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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.16009 | Chunhua Zhang | Chunhua Zhang, Peiyao Liu, Cheng Peng, Lian-Ping Wang, Zhaoli Guo | Upscaling the Navier-Stokes-Cahn-Hilliard model for incompressible
multiphase flow in inhomogeneous porous media | 33 pages, 7 figures | null | null | null | physics.flu-dyn math-ph math.MP physics.comp-ph physics.geo-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present a macroscopic model for the flow of two immiscible
and incompressible fluids in inhomogeneous porous medium. At the pore scale,
the flow is governed by the fully Navier-Stokes equations while the evolution
of the phase interface is captured by the Cahn-Hilliard equation. Using the
volume averaging method, the upscaled equations describing the averaged
behavior of two fluids at the Darcy scale are obtained, with unclosed terms
related to spatial deviations. Then, spatial derivations are carefully modeled
up to some undetermined coefficients, which could be evaluated by solving
simplified closure problems in each representative volume element. In
particular, the wetting behavior is incorporated into the averaged chemical
potential. The differences between the proposed equations and the empirical
two-phase Darcy-type models are discussed. Finally, a phase-field-based lattice
Boltzmann model for the averaged equations is presented, and numerical results
demonstrate the abilities of the proposed model.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:16:19 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zhang",
"Chunhua",
""
],
[
"Liu",
"Peiyao",
""
],
[
"Peng",
"Cheng",
""
],
[
"Wang",
"Lian-Ping",
""
],
[
"Guo",
"Zhaoli",
""
]
] |
2504.16010 | Tuong Manh Vu | Tuong Manh Vu, Ernesto Carrella, Robert Axtell, Omar A. Guerrero | The Formation of Production Networks: How Supply Chains Arise from
Simple Learning with Minimal Information | null | null | null | null | cs.MA cs.LG econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | We develop a model where firms determine the price at which they sell their
differentiable goods, the volume that they produce, and the inputs (types and
amounts) that they purchase from other firms. A steady-state production network
emerges endogenously without resorting to assumptions such as equilibrium or
perfect knowledge about production technologies. Through a simple version of
reinforcement learning, firms with heterogeneous technologies cope with
uncertainty and maximize profits. Due to this learning process, firms can adapt
to shocks such as demand shifts, suppliers/clients closure, productivity
changes, and production technology modifications; effectively reshaping the
production network. To demonstrate the potential of this model, we analyze the
upstream and downstream impact of demand and productivity shocks.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:18:19 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Vu",
"Tuong Manh",
""
],
[
"Carrella",
"Ernesto",
""
],
[
"Axtell",
"Robert",
""
],
[
"Guerrero",
"Omar A.",
""
]
] |
2504.16012 | Hiroki Ishizaka | Hiroki Ishizaka | Interpolation error analysis using a new geometric parameter | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This article presents novel proof methods for estimating interpolation
errors, predicated on the understanding that one has already studied
foundational error analysis using the finite element method.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:20:54 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Ishizaka",
"Hiroki",
""
]
] |
2504.16013 | Markos Poulos | Markos Poulos and Konstantinos Termentzidis | Thermal rectification and phonon properties in partially perforated
graphene | null | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | In this work, a thermal rectification ratio $\eta$ of 18.5% was observed in
partially perforated graphene with the use of Molecular Dynamics (MD)
simulations. In all cases studied here, heat preferentially flows from the
porous to the pristine region and both $\kappa$ and $\eta$ increase upon
increasing the length of the pristine region and upon decreasing the size of
the pores. To interpret the results, the macroscopic "R-Series Model" is
applied, attributing rectification to the different temperature dependence of
$\kappa$ of perforated and pristine graphene. According to the model, $\eta$ is
maximized when the two regions composing the structure have matching thermal
resistances and mismatching temperature-dependence of $\kappa$. The model
agrees qualitatively with the MD results, indicating that the latter is the
principal rectification mechanism, but it can significantly underestimate
$\eta$. Phonon analysis further reveals the appearance of new 'defect' modes
localized around and between pores, resulting in the emergence of a new
prominent peak in the phonon Density of States at 520 $cm^{-1}$. The study
considers key geometric factors such as the length of the pristine region, and
the pore size, shape, alignment, and orientation. Pore shape and alignment
exert minimal influence on $\eta$, although alignment greatly influences
$\kappa$. Eventually, arranged pores are deemed more efficient than randomly
distributed defects for increasing rectification.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:25:01 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Poulos",
"Markos",
""
],
[
"Termentzidis",
"Konstantinos",
""
]
] |
2504.16014 | Kimberley Keri Vos | Robert Fleischer and Martijn van Hamersveld and Tim Kortekaas and
Anders Rehult and K. Keri Vos | Probing New Physics Through CP Violation in $B_{(s)}\to V\mu^+\mu^-$
Decays | 29 pages, 11 figures | null | null | Nkhef-2025-006 | hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Rare decays of the kind $B\to K^*\mu^+\mu^-$ and $B_s\to \phi\mu^+\mu^-$ are
key players for testing the Standard Model. The current experimental data for
their decay rates and angular observables show tensions with the theoretical
predictions that may be indications of New Physics. We present a strategy to
extract the relevant short-distance coefficients in the presence of new sources
of CP violation, utilizing a synergy with $B\to K\mu^+\mu^-$ decays. Using the
current data as a guideline, we illustrate the new method to determine the
complex coefficients $C_9^{(\prime)}$ and $C_{10}^{(\prime)}$ using only four
angular observables. Interestingly, the current experimental picture leaves
significant room for CP-violating New Physics. We discuss also the link to
leptonic $B^0_s\to\mu^+\mu^-$ decays. We are looking forward to the
implementation of these strategies at the future high-precision frontier of
flavour physics.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:26:17 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Fleischer",
"Robert",
""
],
[
"van Hamersveld",
"Martijn",
""
],
[
"Kortekaas",
"Tim",
""
],
[
"Rehult",
"Anders",
""
],
[
"Vos",
"K. Keri",
""
]
] |
2504.16015 | Alexander Markus Maier | Alexander M. Maier, Udo Seifert and Jann van der Meer | From observed transitions to hidden paths in Markov networks | null | null | null | null | cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The number of observable degrees of freedom is typically limited in
experiments. Here, we consider discrete Markov networks in which an observer
has access to a few visible transitions and the waiting times between these
transitions. Focusing on the underlying structure of a discrete network, we
present methods to infer local and global properties of the network from
observed data. First, we derive bounds on the microscopic entropy production
along the hidden paths between two visible transitions, which complement extant
bounds on mean entropy production and affinities of hidden cycles. Second, we
demonstrate how the operationally accessible data encodes information about the
topology of shortest hidden paths, which can be used to identify potential
clusters of states or exclude their existence. Finally, we outline a systematic
way to combine the inferred data, resulting in an algorithm that finds the
candidates for a minimal graph of the underlying network, i.e., a graph that is
part of the original one and compatible with the observations. Our results
highlight the interplay between thermodynamic methods, waiting-time
distributions and topological aspects like network structure, which can be
expected to provide novel insights in other set-ups of coarse graining as well.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:26:29 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Maier",
"Alexander M.",
""
],
[
"Seifert",
"Udo",
""
],
[
"van der Meer",
"Jann",
""
]
] |
2504.16016 | Jianhui Wang | Xinyuan Song, Yangfan He, Sida Li, Jianhui Wang, Hongyang He, Xinhang
Yuan, Ruoyu Wang, Jiaqi Chen, Keqin Li, Kuan Lu, Menghao Huo, Binxu Li, Pei
Liu | Efficient Temporal Consistency in Diffusion-Based Video Editing with
Adaptor Modules: A Theoretical Framework | arXiv admin note: substantial text overlap with arXiv:2501.04606 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adapter-based methods are commonly used to enhance model performance with
minimal additional complexity, especially in video editing tasks that require
frame-to-frame consistency. By inserting small, learnable modules into
pretrained diffusion models, these adapters can maintain temporal coherence
without extensive retraining. Approaches that incorporate prompt learning with
both shared and frame-specific tokens are particularly effective in preserving
continuity across frames at low training cost. In this work, we want to provide
a general theoretical framework for adapters that maintain frame consistency in
DDIM-based models under a temporal consistency loss. First, we prove that the
temporal consistency objective is differentiable under bounded feature norms,
and we establish a Lipschitz bound on its gradient. Second, we show that
gradient descent on this objective decreases the loss monotonically and
converges to a local minimum if the learning rate is within an appropriate
range. Finally, we analyze the stability of modules in the DDIM inversion
procedure, showing that the associated error remains controlled. These
theoretical findings will reinforce the reliability of diffusion-based video
editing methods that rely on adapter strategies and provide theoretical
insights in video generation tasks.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:28:35 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Song",
"Xinyuan",
""
],
[
"He",
"Yangfan",
""
],
[
"Li",
"Sida",
""
],
[
"Wang",
"Jianhui",
""
],
[
"He",
"Hongyang",
""
],
[
"Yuan",
"Xinhang",
""
],
[
"Wang",
"Ruoyu",
""
],
[
"Chen",
"Jiaqi",
""
],
[
"Li",
"Keqin",
""
],
[
"Lu",
"Kuan",
""
],
[
"Huo",
"Menghao",
""
],
[
"Li",
"Binxu",
""
],
[
"Liu",
"Pei",
""
]
] |
2504.16017 | Navonil Neogi | Nabil Iqbal, Navonil Neogi | Wilson lines with endpoints in 3d CFT | 24 pages + appendices | null | null | null | hep-th cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | In Abelian gauge theories with dynamical matter, Wilson lines can end on the
insertions of charged fields. We study the endpoints of Wilson lines in large
$N$ bosonic QED$_3$. at its critical point. We first study the stability of an
infinite Wilson line in the $\mathbb{CP}^{N-1}$ model by computing the
appropriate functional determinant at large $N$. We also compute the conformal
dimension of the lowest-dimension endpoint of the line to first order in
$N^{-1}$. Along the way we calculate the field-strength tensor $F_{\mu\nu}$ in
the presence of the line with endpoint and discuss a state-operator
correspondence for the endpoints, as well as the existence of an OPE that
allows two open-ended Wilson lines to be glued together into a single line.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:28:46 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Iqbal",
"Nabil",
""
],
[
"Neogi",
"Navonil",
""
]
] |
2504.16018 | Alexander Esterov | Alexander Esterov | Engineered complete intersections: eliminating variables and
understanding topology | 28 pages, 1 figure | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We continue the study of engineered complete intersections (ECI) -- an
umbrella generality for a number of important objects in combinatoiral and
applied algebraic geometry (such as nondegenerate toric complete intersections,
critical loci of their projections, hyperplane arrangements, generalized
Calabi--Yau complete intersections, incidence varieties in algebraic
optimization, reaction networks).
In this paper, we work on extending to ECIs several classical results about
toric complete intersections. This includes elimination theory, patchworking
over ${\mathbb R}$, and computing basic geometric invariants over ${\mathbb
C}$.
Our results apply e.g. to eliminating variables in systems of ODEs, such as
reaction networks, computing Newton polytopes of discriminants, constructing
real polynomial maps and reaction networks with prescribed topology.
Along the way, we assign a cohomology ring to an arbitrary tropical fan, and
relate reducible ECIs to arrangements of pairwise intersecting planes.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:28:47 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Esterov",
"Alexander",
""
]
] |
2504.16019 | Saikat Das | Saikat Das, Srijita Hazra, Nayantara Gupta | Cosmic Clues from Amaterasu: Blazar-Driven Ultrahigh-Energy Cosmic Rays? | 8 pages, 3 figures; comments are welcome! | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The detection of the 244 EeV Amaterasu event by the Telescope Array, one of
the most energetic ultrahigh-energy cosmic rays (UHECRs; $E\gtrsim0.1$ EeV)
observed to date, invites scrutiny of its potential source. We investigate
whether the nearby blazar PKS 1717+177, located within $2.5^\circ$ of the
reconstructed arrival direction, could explain the event under a proton-primary
hypothesis. Using a one-zone jet model, we fit the multi-wavelength spectral
energy distribution of the source, incorporating both leptonic and hadronic
cascade emission from photohadronic interactions. Our model supports a
cosmic-ray origin of the very-high-energy ($E\gtrsim 100$ GeV) $\gamma$-ray
flux and predicts a subdominant neutrino flux, an order of magnitude lower than
from TXS 0506+056. Under Lorentz invariance violation, protons above a specific
energy can propagate over hundreds of Mpc without significant energy loss for
certain parameter choices. In such a scenario, our analysis indicates
negligible deflection in the Galactic magnetic field, implying a strong
extragalactic magnetic field, placing a lower bound on the field strength. Our
findings provide a compelling multi-messenger framework linking UHECRs,
$\gamma$ rays, and neutrinos and motivate targeted searches by current and
future high-energy neutrino telescopes during increased $\gamma$-ray or X-ray
activity of this blazar.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:29:32 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Das",
"Saikat",
""
],
[
"Hazra",
"Srijita",
""
],
[
"Gupta",
"Nayantara",
""
]
] |
2504.16021 | Dinithi Dissanayake | Dinithi Dissanayake and Suranga Nanayakkara | Navigating the State of Cognitive Flow: Context-Aware AI Interventions
for Effective Reasoning Support | Presented at the 2025 ACM Workshop on Human-AI Interaction for
Augmented Reasoning, Report Number: CHI25-WS-AUGMENTED-REASONING | Proceedings of the 2025 ACM CHI Workshop on Human-AI Interaction
for Augmented Reasoning | null | CHI25-WS-AUGMENTED-REASONING | cs.HC cs.AI | http://creativecommons.org/licenses/by/4.0/ | Flow theory describes an optimal cognitive state where individuals experience
deep focus and intrinsic motivation when a task's difficulty aligns with their
skill level. In AI-augmented reasoning, interventions that disrupt the state of
cognitive flow can hinder rather than enhance decision-making. This paper
proposes a context-aware cognitive augmentation framework that adapts
interventions based on three key contextual factors: type, timing, and scale.
By leveraging multimodal behavioral cues (e.g., gaze behavior, typing
hesitation, interaction speed), AI can dynamically adjust cognitive support to
maintain or restore flow. We introduce the concept of cognitive flow, an
extension of flow theory in AI-augmented reasoning, where interventions are
personalized, adaptive, and minimally intrusive. By shifting from static
interventions to context-aware augmentation, our approach ensures that AI
systems support deep engagement in complex decision-making and reasoning
without disrupting cognitive immersion.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:35:39 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Dissanayake",
"Dinithi",
""
],
[
"Nanayakkara",
"Suranga",
""
]
] |
2504.16022 | Shaikh Saad | Ilja Dor\v{s}ner, Mijo Matkovi\'c, Shaikh Saad | Non-Renormalizable SU(5) GUTs: Leptoquark-Induced Neutrino Masses | 25 pages + references. Comments are welcome | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | We revisit the doublet-triplet splitting problem within the $SU(5)$ gauge
group framework to advocate a viable regime with the light scalar leptoquark of
the doublet-triplet splitting notoriety that is compatible with the current
experimental bounds on partial proton decay lifetimes. We explicitly
demonstrate, through a consistent use of higher-dimensional operators, how to
implement suppression of baryon number violating interactions of the
aforementioned color triplet. Our study thus offers an alternative approach to
the doublet-triplet splitting problem as it removes a need for an extreme mass
hierarchy between the partners residing in the same representation. We
furthermore pursue two different extensions of two distinct symmetry breaking
scenarios of $SU(5)$, one with a $24$-dimensional representation and the other
one with a $75$-dimensional representation, to produce comparative study of
novel consequences for the gauge coupling unification and the one-loop level
neutrino mass generation. Our results point towards qualitatively novel $SU(5)$
scenarios, where the light scalar leptoquarks, responsible for the neutrino
mass generation, might be even accessible at colliders and thus serve as an
accelerator accessible portal to the high-scale physics.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:39:02 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Doršner",
"Ilja",
""
],
[
"Matković",
"Mijo",
""
],
[
"Saad",
"Shaikh",
""
]
] |
2504.16023 | Song Wang | Song Wang, Xiaolu Liu, Lingdong Kong, Jianyun Xu, Chunyong Hu, Gongfan
Fang, Wentong Li, Jianke Zhu, Xinchao Wang | PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud
Learning | Accepted by CVPR2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-supervised representation learning for point cloud has demonstrated
effectiveness in improving pre-trained model performance across diverse tasks.
However, as pre-trained models grow in complexity, fully fine-tuning them for
downstream applications demands substantial computational and storage
resources. Parameter-efficient fine-tuning (PEFT) methods offer a promising
solution to mitigate these resource requirements, yet most current approaches
rely on complex adapter and prompt mechanisms that increase tunable parameters.
In this paper, we propose PointLoRA, a simple yet effective method that
combines low-rank adaptation (LoRA) with multi-scale token selection to
efficiently fine-tune point cloud models. Our approach embeds LoRA layers
within the most parameter-intensive components of point cloud transformers,
reducing the need for tunable parameters while enhancing global feature
capture. Additionally, multi-scale token selection extracts critical local
information to serve as prompts for downstream fine-tuning, effectively
complementing the global context captured by LoRA. The experimental results
across various pre-trained models and three challenging public datasets
demonstrate that our approach achieves competitive performance with only 3.43%
of the trainable parameters, making it highly effective for
resource-constrained applications. Source code is available at:
https://github.com/songw-zju/PointLoRA.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:41:21 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Wang",
"Song",
""
],
[
"Liu",
"Xiaolu",
""
],
[
"Kong",
"Lingdong",
""
],
[
"Xu",
"Jianyun",
""
],
[
"Hu",
"Chunyong",
""
],
[
"Fang",
"Gongfan",
""
],
[
"Li",
"Wentong",
""
],
[
"Zhu",
"Jianke",
""
],
[
"Wang",
"Xinchao",
""
]
] |
2504.16024 | Michael Sitwell | Michael Sitwell | EnsAI: An Emulator for Atmospheric Chemical Ensembles | 32 pages, 17 figures | null | null | null | physics.ao-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Ensemble-based methods for data assimilation and emission inversions are a
popular way to encode flow-dependency within the model error covariance. While
most ensemble methods do not require the use of an adjoint model, the need to
repeatedly run a geophysical model to generate the ensemble can be a
significant computational burden. In this paper, we introduce EnsAI, a new
AI-based ensemble generation system for atmospheric chemical constituents. When
trained on an existing ensemble for ammonia generated by the GEM-MACH air
quality model, it was shown that the ensembles produced by EnsAI can accurately
reproduce the meteorology-dependent features of the original ensemble, while
generating the ensemble 3,300 times faster than the original GEM-MACH ensemble.
While EnsAI requires an upfront cost for generating an ensemble used for
training, as well as the training itself, the long term computational savings
can greatly exceed these initial computational costs. When used in an emissions
inversion system, EnsAI produced similar inversion results to those in which
the original GEM-MACH ensemble was used while using significantly less
computational resources.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:41:26 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Sitwell",
"Michael",
""
]
] |
2504.16025 | Zihan Zhou | Zihan Zhou, Giovanni Maria Tomaselli, Irvin Mart\'inez-Rodr\'iguez,
Jingping Li | Modeling Tidal Disruptions with Dynamical Tides | 11 pages, 5 figures | null | null | null | astro-ph.HE astro-ph.GA astro-ph.SR gr-qc hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | Tidal disruption events (TDEs) occur when stars pass close enough to
supermassive black holes to be torn apart by tidal forces. Traditionally, these
events are studied with computationally intensive hydrodynamical simulations.
