pure two-qutrit states, we
further find the values of their concurrence and compare them with the values
of their spin-$1$ CHSH parameter, which determines violation or nonviolation by
a two-qutrit state of the CHSH inequality under spin-$1$ measurements. This
comparison indicates that, in contrast to spin-$\frac{1}{2}$ measurements,
where the spin-$\frac{1}{2}$ CHSH parameter of a pure two-qubit state is
increasing monotonically with a growth of its entanglement, for a pure
two-qutrit state, this is not the case. In particular, for the two-qutrit GHZ
state, which is maximally entangled, the spin-$1$ CHSH parameter is equal to
$\sqrt{\frac{8}{9}}$, while, for some separable pure two-qutrit states, this
parameter can be equal to unity. Moreover, for the two-qutrit Horodecki state,
the spin-$1$ CHSH parameter is equal to $4\sqrt{2}/21<1$ regardless of the
entanglement type of this mixed state.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 06:35:21 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 08:27:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hanotel",
"Louis",
""
],
[
"Loubenets",
"Elena R.",
""
]
] |
2503.05446 | Youwei Zhang | Youwei Zhang, Shenchao Jin, Junlei Duan, Klaus M{\o}lmer, Guiying
Zhang, Mingfeng Wang and Yanhong Xiao | Cooperative squeezing of internal and collective spins in an atomic
ensemble | 6 pages, 3 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Creating highly spin-squeezed states for quantum metrology surpassing the
standard quantum limit is a topic of great interest. Spin squeezing has been
achieved by either entangling different atoms in an ensemble, or by controlling
the multi-level internal spin state of an atom. Here, we experimentally
demonstrate combined internal and collective spin squeezing in a hot atomic
ensemble with $\sim 10^{11}$ rubidium atoms. By synergistically combining these
two types of squeezing and carefully aligning their squeezing quadratures, we
have achieved a metrologically relevant spin squeezing of $-6.21\pm0.84$ dB,
significantly outperforming the results obtained by utilizing either type of
squeezing alone. Our approach provides a new perspective on fully harnessing
the degrees of freedom inherent in quantum states of an atomic ensemble.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 14:16:57 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Apr 2025 11:17:56 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Apr 2025 05:55:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Youwei",
""
],
[
"Jin",
"Shenchao",
""
],
[
"Duan",
"Junlei",
""
],
[
"Mølmer",
"Klaus",
""
],
[
"Zhang",
"Guiying",
""
],
[
"Wang",
"Mingfeng",
""
],
[
"Xiao",
"Yanhong",
""
]
] |
2503.05485 | Marco Geraci | Marco Geraci | Estimation of the generalized Laplace distribution and its projection
onto the circle | 14 pages, 6 figures, 2 tables | null | null | null | stat.ME | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The generalized Laplace (GL) distribution, which falls in the larger family
of generalized hyperbolic distributions, provides a versatile model to deal
with a variety of applications thanks to its shape parameters. The elliptically
symmetric GL admits a polar representation that can be used to yield a circular
distribution, which we call projected GL (PGL) distribution. The latter does
not appear to have been considered yet in practical applications. In this
article, we explore an easy-to-implement maximum likelihood estimation strategy
based on Gaussian quadrature for the scale-mixture representation of the GL and
its projection onto the circle. A simulation study is carried out to benchmark
the fitting routine against expectation-maximization and direct maximum
likelihood to assess its feasibility, while the PGL model is contrasted with
the von Mises and projected normal distributions to assess its prospective
utility. The results showed that quadrature-based estimation is more reliable
consistently across selected scenarios and sample sizes than alternative
estimation methods, while the PGL complements other distributions in terms of
flexibility.
| [
{
"version": "v1",
"created": "Fri, 7 Mar 2025 14:54:55 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 10:38:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Geraci",
"Marco",
""
]
] |
2503.06057 | Prince Phillip | Alin Thomas Tharakan, Prince Phillip, Gokulan T., Sumit Kumar, Gaurab
Banerjee | A 2-6 GHz Ultra-Wideband CMOS Transceiver for Radar Applications | null | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a low power, low cost transceiver architecture to
implement radar-on-a-chip. The transceiver comprises of a full ultra-wideband
(UWB) transmitter and a full UWB band receiver. A design methodology to
maximize the tuning range of the voltage-controlled oscillator (VCO) is
presented. At the transmitter side, a sub-harmonic mixer is used for signal
up-conversion. The receiver low noise amplifier (LNA) has a 2 to 6 GHz input
matching bandwidth with a power gain of 9 dB and a noise figure of 2.5 dB. The
transceiver is implemented in Cadence EDA tools using 65nm CMOS technology. The
system achieves a total dc power consumption of 50 mW. Good noise figure
performance; good wide-band matching; gain; high level of integration; low
power; low cost of the proposed UWB radar transceiver front-end make it a
highly competitive SoC solution for low power UWB transceivers.
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 04:52:24 GMT"
},
{
"version": "v2",
"created": "Fri, 4 Apr 2025 07:07:37 GMT"
},
{
"version": "v3",
"created": "Wed, 9 Apr 2025 10:18:44 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Tharakan",
"Alin Thomas",
""
],
[
"Phillip",
"Prince",
""
],
[
"T.",
"Gokulan",
""
],
[
"Kumar",
"Sumit",
""
],
[
"Banerjee",
"Gaurab",
""
]
] |
2503.06730 | Matthew Shen | Matthew Shen, Aliyah Hsu, Abhineet Agarwal, Bin Yu | Adaptive Test-Time Intervention for Concept Bottleneck Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Concept bottleneck models (CBM) aim to improve model interpretability by
predicting human level "concepts" in a bottleneck within a deep learning model
architecture. However, how the predicted concepts are used in predicting the
target still either remains black-box or is simplified to maintain
interpretability at the cost of prediction performance. We propose to use Fast
Interpretable Greedy Sum-Trees (FIGS) to obtain Binary Distillation (BD). This
new method, called FIGS-BD, distills a binary-augmented concept-to-target
portion of the CBM into an interpretable tree-based model, while maintaining
the competitive prediction performance of the CBM teacher. FIGS-BD can be used
in downstream tasks to explain and decompose CBM predictions into interpretable
binary-concept-interaction attributions and guide adaptive test-time
intervention. Across 4 datasets, we demonstrate that our adaptive test-time
intervention identifies key concepts that significantly improve performance for
realistic human-in-the-loop settings that only allow for limited concept
interventions.
| [
{
"version": "v1",
"created": "Sun, 9 Mar 2025 19:03:48 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:39:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shen",
"Matthew",
""
],
[
"Hsu",
"Aliyah",
""
],
[
"Agarwal",
"Abhineet",
""
],
[
"Yu",
"Bin",
""
]
] |
2503.07001 | Jacek Jakimiuk | Jacek Jakimiuk | Stability of Khintchine inequalities with optimal constants between the
second and the $p$-th moment for $p \ge 3$ | The statement of Theorem 2 is noticeably improved compared to the
previous version by slight modification of the proof (observing the identity
(10)) | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give a strengthening of the classical Khintchine inequality between the
second and the $p$-th moment for $p \ge 3$ with optimal constant by adding a
deficit depending on the vector of coefficients of the Rademacher sum.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 07:32:08 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 09:52:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jakimiuk",
"Jacek",
""
]
] |
2503.07117 | Jesus Salas | Yihao Xu, Jes\'us Salas, and Youjin Deng | Finite-size corrections from the subleading magnetic scaling field for
the Ising and Potts models in two dimensions | The document contains the paper (pdflatex, 17 pages), and 10 pdf
figures. Minor changes with respect to v1. Final version | Entropy 27 (2025) 418 | 10.3390/e27040418 | null | cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | In finite-size scaling analyses of critical phenomena, proper consideration
of correction terms, which can come from different sources, plays an important
role. For the Fortuin-Kasteleyn representation of the $Q$-state Potts model in
two dimensions, although the subleading magnetic scaling field, with exactly
known exponent, is theoretically expected to give rise in finite-size-scaling
analyses, numerical observation remains elusive probably due to the mixing of
various corrections. We simulate the O($n$) loop model on the hexagonal
lattice, which is in the same universality class as the $Q=n^2$ Potts model but
has suppressed corrections from other sources, and provides strong numerical
evidence for the attribution of the subleading magnetic field in finite-size
corrections. Interestingly, it is also observed that the corrections in small-
and large-cluster-size regions have opposite magnitudes, and, for the special
$n=2$ case, they compensate with each other in observables like the second
moment of the cluster-size distribution. Our finding reveals that the effect of
the subleading magnetic field should be taken into account in
finite-size-scaling analyses, which was unfortunately ignored in many previous
studies.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 09:46:34 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 10:47:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Yihao",
""
],
[
"Salas",
"Jesús",
""
],
[
"Deng",
"Youjin",
""
]
] |
2503.07592 | Andrei Mironov | A. Mironov, A. Morozov, A. Popolitov, Z. Zakirova | Diamond of triads | 9 pages, LaTeX | null | null | MIPT/TH-04/25; FIAN/TD-03/25; ITEP/TH-04/25; IITP/TH-04/25 | hep-th math-ph math.MP math.QA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The triad refers to embedding of two systems of polynomials, symmetric ones
and those of the Baker-Akhiezer type into a power series of the Noumi-Shiraishi
type. It provides an alternative definition of Macdonald theory and its
extensions. The basic triad is associated with the vector representation of the
Ding-Iohara-Miki (DIM) algebra. We discuss lifting this triad to two elliptic
generalizations and further to the bi-elliptic triad. At the algebraic level,
it corresponds to elliptic and bi-elliptic DIM algebras. This completes the
list of polynomials associated with Seiberg-Witten theory with adjoint matter
in various dimensions.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 17:55:15 GMT"
},
{
"version": "v2",
"created": "Sun, 6 Apr 2025 19:14:21 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 20:29:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mironov",
"A.",
""
],
[
"Morozov",
"A.",
""
],
[
"Popolitov",
"A.",
""
],
[
"Zakirova",
"Z.",
""
]
] |
2503.07901 | Francesco Iodice | Francesco Iodice, Elena De Momi, and Arash Ajoudani | Intelligent Framework for Human-Robot Collaboration: Dynamic Ergonomics
and Adaptive Decision-Making | 15 pagine, 8figure, 3 tabelle, formato conferenza IEEE | null | null | null | cs.RO cs.CV cs.HC cs.LG | http://creativecommons.org/licenses/by/4.0/ | The integration of collaborative robots into industrial environments has
improved productivity, but has also highlighted significant challenges related
to operator safety and ergonomics. This paper proposes an innovative framework
that integrates advanced visual perception, continuous ergonomic monitoring,
and adaptive Behaviour Tree decision-making to overcome the limitations of
traditional methods that typically operate as isolated components. Our approach
synthesizes deep learning models, advanced tracking algorithms, and dynamic
ergonomic assessments into a modular, scalable, and adaptive system.
Experimental validation demonstrates the framework's superiority over existing
solutions across multiple dimensions: the visual perception module outperformed
previous detection models with 72.4% mAP@50:95; the system achieved high
accuracy in recognizing operator intentions (92.5%); it promptly classified
ergonomic risks with minimal latency (0.57 seconds); and it dynamically managed
robotic interventions with exceptionally responsive decision-making
capabilities (0.07 seconds), representing a 56% improvement over benchmark
systems. This comprehensive solution provides a robust platform for enhancing
human-robot collaboration in industrial environments by prioritizing ergonomic
safety, operational efficiency, and real-time adaptability.
| [
{
"version": "v1",
"created": "Mon, 10 Mar 2025 22:43:07 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:02:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Iodice",
"Francesco",
""
],
[
"De Momi",
"Elena",
""
],
[
"Ajoudani",
"Arash",
""
]
] |
2503.07928 | Hunter McNichols | Hunter McNichols, Andrew Lan | The StudyChat Dataset: Student Dialogues With ChatGPT in an Artificial
Intelligence Course | Pre-print v0.2 | null | null | null | cs.AI cs.HC | http://creativecommons.org/licenses/by/4.0/ | The widespread availability of large language models (LLMs), such as ChatGPT,
has significantly impacted education, raising both opportunities and
challenges. Students can frequently interact with LLM-powered, interactive
learning tools, but their usage patterns need to be analyzed to ensure ethical
usage of these tools. To better understand how students interact with LLMs in
an academic setting, we introduce \textbf{StudyChat}, a publicly available
dataset capturing real-world student interactions with an LLM-powered tutoring
chatbot in a semester-long, university-level artificial intelligence (AI)
course. We deploy a web application that replicates ChatGPT's core
functionalities, and use it to log student interactions with the LLM while
working on programming assignments. We collect 1,197 conversations, which we
annotate using a dialogue act labeling schema inspired by observed interaction
patterns and prior research. Additionally, we analyze these interactions,
highlight behavioral trends, and analyze how specific usage patterns relate to
course outcomes. \textbf{StudyChat} provides a rich resource for the learning
sciences and AI in education communities, enabling further research into the
evolving role of LLMs in education.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 00:17:07 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 02:42:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"McNichols",
"Hunter",
""
],
[
"Lan",
"Andrew",
""
]
] |
2503.07936 | Yuanming Wang | Yuanming Wang, Pavan Uttarkar, Ryan Shannon, Yu Wing Joshua Lee,
Dougal Dobie, Ziteng Wang, Keith Bannister, Manisha Caleb, Adam Deller,
Marcin Glowacki, Joscha Jahns-Schindler, Tara Murphy, Reshma Anna-Thomas,
N.D.R. Bhat, Xinping Deng, Vivek Gupta, Akhil Jaini, Clancy James and John
Tuthill | The discovery of a 41s radio pulsar PSR J0311+1402 with ASKAP | 11 pages, 5 figures; accepted for publication in ApJL | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The emerging population of long-period radio transients (LPTs) show both
similarities and differences with normal pulsars. A key difference is that
their radio emission is too bright to be powered solely by rotational energy.