In this paper, we present a fast, physically motivated two-stage model for
TDEs. In the first stage, we model the star's tidal deformation using linear
stellar perturbation theory, treating the star as a collection of driven
harmonic oscillators. When the tidal energy exceeds a fraction $\gamma$ of the
star's gravitational binding energy (with $\gamma \sim \mathcal{O}(1)$), we
transition to the second stage, where we model the disrupted material as free
particles. The parameter $\gamma$ is determined with a one-time calibration to
hydrodynamical simulations. This method enables fast computation of the energy
distribution $\mathrm{d}M/\mathrm{d}E$ and fallback rate
$\mathrm{d}M/\mathrm{d}T$, while offering physical insight into the disruption
process. We apply our model to MESA-generated profiles of middle-age
main-sequence stars. Our code is available on GitHub.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:42:18 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zhou",
"Zihan",
""
],
[
"Tomaselli",
"Giovanni Maria",
""
],
[
"Martínez-Rodríguez",
"Irvin",
""
],
[
"Li",
"Jingping",
""
]
] |
2504.16027 | Ahmed R. Sadik Dr.-Ing. | Ahmed R. Sadik, Siddhata Govind | Benchmarking LLM for Code Smells Detection: OpenAI GPT-4.0 vs
DeepSeek-V3 | null | null | null | null | cs.SE cs.AI cs.LG cs.PL | http://creativecommons.org/licenses/by/4.0/ | Determining the most effective Large Language Model for code smell detection
presents a complex challenge. This study introduces a structured methodology
and evaluation matrix to tackle this issue, leveraging a curated dataset of
code samples consistently annotated with known smells. The dataset spans four
prominent programming languages Java, Python, JavaScript, and C++; allowing for
cross language comparison. We benchmark two state of the art LLMs, OpenAI GPT
4.0 and DeepSeek-V3, using precision, recall, and F1 score as evaluation
metrics. Our analysis covers three levels of detail: overall performance,
category level performance, and individual code smell type performance.
Additionally, we explore cost effectiveness by comparing the token based
detection approach of GPT 4.0 with the pattern-matching techniques employed by
DeepSeek V3. The study also includes a cost analysis relative to traditional
static analysis tools such as SonarQube. The findings offer valuable guidance
for practitioners in selecting an efficient, cost effective solution for
automated code smell detection
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:44:39 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Sadik",
"Ahmed R.",
""
],
[
"Govind",
"Siddhata",
""
]
] |
2504.16028 | Christoph Aoun | Tigran Bakaryan, Christoph Aoun, Ricardo de Lima Ribeiro, Naira
Hovakimyan, and Diogo Gomes | Hessian Riemannian Flow For Multi-Population Wardrop Equilibrium | null | null | null | null | eess.SY cs.MA cs.SY math.OC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we address the problem of optimizing flows on generalized
graphs that feature multiple entry points and multiple populations, each with
varying cost structures. We tackle this problem by considering the
multi-population Wardrop equilibrium, defined through variational inequalities.
We rigorously analyze the existence and uniqueness of the Wardrop equilibrium.
Furthermore, we introduce an efficient numerical method to find the solution.
In particular, we reformulate the equilibrium problem as a distributed
optimization problem over subgraphs and introduce a novel Hessian Riemannian
flow method, a Riemannian-manifold-projected Hessian flow, to efficiently
compute a solution. Finally, we demonstrate the effectiveness of our approach
through examples in urban traffic management, including routing for diverse
vehicle types and strategies for minimizing emissions in congested
environments.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:45:29 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Bakaryan",
"Tigran",
""
],
[
"Aoun",
"Christoph",
""
],
[
"Ribeiro",
"Ricardo de Lima",
""
],
[
"Hovakimyan",
"Naira",
""
],
[
"Gomes",
"Diogo",
""
]
] |
2504.16029 | Heiko Gimperlein | Heiko Gimperlein, Ruma R. Maity, Apala Majumdar, Michael
Oberguggenberger | Bayesian Parameter Identification in the Landau-de Gennes Theory for
Nematic Liquid Crystals | 25 pages | null | null | null | math.NA cond-mat.soft cs.NA math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This manuscript establishes a pathway to reconstruct material parameters from
measurements within the Landau-de Gennes model for nematic liquid crystals. We
present a Bayesian approach to this inverse problem and analyse its properties
using given, simulated data for benchmark problems of a planar bistable nematic
device. In particular, we discuss the accuracy of the Markov chain Monte Carlo
approximations, confidence intervals and the limits of identifiability.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:49:43 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Gimperlein",
"Heiko",
""
],
[
"Maity",
"Ruma R.",
""
],
[
"Majumdar",
"Apala",
""
],
[
"Oberguggenberger",
"Michael",
""
]
] |
2504.16030 | Joya Chen | Joya Chen and Ziyun Zeng and Yiqi Lin and Wei Li and Zejun Ma and Mike
Zheng Shou | LiveCC: Learning Video LLM with Streaming Speech Transcription at Scale | CVPR 2025. If any references are missing, please contact
[email protected] | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent video large language models (Video LLMs) often depend on costly human
annotations or proprietary model APIs (e.g., GPT-4o) to produce training data,
which limits their training at scale. In this paper, we explore large-scale
training for Video LLM with cheap automatic speech recognition (ASR)
transcripts. Specifically, we propose a novel streaming training approach that
densely interleaves the ASR words and video frames according to their
timestamps. Compared to previous studies in vision-language representation with
ASR, our method naturally fits the streaming characteristics of ASR, thus
enabling the model to learn temporally-aligned, fine-grained vision-language
modeling. To support the training algorithm, we introduce a data production
pipeline to process YouTube videos and their closed captions (CC, same as ASR),
resulting in Live-CC-5M dataset for pre-training and Live-WhisperX-526K dataset
for high-quality supervised fine-tuning (SFT). Remarkably, even without SFT,
the ASR-only pre-trained LiveCC-7B-Base model demonstrates competitive general
video QA performance and exhibits a new capability in real-time video
commentary. To evaluate this, we carefully design a new LiveSports-3K
benchmark, using LLM-as-a-judge to measure the free-form commentary.
Experiments show our final LiveCC-7B-Instruct model can surpass advanced 72B
models (Qwen2.5-VL-72B-Instruct, LLaVA-Video-72B) in commentary quality even
working in a real-time mode. Meanwhile, it achieves state-of-the-art results at
the 7B/8B scale on popular video QA benchmarks such as VideoMME and OVOBench,
demonstrating the broad generalizability of our approach. All resources of this
paper have been released at https://showlab.github.io/livecc.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:52:09 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Chen",
"Joya",
""
],
[
"Zeng",
"Ziyun",
""
],
[
"Lin",
"Yiqi",
""
],
[
"Li",
"Wei",
""
],
[
"Ma",
"Zejun",
""
],
[
"Shou",
"Mike Zheng",
""
]
] |
2504.16031 | Ali Arya | Ali Arya, Anthony Scavarelli, Dan Hawes, Luciara Nardon | VR-based Intervention for Perspective Change: A Case to Investigate
Virtual Materiality | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper addresses the concept of materiality in virtual environments,
which we define as being composed of objects that can influence user experience
actively. Such virtual materiality is closely related to its physical
counterpart, which is discussed in theoretical frameworks such as
sociomateriality and actor-network theory. They define phenomena in terms of
the entanglement of human and non-human elements. We report on an early
investigation of virtual materiality within the context of reflection and
perspective change in nature-based virtual environments. We considered the case
of university students reflecting on the planning and management of their
theses and major projects. Inspired by nature's known positive cognitive and
affective effects and repeated questioning processes, we established a virtual
reflection intervention to demonstrate the environmental mechanisms and
material characteristics relevant to virtual materiality. Our work is a
preliminary step toward understanding virtual materiality and its implications
for research and the design of virtual environments.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:53:21 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Arya",
"Ali",
""
],
[
"Scavarelli",
"Anthony",
""
],
[
"Hawes",
"Dan",
""
],
[
"Nardon",
"Luciara",
""
]
] |
2504.16032 | Yazan Otoum | Yazan Otoum, Arghavan Asad, Amiya Nayak | LLMs meet Federated Learning for Scalable and Secure IoT Management | This work has been submitted to the IEEE Global Communications
Conference (GLOBECOM) 2025 for possible publication | null | null | null | cs.LG cs.AI cs.ET | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid expansion of IoT ecosystems introduces severe challenges in
scalability, security, and real-time decision-making. Traditional centralized
architectures struggle with latency, privacy concerns, and excessive resource
consumption, making them unsuitable for modern large-scale IoT deployments.
This paper presents a novel Federated Learning-driven Large Language Model
(FL-LLM) framework, designed to enhance IoT system intelligence while ensuring
data privacy and computational efficiency. The framework integrates Generative
IoT (GIoT) models with a Gradient Sensing Federated Strategy (GSFS),
dynamically optimizing model updates based on real-time network conditions. By
leveraging a hybrid edge-cloud processing architecture, our approach balances
intelligence, scalability, and security in distributed IoT environments.
Evaluations on the IoT-23 dataset demonstrate that our framework improves model
accuracy, reduces response latency, and enhances energy efficiency,
outperforming traditional FL techniques (i.e., FedAvg, FedOpt). These findings
highlight the potential of integrating LLM-powered federated learning into
large-scale IoT ecosystems, paving the way for more secure, scalable, and
adaptive IoT management solutions.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:56:59 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Otoum",
"Yazan",
""
],
[
"Asad",
"Arghavan",
""
],
[
"Nayak",
"Amiya",
""
]
] |
2504.16033 | Simon Chiossi | Adri\'an M. Andrada, Simon G. Chiossi, Alberth J. Nu\~nez | $\eta$-Einstein Sasakian Lie algebras | 32 pages | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | We study $\eta$-Einstein Sasakian structures on Lie algebras, that is,
Sasakian structures whose associated Ricci tensor satisfies an Einstein-like
condition. We divide into the cases in which the Lie algebra's centre is
non-trivial (and necessarily one-dimensional) and those where it is zero. In
the former case we show that any Sasakian structure on a unimodular Lie algebra
is $\eta$-Einstein. As for centreless Sasakian Lie algebras, we devise a
complete characterisation under certain dimensional assumptions regarding the
action of the Reeb vector. Using this result, together with the theory of
normal $j$-algebras and modifications of Hermitian Lie algebras, we construct
new examples of $\eta$-Einstein Sasakian Lie algebras and solvmanifolds, and
provide effective restrictions for their existence.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:57:22 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Andrada",
"Adrián M.",
""
],
[
"Chiossi",
"Simon G.",
""
],
[
"Nuñez",
"Alberth J.",
""
]
] |
2504.16034 | Pasquale Di Nezza | A. Accardi, A. Bacchetta, L. Barion, G. Bedeschi, V. Benesova, S.
Bertelli, V. Bertone, C. Bissolotti, M. Boglione, G. Bozzi, N. Bundaleski, V.
Carassiti, F. G. Celiberto, Z. Chen, G. Ciullo, M. Constantinou, P. Costa
Pinto, A. Courtoy, U. D'Alesio, C. De Angelis, E. De Lucia, I. Denisenko, P.
Di Nezza, M. Diehl, F. Donato, N. Doshita, O. M. N. Duarte Teodoro, M. G.
Echevarria, T. El-Kordy, R. Engels, F. Fabiano, I. P. Fernando, M.
Ferro-Luzzi, C. Flore, L. Gamberg, G. R. Goldstein, J. O. Gonzalez-Hernandez,
B. Gou, A. Gridin, A. Guskov, C. Hadjidakis, V. Hejny, T. Iwata, D. Keller,
N. Koch, A. Kotzinian, J. P. Lansberg, P. Lenisa, X. Li, H. W. Lin, S. Liuti,
R. Longo, M. Maggiora, G. Manca, S. Mariani, J. Matousek, T. Matsuda, A.
Metz, M. Mirazita, Y. Miyachi, A. Movsisyan, F. Murgia, A. Nass, E. R.
Nocera, C. Oppedisano, L. L. Pappalardo, B. Parsamyan, B. Pasquini, M. Pesek,
A. Piccoli, C. Pisano, D. Pitonyak, J. Pretz, A. Prokudin, M. Radici, F.
Rathmann, M. Rotondo, M. Santimaria, G. Schnell, R. Shankar, A. Signori, D.
Sivers, S. Squerzanti, M. Stancari, E. Steffens, L. Sun, H. Suzuki, G.
Tagliente, F. Tessarotto, C. Van Hulse, Q. Xu, Z. Ye, J. Zhang | LHCspin: a Polarized Gas Target for LHC | null | null | null | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | The goal of the LHCspin project is to develop innovative solutions for
measuring the 3D structure of nucleons in high-energy polarized fixed-target
collisions at LHC, exploring new processes and exploiting new probes in a
unique, previously unexplored, kinematic regime. A precise multi-dimensional
description of the hadron structure has, in fact, the potential to deepen our
understanding of the strong interactions and to provide a much more precise
framework for measuring both Standard Model and Beyond Standard Model
observables. This ambitious task poses its basis on the recent experience with
the successful installation and operation of the SMOG2 unpolarized gas target
in front of the LHCb spectrometer. Besides allowing for interesting physics
studies ranging from astrophysics to heavy-ion physics, SMOG2 provides an ideal
benchmark for studying beam-target dynamics at the LHC and demonstrates the
feasibility of simultaneous operation with beam-beam collisions. With the
installation of the proposed polarized target system, LHCb will become the
first experiment to simultaneously collect data from unpolarized beam-beam
collisions at $\sqrt{s}$=14 TeV and polarized and unpolarized beam-target
collisions at $\sqrt{s_{NN}}\sim$100 GeV. LHCspin has the potential to open new
frontiers in physics by exploiting the capabilities of the world's most
powerful collider and one of the most advanced spectrometers. This document
also highlights the need to perform an R\&D campaign and the commissioning of
the apparatus at the LHC Interaction Region 4 during the Run 4, before its
final installation in LHCb. This opportunity could also allow to undertake
preliminary physics measurements with unprecedented conditions.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:58:08 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Accardi",
"A.",
""
],
[
"Bacchetta",
"A.",
""
],
[
"Barion",
"L.",
""
],
[
"Bedeschi",
"G.",
""
],
[
"Benesova",
"V.",
""
],
[
"Bertelli",
"S.",
""
],
[
"Bertone",
"V.",
""
],
[
"Bissolotti",
"C.",
""
],
[
"Boglione",
"M.",
""
],
[
"Bozzi",
"G.",
""
],
[
"Bundaleski",
"N.",
""
],
[
"Carassiti",
"V.",
""
],
[
"Celiberto",
"F. G.",
""
],
[
"Chen",
"Z.",
""
],
[
"Ciullo",
"G.",
""
],
[
"Constantinou",
"M.",
""
],
[
"Pinto",
"P. Costa",
""
],
[
"Courtoy",
"A.",
""
],
[
"D'Alesio",
"U.",
""
],
[
"De Angelis",
"C.",
""
],
[
"De Lucia",
"E.",
""
],
[
"Denisenko",
"I.",
""
],
[
"Di Nezza",
"P.",
""
],
[
"Diehl",
"M.",
""
],
[
"Donato",
"F.",
""
],
[
"Doshita",
"N.",
""
],
[
"Teodoro",
"O. M. N. Duarte",
""
],
[
"Echevarria",
"M. G.",
""
],
[
"El-Kordy",
"T.",
""
],
[
"Engels",
"R.",
""
],
[
"Fabiano",
"F.",
""
],
[
"Fernando",
"I. P.",
""
],
[
"Ferro-Luzzi",
"M.",
""
],
[
"Flore",
"C.",
""
],
[
"Gamberg",
"L.",
""
],
[
"Goldstein",
"G. R.",
""
],
[
"Gonzalez-Hernandez",
"J. O.",
""
],
[
"Gou",
"B.",
""
],
[
"Gridin",
"A.",
""
],
[
"Guskov",
"A.",
""
],
[
"Hadjidakis",
"C.",
""
],
[
"Hejny",
"V.",
""
],
[
"Iwata",
"T.",
""
],
[
"Keller",
"D.",
""
],
[
"Koch",
"N.",
""
],
[
"Kotzinian",
"A.",
""
],
[
"Lansberg",
"J. P.",
""
],
[
"Lenisa",
"P.",
""
],
[
"Li",
"X.",
""
],
[
"Lin",
"H. W.",
""
],
[
"Liuti",
"S.",
""
],
[
"Longo",
"R.",
""
],
[
"Maggiora",
"M.",
""
],
[
"Manca",
"G.",
""
],
[
"Mariani",
"S.",
""
],
[
"Matousek",
"J.",
""
],
[
"Matsuda",
"T.",
""
],
[
"Metz",
"A.",
""
],
[
"Mirazita",
"M.",
""
],
[
"Miyachi",
"Y.",
""
],
[
"Movsisyan",
"A.",
""
],
[
"Murgia",
"F.",
""
],
[
"Nass",
"A.",
""
],
[
"Nocera",
"E. R.",
""
],
[
"Oppedisano",
"C.",
""
],
[
"Pappalardo",
"L. L.",
""
],
[
"Parsamyan",
"B.",
""
],
[
"Pasquini",
"B.",
""
],
[
"Pesek",
"M.",
""
],
[
"Piccoli",
"A.",
""
],
[
"Pisano",
"C.",
""
],
[
"Pitonyak",
"D.",
""
],
[
"Pretz",
"J.",
""
],
[
"Prokudin",
"A.",
""
],
[
"Radici",
"M.",
""
],
[
"Rathmann",
"F.",
""
],
[
"Rotondo",
"M.",
""
],
[
"Santimaria",
"M.",
""
],
[
"Schnell",
"G.",
""
],
[
"Shankar",
"R.",
""
],
[
"Signori",
"A.",
""
],
[
"Sivers",
"D.",
""
],
[
"Squerzanti",
"S.",
""
],
[
"Stancari",
"M.",
""
],
[
"Steffens",
"E.",
""
],
[
"Sun",
"L.",
""
],
[
"Suzuki",
"H.",
""
],
[
"Tagliente",
"G.",
""
],
[
"Tessarotto",
"F.",
""
],
[
"Van Hulse",
"C.",
""
],
[
"Xu",
"Q.",
""
],
[
"Ye",
"Z.",
""
],
[
"Zhang",
"J.",
""
]
] |
2504.16035 | Wenjing Zhang | Wenjing Zhang, Wandi Ding, Huaiping Zhu | Universal differential equations for optimal control problems and its
application on cancer therapy | null | null | null | null | math.OC math.DS | http://creativecommons.org/licenses/by/4.0/ | This paper highlights a parallel between the forward backward sweeping method
for optimal control and deep learning training procedures. We reformulate a
classical optimal control problem, constrained by a differential equation
system, into an optimization framework that uses neural networks to represent
control variables. We demonstrate that this deep learning method adheres to
Pontryagin Maximum Principle and mitigates numerical instabilities by employing
backward propagation instead of a backward sweep for the adjoint equations. As
a case study, we solve an optimal control problem to find the optimal
combination of immunotherapy and chemotherapy. Our approach holds significant
potential across various fields, including epidemiology, ecological modeling,
engineering, and financial mathematics, where optimal control under complex
dynamic constraints is crucial.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:58:37 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zhang",
"Wenjing",
""
],
[
"Ding",
"Wandi",
""
],
[
"Zhu",
"Huaiping",
""
]
] |
2504.16036 | Yang Zhang | Yang Zhang, Shuai Na, Jonathan J. Russin, Karteekeya Sastry, Li Lin,
Junfu Zheng, Yilin Luo, Xin Tong, Yujin An, Peng Hu, Konstantin Maslov,
Tze-Woei Tan, Charles Y. Liu, Lihong V. Wang | Rotational ultrasound and photoacoustic tomography of the human body | null | null | null | null | physics.med-ph eess.SP physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Imaging the human body's morphological and angiographic information is
essential for diagnosing, monitoring, and treating medical conditions.
Ultrasonography performs the morphological assessment of the soft tissue based
on acoustic impedance variations, whereas photoacoustic tomography (PAT) can
visualize blood vessels based on intrinsic hemoglobin absorption.
Three-dimensional (3D) panoramic imaging of the vasculature is generally not
practical in conventional ultrasonography with limited field-of-view (FOV)
probes, and PAT does not provide sufficient scattering-based soft tissue
morphological contrast. Complementing each other, fast panoramic rotational
ultrasound tomography (RUST) and PAT are integrated for hybrid rotational
ultrasound and photoacoustic tomography (RUS-PAT), which obtains 3D ultrasound
structural and PAT angiographic images of the human body quasi-simultaneously.
The RUST functionality is achieved in a cost-effective manner using a
single-element ultrasonic transducer for ultrasound transmission and rotating
arc-shaped arrays for 3D panoramic detection. RUST is superior to conventional
ultrasonography, which either has a limited FOV with a linear array or is
high-cost with a hemispherical array that requires both transmission and
receiving. By switching the acoustic source to a light source, the system is
conveniently converted to PAT mode to acquire angiographic images in the same
region. Using RUS-PAT, we have successfully imaged the human head, breast,
hand, and foot with a 10 cm diameter FOV, submillimeter isotropic resolution,
and 10 s imaging time for each modality. The 3D RUS-PAT is a powerful tool for
high-speed, 3D, dual-contrast imaging of the human body with potential for
rapid clinical translation.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:02:12 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zhang",
"Yang",
""
],
[
"Na",
"Shuai",
""
],
[
"Russin",
"Jonathan J.",
""
],
[
"Sastry",
"Karteekeya",
""
],
[
"Lin",
"Li",
""
],
[
"Zheng",
"Junfu",
""
],
[
"Luo",
"Yilin",
""
],
[
"Tong",
"Xin",
""
],
[
"An",
"Yujin",
""
],
[
"Hu",
"Peng",
""
],
[
"Maslov",
"Konstantin",
""
],
[
"Tan",
"Tze-Woei",
""
],
[
"Liu",
"Charles Y.",
""
],
[
"Wang",
"Lihong V.",
""
]
] |
2504.16037 | Shiliang Zhang | Haolin Liu, Shiliang Zhang, Shangbin Jiao, Xiaohui Zhang, Xuehui Ma,
Yan Yan, Wenchuan Cui, Youmin Zhang | Adaptive Fault-tolerant Control of Underwater Vehicles with Thruster
Failures | null | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper presents a fault-tolerant control for the trajectory tracking of
autonomous underwater vehicles (AUVs) against thruster failures. We formulate
faults in AUV thrusters as discrete switching events during a UAV mission, and
develop a soft-switching approach in facilitating shift of control strategies
across fault scenarios. We mathematically define AUV thruster fault scenarios,
and develop the fault-tolerant control that captures the fault scenario via
Bayesian approach. Particularly, when the AUV fault type switches from one to
another, the developed control captures the fault states and maintains the
control by a linear quadratic tracking controller. With the captured fault
states by Bayesian approach, we derive the control law by aggregating the
control outputs for individual fault scenarios weighted by their Bayesian
posterior probability. The developed fault-tolerant control works in an
adaptive way and guarantees soft-switching across fault scenarios, and requires
no complicated fault detection dedicated to different type of faults. The
entailed soft-switching ensures stable AUV trajectory tracking when fault type
shifts, which otherwise leads to reduced control under hard-switching control
strategies. We conduct numerical simulations with diverse AUV thruster fault
settings. The results demonstrate that the proposed control can provide smooth
transition across thruster failures, and effectively sustain AUV trajectory
tracking control in case of thruster failures and failure shifts.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:02:29 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Liu",
"Haolin",
""
],
[
"Zhang",
"Shiliang",
""
],
[
"Jiao",
"Shangbin",
""
],
[
"Zhang",
"Xiaohui",
""
],
[
"Ma",
"Xuehui",
""
],
[
"Yan",
"Yan",
""
],
[
"Cui",
"Wenchuan",
""
],
[
"Zhang",
"Youmin",
""
]
] |
2504.16038 | Atul Anurag | Atul Anurag, Roy H. Goodman | The Global Phase Space of the Three-Vortex Interaction System | null | null | null | null | math.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We derive a symplectic reduction of the evolution equations for a system of
three point vortices and use the reduced system to succinctly explain a kind of
bifurcation diagram that has appeared in the literature in a form that was
difficult to understand and interpret. Using this diagram, we enumerate and
plot all the global phase-space diagrams that occur as the circulations of the
three vortices are varied. The reduction proceeds in two steps: a reduction to
Jacobi coordinates and a Lie-Poisson reduction. In a recent paper, we used a
different method in the second step. This took two forms depending on a sign
that arose in the calculation. The Lie-Poisson equations unify these into a
single form. The Jacobi coordinate reduction fails when the total circulation
vanishes. We adapt the reduction method to this case and show how it relates to
the non-vanishing case.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:03:03 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Anurag",
"Atul",
""
],
[
"Goodman",
"Roy H.",
""
]
] |
2504.16039 | Ahmet Coskun Dr | Batuhan Kaplan, Samed Ke\c{s}\.ir, Ahmet Faruk Co\c{s}kun | A Comparative and Measurement-Based Study on Real-Time Network KPI
Extraction Methods for 5G and Beyond Applications | 6 pages, 4 figures, 1 table | null | null | null | cs.NI eess.SP | http://creativecommons.org/licenses/by/4.0/ | Key performance indicators (KPIs), which can be extracted from the
standardized interfaces of network equipment defined by current standards,
constitute a primary data source that can be leveraged in the development of
non-standardized new equipment, architectures, and computational tools. In
next-generation technologies, the demand for data has evolved beyond the
conventional log generation or export capabilities provided by existing
licensed network monitoring tools. There is now a growing need to collect such
data at specific time intervals and with defined granularities. At this stage,
the development of real-time KPI extraction methods and enabling their exchange
between both standardized/commercialized and non-standardized components or
tools has become increasingly critical. This study presents a comprehensive
evaluation of three distinct KPI extraction methodologies applied to two
commercially available devices. The analysis aims to uncover the strengths,
weaknesses, and overall efficacy of these approaches under varying conditions,
and highlights the critical insights into the practical capabilities and
limitations. The findings serve as a foundational guide for the seamless
integration and robust testing of novel technologies and approaches within
commercial telecommunication networks. This work aspires to bridge the gap
between technological innovation and real-world applicability, fostering
enhanced decision-making in network deployment and optimization.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:04:02 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Kaplan",
"Batuhan",
""
],
[
"Keşir",
"Samed",
""
],
[
"Coşkun",
"Ahmet Faruk",
""
]
] |
2504.16040 | Jongkuk Kim | Sarif Khan, Jongkuk Kim, Pyungwon Ko | Linking the KM3-230213A Neutrino Event to Dark Matter Decay and
Gravitational Waves Signals | 17 pages, 4 figures | null | null | null | hep-ph astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The KM3NeT collaboration recently reported the detection of an
ultra-high-energy neutrino event, dubbed KM3-230213A. This is the first
observed neutrino event with energy of the order of $\mathcal{O}(100)$PeV, the
origin of which remains unclear. We interprete this high energy neutrino event
results from the Dirac fermion dark matter (DM) $\chi$ decay through the
right-handed (RH) neutrino portal assuming the Type-I seesaw mechanism for
neutrino masses and mixings. Fruthermore, dark matter $\chi$ is assumed to
charged under $U(1)_X$ dark gauge symmetry, which is sponetaneously broken by
the vacuum expectation value of the dark Higgs $\Phi$. In this scenario, the DM
can decay into a pair of Standard Model (SM) particles for $v_\Phi \gg m_\chi$,
which we assume is the case. If the DM mass is around $440$ PeV with a lifetime
$5\times 10^{29}$ sec, it can account for the KM3-230213A event. However, such
heavy DM cannot be produced through the thermal freeze-out mechanism due to
overproduction and violation of unitarity bounds. We focus on the UV freeze-in
production of DM through a dimension-5 operator, which helps in producing the
DM dominantly in the early Universe. We have also found a set of allowed
parameter values that can correctly account for the DM relic density and decay
lifetime required to explain the KM3NeT signal. Moreover, we have generated the
neutrino spectra from the two-body decay using the HDMSpectra package, which
requires the dark Higgs vacuum expectations value (VEV) to be much larger than
the DM mass. Finally, the large value of the dark Higgs field VEV opens up the
possibility of generating GW spectra from cosmic strings. We have found a
reasonable set of parameter values that can address the KM3NeT signal, yield
the correct value of the DM relic density through freeze-in mechanism, and
allow for possible detection of GW at future detectors.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:07:39 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Khan",
"Sarif",
""
],
[
"Kim",
"Jongkuk",
""
],
[
"Ko",
"Pyungwon",
""
]
] |
2504.16041 | Amund Tveit | Amund Tveit and Bj{\o}rn Remseth and Arve Skogvold | Muon Optimizer Accelerates Grokking | 8 pages, 4 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the impact of different optimizers on the grokking
phenomenon, where models exhibit delayed generalization. We conducted
experiments across seven numerical tasks (primarily modular arithmetic) using a
modern Transformer architecture. The experimental configuration systematically
varied the optimizer (Muon vs. AdamW) and the softmax activation function
(standard softmax, stablemax, and sparsemax) to assess their combined effect on
learning dynamics. Our empirical evaluation reveals that the Muon optimizer,
characterized by its use of spectral norm constraints and second-order
information, significantly accelerates the onset of grokking compared to the
widely used AdamW optimizer. Specifically, Muon reduced the mean grokking epoch
from 153.09 to 102.89 across all configurations, a statistically significant
difference (t = 5.0175, p = 6.33e-08). This suggests that the optimizer choice
plays a crucial role in facilitating the transition from memorization to
generalization.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:08:09 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Tveit",
"Amund",
""
],
[
"Remseth",
"Bjørn",
""
],
[
"Skogvold",
"Arve",
""
]
] |
2504.16042 | Isma\"il Baaj | Isma\"il Baaj | Approximate matrices of systems of max-min fuzzy relational equations | null | null | null | null | cs.AI cs.LO | http://creativecommons.org/licenses/by/4.0/ | In this article, we address the inconsistency of a system of max-min fuzzy
relational equations by minimally modifying the matrix governing the system in
order to achieve consistency. Our method yields consistent systems that
approximate the original inconsistent system in the following sense: the
right-hand side vector of each consistent system is that of the inconsistent
system, and the coefficients of the matrix governing each consistent system are
obtained by modifying, exactly and minimally, the entries of the original
matrix that must be corrected to achieve consistency, while leaving all other
entries unchanged.
To obtain a consistent system that closely approximates the considered
inconsistent system, we study the distance (in terms of a norm among $L_1$,
$L_2$ or $L_\infty$) between the matrix of the inconsistent system and the set
formed by the matrices of consistent systems that use the same right-hand side
vector as the inconsistent system. We show that our method allows us to
directly compute matrices of consistent systems that use the same right-hand
side vector as the inconsistent system whose distance in terms of $L_\infty$
norm to the matrix of the inconsistent system is minimal (the computational
costs are higher when using $L_1$ norm or $L_2$ norm). We also give an explicit
analytical formula for computing this minimal $L_\infty$ distance. Finally, we
translate our results for systems of min-max fuzzy relational equations and
present some potential applications.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:09:02 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Baaj",
"Ismaïl",
""
]
] |
2504.16043 | Shankar Dayal Pathak | Gaurav Bhandari, S. D. Pathak, Vikash Kumar Ojha | RGUP Corrections to Scalar and Fermionic Fields | null | null | null | null | gr-qc hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | We investigate the Relativistic Generalized Uncertainty Principle (RGUP)
effects on scalar and fermionic fields using the Stetsko-Tkachuk approximation.
Modified equations of motion, Hamiltonians, and stress-energy tensors are
derived in Minkowski spacetime, incorporating quantum gravitational corrections
that ensure a minimal observable length and revert to standard dynamics when
corrections are absent. For fermionic fields in curved spacetime, spin
connections maintain gravitational consistency. This framework, applicable to
high-energy physics, black hole thermodynamics, and cosmology, integrates
quantum gravity into relativistic field theories.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:10:40 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Bhandari",
"Gaurav",
""
],
[
"Pathak",
"S. D.",
""
],
[
"Ojha",
"Vikash Kumar",
""
]
] |
2504.16044 | Yuzhe Zhang | Julian Walter, Olympia Maliaka, Yuzhe Zhang, John Blanchard, Gary
Centers, Arian Dogan, Martin Engler, Nataniel L. Figueroa, Younggeun Kim,
Derek F. Jackson Kimball, Matthew Lawson, Declan W. Smith, Alexander O.
Sushkov, Dmitry Budker, Hendrik Bekker and Arne Wickenbrock | Search for Axionlike Dark Matter Using Liquid-State Nuclear Magnetic
Resonance | null | null | null | null | hep-ex physics.atom-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We search for dark matter in the form of axionlike particles (ALPs) in the
mass range $5.576741 \,\mathrm{neV/c^2}$ - $5.577733\,\mathrm{neV/c^2}$ by
probing their possible coupling to fermion spins through the ALP field
gradient. This is achieved by performing proton nuclear magnetic resonance
spectroscopy on a sample of methanol as a technical demonstration of the Cosmic
Axion Spin Precession Experiment Gradient (CASPEr-Gradient) Low-Field
apparatus. Searching for spin-coupled ALP dark matter in this mass range with
associated Compton frequencies in a 240 Hz window centered at 1.348570 MHz
resulted in a sensitivity to the ALP-proton coupling constant of
$g_{\mathrm{ap}} \approx 3 \times 10^{-2}\,\mathrm{GeV}^{-1}$. This
narrow-bandwidth search serves as a proof-of-principle and a commissioning
measurement, validating our methodology and demonstrating the experiment's
capabilities. It opens the door to probing large swaths of hitherto unexplored
mass-coupling parameter space in the future by using hyperpolarized samples.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:11:01 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Walter",
"Julian",
""
],
[
"Maliaka",
"Olympia",
""
],
[
"Zhang",
"Yuzhe",
""
],
[
"Blanchard",
"John",
""
],
[
"Centers",
"Gary",
""
],
[
"Dogan",
"Arian",
""
],
[
"Engler",
"Martin",
""
],
[
"Figueroa",
"Nataniel L.",
""
],
[
"Kim",
"Younggeun",
""
],
[
"Kimball",
"Derek F. Jackson",
""
],
[
"Lawson",
"Matthew",
""
],
[
"Smith",
"Declan W.",
""
],
[
"Sushkov",
"Alexander O.",
""
],
[
"Budker",
"Dmitry",
""
],
[
"Bekker",
"Hendrik",
""
],
[
"Wickenbrock",
"Arne",
""
]
] |
2504.16045 | Mao Zeng | Mao Zeng | Reinforcement Learning and Metaheuristics for Feynman Integral Reduction | 11 pages, 7 figures, 1 table | null | null | null | hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | We propose new methods for optimizing the integration-by-parts (IBP)
reduction of Feynman integrals, an important computational bottleneck in modern
perturbative calculations in quantum field theory. Using the simple example of
one-loop massive bubble integrals, we pose the problem of minimizing the number
of arithmetic operations in reducing a target integral to master integrals via
the Laporta algorithm. This is a nontrivial combinatorial optimization problem
over the ordering of IBP equation generation (from pairs of seed integrals and
IBP operators) and the ordering of integral elimination. Our first proposed
method is reinforcement learning, which involves an agent interacting with an
environment in a step-by-step manner and learning the best actions to take
given an observation of the environment (in this case, the current state of the
IBP reduction process). The second method is using metaheuristics, e.g.
simulated annealing, to minimize the computational cost as a black-box function
of numerical priority values that control the orderings. For large-scale
problems, the number of free parameters can be compressed by using a small
neural network to assign priority values. Remarkably, both methods lead to IBP
reduction schemes that are more efficient than human-designed algorithms.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:14:05 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zeng",
"Mao",
""
]
] |
2504.16047 | Frank Li | Frank Li, Hari Trivedi, Bardia Khosravi, Theo Dapamede, Mohammadreza
Chavoshi, Abdulhameed Dere, Rohan Satya Isaac, Aawez Mansuri, Janice Newsome,
Saptarshi Purkayastha, Judy Gichoya | Evaluating Vision Language Models (VLMs) for Radiology: A Comprehensive
Analysis | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Foundation models, trained on vast amounts of data using self-supervised
techniques, have emerged as a promising frontier for advancing artificial
intelligence (AI) applications in medicine. This study evaluates three
different vision-language foundation models (RAD-DINO, CheXagent, and
BiomedCLIP) on their ability to capture fine-grained imaging features for
radiology tasks. The models were assessed across classification, segmentation,
and regression tasks for pneumothorax and cardiomegaly on chest radiographs.
Self-supervised RAD-DINO consistently excelled in segmentation tasks, while
text-supervised CheXagent demonstrated superior classification performance.
BiomedCLIP showed inconsistent performance across tasks. A custom segmentation
model that integrates global and local features substantially improved
performance for all foundation models, particularly for challenging
pneumothorax segmentation. The findings highlight that pre-training methodology
significantly influences model performance on specific downstream tasks. For
fine-grained segmentation tasks, models trained without text supervision
performed better, while text-supervised models offered advantages in
classification and interpretability. These insights provide guidance for
selecting foundation models based on specific clinical applications in
radiology.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:20:34 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Li",
"Frank",
""
],
[
"Trivedi",
"Hari",
""
],
[
"Khosravi",
"Bardia",
""
],
[
"Dapamede",
"Theo",
""
],
[
"Chavoshi",
"Mohammadreza",
""
],
[
"Dere",
"Abdulhameed",
""
],
[
"Isaac",
"Rohan Satya",
""
],
[
"Mansuri",
"Aawez",
""
],
[
"Newsome",
"Janice",
""
],
[
"Purkayastha",
"Saptarshi",
""
],
[
"Gichoya",
"Judy",
""
]
] |
2504.16048 | Nicholas Behr | Nicholas Julian Behr, Mattia Bianchi, Keith Moffat, Saverio Bolognani,
and Florian D\"orfler | PRIME: Fast Primal-Dual Feedback Optimization for Markets with
Application to Optimal Power Flow | Source code available at https://github.com/NicholasBehr/prime | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by-sa/4.0/ | Online Feedback Optimization (OFO) controllers iteratively drive a plant to
an optimal operating point that satisfies input and output constraints, relying
solely on the input-output sensitivity as model information. This paper
introduces PRIME (PRoximal Iterative MarkEts), a novel OFO approach based on
proximal-point iterations. Unlike existing OFO solutions, PRIME admits a
market-based implementation, where self-interested actors are incentivized to
make choices that result in a safe and efficient operation, without
communicating private costs or constraints. Furthermore, PRIME can cope with
non-smooth objective functions, achieve fast convergence rates and rapid
constraint satisfaction, and reject measurement noise. We demonstrate PRIME on
an AC optimal power flow problem, obtaining an efficient real-time nonlinear
local marginal pricing scheme.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:25:40 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Behr",
"Nicholas Julian",
""
],
[
"Bianchi",
"Mattia",
""
],
[
"Moffat",
"Keith",
""
],
[
"Bolognani",
"Saverio",
""
],
[
"Dörfler",
"Florian",
""
]
] |
2504.16049 | Ashutosh Dwibedi DWIBEDI | Ashutosh Dwibedi, Dani Rose J Marattukalam, Nandita Padhan, Arghya
Chatterjee and Sabyasachi Ghosh | Effect of Coriolis Force on the Shear Viscosity of Rotating Nuclear
Medium | null | null | null | null | nucl-th hep-th | http://creativecommons.org/licenses/by/4.0/ | Following the recent observation of non-zero spin polarization and spin
alignment of a few hadrons, the rotational aspect of quark-gluon plasma formed
in heavy ion collisions has attracted considerable interest. The present work
explores the effect of the Coriolis force, arising due to this rotation, on the
shear viscosity of the medium. Using the relaxation time approximation within
the kinetic theory framework, we analyze the parallel (\(\eta_{||}/s\)),
perpendicular (\(\eta_\perp/s\)) and Hall (\(\eta_\times/s\)) components of
shear viscosity to entropy density ratio under rotation. The estimation of
anisotropic shear viscosity components is carried out using hadron resonance
gas degrees of freedom below the critical (transition) temperature and massless
partonic degrees of freedom above this temperature. Our results show that
rotation suppresses the shear viscosity of the medium, with the degree of
suppression depending on the ratio between the relaxation time and the
rotational period. In the context of realistic heavy-ion collision experiments,
the temperature and angular velocity both decrease with time, and one can
establish a connection between them through the standard approximate cooling
law. For a temperature-dependent angular velocity \(\Omega(T)\), we obtain a
traditional valley-like pattern for all components \(\eta_{||}/s\),
\(\eta_\perp/s\) and \(\eta_\times/s\) with reduced magnitudes compared to the
valley-like isotropic $\eta/s$ one encounters in the absence of rotation.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:27:10 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Dwibedi",
"Ashutosh",
""
],
[
"Marattukalam",
"Dani Rose J",
""
],
[
"Padhan",
"Nandita",
""
],
[
"Chatterjee",
"Arghya",
""
],
[
"Ghosh",
"Sabyasachi",
""
]
] |
2504.16050 | Alexandr Dodin | A.V. Dodin, M.A. Burlak, V.A. Kiryukhina, S.A. Lamzin, I.A.