Various models have been proposed (including both white-dwarf or neutron star
origins), and their nature remains uncertain. Known LPTs have minutes to hours
long spin periods, while normal pulsars have periods ranging from milliseconds
to seconds. Here, we report the discovery of PSR J0311+1402, an object with an
intermediate spin period of 41 seconds, bridging the gap between LPTs and
normal pulsars. PSR J0311+1402 exhibits low linear ($\sim25\%$) and circular
polarisation ($\sim5\%$) and a relatively steep spectral index ($\sim-2.3$),
features similar to normal pulsars. However, its observed spin-down properties
place it below the pulsar death line, where pair production and thus radio
emission are expected to cease. The discovery of PSR J0311+1402 suggests the
existence of a previously undetected population within this intermediate period
range, presumably missed due to selection biases in traditional pulsar search
methods. Finding more such objects is important to fill the current gap in
neutron star spin periods, improving our understanding of the relationships
among rotation-powered pulsars and LPTs.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 00:27:38 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 02:15:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Yuanming",
""
],
[
"Uttarkar",
"Pavan",
""
],
[
"Shannon",
"Ryan",
""
],
[
"Lee",
"Yu Wing Joshua",
""
],
[
"Dobie",
"Dougal",
""
],
[
"Wang",
"Ziteng",
""
],
[
"Bannister",
"Keith",
""
],
[
"Caleb",
"Manisha",
""
],
[
"Deller",
"Adam",
""
],
[
"Glowacki",
"Marcin",
""
],
[
"Jahns-Schindler",
"Joscha",
""
],
[
"Murphy",
"Tara",
""
],
[
"Anna-Thomas",
"Reshma",
""
],
[
"Bhat",
"N. D. R.",
""
],
[
"Deng",
"Xinping",
""
],
[
"Gupta",
"Vivek",
""
],
[
"Jaini",
"Akhil",
""
],
[
"James",
"Clancy",
""
],
[
"Tuthill",
"John",
""
]
] |
2503.07954 | Wei Zhang | Wei Zhang, Yunning Zhao, Lin Ma, Shiming Wen, Chaojian Wu, Juanjuan
Ren, Jianjun Chen, Yuzhong Wu, Zhongrui Bai, Yonghui Hou, Yongheng Zhao, and
Hong Wu | Spatially-resolved Galactic HII regions observed by LAMOST
medium-Resolution Spectroscopic Survey of Nebulae (LAMOST MRS-N) | 22 pages, 8 figures, 3 tables, accepted for publication in AJ | 2025, AJ, 169, 257 | 10.3847/1538-3881/adbf8f | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present spatially-resolved spectroscopic observations of 10 isolated
Galactic HII regions using data from the LAMOST Medium-Resolution Spectroscopic
Survey of Nebulae (LAMOST MRS-N). The high spatial resolution of the data
allows us to investigate the 1D radial profiles of emission line fluxes (Ha, [S
II] and [N II]), flux ratios ([N II]/Ha, [S II]/Ha and [S II]/[N II]), and
radial velocities of these three emission lines. Among these regions, two are
ionization-bounded, while the remaining eight are matter-bounded. The
matter-bounded HII regions exhibit shallower slopes in their radial flux
profiles compared to the ionization-bounded ones. In most cases, the [N II]/Ha
and [S II]/Ha ratios increase with distance from the center of the HII regions,
consistent with model predictions that low-ionization emissions dominate the
outer zones of these regions. The two ionization-bounded HII regions have
kinematic ages (t) of 0.2 and 0.3 Myr, while the matter-bounded regions span
ages from 1 to 12 Myr. For the matter-bounded HII regions, the optical emission
flux decreases continuously beyond the photodissociation region (PDR),
extending to approximately 1-4 times the radius of the PDR (r_PDR). The escape
fraction f_esc of ionizing photons, derived from 1D Ha radial flux profiles, is
~ 0% for ionization-bounded HII regions, while it ranges from 50% to 90% for
the matter-bounded HII regions. The correlation between f_esc and t suggests
that evolved HII regions (with t > 1 Myr) contribute more significantly to
ionizing the surrounding diffuse ionized gas compared to younger, newly formed
HII regions.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 01:28:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Wei",
""
],
[
"Zhao",
"Yunning",
""
],
[
"Ma",
"Lin",
""
],
[
"Wen",
"Shiming",
""
],
[
"Wu",
"Chaojian",
""
],
[
"Ren",
"Juanjuan",
""
],
[
"Chen",
"Jianjun",
""
],
[
"Wu",
"Yuzhong",
""
],
[
"Bai",
"Zhongrui",
""
],
[
"Hou",
"Yonghui",
""
],
[
"Zhao",
"Yongheng",
""
],
[
"Wu",
"Hong",
""
]
] |
2503.08252 | Marco Scutari | Marco Scutari and Samir Salah and Delphine Kerob and Jean Krutmann | Causal Networks of Infodemiological Data: Modelling Dermatitis | 10 pages, 2 figures. Proceedings of the 23rd International Conference
on Artificial Intelligence in Medicine (AIME '25) | null | null | null | stat.ME stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Environmental and mental conditions are known risk factors for dermatitis and
symptoms of skin inflammation, but their interplay is difficult to quantify;
epidemiological studies rarely include both, along with possible confounding
factors. Infodemiology leverages large online data sets to address this issue,
but fusing them produces strong patterns of spatial and temporal correlation,
missingness, and heterogeneity.
In this paper, we design a causal network that correctly models these complex
structures in large-scale infodemiological data from Google, EPA, NOAA and US
Census (434 US counties, 134 weeks). Our model successfully captures known
causal relationships between weather patterns, pollutants, mental conditions,
and dermatitis. Key findings reveal that anxiety accounts for 57.4% of
explained variance in dermatitis, followed by NO2 (33.9%), while environmental
factors show significant mediation effects through mental conditions. The model
predicts that reducing PM2.5 emissions by 25% could decrease dermatitis
prevalence by 18%. Through statistical validation and causal inference, we
provide unprecedented insights into the complex interplay between environmental
and mental health factors affecting dermatitis, offering valuable guidance for
public health policies and environmental regulations.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 10:15:36 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:49:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Scutari",
"Marco",
""
],
[
"Salah",
"Samir",
""
],
[
"Kerob",
"Delphine",
""
],
[
"Krutmann",
"Jean",
""
]
] |
2503.08393 | Joey De Pauw | Joey De Pauw and Bart Goethals | Weighted Tensor Decompositions for Context-aware Collaborative Filtering | Workshop on Context-Aware Recommender Systems, September 18, 2023,
Singapore | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Over recent years it has become well accepted that user interest is not
static or immutable. There are a variety of contextual factors, such as time of
day, the weather or the user's mood, that influence the current interests of
the user. Modelling approaches need to take these factors into account if they
want to succeed at finding the most relevant content to recommend given the
situation.
A popular method for context-aware recommendation is to encode context
attributes as extra dimensions of the classic user-item interaction matrix,
effectively turning it into a tensor, followed by applying the appropriate
tensor decomposition methods to learn missing values. However, unlike with
matrix factorization, where all decompositions are essentially a product of
matrices, there exist many more options for decomposing tensors by combining
vector, matrix and tensor products. We study the most successful decomposition
methods that use weighted square loss and categorize them based on their tensor
structure and regularization strategy. Additionally, we further extend the pool
of methods by filling in the missing combinations.
In this paper we provide an overview of the properties of the different
decomposition methods, such as their complexity, scalability, and modelling
capacity. These benefits are then contrasted with the performances achieved in
offline experiments to gain more insight into which method to choose depending
on a specific situation and constraints.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 12:57:24 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:55:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"De Pauw",
"Joey",
""
],
[
"Goethals",
"Bart",
""
]
] |
2503.08643 | Zheng Zhenxin | Zhenxin Zheng, Zhenjie Zheng | Rethinking Diffusion Model in High Dimension | null | null | null | null | stat.ML cs.AI cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Curse of Dimensionality is an unavoidable challenge in statistical
probability models, yet diffusion models seem to overcome this limitation,
achieving impressive results in high-dimensional data generation. Diffusion
models assume that they can learn the statistical properties of the underlying
probability distribution, enabling sampling from this distribution to generate
realistic samples. But is this really how they work? To address this question,
this paper conducts a detailed analysis of the objective function and inference
methods of diffusion models, leading to several important conclusions that help
answer the above question: 1) In high-dimensional sparse scenarios, the target
of the objective function fitting degrades from a weighted sum of multiple
samples to a single sample. 2) The mainstream inference methods can all be
represented within a simple unified framework, without requiring statistical
concepts such as Markov chains and SDE, while aligning with the degraded
objective function. 3) Guided by this simple framework, more efficient
inference methods can be discovered.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 17:36:11 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 06:15:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zheng",
"Zhenxin",
""
],
[
"Zheng",
"Zhenjie",
""
]
] |
2503.08673 | Jinchang Zhang | Jiakai Lin, Jinchang Zhang, and Guoyu Lu | Keypoint Detection and Description for Raw Bayer Images | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Keypoint detection and local feature description are fundamental tasks in
robotic perception, critical for applications such as SLAM, robot localization,
feature matching, pose estimation, and 3D mapping. While existing methods
predominantly operate on RGB images, we propose a novel network that directly
processes raw images, bypassing the need for the Image Signal Processor (ISP).
This approach significantly reduces hardware requirements and memory
consumption, which is crucial for robotic vision systems. Our method introduces
two custom-designed convolutional kernels capable of performing convolutions
directly on raw images, preserving inter-channel information without converting
to RGB. Experimental results show that our network outperforms existing
algorithms on raw images, achieving higher accuracy and stability under large
rotations and scale variations. This work represents the first attempt to
develop a keypoint detection and feature description network specifically for
raw images, offering a more efficient solution for resource-constrained
environments.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 17:54:12 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 03:11:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lin",
"Jiakai",
""
],
[
"Zhang",
"Jinchang",
""
],
[
"Lu",
"Guoyu",
""
]
] |
2503.08676 | Jinchang Zhang | Jinchang Zhang, Zijun Li, Guoyu Lu | Language-Depth Navigated Thermal and Visible Image Fusion | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Depth-guided multimodal fusion combines depth information from visible and
infrared images, significantly enhancing the performance of 3D reconstruction
and robotics applications. Existing thermal-visible image fusion mainly focuses
on detection tasks, ignoring other critical information such as depth. By
addressing the limitations of single modalities in low-light and complex
environments, the depth information from fused images not only generates more
accurate point cloud data, improving the completeness and precision of 3D
reconstruction, but also provides comprehensive scene understanding for robot
navigation, localization, and environmental perception. This supports precise
recognition and efficient operations in applications such as autonomous driving
and rescue missions. We introduce a text-guided and depth-driven infrared and
visible image fusion network. The model consists of an image fusion branch for
extracting multi-channel complementary information through a diffusion model,
equipped with a text-guided module, and two auxiliary depth estimation
branches. The fusion branch uses CLIP to extract semantic information and
parameters from depth-enriched image descriptions to guide the diffusion model
in extracting multi-channel features and generating fused images. These fused
images are then input into the depth estimation branches to calculate
depth-driven loss, optimizing the image fusion network. This framework aims to
integrate vision-language and depth to directly generate color-fused images
from multimodal inputs.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 17:55:22 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 03:22:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Jinchang",
""
],
[
"Li",
"Zijun",
""
],
[
"Lu",
"Guoyu",
""
]
] |
2503.08915 | Matthieu Terris | Matthieu Terris, Samuel Hurault, Maxime Song, Julian Tachella | Reconstruct Anything Model: a lightweight foundation model for
computational imaging | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Most existing learning-based methods for solving imaging inverse problems can
be roughly divided into two classes: iterative algorithms, such as
plug-and-play and diffusion methods, that leverage pretrained denoisers, and
unrolled architectures that are trained end-to-end for specific imaging
problems. Iterative methods in the first class are computationally costly and
often provide suboptimal reconstruction performance, whereas unrolled
architectures are generally specific to a single inverse problem and require
expensive training. In this work, we propose a novel non-iterative, lightweight
architecture that incorporates knowledge about the forward operator
(acquisition physics and noise parameters) without relying on unrolling. Our
model is trained to solve a wide range of inverse problems beyond denoising,
including deblurring, magnetic resonance imaging, computed tomography,
inpainting, and super-resolution. The proposed model can be easily adapted to
unseen inverse problems or datasets with a few fine-tuning steps (up to a few
images) in a self-supervised way, without ground-truth references. Throughout a
series of experiments, we demonstrate state-of-the-art performance from medical
imaging to low-photon imaging and microscopy.
| [
{
"version": "v1",
"created": "Tue, 11 Mar 2025 21:53:58 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:44:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Terris",
"Matthieu",
""
],
[
"Hurault",
"Samuel",
""
],
[
"Song",
"Maxime",
""
],
[
"Tachella",
"Julian",
""
]
] |
2503.08980 | Yuhang Liu | Yuhang Liu, Dong Gong, Erdun Gao, Zhen Zhang, Biwei Huang, Mingming
Gong, Anton van den Hengel, Javen Qinfeng Shi | I Predict Therefore I Am: Is Next Token Prediction Enough to Learn
Human-Interpretable Concepts from Data? | null | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The remarkable achievements of large language models (LLMs) have led many to
conclude that they exhibit a form of intelligence. This is as opposed to
explanations of their capabilities based on their ability to perform relatively
simple manipulations of vast volumes of data. To illuminate the distinction
between these explanations, we introduce a novel generative model that
generates tokens on the basis of human interpretable concepts represented as
latent discrete variables. Under mild conditions, even when the mapping from
the latent space to the observed space is non-invertible, we establish an
identifiability result: the representations learned by LLMs through next-token
prediction can be approximately modeled as the logarithm of the posterior
probabilities of these latent discrete concepts, up to an invertible linear
transformation. This theoretical finding not only provides evidence that LLMs
capture underlying generative factors, but also strongly reinforces the linear
representation hypothesis, which posits that LLMs learn linear representations
of human-interpretable concepts. Empirically, we validate our theoretical
results through evaluations on both simulation data and the Pythia, Llama, and
DeepSeek model families.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 01:21:17 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 11:00:31 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Yuhang",
""
],
[
"Gong",
"Dong",
""
],
[
"Gao",
"Erdun",
""
],
[
"Zhang",
"Zhen",
""
],
[
"Huang",
"Biwei",
""
],
[
"Gong",
"Mingming",
""
],
[
"Hengel",
"Anton van den",
""
],
[
"Shi",
"Javen Qinfeng",
""
]
] |
2503.09017 | Yuhang Liu | Yuhang Liu, Jindou Jia, Zihan Yang, and Kexin Guo | Accurate Control under Voltage Drop for Rotor Drones | null | null | null | null | eess.SY cs.RO cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This letter proposes an anti-disturbance control scheme for rotor drones to
counteract voltage drop (VD) disturbance caused by voltage drop of the battery,
which is a common case for long-time flight or aggressive maneuvers. Firstly,
the refined dynamics of rotor drones considering VD disturbance are presented.