Shaposhnikov, I.A. Strakhov, A.A. Tatarnikov, A.M. Tatarnikov | The jet and circumstellar environment of the young binary DF Tau | 5 pages, 6 figures, accepted for publication in A&A Letters | null | null | null | astro-ph.SR astro-ph.EP astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Jets and disc winds play an important role in the evolution of protoplanetary
discs and the formation of planetary systems. However, there is still a lack of
observational data regarding the presence and parameters of outflows,
especially for close young binaries. In this study, we aim to find the HH flow
near the young sub-arcsecond binary DF Tau and explore its morphology.
Narrow-band H$\alpha$ and H$_2$ 2.12 $\mu$m imaging and spectroscopic
observations of DF Tau and its vicinity were performed. We have discovered
several emission nebulae near the binary, which likely result from the
interaction of gas outflow from the binary components with the surrounding
medium. The outflow appears to occur both in the form of jets, generating
numerous Herbig-Haro objects (HH 1266 flow), and as a weakly collimated wind
responsible for the formation of the ring-like nebula around the binary and the
rim of the cometary globule. We have found that the angle between the jet and
the counter-jet is $168^\circ$ and discuss the complex morphology of the HH
flow.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:27:17 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Dodin",
"A. V.",
""
],
[
"Burlak",
"M. A.",
""
],
[
"Kiryukhina",
"V. A.",
""
],
[
"Lamzin",
"S. A.",
""
],
[
"Shaposhnikov",
"I. A.",
""
],
[
"Strakhov",
"I. A.",
""
],
[
"Tatarnikov",
"A. A.",
""
],
[
"Tatarnikov",
"A. M.",
""
]
] |
2504.16051 | Ulrich Derenthal | Ulrich Derenthal | Manin's conjecture for the chordal cubic fourfold | 21 pages | null | null | null | math.NT math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove the thin set version of Manin's conjecture for the chordal (or:
determinantal) cubic fourfold, which is the secant variety of the Veronese
surface. We reduce this counting problem to a result of Schmidt for quadratic
points in the projective plane by showing that the chordal cubic fourfold is
isomorphic to the symmetric square of the projective plane over the rational
numbers.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:27:23 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Derenthal",
"Ulrich",
""
]
] |
2504.16052 | Ali Sadeghi | A. Aryanpour and Ali Sadeghi | Locally Machine-Learnability of Density of Electronic States | null | null | null | null | cond-mat.mtrl-sci cond-mat.dis-nn | http://creativecommons.org/licenses/by/4.0/ | The electronic density of states (DOS) plays a crucial role in determining
the properties of materials. In this study, we investigate the machine
learnability of additive atomic contributions to the electronic DOS. Our
approach focuses on atom-projected DOS rather than structural DOS. This method
for structure-property mapping is both scalable and transferable, achieving
high prediction accuracy for pure and compound silicon and carbon structures of
varying sizes and configurations. Furthermore, we demonstrate the effectiveness
of our method on the complex Sn-S-Se compound structures. By employing locally
trained DOS, we significantly enhance the accuracy in predicting secondary
material properties, such as band energy, Fermi level, heat capacity, and
magnetic susceptibility. Our findings indicate that directly learning atomic
DOS, rather than structural DOS, improves the efficiency, accuracy, and
interpretability of machine learning in structure-property mapping. This
streamlined approach reduces computational complexity, paving the way for the
examination of electronic structures in materials without the need for
computationally expensive ab initio calculations.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:28:53 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Aryanpour",
"A.",
""
],
[
"Sadeghi",
"Ali",
""
]
] |
2504.16053 | Zhifan Ye | Zhifan Ye, Kejing Xia, Yonggan Fu, Xin Dong, Jihoon Hong, Xiangchi
Yuan, Shizhe Diao, Jan Kautz, Pavlo Molchanov, Yingyan Celine Lin | LongMamba: Enhancing Mamba's Long Context Capabilities via Training-Free
Receptive Field Enlargement | Accepted by ICLR 2025 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | State space models (SSMs) have emerged as an efficient alternative to
Transformer models for language modeling, offering linear computational
complexity and constant memory usage as context length increases. However,
despite their efficiency in handling long contexts, recent studies have shown
that SSMs, such as Mamba models, generally underperform compared to
Transformers in long-context understanding tasks. To address this significant
shortfall and achieve both efficient and accurate long-context understanding,
we propose LongMamba, a training-free technique that significantly enhances the
long-context capabilities of Mamba models. LongMamba builds on our discovery
that the hidden channels in Mamba can be categorized into local and global
channels based on their receptive field lengths, with global channels primarily
responsible for long-context capability. These global channels can become the
key bottleneck as the input context lengthens. Specifically, when input lengths
largely exceed the training sequence length, global channels exhibit
limitations in adaptively extend their receptive fields, leading to Mamba's
poor long-context performance. The key idea of LongMamba is to mitigate the
hidden state memory decay in these global channels by preventing the
accumulation of unimportant tokens in their memory. This is achieved by first
identifying critical tokens in the global channels and then applying token
filtering to accumulate only those critical tokens. Through extensive
benchmarking across synthetic and real-world long-context scenarios, LongMamba
sets a new standard for Mamba's long-context performance, significantly
extending its operational range without requiring additional training. Our code
is available at https://github.com/GATECH-EIC/LongMamba.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:30:36 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Ye",
"Zhifan",
""
],
[
"Xia",
"Kejing",
""
],
[
"Fu",
"Yonggan",
""
],
[
"Dong",
"Xin",
""
],
[
"Hong",
"Jihoon",
""
],
[
"Yuan",
"Xiangchi",
""
],
[
"Diao",
"Shizhe",
""
],
[
"Kautz",
"Jan",
""
],
[
"Molchanov",
"Pavlo",
""
],
[
"Lin",
"Yingyan Celine",
""
]
] |
2504.16054 | Brian Ichter | Physical Intelligence, Kevin Black, Noah Brown, James Darpinian, Karan
Dhabalia, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo
Fusai, Manuel Y. Galliker, Dibya Ghosh, Lachy Groom, Karol Hausman, Brian
Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Devin LeBlanc, Sergey
Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Allen Z.
Ren, Lucy Xiaoyang Shi, Laura Smith, Jost Tobias Springenberg, Kyle
Stachowicz, James Tanner, Quan Vuong, Homer Walke, Anna Walling, Haohuan
Wang, Lili Yu, Ury Zhilinsky | $\pi_{0.5}$: a Vision-Language-Action Model with Open-World
Generalization | null | null | null | null | cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In order for robots to be useful, they must perform practically relevant
tasks in the real world, outside of the lab. While vision-language-action (VLA)
models have demonstrated impressive results for end-to-end robot control, it
remains an open question how far such models can generalize in the wild. We
describe $\pi_{0.5}$, a new model based on $\pi_{0}$ that uses co-training on
heterogeneous tasks to enable broad generalization. $\pi_{0.5}$\ uses data from
multiple robots, high-level semantic prediction, web data, and other sources to
enable broadly generalizable real-world robotic manipulation. Our system uses a
combination of co-training and hybrid multi-modal examples that combine image
observations, language commands, object detections, semantic subtask
prediction, and low-level actions. Our experiments show that this kind of
knowledge transfer is essential for effective generalization, and we
demonstrate for the first time that an end-to-end learning-enabled robotic
system can perform long-horizon and dexterous manipulation skills, such as
cleaning a kitchen or bedroom, in entirely new homes.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:31:29 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Intelligence",
"Physical",
""
],
[
"Black",
"Kevin",
""
],
[
"Brown",
"Noah",
""
],
[
"Darpinian",
"James",
""
],
[
"Dhabalia",
"Karan",
""
],
[
"Driess",
"Danny",
""
],
[
"Esmail",
"Adnan",
""
],
[
"Equi",
"Michael",
""
],
[
"Finn",
"Chelsea",
""
],
[
"Fusai",
"Niccolo",
""
],
[
"Galliker",
"Manuel Y.",
""
],
[
"Ghosh",
"Dibya",
""
],
[
"Groom",
"Lachy",
""
],
[
"Hausman",
"Karol",
""
],
[
"Ichter",
"Brian",
""
],
[
"Jakubczak",
"Szymon",
""
],
[
"Jones",
"Tim",
""
],
[
"Ke",
"Liyiming",
""
],
[
"LeBlanc",
"Devin",
""
],
[
"Levine",
"Sergey",
""
],
[
"Li-Bell",
"Adrian",
""
],
[
"Mothukuri",
"Mohith",
""
],
[
"Nair",
"Suraj",
""
],
[
"Pertsch",
"Karl",
""
],
[
"Ren",
"Allen Z.",
""
],
[
"Shi",
"Lucy Xiaoyang",
""
],
[
"Smith",
"Laura",
""
],
[
"Springenberg",
"Jost Tobias",
""
],
[
"Stachowicz",
"Kyle",
""
],
[
"Tanner",
"James",
""
],
[
"Vuong",
"Quan",
""
],
[
"Walke",
"Homer",
""
],
[
"Walling",
"Anna",
""
],
[
"Wang",
"Haohuan",
""
],
[
"Yu",
"Lili",
""
],
[
"Zhilinsky",
"Ury",
""
]
] |
2504.16055 | Denielle Oliva | Denielle Oliva, Abbie Olszewski, David Feil-Seifer | SAR4SLPs: An Asynchronous Survey of Speech-Language Pathologists'
Perspectives on Socially Assistive Robots | 8 pages, 1 figure, 2 tables | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Socially Assistive Robots (SARs) offer unique opportunities within speech
language pathology (SLP) education and practice by supporting interactive
interventions for children with communication disorders. This paper explores
the implementation of SAR4SLPs (Socially Assistive Robots for Speech-Language
Pathologists) to investigate aspects such as engagement, therapeutic strategy
discipline, and consistent intervention support. We assessed the current
application of technology to clinical and educational settings, especially with
respect to how SLPs might use SAR in their therapeutic work. An asynchronous
remote community (ARC) collaborated with a cohort of practicing SLPs to
consider the feasibility, potential effectiveness, and anticipated challenges
with implementing SARs in day-to-day interventions and as practice
facilitators. We focus in particular on the expressive functionality of SARs,
modeling a foundational strategy that SLPs employ across various intervention
targets. This paper highlights clinician-driven insights and design
implications for developing SARs that support specific treatment goals through
collaborative and iterative design.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:32:09 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Oliva",
"Denielle",
""
],
[
"Olszewski",
"Abbie",
""
],
[
"Feil-Seifer",
"David",
""
]
] |
2504.16056 | Daniel Hendriks | Daniel Hendriks, Philipp Spitzer, Niklas K\"uhl, and Gerhard Satzger | Honey, I Shrunk the Language Model: Impact of Knowledge Distillation
Methods on Performance and Explainability | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Artificial Intelligence (AI) has increasingly influenced modern society,
recently in particular through significant advancements in Large Language
Models (LLMs). However, high computational and storage demands of LLMs still
limit their deployment in resource-constrained environments. Knowledge
distillation addresses this challenge by training a small student model from a
larger teacher model. Previous research has introduced several distillation
methods for both generating training data and for training the student model.
Despite their relevance, the effects of state-of-the-art distillation methods
on model performance and explainability have not been thoroughly investigated
and compared. In this work, we enlarge the set of available methods by applying
critique-revision prompting to distillation for data generation and by
synthesizing existing methods for training. For these methods, we provide a
systematic comparison based on the widely used Commonsense Question-Answering
(CQA) dataset. While we measure performance via student model accuracy, we
employ a human-grounded study to evaluate explainability. We contribute new
distillation methods and their comparison in terms of both performance and
explainability. This should further advance the distillation of small language
models and, thus, contribute to broader applicability and faster diffusion of
LLM technology.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:32:48 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Hendriks",
"Daniel",
""
],
[
"Spitzer",
"Philipp",
""
],
[
"Kühl",
"Niklas",
""
],
[
"Satzger",
"Gerhard",
""
]
] |
2504.16058 | Hendrik M\"uller | Hendrik M\"uller, Sanjay Bhatnagar | How to make CLEAN variants faster? Using clustered components informed
by the autocorrelation function | accepted for A&A | null | null | null | astro-ph.IM astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Deconvolution, imaging and calibration of data from radio interferometers is
a challenging computational (inverse) problem. The upcoming generation of radio
telescopes poses significant challenges to existing, and well proven data
reduction pipelines due to the large data sizes expected from these
experiments, and the high resolution and dynamic range. In this manuscript, we
deal with the deconvolution problem. A variety of multiscalar variants to the
classical CLEAN algorithm (the de-facto standard) have been proposed in the
past, often outperforming CLEAN at the cost of significantly increasing
numerical resources. In this work, we aim to combine some of these ideas for a
new algorithm, Autocorr-CLEAN, to accelerate the deconvolution and prepare the
data reduction pipelines for the data sizes expected by the upcoming generation
of instruments. To this end, we propose to use a cluster of CLEAN components
fitted to the autocorrelation function of the residual in a subminor loop, to
derive continuously changing, and potentially non-radially symmetric, basis
functions for CLEANing the residual. Autocorr-CLEAN allows for the superior
reconstruction fidelity achieved by modern multiscalar approaches, and their
superior convergence speed. It achieves this without utilizing any substep of
super-linear complexity in the minor loops, keeping the single minor loop and
subminor loop iterations at an execution time comparable to CLEAN. Combining
these advantages, Autocorr-CLEAN is found to be up to a magnitude faster than
the classical CLEAN procedure. Autocorr-CLEAN fits well in the algorithmic
framework common for radio interferometry, making it relatively straightforward
to include in future data reduction pipelines. With its accelerated convergence
speed, and smaller residual, Autocorr-CLEAN may be an important asset for the
data analysis in the future.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:33:55 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Müller",
"Hendrik",
""
],
[
"Bhatnagar",
"Sanjay",
""
]
] |
2504.16059 | Debottam Nandi | Debottam Nandi (CCSP, SGT University), Rohan Roy, Simran Yadav, Arnab
Sarkar | Sub-Horizon Amplification of Curvature Perturbations: A New Route to
Primordial Black Holes and Gravitational Waves | 32 pages, 4 figures | null | null | null | astro-ph.CO gr-qc hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | The enhanced primordial scalar power spectrum is a widely studied mechanism
for generating primordial gravitational waves (PGWs), also referred to as
scalar-induced gravitational waves (SIGWs). This process also plays a pivotal
role in facilitating the formation of primordial black holes (PBHs).
Traditionally, the ultra slow-roll (USR) mechanism has been the predominant
approach used in the early universe. In this framework, the second slow-roll
parameter $\epsilon_2$, is typically set to $-6$ or lower for a brief period --
marking a significant departure from the standard slow-roll condition where
$\epsilon_2 \simeq 0$. Such conditions often emerge in models with inflection
points or localized features, such as bumps in the potential. In this paper, we
challenge the conventional assumption that $\epsilon_2 \lesssim -6$ is a
prerequisite for substantial amplification of the scalar power spectrum. We
demonstrate that any negative value of the second slow-roll parameter can
indeed enhance the scalar power spectrum through sub-horizon growth,
establishing this as a necessary and sufficient condition for amplification.
Consequently, this mechanism facilitates the generation of both PGWs and PBHs.
To illustrate this, we examine a standard scenario where a brief USR phase is
embedded between two slow-roll (SR) phases. By systematically varying
$\epsilon_{2}$ values from $-1$ to $-10$ in the USR region, we investigate the
amplification of the power spectrum and its implications for PGWs and PBHs
production, particularly in the context of ongoing and future cosmological
missions.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:36:01 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Nandi",
"Debottam",
"",
"CCSP, SGT University"
],
[
"Roy",
"Rohan",
""
],
[
"Yadav",
"Simran",
""
],
[
"Sarkar",
"Arnab",
""
]
] |
2504.16060 | Ziqiao Ma | Ziqiao Ma, Jing Ding, Xuejun Zhang, Dezhi Luo, Jiahe Ding, Sihan Xu,
Yuchen Huang, Run Peng, Joyce Chai | Vision-Language Models Are Not Pragmatically Competent in Referring
Expression Generation | Homepage: https://vlm-reg.github.io/ | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Referring Expression Generation (REG) is a core task for evaluating the
pragmatic competence of vision-language systems, requiring not only accurate
semantic grounding but also adherence to principles of cooperative
communication (Grice, 1975). However, current evaluations of vision-language
models (VLMs) often overlook the pragmatic dimension, reducing REG to a
region-based captioning task and neglecting Gricean maxims. In this work, we
revisit REG from a pragmatic perspective, introducing a new dataset (RefOI) of
1.5k images annotated with both written and spoken referring expressions.
Through a systematic evaluation of state-of-the-art VLMs, we identify three key
failures of pragmatic competence: (1) failure to uniquely identify the
referent, (2) inclusion of excessive or irrelevant information, and (3)
misalignment with human pragmatic preference, such as the underuse of minimal
spatial cues. We also show that standard automatic evaluations fail to capture
these pragmatic violations, reinforcing superficial cues rather than genuine
referential success. Our findings call for a renewed focus on pragmatically
informed models and evaluation frameworks that align with real human
communication.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:37:16 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Ma",
"Ziqiao",
""
],
[
"Ding",
"Jing",
""
],
[
"Zhang",
"Xuejun",
""
],
[
"Luo",
"Dezhi",
""
],
[
"Ding",
"Jiahe",
""
],
[
"Xu",
"Sihan",
""
],
[
"Huang",
"Yuchen",
""
],
[
"Peng",
"Run",
""
],
[
"Chai",
"Joyce",
""
]
] |
2504.16061 | Sangeet Khemlani | Sangeet Khemlani, Tyler Tran, Nathaniel Gyory, Anthony M. Harrison,
Wallace E. Lawson, Ravenna Thielstrom, Hunter Thompson, Taaren Singh, J.
Gregory Trafton | Vision language models are unreliable at trivial spatial cognition | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Vision language models (VLMs) are designed to extract relevant visuospatial
information from images. Some research suggests that VLMs can exhibit humanlike
scene understanding, while other investigations reveal difficulties in their
ability to process relational information. To achieve widespread applicability,
VLMs must perform reliably, yielding comparable competence across a wide
variety of related tasks. We sought to test how reliable these architectures
are at engaging in trivial spatial cognition, e.g., recognizing whether one
object is left of another in an uncluttered scene. We developed a benchmark
dataset -- TableTest -- whose images depict 3D scenes of objects arranged on a
table, and used it to evaluate state-of-the-art VLMs. Results show that
performance could be degraded by minor variations of prompts that use logically
equivalent descriptions. These analyses suggest limitations in how VLMs may
reason about spatial relations in real-world applications. They also reveal
novel opportunities for bolstering image caption corpora for more efficient
training and testing.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:38:01 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Khemlani",
"Sangeet",
""
],
[
"Tran",
"Tyler",
""
],
[
"Gyory",
"Nathaniel",
""
],
[
"Harrison",
"Anthony M.",
""
],
[
"Lawson",
"Wallace E.",
""
],
[
"Thielstrom",
"Ravenna",
""
],
[
"Thompson",
"Hunter",
""
],
[
"Singh",
"Taaren",
""
],
[
"Trafton",
"J. Gregory",
""
]
] |
2504.16062 | Jiaxu Xing | Hardik Shah, Jiaxu Xing, Nico Messikommer, Boyang Sun, Marc Pollefeys,
Davide Scaramuzza | ForesightNav: Learning Scene Imagination for Efficient Exploration | null | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding how humans leverage prior knowledge to navigate unseen
environments while making exploratory decisions is essential for developing
autonomous robots with similar abilities. In this work, we propose
ForesightNav, a novel exploration strategy inspired by human imagination and
reasoning. Our approach equips robotic agents with the capability to predict
contextual information, such as occupancy and semantic details, for unexplored
regions. These predictions enable the robot to efficiently select meaningful
long-term navigation goals, significantly enhancing exploration in unseen
environments. We validate our imagination-based approach using the Structured3D
dataset, demonstrating accurate occupancy prediction and superior performance
in anticipating unseen scene geometry. Our experiments show that the
imagination module improves exploration efficiency in unseen environments,
achieving a 100% completion rate for PointNav and an SPL of 67% for ObjectNav
on the Structured3D Validation split. These contributions demonstrate the power
of imagination-driven reasoning for autonomous systems to enhance generalizable
and efficient exploration.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:38:38 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Shah",
"Hardik",
""
],
[
"Xing",
"Jiaxu",
""
],
[
"Messikommer",
"Nico",
""
],
[
"Sun",
"Boyang",
""
],
[
"Pollefeys",
"Marc",
""
],
[
"Scaramuzza",
"Davide",
""
]
] |
2504.16063 | Andrea Fronzetti Colladon PhD | A. Fronzetti Colladon, R. Vestrelli | A Python Tool for Reconstructing Full News Text from GDELT | null | null | null | null | cs.CL cs.DB cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | News data have become an essential resource across various disciplines,
including economics, finance, management, social sciences, and computer
science. Researchers leverage newspaper articles to study economic trends,
market dynamics, corporate strategies, public perception, political discourse,
and the evolution of public opinion. Additionally, news datasets have been
instrumental in training large-scale language models, with applications in
sentiment analysis, fake news detection, and automated news summarization.
Despite their significance, access to comprehensive news corpora remains a key
challenge. Many full-text news providers, such as Factiva and LexisNexis,
require costly subscriptions, while free alternatives often suffer from
incomplete data and transparency issues. This paper presents a novel approach
to obtaining full-text newspaper articles at near-zero cost by leveraging data
from the Global Database of Events, Language, and Tone (GDELT). Specifically,
we focus on the GDELT Web News NGrams 3.0 dataset, which provides
high-frequency updates of n-grams extracted from global online news sources. We
provide Python code to reconstruct full-text articles from these n-grams by
identifying overlapping textual fragments and intelligently merging them. Our
method enables researchers to access structured, large-scale newspaper data for
text analysis while overcoming the limitations of existing proprietary
datasets. The proposed approach enhances the accessibility of news data for
empirical research, facilitating applications in economic forecasting,
computational social science, and natural language processing.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:40:42 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Colladon",
"A. Fronzetti",
""
],
[
"Vestrelli",
"R.",
""
]
] |
2504.16064 | Thodoris Kouzelis | Theodoros Kouzelis, Efstathios Karypidis, Ioannis Kakogeorgiou, Spyros
Gidaris, Nikos Komodakis | Boosting Generative Image Modeling via Joint Image-Feature Synthesis | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Latent diffusion models (LDMs) dominate high-quality image generation, yet
integrating representation learning with generative modeling remains a
challenge. We introduce a novel generative image modeling framework that
seamlessly bridges this gap by leveraging a diffusion model to jointly model
low-level image latents (from a variational autoencoder) and high-level
semantic features (from a pretrained self-supervised encoder like DINO). Our
latent-semantic diffusion approach learns to generate coherent image-feature
pairs from pure noise, significantly enhancing both generative quality and
training efficiency, all while requiring only minimal modifications to standard
Diffusion Transformer architectures. By eliminating the need for complex
distillation objectives, our unified design simplifies training and unlocks a
powerful new inference strategy: Representation Guidance, which leverages
learned semantics to steer and refine image generation. Evaluated in both
conditional and unconditional settings, our method delivers substantial
improvements in image quality and training convergence speed, establishing a
new direction for representation-aware generative modeling.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:41:42 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Kouzelis",
"Theodoros",
""
],
[
"Karypidis",
"Efstathios",
""
],
[
"Kakogeorgiou",
"Ioannis",
""
],
[
"Gidaris",
"Spyros",
""
],
[
"Komodakis",
"Nikos",
""
]
] |
2504.16065 | Shyamal Patel | Xi Chen, Shyamal Patel, Rocco A. Servedio | A Mysterious Connection Between Tolerant Junta Testing and Agnostically
Learning Conjunctions | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The main conceptual contribution of this paper is identifying a previously
unnoticed connection between two central problems in computational learning
theory and property testing: agnostically learning conjunctions and tolerantly
testing juntas. Inspired by this connection, the main technical contribution is
a pair of improved algorithms for these two problems.
In more detail,
- We give a distribution-free algorithm for agnostically PAC learning
conjunctions over $\{\pm 1\}^n$ that runs in time $2^{\widetilde{O}(n^{1/3})}$,
for constant excess error $\varepsilon$. This improves on the fastest
previously published algorithm, which runs in time $2^{\widetilde{O}(n^{1/2})}$
[KKMS08].
- Building on the ideas in our agnostic conjunction learner and using
significant additional technical ingredients, we give an adaptive tolerant
testing algorithm for $k$-juntas that makes $2^{\widetilde{O}(k^{1/3})}$
queries, for constant "gap parameter" $\varepsilon$ between the "near" and
"far" cases. This improves on the best previous results, due to [ITW21, NP24],
which make $2^{\widetilde{O}(\sqrt{k})}$ queries. Since there is a known
$2^{\widetilde{\Omega}(\sqrt{k})}$ lower bound for non-adaptive tolerant junta
testers, our result shows that adaptive tolerant junta testing algorithms
provably outperform non-adaptive ones.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:42:52 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Chen",
"Xi",
""
],
[
"Patel",
"Shyamal",
""
],
[
"Servedio",
"Rocco A.",
""
]
] |
2504.16066 | Kejie Fang | Kejie Fang and Elizabeth A. Goldschmidt | Fidelity and efficiency analysis for heralded entanglement swapping in
lossy channels: linear and nonlinear optical approaches | null | null | null | null | quant-ph physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bell state measurements (BSMs) of photonic qubits are used for quantum
networking protocols to herald the distribution and transfer of quantum
information. However, standard BSMs based on linear optics (LO-BSMs) require
identical photons and are susceptible to errors from multiphoton emissions,
leading to reduced fidelity of protocols, particularly in the absence of
postselection. To overcome these challenges, Bell state measurements based on
nonlinear optics (NLO-BSMs) have been proposed and implemented, leveraging
sum-frequency generation (SFG) to filter out multiphoton emissions and improve
the fidelity without postselection. Here we analyze the fidelity of LO-BSM and
NLO-BSM heralded entanglement swapping in lossy channels and compare their
performance under realistic conditions. We also explore the impact of
state-of-the-art nonlinear photonics platforms on SFG efficiency and highlight
how advances in nanophotonics will enable practical, high-fidelity
NLO-BSM-based quantum networking applications.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:43:45 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Fang",
"Kejie",
""
],
[
"Goldschmidt",
"Elizabeth A.",
""
]
] |
2504.16067 | Ruohong Li | Romina Schulz, Ruohong Li, Julius Wessolek, Maryam Mostamand, Peter
Kunz, Jens Lassen | Two-step laser resonant ionization spectroscopy of chromium | 8 pages, 7 figures | null | null | null | physics.atom-ph physics.acc-ph physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | At TRIUMF's off-line laser ion source test stand, stepwise resonant laser
ionization spectroscopy of chromium (Cr) was carried out, to find an efficient
ionization scheme suitable for titanium sapphire (Ti:Sa) laser systems. With
three different first-excitation transitions, 357.971 nm, 359.451 nm, and
360.636 nm, automated continuous laser-frequency scans using a
frequency-doubled, grating-tuned Ti:Sa laser were performed. Rydberg series as
well as autoionizing(AI) states were observed. From these results, the
ionization potential (IP) of Cr was determined as
54575.49(2)$_\text{stat}$(2)$_\text{sys}$ cm$^{-1}$, which is one order of
magnitude more precise than the previously reported 54575.6(3) cm$^{-1}$ in
NIST database. The ionization scheme using the observed AI resonance, with
357.971 nm as the first step and 373.935 nm as the second step was subsequently
deployed to the online delivery of radioactive Cr isotope beams for precision
mass measurements. The online yields of $^{50-59}$Cr have been measured at
TRIUMF-ISAC.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:46:56 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Schulz",
"Romina",
""
],
[
"Li",
"Ruohong",
""
],
[
"Wessolek",
"Julius",
""
],
[
"Mostamand",
"Maryam",
""
],
[
"Kunz",
"Peter",
""
],
[
"Lassen",
"Jens",
""
]
] |
2504.16068 | Albert Musaelian | Chuin Wei Tan, Marc L. Descoteaux, Mit Kotak, Gabriel de Miranda
Nascimento, Se\'an R. Kavanagh, Laura Zichi, Menghang Wang, Aadit Saluja,
Yizhong R. Hu, Tess Smidt, Anders Johansson, William C. Witt, Boris Kozinsky,
Albert Musaelian | High-performance training and inference for deep equivariant interatomic
potentials | null | null | null | null | physics.comp-ph cs.LG physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning interatomic potentials, particularly those based on deep
equivariant neural networks, have demonstrated state-of-the-art accuracy and
computational efficiency in atomistic modeling tasks like molecular dynamics
and high-throughput screening. The size of datasets and demands of downstream
workflows are growing rapidly, making robust and scalable software essential.
This work presents a major overhaul of the NequIP framework focusing on
multi-node parallelism, computational performance, and extensibility. The
redesigned framework supports distributed training on large datasets and
removes barriers preventing full utilization of the PyTorch 2.0 compiler at
train time. We demonstrate this acceleration in a case study by training
Allegro models on the SPICE 2 dataset of organic molecular systems. For
inference, we introduce the first end-to-end infrastructure that uses the
PyTorch Ahead-of-Time Inductor compiler for machine learning interatomic
potentials. Additionally, we implement a custom kernel for the Allegro model's
most expensive operation, the tensor product. Together, these advancements
speed up molecular dynamics calculations on system sizes of practical relevance
by up to a factor of 18.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:47:01 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Tan",
"Chuin Wei",
""
],
[
"Descoteaux",
"Marc L.",
""
],
[
"Kotak",
"Mit",
""
],
[
"Nascimento",
"Gabriel de Miranda",
""
],
[
"Kavanagh",
"Seán R.",
""
],
[
"Zichi",
"Laura",
""
],
[
"Wang",
"Menghang",
""
],
[
"Saluja",
"Aadit",
""
],
[
"Hu",
"Yizhong R.",
""
],
[
"Smidt",
"Tess",
""
],
[
"Johansson",
"Anders",
""
],
[
"Witt",
"William C.",
""
],
[
"Kozinsky",
"Boris",
""
],
[
"Musaelian",
"Albert",
""
]
] |
2504.16069 | Muzi Hong | Minxi He, Muzi Hong, Kyohei Mukaida | Increase of $n_s$ in regularized pole inflation & Einstein-Cartan
gravity | null | null | null | null | astro-ph.CO hep-ph | http://creativecommons.org/licenses/by/4.0/ | We show that the regularization of the second order pole in the pole
inflation can induce the increase of $n_s$, which may be important after the
latest data release of cosmic microwave background (CMB) observation by Atacama
Cosmology Telescope (ACT). Pole inflation is known to provide a unified
description of attractor models that they can generate a flat plateau for
inflation given a general potential. Recent ACT observation suggests that the
constraint on the scalar spectral index $n_s$ at CMB scale may be shifted to a
larger value than the predictions in the Starobinsky model, the Higgs
inflation, and the $\alpha$-attractor model, which motivates us to consider the
modification of the pole inflation. We find that if we regularize the second
order pole in the kinetic term such that the kinetic term becomes regular for
all field range, we can generally increase $n_s$ because the potential in the
large field regime will be lifted. We have explicitly demonstrated that this
type of regularized pole inflation can naturally arise from the Einstein-Cartan
formalism, and the inflationary predictions are consistent with the latest ACT
data without spoiling the success of the $\alpha$-attractor models.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:49:04 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"He",
"Minxi",
""
],
[
"Hong",
"Muzi",
""
],
[
"Mukaida",
"Kyohei",
""
]
] |
2504.16070 | Kumarasamy Sakthivel | T. Sharma, L. Beilina and K. Sakthivel | Reconstruction of source function in a parabolic equation using partial
boundary measurements | null | null | null | null | math.NA cs.NA math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present the analytical and numerical study of the
optimization approach for determining the space-dependent source function in
the parabolic inverse source problem using partial boundary measurements. The
Lagrangian approach for the solution of the optimization problem is presented,
and optimality conditions are derived. The proof of the Fr\'echet
differentiability of the regularized Tikhonov functional and the existence
result for the solution of the inverse source problem are established. A local
stability estimate for the unknown source term is also presented. The numerical
examples justify the theoretical investigations using the conjugate gradient
method (CGM) in 2D and 3D tests with noisy data.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:51:31 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Sharma",
"T.",
""
],
[
"Beilina",
"L.",
""
],
[
"Sakthivel",
"K.",
""
]
] |
2504.16071 | Ahmed Hareedy | Ata Tanr{\i}kulu, Mete Y{\i}ld{\i}r{\i}m, Ahmed Hareedy | A Markov Chain Monte Carlo Method for Efficient Finite-Length LDPC Code
Design | 13 pages (double column), 0 figures, submitted to the IEEE
Transactions on Communications (TCOM) | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Low-density parity-check (LDPC) codes are among the most prominent
error-correction schemes. They find application to fortify various modern
storage, communication, and computing systems. Protograph-based (PB) LDPC codes
offer many degrees of freedom in the code design and enable fast encoding and
decoding. In particular, spatially-coupled (SC) and multi-dimensional (MD)
circulant-based codes are PB-LDPC codes with excellent performance. Efficient
finite-length (FL) algorithms are required in order to effectively exploit the
available degrees of freedom offered by SC partitioning, lifting, and MD
relocations. In this paper, we propose a novel Markov chain Monte Carlo (MCMC
or MC$^2$) method to perform this FL optimization, addressing the removal of
short cycles. While iterating, we draw samples from a defined distribution
where the probability decreases as the number of short cycles from the previous
iteration increases. We analyze our MC$^2$ method theoretically as we prove the
invariance of the Markov chain where each state represents a possible
partitioning or lifting arrangement. Via our simulations, we then fit the
distribution of the number of cycles resulting from a given arrangement on a
Gaussian distribution. We derive estimates for cycle counts that are close to
the actual counts. Furthermore, we derive the order of the expected number of
iterations required by our approach to reach a local minimum as well as the
size of the Markov chain recurrent class. Our approach is compatible with code
design techniques based on gradient-descent. Numerical results show that our
MC$^2$ method generates SC codes with remarkably less number of short cycles
compared with the current state-of-the-art. Moreover, to reach the same number
of cycles, our method requires orders of magnitude less overall time compared
with the available literature methods.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:51:40 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Tanrıkulu",
"Ata",
""
],
[
"Yıldırım",
"Mete",
""
],
[
"Hareedy",
"Ahmed",
""
]
] |
2504.16072 | Long Lian | Long Lian, Yifan Ding, Yunhao Ge, Sifei Liu, Hanzi Mao, Boyi Li, Marco
Pavone, Ming-Yu Liu, Trevor Darrell, Adam Yala, Yin Cui | Describe Anything: Detailed Localized Image and Video Captioning | Project page: https://describe-anything.github.io/ | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Generating detailed and accurate descriptions for specific regions in images
and videos remains a fundamental challenge for vision-language models. We
introduce the Describe Anything Model (DAM), a model designed for detailed
localized captioning (DLC). DAM preserves both local details and global context
through two key innovations: a focal prompt, which ensures high-resolution
encoding of targeted regions, and a localized vision backbone, which integrates
precise localization with its broader context. To tackle the scarcity of
high-quality DLC data, we propose a Semi-supervised learning (SSL)-based Data
Pipeline (DLC-SDP). DLC-SDP starts with existing segmentation datasets and
expands to unlabeled web images using SSL. We introduce DLC-Bench, a benchmark
designed to evaluate DLC without relying on reference captions. DAM sets new
state-of-the-art on 7 benchmarks spanning keyword-level, phrase-level, and
detailed multi-sentence localized image and video captioning.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:51:41 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Lian",
"Long",
""
],
[
"Ding",
"Yifan",
""
],
[
"Ge",
"Yunhao",
""
],
[
"Liu",
"Sifei",
""
],
[
"Mao",
"Hanzi",
""
],
[
"Li",
"Boyi",
""
],
[
"Pavone",
"Marco",
""
],
[
"Liu",
"Ming-Yu",
""
],
[
"Darrell",
"Trevor",
""
],
[
"Yala",
"Adam",
""
],
[
"Cui",
"Yin",
""
]
] |
2504.16073 | Shiyun Xiong | Zhiyuan Hu, Shiyun Xiong, Yifan Zhang, See-Kiong Ng, Anh Tuan Luu, Bo
An, Shuicheng Yan, Bryan Hooi | Guiding VLM Agents with Process Rewards at Inference Time for GUI
Navigation | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in visual language models (VLMs) have notably enhanced
their capabilities in handling complex Graphical User Interface (GUI)
interaction tasks. Despite these improvements, current frameworks often
struggle to generate correct actions in challenging GUI environments.