Based on the dynamics, a voltage drop observer (VDO) is developed to accurately
estimate the VD disturbance by decoupling the disturbance and state information
of the drone, reducing the conservativeness of conventional disturbance
observers. Subsequently, the control scheme integrates the VDO within the
translational loop and a fixed-time sliding mode observer (SMO) within the
rotational loop, enabling it to address force and torque disturbances caused by
voltage drop of the battery. Sufficient real flight experiments are conducted
to demonstrate the effectiveness of the proposed control scheme under VD
disturbance.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 03:13:32 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Apr 2025 09:16:12 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 02:34:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Yuhang",
""
],
[
"Jia",
"Jindou",
""
],
[
"Yang",
"Zihan",
""
],
[
"Guo",
"Kexin",
""
]
] |
2503.09142 | Jian-Qiao Meng | Ba-Lei Tan, Chen Zhang, Qi-Yi Wu, Guo-Hao Dong, Hao Liu, Bo Chen,
Jiao-Jiao Song, Xin-Yi Tian, Ying Zhou, Hai-Yun Liu, Yu-Xia Duan, You-Guo
Shi, and Jian-Qiao Meng | Anisotropic Hybridization Dynamics in the Quasi-One-Dimensional Kondo
Lattice CeCo$_2$Ga$_8$ Revealed by Ultrafast Optical Spectroscopy | 6 pages, 3 figures; Accepted by Frontiers of Physics | Frontiers of Physics 20(4), 044208(2025) | 0.15302/frontphys.2025.044208 | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the ultrafast dynamics of the quasi-one-dimensional Kondo
lattice CeCo$_2$Ga$_8$ using optical pump-probe spectroscopy. Time-resolved
pump-probe reflectivity measurements reveal a strong anisotropy in the
photoinduced response, which is a direct consequence of the material's unique
electronic structure. The temperature dependence of the relaxation dynamics
provides evidence for the formation of two distinct hybridization gaps that
appear at different temperatures in the heavy fermion state. A direct gap of
2$\Delta_{dir}$ $\approx$ 50 meV that persists up to $T^\dag$ $\approx$ 90 K,
well above the coherence temperature $T^*$ $\approx$ 20 K. We attribute this
higher-temperature gap to the hybridization fluctuations. An indirect gap of
2$\Delta_{ind}$ $\approx$ 10 meV opens closer to $T^*$, signifying the
development of long-range coherence in the heavy fermion state. Furthermore, we
find that the hybridization gap can be suppressed with increasing pump fluence,
indicating a delicate interplay between photoexcitation and the coherent heavy
fermion state. Our results provide insights into the interplay of Kondo physics
and low dimensionality in CeCo$_2$Ga$_8$, and establish ultrafast optical
spectroscopy as a sensitive probe of anisotropic hybridization in heavy fermion
materials.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 08:10:14 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Tan",
"Ba-Lei",
""
],
[
"Zhang",
"Chen",
""
],
[
"Wu",
"Qi-Yi",
""
],
[
"Dong",
"Guo-Hao",
""
],
[
"Liu",
"Hao",
""
],
[
"Chen",
"Bo",
""
],
[
"Song",
"Jiao-Jiao",
""
],
[
"Tian",
"Xin-Yi",
""
],
[
"Zhou",
"Ying",
""
],
[
"Liu",
"Hai-Yun",
""
],
[
"Duan",
"Yu-Xia",
""
],
[
"Shi",
"You-Guo",
""
],
[
"Meng",
"Jian-Qiao",
""
]
] |
2503.09329 | Hannes Waclawek | Hannes Waclawek and Stefan Huber | Energy Optimized Piecewise Polynomial Approximation Utilizing Modern
Machine Learning Optimizers | Submitted to AI4IP 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This work explores an extension of machine learning-optimized piecewise
polynomial approximation by incorporating energy optimization as an additional
objective. Traditional closed-form solutions enable continuity and
approximation targets but lack flexibility in accommodating complex
optimization goals. By leveraging modern gradient descent optimizers within
TensorFlow, we introduce a framework that minimizes elastic strain energy in
cam profiles, leading to smoother motion. Experimental results confirm the
effectiveness of this approach, demonstrating its potential to
Pareto-efficiently trade approximation quality against energy consumption.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 12:23:15 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 10:53:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Waclawek",
"Hannes",
""
],
[
"Huber",
"Stefan",
""
]
] |
2503.09849 | Sean Dallas | Sean Dallas (1), Hongjiao Qiang (2), Motaz AbuHijleh (1), Wonse Jo
(2), Kayla Riegner (3), Jon Smereka (3), Lionel Robert (2), Wing-Yue Louie
(1), and Dawn M. Tilbury (2) ((1) Oakland University, (2) University of
Michigan, (3) Ground Vehicle Systems Center (GVSC)) | Training Human-Robot Teams by Improving Transparency Through a Virtual
Spectator Interface | 7 pages, 4 figures, Accepted to ICRA 2025 | null | 10.7302/25266 | null | cs.HC cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | After-action reviews (AARs) are professional discussions that help operators
and teams enhance their task performance by analyzing completed missions with
peers and professionals. Previous studies that compared different formats of
AARs have mainly focused on human teams. However, the inclusion of robotic
teammates brings along new challenges in understanding teammate intent and
communication. Traditional AAR between human teammates may not be satisfactory
for human-robot teams. To address this limitation, we propose a new training
review (TR) tool, called the Virtual Spectator Interface (VSI), to enhance
human-robot team performance and situational awareness (SA) in a simulated
search mission. The proposed VSI primarily utilizes visual feedback to review
subjects' behavior. To examine the effectiveness of VSI, we took elements from
AAR to conduct our own TR, designed a 1 x 3 between-subjects experiment with
experimental conditions: TR with (1) VSI, (2) screen recording, and (3)
non-technology (only verbal descriptions). The results of our experiments
demonstrated that the VSI did not result in significantly better team
performance than other conditions. However, the TR with VSI led to more
improvement in the subjects SA over the other conditions.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 21:13:34 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 22:20:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dallas",
"Sean",
""
],
[
"Qiang",
"Hongjiao",
""
],
[
"AbuHijleh",
"Motaz",
""
],
[
"Jo",
"Wonse",
""
],
[
"Riegner",
"Kayla",
""
],
[
"Smereka",
"Jon",
""
],
[
"Robert",
"Lionel",
""
],
[
"Louie",
"Wing-Yue",
""
],
[
"Tilbury",
"Dawn M.",
""
]
] |
2503.09909 | Hyuga Yoshizaki | Yoshinori Kanamura, Hyuga Yoshizaki | On some periodic continued fractions along the $\mathbb{Z}_2$ extension
over $\mathbb{Q}$ | 15 pages, 2 figures | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In 2021, Brock, Elkies, and Jordan generalized the theory of periodic
continued fractions (PCFs) over $\mathbb{Z}$ to the ring of integers in a
number field. In particular, they considered the case where the number field is
an intermediate field of the $\mathbb{Z}_2$-extension over $\mathbb{Q}$ and
asked whether a $(N, \ell)$-type PCF for $X_n = 2\cos(2\pi/2^{n+2})$ exists. In
this paper, we construct $(1,2)$ and $(0,3)$-type PCFs for $X_n$ for all
$n\geq1$. To the best of our knowledge, this is the first explicit construction
of type (0,3) continued fractions for all $n\geq1$. To obtain such results, for
each type, we construct a bijection between a certain subset of the group of
relative units in each layer of the $\mathbb{Z}_2$-extension and the set of
PCFs for $X_n$. While our result confirms the existence of such PCFs for all
$n\geq1$ in types $(1,2)$ and $(0,3)$, determining all PCFs remains an open
problem. The bijections constructed in our result translate this problem into
the study of the subsets of the relative units. As a second main result, we
give explicit bounds for the logarithms of the relative units corresponding to
$(1,2)$ or $(0,3)$-type PCFs for $X_n$. These bounds allow us to explain
interesting phenomena observed in the distribution of such points.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 23:57:34 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 14:46:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kanamura",
"Yoshinori",
""
],
[
"Yoshizaki",
"Hyuga",
""
]
] |
2503.11297 | YuHao Du | Yuhao Du, Hui Liu, Haoxiang Peng, Xinyuan Cheng, Chenrong Wu, Jiankai
Zhang | GMG: A Video Prediction Method Based on Global Focus and Motion Guided | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent years, weather forecasting has gained significant attention. However,
accurately predicting weather remains a challenge due to the rapid variability
of meteorological data and potential teleconnections. Current spatiotemporal
forecasting models primarily rely on convolution operations or sliding windows
for feature extraction. These methods are limited by the size of the
convolutional kernel or sliding window, making it difficult to capture and
identify potential teleconnection features in meteorological data.
Additionally, weather data often involve non-rigid bodies, whose motion
processes are accompanied by unpredictable deformations, further complicating
the forecasting task. In this paper, we propose the GMG model to address these
two core challenges. The Global Focus Module, a key component of our model,
enhances the global receptive field, while the Motion Guided Module adapts to
the growth or dissipation processes of non-rigid bodies. Through extensive
evaluations, our method demonstrates competitive performance across various
complex tasks, providing a novel approach to improving the predictive accuracy
of complex spatiotemporal data.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 11:06:49 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 14:06:13 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Du",
"Yuhao",
""
],
[
"Liu",
"Hui",
""
],
[
"Peng",
"Haoxiang",
""
],
[
"Cheng",
"Xinyuan",
""
],
[
"Wu",
"Chenrong",
""
],
[
"Zhang",
"Jiankai",
""
]
] |
2503.11562 | Franco Caspe | Franco Caspe, Jordie Shier, Mark Sandler, Charalampos Saitis and
Andrew McPherson | Designing Neural Synthesizers for Low-Latency Interaction | See website at fcaspe.github.io/brave - 13 pages, 5 figures, accepted
to the Journal of the Audio Engineering Society, LaTeX; Corrected typos,
added hyphen to title to reflect JAES version | null | null | null | cs.SD cs.AI cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | Neural Audio Synthesis (NAS) models offer interactive musical control over
high-quality, expressive audio generators. While these models can operate in
real-time, they often suffer from high latency, making them unsuitable for
intimate musical interaction. The impact of architectural choices in deep
learning models on audio latency remains largely unexplored in the NAS
literature. In this work, we investigate the sources of latency and jitter
typically found in interactive NAS models. We then apply this analysis to the
task of timbre transfer using RAVE, a convolutional variational autoencoder for
audio waveforms introduced by Caillon et al. in 2021. Finally, we present an
iterative design approach for optimizing latency. This culminates with a model
we call BRAVE (Bravely Realtime Audio Variational autoEncoder), which is
low-latency and exhibits better pitch and loudness replication while showing
timbre modification capabilities similar to RAVE. We implement it in a
specialized inference framework for low-latency, real-time inference and
present a proof-of-concept audio plugin compatible with audio signals from
musical instruments. We expect the challenges and guidelines described in this
document to support NAS researchers in designing models for low-latency
inference from the ground up, enriching the landscape of possibilities for
musicians.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 16:30:31 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 18:00:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Caspe",
"Franco",
""
],
[
"Shier",
"Jordie",
""
],
[
"Sandler",
"Mark",
""
],
[
"Saitis",
"Charalampos",
""
],
[
"McPherson",
"Andrew",
""
]
] |
2503.11641 | William Simon | William A. Simon, Carter M. Gustin, Kamil Serafin, Alexis Ralli, Gary
R. Goldstein, Peter J. Love | Ladder Operator Block-Encoding | edit: added or elaborated references; added Mason Rhodes to
acknowledgments; opted for use of "ladder operator" instead of
second-quantized operator for precision | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe and analyze LOBE (Ladder Operator Block-Encoding), a framework
for block-encoding second-quantized ladder operators that act upon fermionic
and bosonic modes. We numerically benchmark these constructions using models
arising in quantum field theories including the quartic oscillator, and
$\phi^4$ and Yukawa Hamiltonians on the light front. These benchmarks show that
LOBE produces block-encodings with fewer non-Clifford operations, fewer
block-encoding ancillae and overall number of qubits, and lower rescaling
factors for various second-quantized operators as compared to block-encoding
frameworks that expand the ladder operators in the Pauli basis. The LOBE
constructions also demonstrate favorable scaling with respect to key
parameters, including the maximum occupation of bosonic modes, the total number
of fermionic and bosonic modes, and the locality of the operators. LOBE is
implemented as an open source python package to enable further applications.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 17:57:21 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:46:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Simon",
"William A.",
""
],
[
"Gustin",
"Carter M.",
""
],
[
"Serafin",
"Kamil",
""
],
[
"Ralli",
"Alexis",
""
],
[
"Goldstein",
"Gary R.",
""
],
[
"Love",
"Peter J.",
""
]
] |
2503.11736 | Onur Beker | Onur Beker, Nico G\"urtler, Ji Shi, A. Ren\'e Geist, Amirreza Razmjoo,
Georg Martius, Sylvain Calinon | A Smooth Analytical Formulation of Collision Detection and Rigid Body
Dynamics With Contact | Added references to point-based implicit surface representations | null | null | null | cs.RO math.OC | http://creativecommons.org/licenses/by/4.0/ | Generating intelligent robot behavior in contact-rich settings is a research
problem where zeroth-order methods currently prevail. A major contributor to
the success of such methods is their robustness in the face of non-smooth and
discontinuous optimization landscapes that are characteristic of contact
interactions, yet zeroth-order methods remain computationally inefficient. It
is therefore desirable to develop methods for perception, planning and control
in contact-rich settings that can achieve further efficiency by making use of
first and second order information (i.e., gradients and Hessians). To
facilitate this, we present a joint formulation of collision detection and
contact modelling which, compared to existing differentiable simulation
approaches, provides the following benefits: i) it results in forward and
inverse dynamics that are entirely analytical (i.e. do not require solving
optimization or root-finding problems with iterative methods) and smooth (i.e.
twice differentiable), ii) it supports arbitrary collision geometries without
needing a convex decomposition, and iii) its runtime is independent of the
number of contacts. Through simulation experiments, we demonstrate the validity
of the proposed formulation as a "physics for inference" that can facilitate
future development of efficient methods to generate intelligent contact-rich
behavior.
| [
{
"version": "v1",
"created": "Fri, 14 Mar 2025 14:02:12 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:57:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Beker",
"Onur",
""
],
[
"Gürtler",
"Nico",
""
],
[
"Shi",
"Ji",
""
],
[
"Geist",
"A. René",
""
],
[
"Razmjoo",
"Amirreza",
""
],
[
"Martius",
"Georg",
""
],
[
"Calinon",
"Sylvain",
""
]
] |
2503.12047 | Hangrui Xu | Hangrui Xu, Chuanrui Zhang, Zhengxian Wu, Peng Jiao, Haoqian Wang | PSGait: Gait Recognition using Parsing Skeleton | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gait recognition has emerged as a robust biometric modality due to its
non-intrusive nature and resilience to occlusion. Conventional gait recognition
methods typically rely on silhouettes or skeletons. Despite their success in
gait recognition for controlled laboratory environments, they usually fail in
real-world scenarios due to their limited information entropy for gait
representations. To achieve accurate gait recognition in the wild, we propose a
novel gait representation, named Parsing Skeleton. This representation
innovatively introduces the skeleton-guided human parsing method to capture
fine-grained body dynamics, so they have much higher information entropy to
encode the shapes and dynamics of fine-grained human parts during walking.
Moreover, to effectively explore the capability of the Parsing Skeleton
representation, we propose a novel Parsing Skeleton-based gait recognition
framework, named PSGait, which takes Parsing Skeletons and silhouettes as
input. By fusing these two modalities, the resulting image sequences are fed
into gait recognition models for enhanced individual differentiation. We
conduct comprehensive benchmarks on various datasets to evaluate our model.
PSGait outperforms existing state-of-the-art multimodal methods that utilize
both skeleton and silhouette inputs while significantly reducing computational
resources. Furthermore, as a plug-and-play method, PSGait leads to a maximum
improvement of 10.9% in Rank-1 accuracy across various gait recognition models.
These results demonstrate that Parsing Skeleton offers a lightweight,
effective, and highly generalizable representation for gait recognition in the
wild.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 08:38:47 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 15:46:04 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xu",
"Hangrui",
""
],
[
"Zhang",
"Chuanrui",
""
],
[
"Wu",
"Zhengxian",
""
],
[
"Jiao",
"Peng",
""
],
[
"Wang",
"Haoqian",
""
]
] |
2503.12143 | Maryam Daniali | Maryam Daniali, Shivaram Karandikar, Dabriel Zimmerman, J. Eric
Schmitt, Matthew J. Buczek, Benjamin Jung, Laura Mercedes, Jakob Seidlitz,
Vanessa Troiani, Lena Dorfschmidt, Eren Kafadar, Remo Williams, Susan
Sotardi, Arastoo Vossough, Scott Haag, Jenna M. Schabdach, Aaron
Alexander-Bloch | Language Models for Automated Classification of Brain MRI Reports and
Growth Chart Generation | null | null | null | null | eess.IV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Clinically acquired brain MRIs and radiology reports are valuable but
underutilized resources due to the challenges of manual analysis and data
heterogeneity. We developed fine-tuned language models (LMs) to classify brain
MRI reports as normal (reports with limited pathology) or abnormal, fine-tuning
BERT, BioBERT, ClinicalBERT, and RadBERT on 44,661 reports. We also explored
the reasoning capabilities of a leading LM, Gemini 1.5-Pro, for normal report
categorization. Automated image processing and modeling generated brain growth
charts from LM-classified normal scans, comparing them to human-derived charts.
Fine-tuned LMs achieved high classification performance (F1-Score >97%), with
unbalanced training mitigating class imbalance. Performance was robust on
out-of-distribution data, with full text outperforming summary (impression)
sections. Gemini 1.5-Pro showed a promising categorization performance,
especially with clinical inference. LM-derived brain growth charts were nearly
identical to human-annotated charts (r = 0.99, p < 2.2e-16). Our LMs offer
scalable analysis of radiology reports, enabling automated classification of
brain MRIs in large datasets. One application is automated generation of brain
growth charts for benchmarking quantitative image features. Further research is
needed to address data heterogeneity and optimize LM reasoning.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 13:59:44 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Daniali",
"Maryam",
""
],
[
"Karandikar",
"Shivaram",
""
],
[
"Zimmerman",
"Dabriel",
""
],
[
"Schmitt",
"J. Eric",
""
],
[
"Buczek",
"Matthew J.",
""
],
[
"Jung",
"Benjamin",
""
],
[
"Mercedes",
"Laura",
""
],
[
"Seidlitz",
"Jakob",
""
],
[
"Troiani",
"Vanessa",
""
],
[
"Dorfschmidt",
"Lena",
""
],
[
"Kafadar",
"Eren",
""
],
[
"Williams",
"Remo",
""
],
[
"Sotardi",
"Susan",
""
],
[
"Vossough",
"Arastoo",
""
],
[
"Haag",
"Scott",
""
],
[
"Schabdach",
"Jenna M.",
""
],
[
"Alexander-Bloch",
"Aaron",
""
]
] |
2503.12238 | V Varagapriya | V Varagapriya | Transition Uncertainties in Constrained Markov Decision Models: A Robust
Optimization Approach | null | null | null | null | math.OC | http://creativecommons.org/publicdomain/zero/1.0/ | We examine a constrained Markov decision process under uncertain transition
probabilities, with the uncertainty modeled as deviations from observed
transition probabilities. We construct the uncertainty set associated with the
deviations using polyhedral and second-order cone constraints and employ a
robust optimization framework. We demonstrate that each inner optimization
problem of the robust model can be equivalently transformed into a second-order
cone programming problem. Using strong duality arguments, we show that the
resulting robust problem can be equivalently reformulated into a second-order
cone programming problem with bilinear constraints. In the numerical
experiments, we study a machine replacement problem and explore potential
sources of uncertainty in the transition probabilities. We examine how the
optimal values and solutions differ as we vary the feasible region of the
uncertainty set, considering only polyhedral constraints and a combination of
polyhedral and second-order cone constraints. Furthermore, we analyze the
impact of the number of states, the discount factor, and variations in the
feasible region of the uncertainty set on the optimal values.
| [
{
"version": "v1",
"created": "Sat, 15 Mar 2025 19:23:54 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 12:51:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Varagapriya",
"V",
""
]
] |
2503.12414 | YaBing Zuo | Ya-Bing Zuo, Jia-Yu Zou, Shi-Yu Liang, Ming-Ge Li, Shan-Shan Hu | $B \rightarrow TT$ decays in the QCD factorization approach | 23 pages,13 tables | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | In this study, the nonleptonic two-body $B$ decays into two tensor mesons
(including $a_2(1320)$, $K^*_2(1430)$, $f_2(1270)$, $f^\prime_2(1525)$, denoted
generically as $T$) are investigated in the QCD factorization approach. The
branching ratios, longitudinal polarization fractions, and CP asymmetries are
predicted systematically. It is found that, from the perspective of central
values, the branching ratios for the decays $B \rightarrow \{a_2 f^\prime_2,
f_2 f^\prime_2\}$, $\{a_2f_2, f_2 f_2\}$ and $\{f_2 K^*_2, K^*_2f^\prime_2\}$
are at the order of $10^{-8}$, $10^{-7}$ and $10^{-6}$, respectively. The
branching ratios for the decays into $\{a_2 a_2, K^*_2 K^*_2\}$ and $a_2 K^*_2$
are at the order of $10^{-8} - 10^{-7}$ and $10^{-7} - 10^{-6}$, respectively.
The longitudinal polarization fractions are approximately $0.7-1$. Especially,
for the $\bar{B}^0/B^0 \rightarrow \{K^{*-}_2 K^{*+}_2, f^\prime_2
f^\prime_2\}$ decays, the longitudinal polarization fractions are equal to $1$.
The CP asymmetry for the $B^\pm \rightarrow a^\pm_2 f_2$ modes is most
significant, roughly $-22\%$. The CP asymmetries for the decays $B \rightarrow
\{a_2 f^\prime_2, f_2 f^\prime_2, f^\prime_2 f^\prime_2\}$ and $\bar{B}^0/B^0
\rightarrow K^{*-}_2 K^{*+}_2$ are equal to zero. Our results may be tested by
more precise experiments in the future.
| [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 08:53:33 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 03:52:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zuo",
"Ya-Bing",
""
],
[
"Zou",
"Jia-Yu",
""
],
[
"Liang",
"Shi-Yu",
""
],
[
"Li",
"Ming-Ge",
""
],
[
"Hu",
"Shan-Shan",
""
]
] |
2503.12419 | Kailun Yang | Luming Wang, Hao Shi, Xiaoting Yin, Kailun Yang, Kaiwei Wang, Jian Bai | EgoEvGesture: Gesture Recognition Based on Egocentric Event Camera | The dataset and models are made available at
https://github.com/3190105222/EgoEv_Gesture | null | null | null | cs.CV cs.RO eess.IV physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Egocentric gesture recognition is a pivotal technology for enhancing natural
human-computer interaction, yet traditional RGB-based solutions suffer from
motion blur and illumination variations in dynamic scenarios. While event
cameras show distinct advantages in handling high dynamic range with ultra-low
power consumption, existing RGB-based architectures face inherent limitations
in processing asynchronous event streams due to their synchronous frame-based
nature. Moreover, from an egocentric perspective, event cameras record data
that includes events generated by both head movements and hand gestures,
thereby increasing the complexity of gesture recognition. To address this, we
propose a novel network architecture specifically designed for event data
processing, incorporating (1) a lightweight CNN with asymmetric depthwise
convolutions to reduce parameters while preserving spatiotemporal features, (2)
a plug-and-play state-space model as context block that decouples head movement
noise from gesture dynamics, and (3) a parameter-free Bins-Temporal Shift
Module (BSTM) that shifts features along bins and temporal dimensions to fuse
sparse events efficiently. We further establish the EgoEvGesture dataset, the
first large-scale dataset for egocentric gesture recognition using event
cameras. Experimental results demonstrate that our method achieves 62.7%
accuracy tested on unseen subjects with only 7M parameters, 3.1% higher than
state-of-the-art approaches. Notable misclassifications in freestyle motions
stem from high inter-personal variability and unseen test patterns differing
from training data. Moreover, our approach achieved a remarkable accuracy of
97.0% on the DVS128 Gesture, demonstrating the effectiveness and generalization
capability of our method on public datasets. The dataset and models are made
available at https://github.com/3190105222/EgoEv_Gesture.
| [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 09:08:02 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 02:44:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Luming",
""
],
[
"Shi",
"Hao",
""
],
[
"Yin",
"Xiaoting",
""
],
[
"Yang",
"Kailun",
""
],
[
"Wang",
"Kaiwei",
""
],
[
"Bai",
"Jian",
""
]
] |
2503.12427 | Bocheng Wang | Bocheng Wang, Chusheng Zeng, Mulin Chen, Xuelong Li | Towards Learnable Anchor for Deep Multi-View Clustering | Accepted by AAAI25 | Vol. 39, No. 20, 2025 | 10.1609/aaai.v39i20.35401 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Deep multi-view clustering incorporating graph learning has presented
tremendous potential. Most methods encounter costly square time consumption
w.r.t. data size. Theoretically, anchor-based graph learning can alleviate this
limitation, but related deep models mainly rely on manual discretization
approaches to select anchors, which indicates that 1) the anchors are fixed
during model training and 2) they may deviate from the true cluster
distribution. Consequently, the unreliable anchors may corrupt clustering
results. In this paper, we propose the Deep Multi-view Anchor Clustering (DMAC)
model that performs clustering in linear time. Concretely, the initial anchors
are intervened by the positive-incentive noise sampled from Gaussian
distribution, such that they can be optimized with a newly designed anchor
learning loss, which promotes a clear relationship between samples and anchors.