State-of-the-art commercial VLMs are black-boxes, and fine-tuning open-source
VLMs for GUI tasks requires significant resources. Additionally, existing
trajectory-level evaluation and refinement techniques frequently fall short due
to delayed feedback and local optimization issues. To address these challenges,
we propose an approach that guides VLM agents with process supervision by a
reward model during GUI navigation and control at inference time. This guidance
allows the VLM agent to optimize actions at each inference step, thereby
improving performance in both static and dynamic environments. In particular,
our method demonstrates significant performance gains in three GUI navigation
tasks, achieving a 3.4% improvement in single step action accuracy for static
environments, along with a around 33% increase in task success rate in one
dynamic environment. With further integration of trajectory reflection and
retry mechanisms, we also demonstrate even greater enhancement in task success.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:52:42 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Hu",
"Zhiyuan",
""
],
[
"Xiong",
"Shiyun",
""
],
[
"Zhang",
"Yifan",
""
],
[
"Ng",
"See-Kiong",
""
],
[
"Luu",
"Anh Tuan",
""
],
[
"An",
"Bo",
""
],
[
"Yan",
"Shuicheng",
""
],
[
"Hooi",
"Bryan",
""
]
] |
2504.16074 | Zhuo-Yang Song | Shi Qiu, Shaoyang Guo, Zhuo-Yang Song, Yunbo Sun, Zeyu Cai, Jiashen
Wei, Tianyu Luo, Yixuan Yin, Haoxu Zhang, Yi Hu, Chenyang Wang, Chencheng
Tang, Haoling Chang, Qi Liu, Ziheng Zhou, Tianyu Zhang, Jingtian Zhang,
Zhangyi Liu, Minghao Li, Yuku Zhang, Boxuan Jing, Xianqi Yin, Yutong Ren,
Zizhuo Fu, Weike Wang, Xudong Tian, Anqi Lv, Laifu Man, Jianxiang Li, Feiyu
Tao, Qihua Sun, Zhou Liang, Yushu Mu, Zhongxuan Li, Jing-Jun Zhang, Shutao
Zhang, Xiaotian Li, Xingqi Xia, Jiawei Lin, Zheyu Shen, Jiahang Chen, Qiuhao
Xiong, Binran Wang, Fengyuan Wang, Ziyang Ni, Bohan Zhang, Fan Cui, Changkun
Shao, Qing-Hong Cao, Ming-xing Luo, Muhan Zhang, Hua Xing Zhu | PHYBench: Holistic Evaluation of Physical Perception and Reasoning in
Large Language Models | 21 pages ,8 figures, 4 tables | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce PHYBench, a novel, high-quality benchmark designed for
evaluating reasoning capabilities of large language models (LLMs) in physical
contexts. PHYBench consists of 500 meticulously curated physics problems based
on real-world physical scenarios, designed to assess the ability of models to
understand and reason about realistic physical processes. Covering mechanics,
electromagnetism, thermodynamics, optics, modern physics, and advanced physics,
the benchmark spans difficulty levels from high school exercises to
undergraduate problems and Physics Olympiad challenges. Additionally, we
propose the Expression Edit Distance (EED) Score, a novel evaluation metric
based on the edit distance between mathematical expressions, which effectively
captures differences in model reasoning processes and results beyond
traditional binary scoring methods. We evaluate various LLMs on PHYBench and
compare their performance with human experts. Our results reveal that even
state-of-the-art reasoning models significantly lag behind human experts,
highlighting their limitations and the need for improvement in complex physical
reasoning scenarios. Our benchmark results and dataset are publicly available
at https://phybench-official.github.io/phybench-demo/.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:53:29 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Qiu",
"Shi",
""
],
[
"Guo",
"Shaoyang",
""
],
[
"Song",
"Zhuo-Yang",
""
],
[
"Sun",
"Yunbo",
""
],
[
"Cai",
"Zeyu",
""
],
[
"Wei",
"Jiashen",
""
],
[
"Luo",
"Tianyu",
""
],
[
"Yin",
"Yixuan",
""
],
[
"Zhang",
"Haoxu",
""
],
[
"Hu",
"Yi",
""
],
[
"Wang",
"Chenyang",
""
],
[
"Tang",
"Chencheng",
""
],
[
"Chang",
"Haoling",
""
],
[
"Liu",
"Qi",
""
],
[
"Zhou",
"Ziheng",
""
],
[
"Zhang",
"Tianyu",
""
],
[
"Zhang",
"Jingtian",
""
],
[
"Liu",
"Zhangyi",
""
],
[
"Li",
"Minghao",
""
],
[
"Zhang",
"Yuku",
""
],
[
"Jing",
"Boxuan",
""
],
[
"Yin",
"Xianqi",
""
],
[
"Ren",
"Yutong",
""
],
[
"Fu",
"Zizhuo",
""
],
[
"Wang",
"Weike",
""
],
[
"Tian",
"Xudong",
""
],
[
"Lv",
"Anqi",
""
],
[
"Man",
"Laifu",
""
],
[
"Li",
"Jianxiang",
""
],
[
"Tao",
"Feiyu",
""
],
[
"Sun",
"Qihua",
""
],
[
"Liang",
"Zhou",
""
],
[
"Mu",
"Yushu",
""
],
[
"Li",
"Zhongxuan",
""
],
[
"Zhang",
"Jing-Jun",
""
],
[
"Zhang",
"Shutao",
""
],
[
"Li",
"Xiaotian",
""
],
[
"Xia",
"Xingqi",
""
],
[
"Lin",
"Jiawei",
""
],
[
"Shen",
"Zheyu",
""
],
[
"Chen",
"Jiahang",
""
],
[
"Xiong",
"Qiuhao",
""
],
[
"Wang",
"Binran",
""
],
[
"Wang",
"Fengyuan",
""
],
[
"Ni",
"Ziyang",
""
],
[
"Zhang",
"Bohan",
""
],
[
"Cui",
"Fan",
""
],
[
"Shao",
"Changkun",
""
],
[
"Cao",
"Qing-Hong",
""
],
[
"Luo",
"Ming-xing",
""
],
[
"Zhang",
"Muhan",
""
],
[
"Zhu",
"Hua Xing",
""
]
] |
2504.16075 | Joshua Harvey | Joshua S. Harvey, Joshua Rosaler, Mingshu Li, Dhruv Desai, Dhagash
Mehta | Explainable Unsupervised Anomaly Detection with Random Forest | 14 pages, 5 figures | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We describe the use of an unsupervised Random Forest for similarity learning
and improved unsupervised anomaly detection. By training a Random Forest to
discriminate between real data and synthetic data sampled from a uniform
distribution over the real data bounds, a distance measure is obtained that
anisometrically transforms the data, expanding distances at the boundary of the
data manifold. We show that using distances recovered from this transformation
improves the accuracy of unsupervised anomaly detection, compared to other
commonly used detectors, demonstrated over a large number of benchmark
datasets. As well as improved performance, this method has advantages over
other unsupervised anomaly detection methods, including minimal requirements
for data preprocessing, native handling of missing data, and potential for
visualizations. By relating outlier scores to partitions of the Random Forest,
we develop a method for locally explainable anomaly predictions in terms of
feature importance.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:54:44 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Harvey",
"Joshua S.",
""
],
[
"Rosaler",
"Joshua",
""
],
[
"Li",
"Mingshu",
""
],
[
"Desai",
"Dhruv",
""
],
[
"Mehta",
"Dhagash",
""
]
] |
2504.16076 | Claire Lamman | Claire Lamman, Jonathan Blazek, Daniel J. Eisenstein | Optimal intrinsic alignment estimators in the presence of redshift-space
distortions | For an accessible summary of this paper, see
https://cmlamman.github.io/doc/optimalRSDstats-annotation.pdf | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We present estimators for quantifying intrinsic alignments in large
spectroscopic surveys that efficiently capture line-of-sight (LOS) information
while being relatively insensitive to redshift-space distortions (RSD). We
demonstrate that changing the LOS integration range, {\Pi}max, as a function of
transverse separation outperforms the conventional choice of a single {\Pi}max
value. This is further improved by replacing the flat {\Pi}max cut with a LOS
weighting based on shape projection and RSD. Although these estimators
incorporate additional LOS information, they are projected correlations that
exhibit signal-to-noise ratios comparable to 3D correlation functions, such as
the IA quadrupole. Using simulations from Abacus Summit, we evaluate these
estimators and provide recommended {\Pi}max values and weights for projected
separations of 1 - 100 Mpc/h. These will improve measurements of intrinsic
alignments in large cosmological surveys and the constraints they provide for
both weak lensing and direct cosmological applications.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:55:52 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Lamman",
"Claire",
""
],
[
"Blazek",
"Jonathan",
""
],
[
"Eisenstein",
"Daniel J.",
""
]
] |
2504.16078 | Thomas Schmied | Thomas Schmied, J\"org Bornschein, Jordi Grau-Moya, Markus Wulfmeier,
Razvan Pascanu | LLMs are Greedy Agents: Effects of RL Fine-tuning on Decision-Making
Abilities | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The success of Large Language Models (LLMs) has sparked interest in various
agentic applications. A key hypothesis is that LLMs, leveraging common sense
and Chain-of-Thought (CoT) reasoning, can effectively explore and efficiently
solve complex domains. However, LLM agents have been found to suffer from
sub-optimal exploration and the knowing-doing gap, the inability to effectively
act on knowledge present in the model. In this work, we systematically study
why LLMs perform sub-optimally in decision-making scenarios. In particular, we
closely examine three prevalent failure modes: greediness, frequency bias, and
the knowing-doing gap. We propose mitigation of these shortcomings by
fine-tuning via Reinforcement Learning (RL) on self-generated CoT rationales.
Our experiments across multi-armed bandits, contextual bandits, and
Tic-tac-toe, demonstrate that RL fine-tuning enhances the decision-making
abilities of LLMs by increasing exploration and narrowing the knowing-doing
gap. Finally, we study both classic exploration mechanisms, such as
$\epsilon$-greedy, and LLM-specific approaches, such as self-correction and
self-consistency, to enable more effective fine-tuning of LLMs for
decision-making.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:57:14 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Schmied",
"Thomas",
""
],
[
"Bornschein",
"Jörg",
""
],
[
"Grau-Moya",
"Jordi",
""
],
[
"Wulfmeier",
"Markus",
""
],
[
"Pascanu",
"Razvan",
""
]
] |
2504.16079 | Xiangyu Zhang | Sara Algeri, Xiangyu Zhang, Erik Floden, Hongru Zhao, Galin L. Jones,
Vuk Mandic, Jesse Miller | A Distribution-Free Approach to Testing Models for Angular Power Spectra | null | null | null | null | physics.data-an astro-ph.IM | http://creativecommons.org/licenses/by/4.0/ | A novel goodness-of-fit strategy is introduced for testing models for angular
power spectra characterized by unknown parameters. Using this strategy, it is
possible to assess the validity of such models without specifying the
distribution of the estimators of the angular power spectrum being used. This
holds under general conditions, ensuring the applicability of the method across
diverse scenarios. Moreover, the proposed solution overcomes the need for
case-by-case simulations when testing different models -- leading to notable
computational advantages.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:57:37 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Algeri",
"Sara",
""
],
[
"Zhang",
"Xiangyu",
""
],
[
"Floden",
"Erik",
""
],
[
"Zhao",
"Hongru",
""
],
[
"Jones",
"Galin L.",
""
],
[
"Mandic",
"Vuk",
""
],
[
"Miller",
"Jesse",
""
]
] |
2504.16080 | Le Zhuo | Le Zhuo, Liangbing Zhao, Sayak Paul, Yue Liao, Renrui Zhang, Yi Xin,
Peng Gao, Mohamed Elhoseiny, Hongsheng Li | From Reflection to Perfection: Scaling Inference-Time Optimization for
Text-to-Image Diffusion Models via Reflection Tuning | All code, checkpoints, and datasets are available at
\url{https://diffusion-cot.github.io/reflection2perfection} | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent text-to-image diffusion models achieve impressive visual quality
through extensive scaling of training data and model parameters, yet they often
struggle with complex scenes and fine-grained details. Inspired by the
self-reflection capabilities emergent in large language models, we propose
ReflectionFlow, an inference-time framework enabling diffusion models to
iteratively reflect upon and refine their outputs. ReflectionFlow introduces
three complementary inference-time scaling axes: (1) noise-level scaling to
optimize latent initialization; (2) prompt-level scaling for precise semantic
guidance; and most notably, (3) reflection-level scaling, which explicitly
provides actionable reflections to iteratively assess and correct previous
generations. To facilitate reflection-level scaling, we construct GenRef, a
large-scale dataset comprising 1 million triplets, each containing a
reflection, a flawed image, and an enhanced image. Leveraging this dataset, we
efficiently perform reflection tuning on state-of-the-art diffusion
transformer, FLUX.1-dev, by jointly modeling multimodal inputs within a unified
framework. Experimental results show that ReflectionFlow significantly
outperforms naive noise-level scaling methods, offering a scalable and
compute-efficient solution toward higher-quality image synthesis on challenging
tasks.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:58:07 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zhuo",
"Le",
""
],
[
"Zhao",
"Liangbing",
""
],
[
"Paul",
"Sayak",
""
],
[
"Liao",
"Yue",
""
],
[
"Zhang",
"Renrui",
""
],
[
"Xin",
"Yi",
""
],
[
"Gao",
"Peng",
""
],
[
"Elhoseiny",
"Mohamed",
""
],
[
"Li",
"Hongsheng",
""
]
] |
2504.16081 | Yimu Wang | Yimu Wang, Xuye Liu, Wei Pang, Li Ma, Shuai Yuan, Paul Debevec, Ning
Yu | Survey of Video Diffusion Models: Foundations, Implementations, and
Applications | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in diffusion models have revolutionized video generation,
offering superior temporal consistency and visual quality compared to
traditional generative adversarial networks-based approaches. While this
emerging field shows tremendous promise in applications, it faces significant
challenges in motion consistency, computational efficiency, and ethical
considerations. This survey provides a comprehensive review of diffusion-based
video generation, examining its evolution, technical foundations, and practical
applications. We present a systematic taxonomy of current methodologies,
analyze architectural innovations and optimization strategies, and investigate
applications across low-level vision tasks such as denoising and
super-resolution. Additionally, we explore the synergies between diffusionbased
video generation and related domains, including video representation learning,
question answering, and retrieval. Compared to the existing surveys (Lei et
al., 2024a;b; Melnik et al., 2024; Cao et al., 2023; Xing et al., 2024c) which
focus on specific aspects of video generation, such as human video synthesis
(Lei et al., 2024a) or long-form content generation (Lei et al., 2024b), our
work provides a broader, more updated, and more fine-grained perspective on
diffusion-based approaches with a special section for evaluation metrics,
industry solutions, and training engineering techniques in video generation.
This survey serves as a foundational resource for researchers and practitioners
working at the intersection of diffusion models and video generation, providing
insights into both the theoretical frameworks and practical implementations
that drive this rapidly evolving field. A structured list of related works
involved in this survey is also available on
https://github.com/Eyeline-Research/Survey-Video-Diffusion.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:59:17 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Wang",
"Yimu",
""
],
[
"Liu",
"Xuye",
""
],
[
"Pang",
"Wei",
""
],
[
"Ma",
"Li",
""
],
[
"Yuan",
"Shuai",
""
],
[
"Debevec",
"Paul",
""
],
[
"Yu",
"Ning",
""
]
] |
2504.16082 | Ziqi Pang | Ziqi Pang, Yu-Xiong Wang | MR. Video: "MapReduce" is the Principle for Long Video Understanding | Preprint | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We propose MR. Video, an agentic long video understanding framework that
demonstrates the simple yet effective MapReduce principle for processing long
videos: (1) Map: independently and densely perceiving short video clips, and
(2) Reduce: jointly aggregating information from all clips. Compared with
sequence-to-sequence vision-language models (VLMs), MR. Video performs detailed
short video perception without being limited by context length. Compared with
existing video agents that typically rely on sequential key segment selection,
the Map operation enables simpler and more scalable sequence parallel
perception of short video segments. Its Reduce step allows for more
comprehensive context aggregation and reasoning, surpassing explicit key
segment retrieval. This MapReduce principle is applicable to both VLMs and
video agents, and we use LLM agents to validate its effectiveness.
In practice, MR. Video employs two MapReduce stages: (A) Captioning:
generating captions for short video clips (map), then standardizing repeated
characters and objects into shared names (reduce); (B) Analysis: for each user
question, analyzing relevant information from individual short videos (map),
and integrating them into a final answer (reduce). MR. Video achieves over 10%
accuracy improvement on the challenging LVBench compared to state-of-the-art
VLMs and video agents.
Code is available at: https://github.com/ziqipang/MR-Video
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:59:41 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Pang",
"Ziqi",
""
],
[
"Wang",
"Yu-Xiong",
""
]
] |
2504.16083 | Huiqiang Jiang | Yucheng Li, Huiqiang Jiang, Chengruidong Zhang, Qianhui Wu, Xufang
Luo, Surin Ahn, Amir H. Abdi, Dongsheng Li, Jianfeng Gao, Yuqing Yang, Lili
Qiu | MMInference: Accelerating Pre-filling for Long-Context VLMs via
Modality-Aware Permutation Sparse Attention | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The integration of long-context capabilities with visual understanding
unlocks unprecedented potential for Vision Language Models (VLMs). However, the
quadratic attention complexity during the pre-filling phase remains a
significant obstacle to real-world deployment. To overcome this limitation, we
introduce MMInference (Multimodality Million tokens Inference), a dynamic
sparse attention method that accelerates the prefilling stage for long-context
multi-modal inputs. First, our analysis reveals that the temporal and spatial
locality of video input leads to a unique sparse pattern, the Grid pattern.
Simultaneously, VLMs exhibit markedly different sparse distributions across
different modalities. We introduce a permutation-based method to leverage the
unique Grid pattern and handle modality boundary issues. By offline search the
optimal sparse patterns for each head, MMInference constructs the sparse
distribution dynamically based on the input. We also provide optimized GPU
kernels for efficient sparse computations. Notably, MMInference integrates
seamlessly into existing VLM pipelines without any model modifications or
fine-tuning. Experiments on multi-modal benchmarks-including Video QA,
Captioning, VisionNIAH, and Mixed-Modality NIAH-with state-of-the-art
long-context VLMs (LongVila, LlavaVideo, VideoChat-Flash, Qwen2.5-VL) show that
MMInference accelerates the pre-filling stage by up to 8.3x at 1M tokens while
maintaining accuracy. Our code is available at https://aka.ms/MMInference.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:59:51 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Li",
"Yucheng",
""
],
[
"Jiang",
"Huiqiang",
""
],
[
"Zhang",
"Chengruidong",
""
],
[
"Wu",
"Qianhui",
""
],
[
"Luo",
"Xufang",
""
],
[
"Ahn",
"Surin",
""
],
[
"Abdi",
"Amir H.",
""
],
[
"Li",
"Dongsheng",
""
],
[
"Gao",
"Jianfeng",
""
],
[
"Yang",
"Yuqing",
""
],
[
"Qiu",
"Lili",
""
]
] |
2504.16084 | Kaiyan Zhang | Yuxin Zuo, Kaiyan Zhang, Shang Qu, Li Sheng, Xuekai Zhu, Biqing Qi,
Youbang Sun, Ganqu Cui, Ning Ding, Bowen Zhou | TTRL: Test-Time Reinforcement Learning | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper investigates Reinforcement Learning (RL) on data without explicit
labels for reasoning tasks in Large Language Models (LLMs). The core challenge
of the problem is reward estimation during inference while not having access to
ground-truth information. While this setting appears elusive, we find that
common practices in Test-Time Scaling (TTS), such as majority voting, yield
surprisingly effective rewards suitable for driving RL training. In this work,
we introduce Test-Time Reinforcement Learning (TTRL), a novel method for
training LLMs using RL on unlabeled data. TTRL enables self-evolution of LLMs
by utilizing the priors in the pre-trained models. Our experiments demonstrate
that TTRL consistently improves performance across a variety of tasks and
models. Notably, TTRL boosts the pass@1 performance of Qwen-2.5-Math-7B by
approximately 159% on the AIME 2024 with only unlabeled test data. Furthermore,
although TTRL is only supervised by the Maj@N metric, TTRL has demonstrated
performance to consistently surpass the upper limit of the initial model, and
approach the performance of models trained directly on test data with
ground-truth labels. Our experimental findings validate the general
effectiveness of TTRL across various tasks, and highlight TTRL's potential for
broader tasks and domains. GitHub: https://github.com/PRIME-RL/TTRL
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:59:56 GMT"
}
] | 2025-04-23T00:00:00 | [
[
"Zuo",
"Yuxin",
""
],
[
"Zhang",
"Kaiyan",
""
],
[
"Qu",
"Shang",
""
],
[
"Sheng",
"Li",
""
],
[
"Zhu",
"Xuekai",
""
],
[
"Qi",
"Biqing",
""
],
[
"Sun",
"Youbang",
""
],
[
"Cui",
"Ganqu",
""
],
[
"Ding",
"Ning",
""
],
[
"Zhou",
"Bowen",
""
]
] |
1404.3083 | Valery B Morozov | V.B. Morozov | The gravitational field energy momentum tensor in the Einstein equation.
Construction of the gravitational field equation | 5 pages, 1 figure | Parana J. Sci. Educ., v.8, n.7, (1-5), October 7, 2022 | 10.5281/zenodo.7186630 | null | physics.gen-ph | http://creativecommons.org/licenses/by/4.0/ | Einstein, when he began working on the general theory of relativity, believed
that energy of any kind is the source of the gravitational field. Therefore,
the energy of gravity, like any energy, must be the source of the field. It was
previously discovered that the energy-momentum tensor of the gravitational
field is already contained in the Ricci tensor. This hypothesis is used to
construct a new equation of the gravitational field.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2014 11:58:31 GMT"
},
{
"version": "v2",
"created": "Sat, 31 May 2014 08:28:12 GMT"
},
{
"version": "v3",
"created": "Tue, 3 Jun 2014 10:09:49 GMT"
},
{
"version": "v4",
"created": "Mon, 11 Sep 2017 23:42:34 GMT"
},
{
"version": "v5",
"created": "Mon, 19 Nov 2018 20:19:49 GMT"
},
{
"version": "v6",
"created": "Wed, 12 Oct 2022 20:24:25 GMT"
},
{
"version": "v7",
"created": "Mon, 21 Apr 2025 10:59:40 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Morozov",
"V. B.",
""
]
] |
1507.01534 | Leila Schneps | Leila Schneps | ARI, GARI, Zig and Zag: An introduction to Ecalle's theory of multiple
zeta values | 79 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This text has two goals. The first is to give an introduction to Ecalle's
work on mould theory, multiple zeta values and double shuffle theory and relate
this work explicitly to the classical theory of multiple zeta values and double
shuffle expressed in the usual terms of non-commutative variables. The second
is to provide complete proofs of those of his main results and identities which
are strictly useful in the context of (non-colored) multiple zeta values. Many
of these proofs are difficult, laborious and not enlightening and have been
relegated to appendices. The emphasis in the text is to provide an easily
approachable introduction to Ecalle's language while placing it almost from the
start in the context of multiple zeta value theory.