Afterwards, anchor graph convolution is devised to model the cluster structure
formed by the anchors, and the mutual information maximization loss is built to
provide cross-view clustering guidance. In this way, the learned anchors can
better represent clusters. With the optimal anchors, the full sample graph is
calculated to derive a discriminative embedding for clustering. Extensive
experiments on several datasets demonstrate the superior performance and
efficiency of DMAC compared to state-of-the-art competitors.
| [
{
"version": "v1",
"created": "Sun, 16 Mar 2025 09:38:11 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Bocheng",
""
],
[
"Zeng",
"Chusheng",
""
],
[
"Chen",
"Mulin",
""
],
[
"Li",
"Xuelong",
""
]
] |
2503.12721 | Luca Collini | Luca Collini, Andrew Hennessee, Ramesh Karri, Siddharth Garg | Can Reasoning Models Reason about Hardware? An Agentic HLS Perspective | 7 pages, submitted for peer review | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recent Large Language Models (LLMs) such as OpenAI o3-mini and DeepSeek-R1
use enhanced reasoning through Chain-of-Thought (CoT). Their potential in
hardware design, which relies on expert-driven iterative optimization, remains
unexplored. This paper investigates whether reasoning LLMs can address
challenges in High-Level Synthesis (HLS) design space exploration and
optimization. During HLS, engineers manually define pragmas/directives to
balance performance and resource constraints. We propose an LLM-based
optimization agentic framework that automatically restructures code, inserts
pragmas, and identifies optimal design points via feedback from HLs tools and
access to integer-linear programming (ILP) solvers. Experiments compare
reasoning models against conventional LLMs on benchmarks using success rate,
efficiency, and design quality (area/latency) metrics, and provide the
first-ever glimpse into the CoTs produced by a powerful open-source reasoning
model like DeepSeek-R1.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 01:21:39 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 00:39:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Collini",
"Luca",
""
],
[
"Hennessee",
"Andrew",
""
],
[
"Karri",
"Ramesh",
""
],
[
"Garg",
"Siddharth",
""
]
] |
2503.12768 | Kanji Tanaka | Tatsuro Sakai, Kanji Tanaka, Jonathan Tay Yu Liang, Muhammad Adil
Luqman, Daiki Iwata | Dynamic-Dark SLAM: RGB-Thermal Cooperative Robot Vision Strategy for
Multi-Person Tracking in Both Well-Lit and Low-Light Scenes | 10 pages, 9 figures, technical report | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In robot vision, thermal cameras hold great potential for recognizing humans
even in complete darkness. However, their application to multi-person tracking
(MPT) has been limited due to data scarcity and the inherent difficulty of
distinguishing individuals. In this study, we propose a cooperative MPT system
that utilizes co-located RGB and thermal cameras, where pseudo-annotations
(bounding boxes and person IDs) are used to train both RGB and thermal
trackers. Evaluation experiments demonstrate that the thermal tracker performs
robustly in both bright and dark environments. Moreover, the results suggest
that a tracker-switching strategy -- guided by a binary brightness classifier
-- is more effective for information integration than a tracker-fusion
approach. As an application example, we present an image change pattern
recognition (ICPR) method, the ``human-as-landmark,'' which combines two key
properties: the thermal recognizability of humans in dark environments and the
rich landmark characteristics -- appearance, geometry, and semantics -- of
static objects (occluders). Whereas conventional SLAM focuses on mapping static
landmarks in well-lit environments, the present study takes a first step toward
a new Human-Only SLAM paradigm, ``DD-SLAM,'' which aims to map even dynamic
landmarks in complete darkness.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 03:05:21 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 01:26:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sakai",
"Tatsuro",
""
],
[
"Tanaka",
"Kanji",
""
],
[
"Liang",
"Jonathan Tay Yu",
""
],
[
"Luqman",
"Muhammad Adil",
""
],
[
"Iwata",
"Daiki",
""
]
] |
2503.12885 | Dewei Zhou | Dewei Zhou, Mingwei Li, Zongxin Yang, Yi Yang | DreamRenderer: Taming Multi-Instance Attribute Control in Large-Scale
Text-to-Image Models | 11 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image-conditioned generation methods, such as depth- and canny-conditioned
approaches, have demonstrated remarkable abilities for precise image synthesis.
However, existing models still struggle to accurately control the content of
multiple instances (or regions). Even state-of-the-art models like FLUX and
3DIS face challenges, such as attribute leakage between instances, which limits
user control. To address these issues, we introduce DreamRenderer, a
training-free approach built upon the FLUX model. DreamRenderer enables users
to control the content of each instance via bounding boxes or masks, while
ensuring overall visual harmony. We propose two key innovations: 1) Bridge
Image Tokens for Hard Text Attribute Binding, which uses replicated image
tokens as bridge tokens to ensure that T5 text embeddings, pre-trained solely
on text data, bind the correct visual attributes for each instance during Joint
Attention; 2) Hard Image Attribute Binding applied only to vital layers.
Through our analysis of FLUX, we identify the critical layers responsible for
instance attribute rendering and apply Hard Image Attribute Binding only in
these layers, using soft binding in the others. This approach ensures precise
control while preserving image quality. Evaluations on the COCO-POS and
COCO-MIG benchmarks demonstrate that DreamRenderer improves the Image Success
Ratio by 17.7% over FLUX and enhances the performance of layout-to-image models
like GLIGEN and 3DIS by up to 26.8%. Project Page:
https://limuloo.github.io/DreamRenderer/.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 07:30:16 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 13:35:37 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhou",
"Dewei",
""
],
[
"Li",
"Mingwei",
""
],
[
"Yang",
"Zongxin",
""
],
[
"Yang",
"Yi",
""
]
] |
2503.12899 | Jian Gu | Jian Gu, Aldeida Aleti, Chunyang Chen, Hongyu Zhang | A Semantic-based Optimization Approach for Repairing LLMs: Case Study on
Code Generation | 12 pages, 6 figure, 6 tables, under peer-review | null | null | null | cs.SE cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Language Models (LMs) are widely used in software engineering for code
generation, but they may produce code with errors. Rather than repairing the
generated code, an alternative way is to address the underlying failures of
models. LM repair offers a lightweight solution to this challenge: it requires
minimal data, reduces computational costs, and reduces the side effects. Unlike
retraining, LM repair focuses on applying tailored updates to targeted neurons,
making it ideal for scenarios with limited resources, high-performance demands,
or strict safety requirements. In this paper, we propose \ul{S}emantic
\ul{T}argeting for \ul{A}nalytical \ul{R}epair (\textsc{STAR}), a pioneering
and novel semantic-based optimization approach for repairing LLMs.
\textsc{STAR} realizes main operations in LM repair methods in an optimization
process, including locating ``buggy neurons'', solving ``neuron patches'', and
patching ``buggy neurons''. Correspondingly, it computes the deltas of weight
matrix as the prior information to guide optimization; and attributes the
targeted layers and neurons leveraging statistical insights. The neuron patches
are computed with a solid semantic-based analytical formula, which directly
bridges the changes to logits with the deltas of neurons, by steering latent
representations. Compared to the prior work of LM repair (\textsc{MINT}) and
optimization methods (\textsc{SGD}), \textsc{STAR} integrates their strengths
while mitigating their limitations. \textsc{STAR} supports solving multiple
failures together, significantly improving the usefulness. Evaluated on three
code generation tasks using popular code LMs, \textsc{STAR} demonstrates
superior effectiveness. Additionally, \textsc{STAR} exhibits better efficiency.
In terms of side effects, namely the balance between generalization and
specificity, \textsc{STAR} outperforms prior work by a significant margin.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 07:59:42 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 13:57:28 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gu",
"Jian",
""
],
[
"Aleti",
"Aldeida",
""
],
[
"Chen",
"Chunyang",
""
],
[
"Zhang",
"Hongyu",
""
]
] |
2503.13061 | Partha Nandi | Partha Nandi, Tiasha Bhattacharyya, A. S. Majumdar, Graeme Pleasance,
Francesco Petruccione | Neutrino Decoherence in kappa-Minkowski Quantum Spacetime: An Open
Quantum Systems Paradigm | This manuscript spans 13 pages and features 4 figures that illustrate
our results. We have also added new references | null | null | null | hep-th gr-qc hep-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate neutrino decoherence within the framework of quantum
spacetime, focusing on the $\kappa$-Minkowski model. We show that stochastic
fluctuations in quantum spacetime induce an energy-dependent decoherence
effect, where the decoherence rate scales as $E^{-4}$. This result aligns with
recent IceCube observations, indicating that quantum gravity does not induce
significant decoherence for high-energy neutrinos. Additionally, we establish
conditions under which quantum spacetime effects could influence relic
neutrinos, such as those in the cosmic neutrino background ($C\nu B$). Our
results shed light on how quantum spacetime fluctuations impact neutrino
oscillation physics.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 11:08:14 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 10:33:11 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Nandi",
"Partha",
""
],
[
"Bhattacharyya",
"Tiasha",
""
],
[
"Majumdar",
"A. S.",
""
],
[
"Pleasance",
"Graeme",
""
],
[
"Petruccione",
"Francesco",
""
]
] |
2503.13116 | Zeng Wang | Zeng Wang, Minghao Shao, Mohammed Nabeel, Prithwish Basu Roy, Likhitha
Mankali, Jitendra Bhandari, Ramesh Karri, Ozgur Sinanoglu, Muhammad Shafique,
Johann Knechtel | VeriLeaky: Navigating IP Protection vs Utility in Fine-Tuning for
LLM-Driven Verilog Coding | null | null | null | null | cs.CR cs.AR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) offer significant potential for coding, yet
fine-tuning (FT) with curated data is essential for niche languages like
Verilog. Using proprietary intellectual property (IP) for FT presents a serious
risk, as FT data can be leaked through LLM inference. This leads to a critical
dilemma for design houses: seeking to build externally accessible LLMs offering
competitive Verilog coding, how can they leverage in-house IP to enhance FT
utility while ensuring IP protection?
For the first time in the literature, we study this dilemma. Using LLaMA
3.1-8B, we conduct in-house FT on a baseline Verilog dataset (RTLCoder)
supplemented with our own in-house IP, which is validated through multiple
tape-outs. To rigorously assess IP leakage, we quantify structural similarity
(AST/Dolos) and functional equivalence (Synopsys Formality) between generated
codes and our in-house IP. We show that our IP can indeed be leaked, confirming
the threat. As defense, we evaluate logic locking of Verilog codes (ASSURE).
This offers some level of protection, yet reduces the IP's utility for FT and
degrades the LLM's performance. Our study shows the need for novel strategies
that are both effective and minimally disruptive to FT, an essential effort for
enabling design houses to fully utilize their proprietary IP toward LLM-driven
Verilog coding.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:38:03 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 11:28:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Zeng",
""
],
[
"Shao",
"Minghao",
""
],
[
"Nabeel",
"Mohammed",
""
],
[
"Roy",
"Prithwish Basu",
""
],
[
"Mankali",
"Likhitha",
""
],
[
"Bhandari",
"Jitendra",
""
],
[
"Karri",
"Ramesh",
""
],
[
"Sinanoglu",
"Ozgur",
""
],
[
"Shafique",
"Muhammad",
""
],
[
"Knechtel",
"Johann",
""
]
] |
2503.13128 | Martin Roetteler | Willie Aboumrad, Daiwei Zhu, Claudio Girotto, Fran\c{c}ois-Henry
Rouet, Jezer Jojo, Robert Lucas, Jay Pathak, Ananth Kaushik, Martin Roetteler | Accelerating large-scale linear algebra using variational quantum
imaginary time evolution | 14 pages, 13 figures, 2 tables | null | null | null | quant-ph cs.ET | http://creativecommons.org/licenses/by/4.0/ | The solution of large sparse linear systems via factorization methods such as
LU or Cholesky decomposition, can be computationally expensive due to the
introduction of non-zero elements, or ``fill-in.'' Graph partitioning can be
used to reduce the ``fill-in,'' thereby speeding up the solution of the linear
system. We introduce a quantum approach to the graph partitioning problem based
on variational quantum imaginary time evolution (VarQITE). We develop a hybrid
quantum/classical method to accelerate Finite Element Analysis (FEA) by using
VarQITE in Ansys's LS-DYNA multiphysics simulation software. This allows us to
study different types of FEA problems, from mechanical engineering to
computational fluid dynamics in simulations and on quantum hardware (IonQ Aria
and IonQ Forte).
We demonstrate that VarQITE has the potential to impact LS-DYNA workflows by
measuring the wall-clock time to solution of FEA problems. We report
performance results for our hybrid quantum/classical workflow on selected FEA
problem instances, including simulation of blood pumps, automotive roof crush,
and vibration analysis of car bodies on meshes of up to six million elements.
We find that the LS-DYNA wall clock time can be improved by up to 12\% for some
problems. Finally, we introduce a classical heuristic inspired by
Fiduccia-Mattheyses to improve the quality of VarQITE solutions obtained from
hardware runs. Our results highlight the potential impact of quantum computing
on large-scale FEA problems in the NISQ era.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:52:25 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:28:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Aboumrad",
"Willie",
""
],
[
"Zhu",
"Daiwei",
""
],
[
"Girotto",
"Claudio",
""
],
[
"Rouet",
"François-Henry",
""
],
[
"Jojo",
"Jezer",
""
],
[
"Lucas",
"Robert",
""
],
[
"Pathak",
"Jay",
""
],
[
"Kaushik",
"Ananth",
""
],
[
"Roetteler",
"Martin",
""
]
] |
2503.13202 | Zhibing Zhang | Zhibing Zhang | New Liouville type theorems for 3D steady incompressible MHD equations
and Hall-MHD equations | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study Liouville type results for the three-dimensional
stationary incompressible MHD equations and Hall-MHD equations. By a new
iteration argument, we establish Liouville type theorems if the velocity and
magnetic field satisfy certain growth conditions of Lebesgue norms on the
annulus. In our iteration procedure, the $L^2$ norms of the gradients of the
velocity and magnetic field and the $L^6$ norms of the velocity and magnetic
field are iterated together. The conditions imposed on the magnetic field are
weaker than the velocity field in certain sense. As a consequence, we show that
the velocity and magnetic field are trivial provided that they belong to some
Lebesgue spaces or satisfy some decay conditions at infinity. Our results
extend and improve the recent works of Chae-Lee (2024 Nonlinearity 37 095006)
and Cho-Neustupa-Yang (2024 Nonlinearity 37 035007).
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:10:18 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 04:00:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Zhibing",
""
]
] |
2503.13208 | Fan Sinan | Sinan Fan, Liang Xie, Chen Shen, Ge Teng, Xiaosong Yuan, Xiaofeng
Zhang, Chenxi Huang, Wenxiao Wang, Xiaofei He, Jieping Ye | Improving Complex Reasoning with Dynamic Prompt Corruption: A soft
prompt Optimization Approach | Accepted by ICLR 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Prompt-tuning (PT) for large language models (LLMs) can facilitate the
performance on various conventional NLP tasks with significantly fewer
trainable parameters. However, our investigation reveals that PT provides
limited improvement and may even degrade the primitive performance of LLMs on
complex reasoning tasks. Such a phenomenon suggests that soft prompts can
positively impact certain instances while negatively affecting others,
particularly during the later phases of reasoning. To address these challenges,
We first identify an information accumulation within the soft prompts. Through
detailed analysis, we demonstrate that this phenomenon is often accompanied by
erroneous information flow patterns in the deeper layers of the model, which
ultimately lead to incorrect reasoning outcomes. we propose a novel method
called Dynamic Prompt Corruption (DPC) to take better advantage of soft prompts
in complex reasoning tasks, which dynamically adjusts the influence of soft
prompts based on their impact on the reasoning process. Specifically, DPC
consists of two stages: Dynamic Trigger and Dynamic Corruption. First, Dynamic
Trigger measures the impact of soft prompts, identifying whether beneficial or
detrimental. Then, Dynamic Corruption mitigates the negative effects of soft
prompts by selectively masking key tokens that interfere with the reasoning
process. We validate the proposed approach through extensive experiments on
various LLMs and reasoning tasks, including GSM8K, MATH, and AQuA. Experimental
results demonstrate that DPC can consistently enhance the performance of PT,
achieving 4%-8% accuracy gains compared to vanilla prompt tuning, highlighting
the effectiveness of our approach and its potential to enhance complex
reasoning in LLMs.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:20:48 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 07:04:25 GMT"
},
{
"version": "v3",
"created": "Sun, 13 Apr 2025 12:38:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fan",
"Sinan",
""
],
[
"Xie",
"Liang",
""
],
[
"Shen",
"Chen",
""
],
[
"Teng",
"Ge",
""
],
[
"Yuan",
"Xiaosong",
""
],
[
"Zhang",
"Xiaofeng",
""
],
[
"Huang",
"Chenxi",
""
],
[
"Wang",
"Wenxiao",
""
],
[
"He",
"Xiaofei",
""
],
[
"Ye",
"Jieping",
""
]
] |
2503.13239 | YanGuang Yue | Yan-Guang Yue, Shuai Yang, Zheng-Xin Liu, and Yan Chen | Universal Rule for Topological Hopf Term via Dirac-Spin Coupling | null | null | null | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is known that the topological Hopf term in two-dimensional (2D) spin
systems can be derived by coupling to massless Dirac fermions. We establish a
universal rule governing the generation of Hopf terms in 2D quantum spin
systems coupled to Dirac fermions. The key insight identifies the Hopf
coefficient as the oriented volume in the $\mathfrak{su}(2)$ Lie algebra space
formed by Dirac cone matrix bases. This geometric interpretation allows direct
determination of Hopf term without path integral computations. Applying this
framework, we demonstrate nontrivial Hopf term in tailored checkerboard lattice
models and recover known results in graphene-based systems.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 14:52:48 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 14:48:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yue",
"Yan-Guang",
""
],
[
"Yang",
"Shuai",
""
],
[
"Liu",
"Zheng-Xin",
""
],
[
"Chen",
"Yan",
""
]
] |
2503.13423 | Alisa Liu | Alisa Liu, Jonathan Hayase, Valentin Hofmann, Sewoong Oh, Noah A.
Smith, Yejin Choi | SuperBPE: Space Travel for Language Models | updated related work | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The assumption across nearly all language model (LM) tokenization schemes is
that tokens should be subwords, i.e., contained within word boundaries. While
providing a seemingly reasonable inductive bias, is this common practice
limiting the potential of modern LMs? Whitespace is not a reliable delimiter of
meaning, as evidenced by multi-word expressions (e.g., "by the way"),
crosslingual variation in the number of words needed to express a concept
(e.g., "spacesuit helmet" in German is "raumanzughelm"), and languages that do
not use whitespace at all (e.g., Chinese). To explore the potential of
tokenization beyond subwords, we introduce a "superword" tokenizer, SuperBPE,
which incorporates a simple pretokenization curriculum into the byte-pair
encoding (BPE) algorithm to first learn subwords, then superwords that bridge
whitespace. This brings dramatic improvements in encoding efficiency: when
fixing the vocabulary size to 200k, SuperBPE encodes a fixed piece of text with
up to 33% fewer tokens than BPE on average. In experiments, we pretrain 8B
transformer LMs from scratch while fixing the model size, vocabulary size, and
train compute, varying *only* the algorithm for learning the vocabulary. Our
model trained with SuperBPE achieves an average +4.0% absolute improvement over
the BPE baseline across 30 downstream tasks (including +8.2% on MMLU), while
simultaneously requiring 27% less compute at inference time. In analysis, we
find that SuperBPE results in segmentations of text that are more uniform in
per-token difficulty. Qualitatively, this may be because SuperBPE tokens often
capture common multi-word expressions that function semantically as a single
unit. SuperBPE is a straightforward, local modification to tokenization that
improves both encoding efficiency and downstream performance, yielding better
language models overall.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 17:53:23 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:59:35 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Alisa",
""
],
[
"Hayase",
"Jonathan",
""
],
[
"Hofmann",
"Valentin",
""
],
[
"Oh",
"Sewoong",
""
],
[
"Smith",
"Noah A.",
""
],
[
"Choi",
"Yejin",
""
]
] |
2503.13572 | Minghao Shao | Zeng Wang, Minghao Shao, Jitendra Bhandari, Likhitha Mankali, Ramesh
Karri, Ozgur Sinanoglu, Muhammad Shafique, Johann Knechtel | VeriContaminated: Assessing LLM-Driven Verilog Coding for Data
Contamination | null | null | null | null | cs.AR cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) have revolutionized code generation, achieving
exceptional results on various established benchmarking frameworks. However,
concerns about data contamination - where benchmark data inadvertently leaks
into pre-training or fine-tuning datasets - raise questions about the validity
of these evaluations. While this issue is known, limiting the industrial
adoption of LLM-driven software engineering, hardware coding has received
little to no attention regarding these risks. For the first time, we analyze
state-of-the-art (SOTA) evaluation frameworks for Verilog code generation
(VerilogEval and RTLLM), using established methods for contamination detection
(CCD and Min-K% Prob). We cover SOTA commercial and open-source LLMs
(CodeGen2.5, Minitron 4b, Mistral 7b, phi-4 mini, LLaMA-{1,2,3.1},
GPT-{2,3.5,4o}, Deepseek-Coder, and CodeQwen 1.5), in baseline and fine-tuned
models (RTLCoder and Verigen). Our study confirms that data contamination is a
critical concern. We explore mitigations and the resulting trade-offs for code
quality vs fairness (i.e., reducing contamination toward unbiased
benchmarking).
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 12:26:49 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 10:02:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Zeng",
""
],
[
"Shao",
"Minghao",
""
],
[
"Bhandari",
"Jitendra",
""
],
[
"Mankali",
"Likhitha",
""
],
[
"Karri",
"Ramesh",
""
],
[
"Sinanoglu",
"Ozgur",
""
],
[
"Shafique",
"Muhammad",
""
],
[
"Knechtel",
"Johann",
""
]
] |
2503.13779 | Hao Chen | Hao Chen, Julian Najera, Dagmawit Geresu, Meenal Datta, Cody Smith,
Scott Howard | Zero-Shot Denoising for Fluorescence Lifetime Imaging Microscopy with
Intensity-Guided Learning | 9 pages,4 figures and 2 tables | null | null | null | eess.IV | http://creativecommons.org/licenses/by/4.0/ | Multimodal and multi-information microscopy techniques such as Fluorescence
Lifetime Imaging Microscopy (FLIM) extend the informational channels beyond
intensity-based fluorescence microscopy but suffer from reduced image quality
due to complex noise patterns. For FLIM, the intrinsic relationship between
intensity and lifetime information means noise in each channel is a
multivariate function across channels without necessarily sharing structural
features. Based on this, we present a novel Zero-Shot Denoising Framework with
an Intensity-Guided Learning approach. Our correlation-preserving strategy
maintains important biological information that might be lost when channels are
processed independently. Our framework implements separate processing paths for
each channel and utilizes a pre-trained intensity denoising prior to guide the
refinement of lifetime components across multiple channels. Through experiments
on real-world FLIM-acquired biological samples, we show that our approach
outperforms existing methods in both noise reduction and lifetime preservation,
thereby enabling more reliable extraction of physiological and molecular
information.
| [
{
"version": "v1",
"created": "Mon, 17 Mar 2025 23:52:03 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 03:00:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Hao",
""
],
[
"Najera",
"Julian",
""
],
[
"Geresu",
"Dagmawit",
""
],
[
"Datta",
"Meenal",
""
],
[
"Smith",
"Cody",
""
],
[
"Howard",
"Scott",
""
]
] |
2503.13829 | Alex Elzenaar | Alex Elzenaar | From disc patterns in the plane to character varieties of knot groups | 29 pages, 17 figures. Comments welcomed. Second version fixes a
number of minor typos and errors, no numbering changes | null | null | null | math.GT math.GR math.MG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Motivated by an experimental study of groups generated by reflections in
planar patterns of tangent circles, we describe some methods for constructing
and studying representation spaces of holonomy groups of infinite volume
hyperbolic $3$-manifolds that arise from unknotting tunnels of links. We
include full descriptions of our computational methods, which were guided by
simplicity and generality rather than by being particularly efficient in
special cases. This makes them easy for non-experts to understand and implement
to produce visualisations that can suggest conjectures and support algebraic
calculations in the character variety. Throughout, we have tried to make the
exposition clear and understandable for graduate students in geometric topology
and related fields.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 02:11:05 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 10:16:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Elzenaar",
"Alex",
""
]
] |
2503.13846 | Shiji Lyu | Shiji Lyu | Uniform bounds in excellent $\mathbf{F}_p$-algebras and applications to
semi-continuity | 37 pages. Comments welcome! v4 - Remark 6.2.11 is changed and moved
to Remark A.2.5 | null | null | null | math.AC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study two important numerical invariants, Hilbert--Kunz multiplicity and
$F$-signature, on the spectrum of a Noetherian $\mathbf{F}_p$-algebra $R$ that
is not necessarily $F$-finite. When $R$ is excellent, we show that the limits
defining the invariants are uniform. As a consequence, we show that the
$F$-signature is lower semi-continuous, and the Hilbert--Kunz multiplicity is
upper semi-continuous provided $R$ is locally equidimensional. Uniform
convergence is achieved via a uniform version of Cohen--Gabber theorem. We
prove the results under weaker conditions than excellence.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 02:38:38 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Mar 2025 18:38:25 GMT"
},
{
"version": "v3",
"created": "Sun, 6 Apr 2025 13:59:31 GMT"
},
{
"version": "v4",
"created": "Mon, 14 Apr 2025 05:55:56 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lyu",
"Shiji",
""
]
] |
2503.14212 | Bharath Srivathsan | Bharath Srivathsan, Rafal Gartman, Robert J. A. Francis-Jones, Peter
Mosley, Joshua Nunn | Scalable Linear-Cavity Enhanced Quantum Memory | 8 pages, 5 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Coherent storage and retrieval of single photons in a quantum memory enables
scalable growth of photonic entangled states via linear optics. The resulting
increase in power of photonic quantum computers will unlock new applications on
the pathway to fault tolerance. Quantum memories based on off-resonant cascaded
absorption (ORCA) in rubidium vapour allow this storage to be broadband,
noise-free, and high efficiency. Through algorithmic pulse-shape optimisation,
we have previously demonstrated single-pass memory efficiency of almost 90\%.