Disclaimer: This text is not final and is not submitted for publication. The
intention is to continue to add to and complete it over time.
| [
{
"version": "v1",
"created": "Mon, 6 Jul 2015 16:41:46 GMT"
},
{
"version": "v2",
"created": "Mon, 9 Dec 2024 13:28:31 GMT"
},
{
"version": "v3",
"created": "Sat, 5 Apr 2025 10:11:22 GMT"
},
{
"version": "v4",
"created": "Sat, 19 Apr 2025 13:41:10 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Schneps",
"Leila",
""
]
] |
1603.09665 | Shokir Davlatov | Shokir Davlatov | A Proof of the truth of the Riemann hypothesis | in English | null | null | null | math.GM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The present work is devoted to the proof of the truth of the Riemann
hypothesis.
| [
{
"version": "v1",
"created": "Fri, 25 Mar 2016 09:29:05 GMT"
},
{
"version": "v2",
"created": "Fri, 30 Dec 2016 05:27:12 GMT"
},
{
"version": "v3",
"created": "Wed, 30 Dec 2020 06:30:25 GMT"
},
{
"version": "v4",
"created": "Thu, 3 Apr 2025 07:53:04 GMT"
},
{
"version": "v5",
"created": "Sat, 19 Apr 2025 07:10:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Davlatov",
"Shokir",
""
]
] |
1707.02612 | Mat\v{e}j Kone\v{c}n\'y | Andr\'es Aranda, David Bradley-Williams, Jan Hubi\v{c}ka, Miltiadis
Karamanlis, Michael Kompatscher, Mat\v{e}j Kone\v{c}n\'y, Micheal Pawliuk | Ramsey expansions of metrically homogeneous graphs | 59 pages, 14 figures. Minor revision, fixed statements of the results
for infinite-diameter classes with Henson constraints | null | null | null | math.CO cs.DM math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate Ramsey expansions, the coherent extension property for partial
isometries (EPPA), and the existence of a stationary independence relation for
all classes of metrically homogeneous graphs from Cherlin's catalogue. We show
that, with the exception of tree-like graphs, all metric spaces in the
catalogue have precompact Ramsey expansions (or lifts) with the expansion
property. With two exceptions we can also characterise the existence of a
stationary independence relation and coherent EPPA.
Our results are a contribution to Ne\v{s}et\v{r}il's classification programme
of Ramsey classes and can be seen as empirical evidence of the recent
convergence in techniques employed to establish the Ramsey property, the
expansion property, EPPA and the existence of a stationary independence
relation. At the heart of our proof is a canonical way of completing
edge-labelled graphs to metric spaces in Cherlin's classes. The existence of
such a ``completion algorithm'' then allows us to apply several strong results
in the areas that imply EPPA or the Ramsey property.
The main results have numerous consequences for the automorphism groups of
the Fraisse limits of the classes. As corollaries, we prove amenability, unique
ergodicity, existence of universal minimal flows, ample generics, small index
property, 21-Bergman property and Serre's property (FA).
| [
{
"version": "v1",
"created": "Sun, 9 Jul 2017 18:16:18 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Dec 2017 08:11:41 GMT"
},
{
"version": "v3",
"created": "Mon, 27 Mar 2023 13:17:01 GMT"
},
{
"version": "v4",
"created": "Fri, 8 Mar 2024 20:30:05 GMT"
},
{
"version": "v5",
"created": "Sun, 20 Apr 2025 14:01:34 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Aranda",
"Andrés",
""
],
[
"Bradley-Williams",
"David",
""
],
[
"Hubička",
"Jan",
""
],
[
"Karamanlis",
"Miltiadis",
""
],
[
"Kompatscher",
"Michael",
""
],
[
"Konečný",
"Matěj",
""
],
[
"Pawliuk",
"Micheal",
""
]
] |
1808.07663 | Jeongseok Oh | Jeongseok Oh | Multiplicative property of localized Chern characters for 2-periodic
complexes | 15 pages | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove the multiplicative property of localized Chern characters. As a
direct consequence, a localized Chern character gives rise to a ring
homomorphism from the K-group of periodic complexes to the bivariant Chow
cohomology group. As an application, we prove the functoriality of Kiem-Li's
cosection-localized intersection homomorphisms.
| [
{
"version": "v1",
"created": "Thu, 23 Aug 2018 08:30:29 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Sep 2018 05:01:09 GMT"
},
{
"version": "v3",
"created": "Mon, 15 Nov 2021 13:26:18 GMT"
},
{
"version": "v4",
"created": "Mon, 21 Apr 2025 00:11:18 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Oh",
"Jeongseok",
""
]
] |
1912.00518 | Krzysztof Argasinski | Krzysztof Argasinski and Ryszard Rudnicki | Beyond classical Hamilton's Rule. State distribution asymmetry and the
dynamics of altruism | null | null | null | null | q-bio.PE nlin.AO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper analyzes the relationships between demographic and state-based
evolutionary games and Hamilton's rule. It is shown that the classical
Hamilton's rule (counterfactual method), combined with demographic payoffs,
leads to easily testable models. It works well when the roles of donor and
receiver are randomly drawn during each interaction event. This is illustrated
by the alarm call example. However, we can imagine situations in which
role-switching results from external mechanism, such as, fluxes of individuals
between the border and the interior of the habitat, when only border
individuals may spot the threat and warn their neighbors. To cover these cases,
a new model is extended to the case with explicit dynamics of the role
distributions among carriers of different strategies, driven by some general
mechanisms. It is shown that even in the case when fluxes between roles are
driven by neutral mechanisms (acting in the same way on all strategies),
differences in mortality in the focal interaction lead to different
distributions of roles for different strategies. This leads to a more complex
rule for cooperation than the classical Hamilton's rule. In addition to the
cost and benefit components, the new rule contains a third component weighted
by the difference in proportions of the donors among carriers of both
strategies. Depending on the sign, this component can be termed the "survival
surplus", when the donors survival have greater survival than receivers, or the
"sacrifice cost" (when it decreases the benefit), when the receiver's survival
exceeds that of the helping donor. When we allow different role-switching rates
for different strategies, cooperators can win even in the case when the
assortment mechanism is inefficient (i.e., the probability of receiving help
for noncooperators is slightly greater than for cooperators), which is
impossible in classical Hamilton's rule.
| [
{
"version": "v1",
"created": "Sun, 1 Dec 2019 23:18:21 GMT"
},
{
"version": "v2",
"created": "Wed, 26 Jun 2024 16:55:20 GMT"
},
{
"version": "v3",
"created": "Mon, 21 Apr 2025 14:06:30 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Argasinski",
"Krzysztof",
""
],
[
"Rudnicki",
"Ryszard",
""
]
] |
1912.01067 | Yu Guo | Yu Guo, Milos Hasan, Lingqi Yan, Shuang Zhao | A Bayesian Inference Framework for Procedural Material Parameter
Estimation | 12 pages, 13 figures, Pacific Graphics 2020 | null | 10.1111/cgf.14142 | null | cs.GR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Procedural material models have been gaining traction in many applications
thanks to their flexibility, compactness, and easy editability. We explore the
inverse rendering problem of procedural material parameter estimation from
photographs, presenting a unified view of the problem in a Bayesian framework.
In addition to computing point estimates of the parameters by optimization, our
framework uses a Markov Chain Monte Carlo approach to sample the space of
plausible material parameters, providing a collection of plausible matches that
a user can choose from, and efficiently handling both discrete and continuous
model parameters. To demonstrate the effectiveness of our framework, we fit
procedural models of a range of materials---wall plaster, leather, wood,
anisotropic brushed metals and layered metallic paints---to both synthetic and
real target images.
| [
{
"version": "v1",
"created": "Mon, 2 Dec 2019 20:19:07 GMT"
},
{
"version": "v2",
"created": "Thu, 5 Dec 2019 21:58:33 GMT"
},
{
"version": "v3",
"created": "Tue, 16 Jun 2020 00:41:02 GMT"
},
{
"version": "v4",
"created": "Tue, 29 Sep 2020 12:07:55 GMT"
},
{
"version": "v5",
"created": "Wed, 4 Nov 2020 01:23:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Guo",
"Yu",
""
],
[
"Hasan",
"Milos",
""
],
[
"Yan",
"Lingqi",
""
],
[
"Zhao",
"Shuang",
""
]
] |
1912.08600 | Vanderson Lima | Tiarlos Cruz, Vanderson Lima and Alexandre de Sousa | Min-max minimal surfaces, horizons and electrostatic systems | Final version. Accepted for publication in Journal of Differential
Geometry. We extended one of the main results to the case where the boundary
has degenerate minimal surfaces (Theorem C). Also, we now only use the
Simon-Smith version of min-max theory. Finnaly, we improved the presentation
of the paper, following the suggestions of the referee | J. Differential Geom. 128 (2) 583 - 637, October 2024 | 10.4310/jdg/1727712890 | null | math.DG gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a connection between minimal surfaces of index one and General
Relativity. First, we show that for a certain class of (electro)static systems,
each of its unstable horizons is the solution of a one-parameter min-max
problem for the area functional, in particular it has index one. We also obtain
an inequality relating the area and the charge of a minimal surface of index
one in a Cauchy data satisfying the Dominant Energy Condition for
non-electromagnetic matter fields. Moreover, we explore a global version of
this inequality, and the rigidity in the case of the equality, using a result
proved by Marques and Neves.
| [
{
"version": "v1",
"created": "Wed, 18 Dec 2019 13:43:18 GMT"
},
{
"version": "v2",
"created": "Tue, 7 Apr 2020 21:09:29 GMT"
},
{
"version": "v3",
"created": "Fri, 17 Jun 2022 21:07:44 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Cruz",
"Tiarlos",
""
],
[
"Lima",
"Vanderson",
""
],
[
"de Sousa",
"Alexandre",
""
]
] |
1912.10160 | Gaurav Kumar | Gaurav Kumar, Rishabh Joshi, Jaspreet Singh, Promod Yenigalla | AMUSED: A Multi-Stream Vector Representation Method for Use in Natural
Dialogue | null | Proceedings of the 12th Conference on Language Resources and
Evaluation, pp. 750-758, 2020 | null | null | cs.CL cs.AI cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The problem of building a coherent and non-monotonous conversational agent
with proper discourse and coverage is still an area of open research. Current
architectures only take care of semantic and contextual information for a given
query and fail to completely account for syntactic and external knowledge which
are crucial for generating responses in a chit-chat system. To overcome this
problem, we propose an end to end multi-stream deep learning architecture which
learns unified embeddings for query-response pairs by leveraging contextual
information from memory networks and syntactic information by incorporating
Graph Convolution Networks (GCN) over their dependency parse. A stream of this
network also utilizes transfer learning by pre-training a bidirectional
transformer to extract semantic representation for each input sentence and
incorporates external knowledge through the the neighborhood of the entities
from a Knowledge Base (KB). We benchmark these embeddings on next sentence
prediction task and significantly improve upon the existing techniques.
Furthermore, we use AMUSED to represent query and responses along with its
context to develop a retrieval based conversational agent which has been
validated by expert linguists to have comprehensive engagement with humans.
| [
{
"version": "v1",
"created": "Wed, 4 Dec 2019 17:35:03 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Kumar",
"Gaurav",
""
],
[
"Joshi",
"Rishabh",
""
],
[
"Singh",
"Jaspreet",
""
],
[
"Yenigalla",
"Promod",
""
]
] |
2001.08265 | Rafael Lucena | Davi Lima and Rafael Lucena | Lipschitz regularity of the invariant measure of Random Dynamical
Systems | This version has been completely rewritten, resulting in a
significant improvement in its readability | null | 10.1007/s11784-025-01196-1 | null | math.DS math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article we derive a regularity result for the disintegration of the
invariant measure associated to a class of Random Dynamical Systems - RDS. The
results of this work are obtained by constructing a suitable anisotropic normed
space defined by the Wasserstein-Kantorovich-like metric and understanding the
dynamics of the associated transfer operator in a neighborhood of its fixed
point. Precisely, we employ functional analytic techniques to demonstrate a
spectral gap for its action on suitable spaces of signed measures. We apply
this analysis to prove an exponential decay of correlation statement for
Lipschitz observables and statistical properties of the RDS.
| [
{
"version": "v1",
"created": "Wed, 22 Jan 2020 20:25:47 GMT"
},
{
"version": "v2",
"created": "Sun, 26 Jan 2020 02:21:53 GMT"
},
{
"version": "v3",
"created": "Wed, 29 Jan 2020 21:45:08 GMT"
},
{
"version": "v4",
"created": "Sun, 29 Mar 2020 05:40:15 GMT"
},
{
"version": "v5",
"created": "Mon, 1 Jun 2020 15:16:36 GMT"
},
{
"version": "v6",
"created": "Wed, 15 May 2024 00:22:19 GMT"
},
{
"version": "v7",
"created": "Mon, 21 Apr 2025 17:59:40 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Lima",
"Davi",
""
],
[
"Lucena",
"Rafael",
""
]
] |
2002.10712 | Takayuki Kihara | Takayuki Kihara | Degrees of incomputability, realizability and constructive reverse
mathematics | null | null | null | null | math.LO cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is a way of assigning a realizability notion to each degree of
incomputability. In our setting, we make use of Weihrauch degrees (degrees of
incomputability/discontinuity of partial multi-valued functions) to obtain
Lifschitz-like relative realizability predicates. In this note, we present
sample examples on how to lift some separation results on Weihrauch degrees to
those over intuitionistic Zermelo-Fraenkel set theory ${\bf IZF}$.
| [
{
"version": "v1",
"created": "Tue, 25 Feb 2020 07:55:08 GMT"
},
{
"version": "v2",
"created": "Sat, 19 Apr 2025 06:40:37 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Kihara",
"Takayuki",
""
]
] |
2003.04341 | Denis Ricardo Candido Sr. | Denis R. Candido, Gregory D. Fuchs, Ezekiel Johnston-Halperin, Michael
E. Flatt\'e | Predicted strong coupling of solid-state spins via a single magnon mode | Main text (15 pages + 4 figures) and Supplemental Material | 2021 Mater. Quantum. Technol. 1 011001 | 10.1088/2633-4356/ab9a55 | null | cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an approach to realize a hybrid quantum system composed of a
diamond nitrogen-vacancy (NV) center spin coupled to a magnon mode of the
low-damping, low-moment organic ferrimagnet vanadium tetracyanoethylene. We
derive an analytical expression for the spin-magnon cooperativity as a function
of NV position under a micron-scale perpendicularly magnetized disk, and show
that, surprisingly, the cooperativity will be higher using this magnetic
material than in more conventional materials with larger magnetic moments, due
to in part to the reduced demagnetization field. For reasonable experimental
parameters, we predict that the spin-magnon-mode coupling strength is $g\sim
10$ kHz. For isotopically pure $^{12}$C diamond we predict strong coupling of
an NV spin to the unoccupied magnon mode, with cooperativity $\mathcal C=6$ for
a wide range of NV spin locations within the diamond, well within the spatial
precision of NV center implantation. Thus our proposal describes a practical
pathway for single-spin-state-to-single-magnon-occupancy transduction and for
entangling NV centers over micron length scales.
| [
{
"version": "v1",
"created": "Mon, 9 Mar 2020 18:16:31 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Sep 2020 18:35:01 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Candido",
"Denis R.",
""
],
[
"Fuchs",
"Gregory D.",
""
],
[
"Johnston-Halperin",
"Ezekiel",
""
],
[
"Flatté",
"Michael E.",
""
]
] |
2004.12908 | Jayanta Dey | Jayanta Dey, Joshua T. Vogelstein, Hayden S. Helm, Will LeVine, Ronak
D. Mehta, Tyler M. Tomita, Haoyin Xu, Ali Geisa, Qingyang Wang, Gido M. van
de Ven, Chenyu Gao, Bryan Tower, Jonathan Larson, Christopher M. White, and
Carey E. Priebe | Simple Lifelong Learning Machines | null | null | null | null | cs.AI cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In lifelong learning, data are used to improve performance not only on the
present task, but also on past and future (unencountered) tasks. While typical
transfer learning algorithms can improve performance on future tasks, their
performance on prior tasks degrades upon learning new tasks (called
forgetting). Many recent approaches for continual or lifelong learning have
attempted to maintain performance on old tasks given new tasks. But striving to
avoid forgetting sets the goal unnecessarily low. The goal of lifelong learning
should be to use data to improve performance on both future tasks (forward
transfer) and past tasks (backward transfer). In this paper, we show that a
simple approach -- representation ensembling -- demonstrates both forward and
backward transfer in a variety of simulated and benchmark data scenarios,
including tabular, vision (CIFAR-100, 5-dataset, Split Mini-Imagenet, and
Food1k), and speech (spoken digit), in contrast to various reference
algorithms, which typically failed to transfer either forward or backward, or
both. Moreover, our proposed approach can flexibly operate with or without a
computational budget.
| [
{
"version": "v1",
"created": "Mon, 27 Apr 2020 16:16:30 GMT"
},
{
"version": "v10",
"created": "Tue, 21 Sep 2021 00:50:21 GMT"
},
{
"version": "v11",
"created": "Thu, 9 Dec 2021 18:17:20 GMT"
},
{
"version": "v12",
"created": "Tue, 11 Jan 2022 04:25:38 GMT"
},
{
"version": "v13",
"created": "Sat, 14 May 2022 14:40:41 GMT"
},
{
"version": "v14",
"created": "Wed, 5 Oct 2022 01:34:23 GMT"
},
{
"version": "v15",
"created": "Sat, 3 Dec 2022 15:33:44 GMT"
},
{
"version": "v16",
"created": "Thu, 23 Mar 2023 00:08:05 GMT"
},
{
"version": "v17",
"created": "Sun, 25 Jun 2023 13:46:43 GMT"
},
{
"version": "v18",
"created": "Fri, 2 Feb 2024 17:10:42 GMT"
},
{
"version": "v19",
"created": "Tue, 11 Jun 2024 17:04:38 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Apr 2020 17:42:48 GMT"
},
{
"version": "v20",
"created": "Fri, 11 Apr 2025 04:40:23 GMT"
},
{
"version": "v21",
"created": "Sun, 20 Apr 2025 16:25:35 GMT"
},
{
"version": "v3",
"created": "Mon, 29 Jun 2020 19:10:05 GMT"
},
{
"version": "v4",
"created": "Thu, 9 Jul 2020 19:22:48 GMT"
},
{
"version": "v5",
"created": "Thu, 20 Aug 2020 14:34:24 GMT"
},
{
"version": "v6",
"created": "Wed, 3 Mar 2021 15:46:10 GMT"
},
{
"version": "v7",
"created": "Mon, 14 Jun 2021 15:35:21 GMT"
},
{
"version": "v8",
"created": "Fri, 20 Aug 2021 22:29:54 GMT"
},
{
"version": "v9",
"created": "Sat, 18 Sep 2021 15:04:04 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Dey",
"Jayanta",
""
],
[
"Vogelstein",
"Joshua T.",
""
],
[
"Helm",
"Hayden S.",
""
],
[
"LeVine",
"Will",
""
],
[
"Mehta",
"Ronak D.",
""
],
[
"Tomita",
"Tyler M.",
""
],
[
"Xu",
"Haoyin",
""
],
[
"Geisa",
"Ali",
""
],
[
"Wang",
"Qingyang",
""
],
[
"van de Ven",
"Gido M.",
""
],
[
"Gao",
"Chenyu",
""
],
[
"Tower",
"Bryan",
""
],
[
"Larson",
"Jonathan",
""
],
[
"White",
"Christopher M.",
""
],
[
"Priebe",
"Carey E.",
""
]
] |
2006.12942 | Jean-Yves Charbonnel | Jean-Yves Charbonnel (IMJ-PRG) | Projective dimension and commuting variety of a reductive Lie algebra | null | null | null | null | math.AG math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The commuting variety of a reductive Lie algebra $\mathfrak{g}$ is the
underlying variety of a well defined subscheme of
$\mathfrak{g}\times\mathfrak{g}$. In this note, it is proved that this scheme
is normal and Cohen-Macaulay. In particular, its ideal of definition is a prime
ideal. As a matter of fact, this theorem results from a so called Property (P)
for a simple Lie algebra. This property says that some cohomology complexes are
exact.
| [
{
"version": "v1",
"created": "Tue, 23 Jun 2020 12:37:48 GMT"
},
{
"version": "v2",
"created": "Fri, 18 Apr 2025 18:21:18 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Charbonnel",
"Jean-Yves",
"",
"IMJ-PRG"
]
] |
2008.13656 | Benjamin Hoffman | Benjamin Hoffman and Jeremy Lane | Stratified Gradient Hamiltonian Vector Fields and Collective Integrable
Systems | Final accepted version | null | null | null | math.SG math.AG math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct completely integrable systems on the dual of the Lie algebra of
any compact Lie group $K$ with respect to the standard Lie-Poisson structure.