Here, we implement a cavity-enhanced GHz-bandwidth ORCA memory with smaller
footprint and reduced power requirements compared to conventional single-pass
schemes. By combining a strong magnetic field with polarisation control, we
maintain a Doppler-free two-photon interaction and eliminate the need for
optical pumping. The performance provided by the cavity establishes the
feasibility of large arrays of ultra-compact, high-efficiency, room-temperature
quantum memories, while the low control power requirements highlight a route to
single-photon-level nonlinearities.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 12:45:56 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 21:15:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Srivathsan",
"Bharath",
""
],
[
"Gartman",
"Rafal",
""
],
[
"Francis-Jones",
"Robert J. A.",
""
],
[
"Mosley",
"Peter",
""
],
[
"Nunn",
"Joshua",
""
]
] |
2503.14221 | Yuxuan Guo | Yuxuan Guo, Sheng Yang and Xue-Jia Yu | Quantum Strong-to-Weak Spontaneous Symmetry Breaking in Decohered One
Dimensional Critical States | 19pages. 9 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Symmetry breaking has been a central theme in classifying quantum phases and
phase transitions. Recently, this concept has been extended to the mixed states
of open systems, attracting considerable attention due to the emergence of
novel physics beyond closed systems. In this Letter, we reveal a new type of
phase transition in mixed states, termed \emph{quantum} strong-to-weak
spontaneous symmetry breaking (SWSSB). Using a combination of field theory
calculations and large-scale matrix product state simulations, we map out the
global phase diagram of the XXZ critical spin chain under two-site XX
decoherence, which features an SWSSB phase and a trivial Luttinger liquid
phase, separated by a straight critical line that belongs to the boundary
Berezinskii-Kosterlitz-Thouless universality class with a varying effective
central charge. Remarkably, the SWSSB transition in our case is \emph{purely
quantum} in the sense that it can only be driven by tuning the Hamiltonian
parameter even under arbitrarily small decoherence strength, fundamentally
distinguishing it from the decoherence-driven SWSSB transitions extensively
discussed in previous literature. Conversely, no such phase transition occurs
under ZZ decoherence. Finally, we also discuss the experimental relevance of
our theory on quantum simulator platforms.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 12:59:17 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Mar 2025 02:05:00 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Apr 2025 08:17:39 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Guo",
"Yuxuan",
""
],
[
"Yang",
"Sheng",
""
],
[
"Yu",
"Xue-Jia",
""
]
] |
2503.14301 | Kamil Ksi\k{a}\.zek | Kamil Ksi\k{a}\.zek, Hubert Jastrz\k{e}bski, Bartosz Trojan, Krzysztof
Pniaczek, Micha{\l} Karp, Jacek Tabor | FeNeC: Enhancing Continual Learning via Feature Clustering with
Neighbor- or Logit-Based Classification | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ability of deep learning models to learn continuously is essential for
adapting to new data categories and evolving data distributions. In recent
years, approaches leveraging frozen feature extractors after an initial
learning phase have been extensively studied. Many of these methods estimate
per-class covariance matrices and prototypes based on backbone-derived feature
representations. Within this paradigm, we introduce FeNeC (Feature Neighborhood
Classifier) and FeNeC-Log, its variant based on the log-likelihood function.
Our approach generalizes the existing concept by incorporating data clustering
to capture greater intra-class variability. Utilizing the Mahalanobis distance,
our models classify samples either through a nearest neighbor approach or
trainable logit values assigned to consecutive classes. Our proposition may be
reduced to the existing approaches in a special case while extending them with
the ability of more flexible adaptation to data. We demonstrate that two FeNeC
variants achieve competitive performance in scenarios where task identities are
unknown and establish state-of-the-art results on several benchmarks.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 14:42:38 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:20:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Książek",
"Kamil",
""
],
[
"Jastrzębski",
"Hubert",
""
],
[
"Trojan",
"Bartosz",
""
],
[
"Pniaczek",
"Krzysztof",
""
],
[
"Karp",
"Michał",
""
],
[
"Tabor",
"Jacek",
""
]
] |
2503.14480 | Jonathan Klos | Jonathan H. Klos, Andreas Just, Evgeny V. Polyachenko, Peter Berczik,
Marina Ishchenko | Global survey of star clusters in the Milky Way VIII. Cluster formation
and evolution | Accepted for publication in A&A, 21 pages, 14 figures; language
edited version | null | 10.1051/0004-6361/202449418 | null | astro-ph.GA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We consider tidal masses and ages of Milky Way open clusters, as well as a
simple model of their distribution. Our aim is to investigate the space of
model parameters and the correspondence between modelled and observed
two-dimensional cluster age-mass distributions. The model for cluster evolution
is comprised of a two-section cluster initial mass function, constant cluster
formation rate, and a mass loss function. This mass loss function represents a
supervirial phase after gas expulsion, mass loss due to stellar evolution, and
gradual dissolution driven by internal dynamics and the Galactic tidal field.
We construct different estimators of model fitness based on
$\chi^2$-statistics, the Kullback-Leibler divergence (KLD) and a
maximum-likelihood approach. Using these estimators and Markov Chain Monte
Carlo sampling, we obtain best-fit values and posterior distributions for a
selection of model parameters. The KLD returns a superior model compared to the
other statistics. The cluster initial mass function is well constrained and we
find a clear signature of an enhanced cluster mass loss in the first 50 Myr. In
the KLD best model, clusters lose 72% of their initial mass in the violent
relaxation phase, after which cluster mass loss slows down, allowing for a
relatively low rate of cluster formation of $0.088\mathrm{M_\odot kpc^{-2}
Gyr^{-1}}$. The observed upper limit of cluster ages at approx. 5 Gyr is
reflected in the model by a shallow lifetime-mass relation for clusters with
initial masses above $1000\mathrm{M_\odot}$. The application of the model to an
independent cluster sample based on Gaia DR3 data yields similar results except
for a systematic shift in age. The observed cluster age-mass distribution is
compatible with a constant cluster formation rate. The enhanced number of young
massive clusters observed requires an early violent relaxation phase of strong
mass loss.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 17:52:28 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 07:34:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Klos",
"Jonathan H.",
""
],
[
"Just",
"Andreas",
""
],
[
"Polyachenko",
"Evgeny V.",
""
],
[
"Berczik",
"Peter",
""
],
[
"Ishchenko",
"Marina",
""
]
] |
2503.14721 | Aaron Dunbrack | Aaron Dunbrack, Pauli Virtanen, Tero T. Heikkil\"a | Quantum Geometry of Time-Reversal Symmetry Breaking in Flat-Band
Superconductivity | null | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Several physical phenomena in superconductors, such as helical
superconductivity and the diode effect, rely on breaking time-reversal
symmetry. This symmetry-breaking is usually accounted for via the Lifshitz
invariant, a contribution to the free energy which is linear in the phase
gradient of the order parameter. In dispersive single-band superconductors with
conventional pairing, the Lifshitz invariant can be computed from the
asymmetries of the spectrum near the Fermi surface. We show that in multi-band
superconductors, the quantum geometry also contributes to the Lifshitz
invariant, and this is the dominant contribution when the low-energy bands are
flat.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 20:49:26 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 21:01:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dunbrack",
"Aaron",
""
],
[
"Virtanen",
"Pauli",
""
],
[
"Heikkilä",
"Tero T.",
""
]
] |
2503.14911 | Siyuan Yan | Siyuan Yan, Ming Hu, Yiwen Jiang, Xieji Li, Hao Fei, Philipp Tschandl,
Harald Kittler, Zongyuan Ge | Derm1M: A Million-scale Vision-Language Dataset Aligned with Clinical
Ontology Knowledge for Dermatology | Our dataset and code will be publicly available at
https://github.com/SiyuanYan1/Derm1M | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The emergence of vision-language models has transformed medical AI, enabling
unprecedented advances in diagnostic capability and clinical applications.
However, progress in dermatology has lagged behind other medical domains due to
the lack of standard image-text pairs. Existing dermatological datasets are
limited in both scale and depth, offering only single-label annotations across
a narrow range of diseases instead of rich textual descriptions, and lacking
the crucial clinical context needed for real-world applications. To address
these limitations, we present Derm1M, the first large-scale vision-language
dataset for dermatology, comprising 1,029,761 image-text pairs. Built from
diverse educational resources and structured around a standard ontology
collaboratively developed by experts, Derm1M provides comprehensive coverage
for over 390 skin conditions across four hierarchical levels and 130 clinical
concepts with rich contextual information such as medical history, symptoms,
and skin tone. To demonstrate Derm1M potential in advancing both AI research
and clinical application, we pretrained a series of CLIP-like models,
collectively called DermLIP, on this dataset. The DermLIP family significantly
outperforms state-of-the-art foundation models on eight diverse datasets across
multiple tasks, including zero-shot skin disease classification, clinical and
artifacts concept identification, few-shot/full-shot learning, and cross-modal
retrieval. Our dataset and code will be publicly available at
https://github.com/SiyuanYan1/Derm1M upon acceptance.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 05:30:01 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 05:27:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yan",
"Siyuan",
""
],
[
"Hu",
"Ming",
""
],
[
"Jiang",
"Yiwen",
""
],
[
"Li",
"Xieji",
""
],
[
"Fei",
"Hao",
""
],
[
"Tschandl",
"Philipp",
""
],
[
"Kittler",
"Harald",
""
],
[
"Ge",
"Zongyuan",
""
]
] |
2503.14918 | Varun Sivashankar | Matija Buci\'c, Vanshika Jain, Varun Sivashankar | Intersecting hypergraphs with large cover number | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | In their famous 1974 paper introducing the local lemma, Erd\H{o}s and
Lov\'asz posed a question-later referred by Erd\H{o}s as one of his three
favorite open problems: What is the minimum number of edges in an $r$-uniform,
intersecting hypergraph with cover number $r$? This question was solved up to a
constant factor in Kahn's remarkable 1994 paper. More recently, motivated by
applications to Bollob\'as' ''power of many colours'' problem, Alon, Buci\'c,
Christoph, and Krivelevich introduced a natural generalization by imposing a
space constraint that limits the hypergraph to use only $n$ vertices. In this
note we settle this question asymptotically, up to a logarithmic factor in
$n/r$ in the exponent, for the entire range.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 05:53:07 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:58:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bucić",
"Matija",
""
],
[
"Jain",
"Vanshika",
""
],
[
"Sivashankar",
"Varun",
""
]
] |
2503.15166 | \`Alex Pujol Vidal | \`Alex Pujol Vidal, Sergio Escalera, Kamal Nasrollahi, Thomas B.
Moeslund | Machine Unlearning in Hyperbolic vs. Euclidean Multimodal Contrastive
Learning: Adapting Alignment Calibration to MERU | Preprint | null | null | null | cs.CV cs.AI cs.CL cs.LG cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine unlearning methods have become increasingly important for selective
concept removal in large pre-trained models. While recent work has explored
unlearning in Euclidean contrastive vision-language models, the effectiveness
of concept removal in hyperbolic spaces remains unexplored. This paper
investigates machine unlearning in hyperbolic contrastive learning by adapting
Alignment Calibration to MERU, a model that embeds images and text in
hyperbolic space to better capture semantic hierarchies. Through systematic
experiments and ablation studies, we demonstrate that hyperbolic geometry
offers distinct advantages for concept removal, achieving near perfect
forgetting with reasonable performance on retained concepts, particularly when
scaling to multiple concept removal. Our approach introduces
hyperbolic-specific components including entailment calibration and norm
regularization that leverage the unique properties of hyperbolic space.
Comparative analysis with Euclidean models reveals fundamental differences in
unlearning dynamics, with hyperbolic unlearning reorganizing the semantic
hierarchy while Euclidean approaches merely disconnect cross-modal
associations. These findings not only advance machine unlearning techniques but
also provide insights into the geometric properties that influence concept
representation and removal in multimodal models. Source code available at
https://github.com/alex-pv01/HAC
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 12:47:37 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:38:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Vidal",
"Àlex Pujol",
""
],
[
"Escalera",
"Sergio",
""
],
[
"Nasrollahi",
"Kamal",
""
],
[
"Moeslund",
"Thomas B.",
""
]
] |
2503.15199 | Luca De Martini | Luca De Martini, Dario d'Abate, Alessandro Margara and Gianpaolo
Cugola | Radon: a Programming Model and Platform for Computing Continuum Systems | Submitted to EDCCS 2025 | null | null | null | cs.DC cs.SE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Emerging compute continuum environments pose new challenges that traditional
cloud-centric architectures struggle to address. Latency, bandwidth
constraints, and the heterogeneity of edge environments hinder the efficiency
of centralized cloud solutions. While major cloud providers extend their
platforms to the edge, these approaches often overlook its unique
characteristics, limiting its potential.
To tackle these challenges, we introduce Radon, a flexible programming model
and platform designed for the edge-to-cloud continuum. Radon applications are
structured as atoms, isolated stateful entities that communicate through
messaging and can be composed into complex systems. The Radon runtime, based on
WebAssembly (WASM), enables language- and deployment-independent execution,
ensuring portability and adaptability across heterogeneous environments. This
decoupling allows developers to focus on application logic while the runtime
optimizes for diverse infrastructure conditions.
We present a prototype implementation of Radon and evaluate its effectiveness
through a distributed key-value store case study. We analyze the implementation
in terms of code complexity and performance. Our results demonstrate that Radon
facilitates the development and operation of scalable applications across the
edge-to-cloud continuum advancing the current state-of-the-art.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 13:38:25 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 15:52:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"De Martini",
"Luca",
""
],
[
"d'Abate",
"Dario",
""
],
[
"Margara",
"Alessandro",
""
],
[
"Cugola",
"Gianpaolo",
""
]
] |
2503.15239 | Sylvain Vanneste | Julian Auler, Utkarsh Bajpai, Martin Engler, Viktoria Ermuth, Martin
Fertl, Konrad Franz, Werner Heil, Simon Kaufmann, Bernhard Lauss, Niklas
Pfeifer, Dieter Ries, Sylvain Vanneste, Noah Yazdandoost | Ultra-cold neutron simulation framework for the free neutron lifetime
experiment $\tau$SPECT | null | null | null | null | physics.ins-det nucl-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The precise determination of the free neutron lifetime is of great
significance in modern precision physics. This key observable is linked to the
mixing of up and down quarks via the Cabibbo-Kobayashi-Maskawa matrix element
$V_{ud}$, and the abundance of primordial elements after the Big-Bang
Nucleosynthesis. However, the two leading measurement techniques for the
neutron lifetime currently yield incompatible results, a discrepancy referred
to as the neutron lifetime puzzle. To address the systematic uncertainties
arising from neutron interactions with material walls, the $\tau$SPECT
experiment employs a fully magnetic trap for ultra-cold neutrons (UCNs).
UCNs are extremely low-energy neutrons with typical velocities below
$8\,\textrm{m/s}$, which can be manipulated using magnetic fields, gravity, and
suitable material guides, whose surface can reflect them at any angle of
incidence. To precisely study and characterize UCN behavior during production,
guidance, storage, and detection in $\tau$SPECT, we have developed a dedicated
simulation framework. This framework is built upon the externally developed UCN
Monte Carlo software package PENTrack and is enhanced with two companion tools:
one for flexible and parametrizable upstream configuration of PENTrack such
that the simulation's input settings can be adjusted to reproduce the
experimental observations. The second package is used for analyzing,
visualizing, and animating simulation data.
The simulation results align well with experimental data obtained with
$\tau$SPECT at the Paul Scherrer Institute and serve as a powerful resource for
identifying systematic uncertainties and guiding future improvements to the
current experimental setup.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 14:15:34 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 13:37:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Auler",
"Julian",
""
],
[
"Bajpai",
"Utkarsh",
""
],
[
"Engler",
"Martin",
""
],
[
"Ermuth",
"Viktoria",
""
],
[
"Fertl",
"Martin",
""
],
[
"Franz",
"Konrad",
""
],
[
"Heil",
"Werner",
""
],
[
"Kaufmann",
"Simon",
""
],
[
"Lauss",
"Bernhard",
""
],
[
"Pfeifer",
"Niklas",
""
],
[
"Ries",
"Dieter",
""
],
[
"Vanneste",
"Sylvain",
""
],
[
"Yazdandoost",
"Noah",
""
]
] |
2503.15638 | Kaitlin Gili | Kaitlin Gili, Kyle Heuton, Astha Shah, Michael C. Hughes | Using machine learning to measure evidence of students' sensemaking in
physics courses | null | null | null | null | physics.ed-ph cs.LG | http://creativecommons.org/licenses/by/4.0/ | In the education system, problem-solving correctness is often inappropriately
conflated with student learning. Advances in both Physics Education Research
(PER) and Machine Learning (ML) provide the initial tools to develop a more
meaningful and efficient measurement scheme for whether physics students are
engaging in sensemaking: a learning process of figuring out the how and why for
a particular phenomena. In this work, we contribute such a measurement scheme,
which quantifies the evidence of students' physical sensemaking given their
written explanations for their solutions to physics problems. We outline how
the proposed human annotation scheme can be automated into a deployable ML
model using language encoders and shared probabilistic classifiers. The
procedure is scalable for a large number of problems and students. We implement
three unique language encoders with logistic regression, and provide a
deployability analysis on 385 real student explanations from the 2023
Introduction to Physics course at Tufts University. Furthermore, we compute
sensemaking scores for all students, and analyze these measurements alongside
their corresponding problem-solving accuracies. We find no linear relationship
between these two variables, supporting the hypothesis that one is not a
reliable proxy for the other. We discuss how sensemaking scores can be used
alongside problem-solving accuracies to provide a more nuanced snapshot of
student performance in physics class.
| [
{
"version": "v1",
"created": "Wed, 19 Mar 2025 18:49:21 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 19:10:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gili",
"Kaitlin",
""
],
[
"Heuton",
"Kyle",
""
],
[
"Shah",
"Astha",
""
],
[
"Hughes",
"Michael C.",
""
]
] |
2503.16316 | Zhanpeng Zhou | Zhanpeng Zhou, Yongyi Yang, Jie Ren, Mahito Sugiyama, Junchi Yan | On the Cone Effect in the Learning Dynamics | Accepted by ICLR 2025 workshop DeLTa | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Understanding the learning dynamics of neural networks is a central topic in
the deep learning community. In this paper, we take an empirical perspective to
study the learning dynamics of neural networks in real-world settings.
Specifically, we investigate the evolution process of the empirical Neural
Tangent Kernel (eNTK) during training. Our key findings reveal a two-phase
learning process: i) in Phase I, the eNTK evolves significantly, signaling the
rich regime, and ii) in Phase II, the eNTK keeps evolving but is constrained in
a narrow space, a phenomenon we term the cone effect. This two-phase framework
builds on the hypothesis proposed by Fort et al. (2020), but we uniquely
identify the cone effect in Phase II, demonstrating its significant performance
advantages over fully linearized training.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 16:38:25 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 10:24:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhou",
"Zhanpeng",
""
],
[
"Yang",
"Yongyi",
""
],
[
"Ren",
"Jie",
""
],
[
"Sugiyama",
"Mahito",
""
],
[
"Yan",
"Junchi",
""
]
] |
2503.16340 | Wei-Chen Wang | Wei-Chen Wang, Antoine De Comite, Alexandra Voloshina, Monica Daley,
Nidhi Seethapathi | Deep learning framework for action prediction reveals multi-timescale
locomotor control | null | null | null | null | cs.LG cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Modeling human movement in real-world tasks is a fundamental goal for motor
control, biomechanics, and rehabilitation engineering. However, existing models
of essential tasks like locomotion are not applicable across varying terrain,
mechanical conditions, and sensory contexts. This is at least in part due to
simplifying assumptions like linear and fixed timescales mappings between
inputs and future actions, which may not be broadly applicable. Here, we
develop a deep learning-based framework for action prediction, outperforming
traditional models across multiple contexts (walking and running, treadmill and
overground, varying terrains) and input modalities (multiple body states,
visual gaze). We find that neural network architectures with flexible input
history-dependence, like GRU and Transformer, and with architecture-dependent
trial embeddings perform best overall. By quantifying the model's predictions
relative to an autoregressive baseline, we identify context- and
modality-dependent timescales. These analyses reveal that there is greater
reliance on fast-timescale predictions in complex terrain, gaze predicts future
foot placement before body states, and the full-body state predictions precede
those by center-of-mass states. This deep learning framework for human action
prediction provides quantifiable insights into the control of real-world
locomotion and can be extended to other actions, contexts, and populations.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 16:57:15 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Mar 2025 20:22:39 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Mar 2025 04:50:17 GMT"
},
{
"version": "v4",
"created": "Mon, 14 Apr 2025 04:08:16 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wang",
"Wei-Chen",
""
],
[
"De Comite",
"Antoine",
""
],
[
"Voloshina",
"Alexandra",
""
],
[
"Daley",
"Monica",
""
],
[
"Seethapathi",
"Nidhi",
""
]
] |
2503.16403 | Luis Ferroni | Luis Ferroni, Alejandro H. Morales, Greta Panova | Skew shapes, Ehrhart positivity and beyond | 24 pages. 10 figures. Added positivity for circular fences | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A classical result by Kreweras (1965) allows one to compute the number of
plane partitions of a given skew shape and bounded parts as certain
determinants. We prove that these determinants expand as polynomials with
nonnegative coefficients. This result can be reformulated in terms of order
polynomials of cell posets of skew shapes, and explains important positivity
phenomena about the Ehrhart polynomials of shard polytopes, matroids, and order
polytopes. Among other applications, we generalize a positivity statement from
Schubert calculus by Fomin and Kirillov (1997) from straight shapes to skew
shapes. We show that all shard polytopes are Ehrhart positive and, stronger,
that all fence posets, including the zig-zag poset, and all circular fence
posets have order polynomials with nonnegative coefficients. We discuss a
general method for proving positivity which reduces to showing positivity of
the linear terms of the order polynomials. We propose positivity conjectures on
other relevant classes of posets.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 17:55:18 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 19:32:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ferroni",
"Luis",
""
],
[
"Morales",
"Alejandro H.",
""
],
[
"Panova",
"Greta",
""
]
] |
2503.16535 | Jinchang Zhang | Jinchang Zhang, Guoyu Lu | Vision-Language Embodiment for Monocular Depth Estimation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Depth estimation is a core problem in robotic perception and vision tasks,
but 3D reconstruction from a single image presents inherent uncertainties.
Current depth estimation models primarily rely on inter-image relationships for
supervised training, often overlooking the intrinsic information provided by
the camera itself. We propose a method that embodies the camera model and its
physical characteristics into a deep learning model, computing embodied scene
depth through real-time interactions with road environments. The model can
calculate embodied scene depth in real-time based on immediate environmental
changes using only the intrinsic properties of the camera, without any
additional equipment. By combining embodied scene depth with RGB image
features, the model gains a comprehensive perspective on both geometric and
visual details. Additionally, we incorporate text descriptions containing
environmental content and depth information as priors for scene understanding,
enriching the model's perception of objects. This integration of image and
language - two inherently ambiguous modalities - leverages their complementary
strengths for monocular depth estimation. The real-time nature of the embodied
language and depth prior model ensures that the model can continuously adjust
its perception and behavior in dynamic environments. Experimental results show
that the embodied depth estimation method enhances model performance across
different scenes.
| [
{
"version": "v1",
"created": "Tue, 18 Mar 2025 18:05:16 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 03:26:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Jinchang",
""
],
[
"Lu",
"Guoyu",
""
]
] |
2503.16766 | Kewei Zhang | Kewei Zhang | Quantized volume comparison for Fano manifolds | v3: final version, to appear in Proc. Amer. Math. Soc | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A result of Kento Fujita says that the volume of a K\"ahler-Einstein Fano
manifold is bounded from above by the volume of the projective space. In this
short note we establish quantized versions of Fujita's result.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 00:40:24 GMT"
},
{
"version": "v2",
"created": "Sun, 30 Mar 2025 04:04:02 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Apr 2025 12:18:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Kewei",
""
]
] |
2503.16864 | Guo-Peng Li | Guo-Peng Li and Xi-Long Fan | A multi-messenger hierarchical triple merger gravitational-wave event
pair GW190514-GW190521 inside AGN J124942.3 + 344929 | 11 pages, 3 figures, 1 table; accepted for publication in Physical
Review D | null | null | null | astro-ph.HE astro-ph.GA astro-ph.SR gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is a candidate electromagnetic counterpart to the binary black hole
merger GW190521, identified as ZTF19abanrhr within AGN J124942.3 + 344929.
Additionally, GW190514 is proposed as a plausible precursor merger to GW190521
within a hierarchical merger scenario. In this study, we investigate the
potential association between GW190514 and GW190521 as a hierarchical triple
merger associated with ZTF19abanrhr, taking into account of sky position,
distance, and mass of the sources using a Bayesian criterion. Our analysis
reveals that the association is favored over a random coincidence, with a log
Bayes factor of 16.8, corresponding to an odds ratio of $\sim$$199:1$, assuming
an astrophysical prior odds of $10^{-5}$. Notably, when accounting for the
primary masses of the two gravitational wave events as potential products of
mergers in the AGN formation channel, the Bayes factor increases significantly,
further enhancing the preference for this association by a factor of
$\sim$$10^2$, corresponding to a log Bayes factor of 21.5 and an odds ratio of
$\sim$$2\times10^4:1$. Our results suggest strong evidence for the first
hierarchical triple merger associated with an electromagnetic counterpart in
the AGN formation channel. This work is crucial for understanding the formation
mechanisms of massive black holes, the role of AGNs in hierarchical mergers,
and the implications of multi-messenger astronomy.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 05:32:59 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 03:50:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Guo-Peng",
""
],
[
"Fan",
"Xi-Long",
""
]
] |
2503.17039 | Jeremy Barnes | Jeremy Barnes, Naiara Perez, Alba Bonet-Jover, Bego\~na Altuna | Summarization Metrics for Spanish and Basque: Do Automatic Scores and
LLM-Judges Correlate with Humans? | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Studies on evaluation metrics and LLM-as-a-Judge models for automatic text
summarization have largely been focused on English, limiting our understanding
of their effectiveness in other languages. Through our new dataset BASSE
(BAsque and Spanish Summarization Evaluation), we address this situation by
collecting human judgments on 2,040 abstractive summaries in Basque and
Spanish, generated either manually or by five LLMs with four different prompts.
For each summary, annotators evaluated five criteria on a 5-point Likert scale:
coherence, consistency, fluency, relevance, and 5W1H. We use these data to
reevaluate traditional automatic metrics used for evaluating summaries, as well
as several LLM-as-a-Judge models that show strong performance on this task in
English. Our results show that currently proprietary judge LLMs have the
highest correlation with human judgments, followed by criteria-specific
automatic metrics, while open-sourced judge LLMs perform poorly. We release
BASSE and our code publicly, along with the first large-scale Basque
summarization dataset containing 22,525 news articles with their subheads.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 10:52:20 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:25:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Barnes",
"Jeremy",
""
],
[
"Perez",
"Naiara",
""
],
[
"Bonet-Jover",
"Alba",
""
],
[
"Altuna",
"Begoña",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.