These systems generalize key properties of Gelfand-Zeitlin systems: A) the
pullback to any Hamiltonian $K$-manifold defines a Hamiltonian torus action on
an open dense subset, B) if the $K$-manifold is multiplicity-free, then the
resulting torus action is \textit{completely} integrable, and C) the collective
moment map has convexity and fiber connectedness properties. These systems
generalize the relationship between Gelfand-Zeitlin systems and Gelfand-Zeitlin
canonical bases via geometric quantization by a real polarization.
To construct these systems, we generalize Harada and Kaveh's construction of
integrable systems by toric degeneration on smooth projective varieties to
singular quasi-projective varieties. Under certain conditions, we show that the
stratified-gradient Hamiltonian vector field of such a degeneration, which is
defined piece-wise, has a flow whose limit exists and defines continuous
degeneration map.
| [
{
"version": "v1",
"created": "Mon, 31 Aug 2020 14:49:00 GMT"
},
{
"version": "v2",
"created": "Thu, 1 Oct 2020 14:21:50 GMT"
},
{
"version": "v3",
"created": "Fri, 11 Jun 2021 21:53:44 GMT"
},
{
"version": "v4",
"created": "Fri, 27 May 2022 13:06:52 GMT"
},
{
"version": "v5",
"created": "Mon, 26 Jun 2023 23:09:49 GMT"
},
{
"version": "v6",
"created": "Tue, 12 Sep 2023 02:19:09 GMT"
},
{
"version": "v7",
"created": "Mon, 21 Apr 2025 17:22:37 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Hoffman",
"Benjamin",
""
],
[
"Lane",
"Jeremy",
""
]
] |
2010.00297 | Daniil Ryabko | Daniil Ryabko | Universal time-series forecasting with mixture predictors | This is the author's version of the book published by Springer under
the same name. The final authenticated version is available online at:
https://doi.org/10.1007/978-3-030-54304-4 . Further updates and corrections
may be made here. The latest version fixes a numbering issue in Theorem 3.3 | null | 10.1007/978-3-030-54304-4 | null | cs.LG cs.AI cs.IT math.IT math.ST stat.ML stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This book is devoted to the problem of sequential probability forecasting,
that is, predicting the probabilities of the next outcome of a growing sequence
of observations given the past. This problem is considered in a very general
setting that unifies commonly used probabilistic and non-probabilistic
settings, trying to make as few as possible assumptions on the mechanism
generating the observations. A common form that arises in various formulations
of this problem is that of mixture predictors, which are formed as a
combination of a finite or infinite set of other predictors attempting to
combine their predictive powers. The main subject of this book are such mixture
predictors, and the main results demonstrate the universality of this method in
a very general probabilistic setting, but also show some of its limitations.
While the problems considered are motivated by practical applications,
involving, for example, financial, biological or behavioural data, this
motivation is left implicit and all the results exposed are theoretical.
The book targets graduate students and researchers interested in the problem
of sequential prediction, and, more generally, in theoretical analysis of
problems in machine learning and non-parametric statistics, as well as
mathematical and philosophical foundations of these fields.
The material in this volume is presented in a way that presumes familiarity
with basic concepts of probability and statistics, up to and including
probability distributions over spaces of infinite sequences. Familiarity with
the literature on learning or stochastic processes is not required.
| [
{
"version": "v1",
"created": "Thu, 1 Oct 2020 10:56:23 GMT"
},
{
"version": "v2",
"created": "Sat, 19 Apr 2025 14:07:27 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ryabko",
"Daniil",
""
]
] |
2010.05366 | Erlend Grong | Erlend Grong | Canonical connections on sub-Riemannian manifolds with constant symbol | 27 pages | null | null | null | math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As a tool to address the equivalence problem in sub-Riemannian geometry, we
introduce a canonical choice of grading and compatible affine connection,
available on any sub-Riemannian manifold with constant symbol. We completely
compute these structures for contact manifolds of constant symbol, including
the cases where the connections of Tanaka-Webster-Tanno are not defined. We
also give an original intrinsic grading on sub-Riemannian (2,3,5)-manifolds,
and use this to present the first flatness theorem in this setting.
| [
{
"version": "v1",
"created": "Sun, 11 Oct 2020 22:31:28 GMT"
},
{
"version": "v2",
"created": "Sat, 6 Mar 2021 10:23:37 GMT"
},
{
"version": "v3",
"created": "Mon, 21 Apr 2025 08:46:20 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Grong",
"Erlend",
""
]
] |
2010.11622 | Yilong Zhang | Yilong Zhang | Hilbert Scheme of a Pair of Skew Lines on Cubic Threefolds | 22 pages, final version, to appear in Journal of Pure and Applied
Algebra | null | 10.1016/j.jpaa.2025.107971 | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A pair of disjoint lines on a smooth cubic threefold determines an
irreducible component of the Hilbert scheme. We prove that this component is
smooth and isomorphic to the blow-up of the symmetric product of Fano varieties
of lines on the diagonal. We also study its relation to the geometry of lines
and singularities on the hyperplane sections and its relation to Bridgeland
moduli spaces.
| [
{
"version": "v1",
"created": "Thu, 22 Oct 2020 11:39:22 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Jun 2021 07:56:38 GMT"
},
{
"version": "v3",
"created": "Mon, 17 Apr 2023 07:24:47 GMT"
},
{
"version": "v4",
"created": "Sun, 20 Aug 2023 14:33:03 GMT"
},
{
"version": "v5",
"created": "Fri, 18 Apr 2025 18:00:23 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Zhang",
"Yilong",
""
]
] |
2011.00583 | Dr. Yaodong Yang | Yaodong Yang, Chengdong Ma, Zihan Ding, Stephen McAleer, Chi Jin, Jun
Wang | Game-Theoretic Multiagent Reinforcement Learning | null | null | null | null | cs.MA cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Following the remarkable success of the AlphaGo series, significant advances
in multi-agent reinforcement learning (MARL) techniques have been witnessed.
MARL corresponds to the learning problem in a multi-agent system in which
multiple agents learn simultaneously. It is an interdisciplinary domain with a
long history that includes game theory, machine learning, stochastic control,
psychology, and optimisation. Although MARL has achieved considerable empirical
success in solving real-world games, there is a lack of a self-contained
overview in the literature that elaborates the game theoretical foundations of
modern MARL methods and summarises the recent advances. In fact, the majority
of existing surveys are outdated and do not fully cover the recent developments
since 2010. In this work, we provide a monograph on MARL that covers both the
fundamentals and the latest developments in the research frontier. The goal of
our monograph is to provide a self-contained assessment of the current
state-of-the-art MARL techniques from a game theoretical perspective. We expect
this work to serve as a stepping stone for both new researchers who are about
to enter this fast-growing domain and existing domain experts who want to
obtain a panoramic view and identify new directions based on recent advances.
| [
{
"version": "v1",
"created": "Sun, 1 Nov 2020 18:24:40 GMT"
},
{
"version": "v2",
"created": "Thu, 3 Dec 2020 17:50:27 GMT"
},
{
"version": "v3",
"created": "Thu, 18 Mar 2021 01:43:32 GMT"
},
{
"version": "v4",
"created": "Sun, 20 Apr 2025 06:50:24 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Yang",
"Yaodong",
""
],
[
"Ma",
"Chengdong",
""
],
[
"Ding",
"Zihan",
""
],
[
"McAleer",
"Stephen",
""
],
[
"Jin",
"Chi",
""
],
[
"Wang",
"Jun",
""
]
] |
2011.06416 | Sami Stouli | Richard Spady, Sami Stouli | Gaussian Transforms Modeling and the Estimation of Distributional
Regression Functions | 51 pages, 4 figures. This is the scientifically accepted version | null | null | null | econ.EM stat.ME | http://creativecommons.org/licenses/by/4.0/ | We propose flexible Gaussian representations for conditional cumulative
distribution functions and give a concave likelihood criterion for their
estimation. Optimal representations satisfy the monotonicity property of
conditional cumulative distribution functions, including in finite samples and
under general misspecification. We use these representations to provide a
unified framework for the flexible Maximum Likelihood estimation of conditional
density, cumulative distribution, and quantile functions at parametric rate.
Our formulation yields substantial simplifications and finite sample
improvements over related methods. An empirical application to the gender wage
gap in the United States illustrates our framework.
| [
{
"version": "v1",
"created": "Thu, 12 Nov 2020 14:34:57 GMT"
},
{
"version": "v2",
"created": "Sun, 20 Apr 2025 10:54:59 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Spady",
"Richard",
""
],
[
"Stouli",
"Sami",
""
]
] |
2012.14793 | Gabriele Rembado | Giovanni Felder and Gabriele Rembado | Singular modules for affine Lie algebras, and applications to irregular
WZNW conformal blocks | v3: small improvements and added references, comments welcome! | Selecta Mathematica (2023) 29:15 | 10.1007/s00029-022-00821-y | null | math.QA math-ph math.MP math.RA math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give a mathematical definition of spaces of irregular vacua/covacua in
genus zero, for any simple Lie algebra, working at generic noncritical level.
This uses coinvariants of affine-Lie-algebra modules whose parameters match up
with those of moduli spaces of irregular-singular meromorphic connections: the
de Rham spaces. The Segal--Sugawara representation of the Virasoro algebra is
used to show that the spaces of irregular conformal blocks assemble into a flat
vector bundle over the space of somonodromy times \`a la Klar\`es, and we
provide a universal version of the resulting flat connection generalising the
irregular KZ connection of Reshetikhin and the dynamical KZ connection of
Felder--Markov--Tarasov--Varchenko.
| [
{
"version": "v1",
"created": "Tue, 29 Dec 2020 15:07:55 GMT"
},
{
"version": "v2",
"created": "Thu, 5 Jan 2023 15:06:53 GMT"
},
{
"version": "v3",
"created": "Sat, 19 Apr 2025 15:32:20 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Felder",
"Giovanni",
""
],
[
"Rembado",
"Gabriele",
""
]
] |
2101.08908 | Yutao Chen | Yutao Chen, Anthony Ephremides | Minimizing Age of Incorrect Information for Unreliable Channel with
Power Constraint | null | Proc. IEEE Global Communications Conference (GLOBECOM), pp. 1-6,
2021 | 10.1109/GLOBECOM46510.2021.9686027 | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Age of Incorrect Information (AoII) is a newly introduced performance metric
that considers communication goals. Therefore, comparing with traditional
performance metrics and the recently introduced metric - Age of Information
(AoI), AoII achieves better performance in many real-life applications.
However, the fundamental nature of AoII has been elusive so far. In this paper,
we consider the AoII in a system where a transmitter sends updates about a
multi-state Markovian source to a remote receiver through an unreliable
channel. The communication goal is to minimize AoII subject to a power
constraint. We cast the problem into a Constrained Markov Decision Process
(CMDP) and prove that the optimal policy is a mixture of two deterministic
threshold policies. Afterward, by leveraging the notion of Relative Value
Iteration (RVI) and the structural properties of threshold policy, we propose
an efficient algorithm to find the threshold policies as well as the mixing
coefficient. Lastly, numerical results are laid out to highlight the
performance of AoII-optimal policy.
| [
{
"version": "v1",
"created": "Fri, 22 Jan 2021 01:24:38 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Sep 2021 14:31:19 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Chen",
"Yutao",
""
],
[
"Ephremides",
"Anthony",
""
]
] |
2101.10016 | Claudia Pinzari | Sergio Ciamprone, Marco Valerio Giannone, Claudia Pinzari | Weak quasi-Hopf algebras, C*-tensor categories and conformal field
theory, and the Kazhdan-Lusztig-Finkelberg theorem | 294 pages. The main modifications of this version regard the abstract
and the first 4 introductory sections. We have extracted a section that was
recently published as a separate paper and added some more details | null | null | null | math.QA hep-th math-ph math.MP math.OA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper deals with some related problems posed by Doplicher-Roberts,
Seiberg-Moore, Frenkel-Zhu, Huang regarding relations between quantum groups at
roots of unity and categories arising from conformal field theory. We construct
unitary rigid ribbon braided tensor category structures on module categories of
affine vertex operator algebras at positive integer levels for all Lie types,
derived from fusion categories of quantum groups via the construction of
certain semisimple weak Hopf algebras having roots in the work by Mack and
Schomerus, and a Drinfeld twist to the Zhu algebra analogous to Drinfeld twist
for the quantized universal enveloping algebra U_h(g) and Drinfeld quasi-Hopf
algebra structure over U(g). We use our construction to give a direct proof of
an equivalence theorem (going back to Kazhdan-Lusztig and Finkelberg for affine
Lie algebras) between fusion categories of quantum groups and affine vertex
operator algebras at positive integer levels, with respect to Huang-Lepowsky
ribbon braided tensor structure, for the classical Lie types and G_2. We apply
this result to unitarize ribbon braided module categories of vertex operator
algebras. We classify fusion categories with type A fusion rules by their
ribbon structure. In our work, the unitary structure of quantum group fusion
categories obtained by Wenzl and an associated fibre functor to the Hilbert
spaces play a main role to construct naturally associated semisimple weak Hopf
algebras.
| [
{
"version": "v1",
"created": "Mon, 25 Jan 2021 11:32:05 GMT"
},
{
"version": "v10",
"created": "Wed, 29 Jan 2025 17:23:42 GMT"
},
{
"version": "v11",
"created": "Mon, 24 Mar 2025 06:36:28 GMT"
},
{
"version": "v12",
"created": "Sun, 20 Apr 2025 06:52:32 GMT"
},
{
"version": "v2",
"created": "Mon, 22 Feb 2021 10:40:20 GMT"
},
{
"version": "v3",
"created": "Tue, 27 Apr 2021 16:34:51 GMT"
},
{
"version": "v4",
"created": "Thu, 29 Apr 2021 16:19:27 GMT"
},
{
"version": "v5",
"created": "Mon, 12 Jul 2021 14:57:59 GMT"
},
{
"version": "v6",
"created": "Tue, 27 Jul 2021 17:21:48 GMT"
},
{
"version": "v7",
"created": "Wed, 31 Aug 2022 16:42:22 GMT"
},
{
"version": "v8",
"created": "Tue, 17 Jan 2023 19:27:37 GMT"
},
{
"version": "v9",
"created": "Mon, 29 Jul 2024 07:05:09 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ciamprone",
"Sergio",
""
],
[
"Giannone",
"Marco Valerio",
""
],
[
"Pinzari",
"Claudia",
""
]
] |
2102.02515 | Yuanyuan Chen | Yuanyuan Chen, Boyang Li, Han Yu, Pengcheng Wu, Chunyan Miao | HYDRA: Hypergradient Data Relevance Analysis for Interpreting Deep
Neural Networks | null | Proceedings of the AAAI Conference on Artificial Intelligence 2021 | 10.1609/aaai.v35i8.16871 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The behaviors of deep neural networks (DNNs) are notoriously resistant to
human interpretations. In this paper, we propose Hypergradient Data Relevance
Analysis, or HYDRA, which interprets the predictions made by DNNs as effects of
their training data. Existing approaches generally estimate data contributions
around the final model parameters and ignore how the training data shape the
optimization trajectory. By unrolling the hypergradient of test loss w.r.t. the
weights of training data, HYDRA assesses the contribution of training data
toward test data points throughout the training trajectory. In order to
accelerate computation, we remove the Hessian from the calculation and prove
that, under moderate conditions, the approximation error is bounded.
Corroborating this theoretical claim, empirical results indicate the error is
indeed small. In addition, we quantitatively demonstrate that HYDRA outperforms
influence functions in accurately estimating data contribution and detecting
noisy data labels. The source code is available at
https://github.com/cyyever/aaai_hydra_8686.
| [
{
"version": "v1",
"created": "Thu, 4 Feb 2021 10:00:13 GMT"
},
{
"version": "v2",
"created": "Mon, 1 Mar 2021 09:48:05 GMT"
},
{
"version": "v3",
"created": "Wed, 24 Mar 2021 12:46:06 GMT"
},
{
"version": "v4",
"created": "Sat, 5 Feb 2022 13:50:34 GMT"
},
{
"version": "v5",
"created": "Fri, 30 Dec 2022 04:14:24 GMT"
},
{
"version": "v6",
"created": "Sun, 20 Apr 2025 13:29:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Chen",
"Yuanyuan",
""
],
[
"Li",
"Boyang",
""
],
[
"Yu",
"Han",
""
],
[
"Wu",
"Pengcheng",
""
],
[
"Miao",
"Chunyan",
""
]
] |
2103.12374 | Shoya Ishimaru | Shoya Ishimaru | What Do We Get from Two-Way Fixed Effects Regressions? Implications from
Numerical Equivalence | null | null | null | null | econ.EM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In any multiperiod panel, a two-way fixed effects (TWFE) regression is
numerically equivalent to a first-difference (FD) regression that pools all
possible between-period gaps. Building on this observation, this paper develops
numerical and causal interpretations of the TWFE coefficient. At the sample
level, the TWFE coefficient is a weighted average of FD coefficients with
different between-period gaps. This decomposition improves transparency by
revealing the sources of variation that the TWFE coefficient captures. At the
population level, causal interpretation of the TWFE coefficient relies on a
common trends assumption for any between-period gap, conditional on changes,
not levels, of time-varying covariates. I propose a simple modification to the
TWFE approach that naturally eases this assumption.
| [
{
"version": "v1",
"created": "Tue, 23 Mar 2021 08:16:58 GMT"
},
{
"version": "v2",
"created": "Sun, 9 May 2021 02:08:01 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Jan 2022 03:59:48 GMT"
},
{
"version": "v4",
"created": "Tue, 18 Oct 2022 03:21:45 GMT"
},
{
"version": "v5",
"created": "Mon, 8 Jan 2024 05:15:55 GMT"
},
{
"version": "v6",
"created": "Fri, 18 Oct 2024 06:34:47 GMT"
},
{
"version": "v7",
"created": "Mon, 21 Apr 2025 07:18:31 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Ishimaru",
"Shoya",
""
]
] |
2103.15093 | Lifeng Gu | Lifeng Gu | Representation Learning by Ranking across multiple tasks | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In recent years, representation learning has become the research focus of the
machine learning community. Large-scale neural networks are a crucial step
toward achieving general intelligence, with their success largely attributed to
their ability to learn abstract representations of data. Several learning
fields are actively discussing how to learn representations, yet there is a
lack of a unified perspective. We convert the representation learning problem
under different tasks into a ranking problem. By adopting the ranking problem
as a unified perspective, representation learning tasks can be solved in a
unified manner by optimizing the ranking loss. Experiments under various
learning tasks, such as classification, retrieval, multi-label learning, and
regression, prove the superiority of the representation learning by ranking
framework. Furthermore, experiments under self-supervised learning tasks
demonstrate the significant advantage of the ranking framework in processing
unsupervised training data, with data augmentation techniques further enhancing
its performance.
| [
{
"version": "v1",
"created": "Sun, 28 Mar 2021 09:36:36 GMT"
},
{
"version": "v2",
"created": "Sat, 19 Apr 2025 12:15:13 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Gu",
"Lifeng",
""
]
] |
2103.15114 | Lifeng Gu | Lifeng Gu | Explaining Representation by Mutual Information | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | As interpretability gains attention in machine learning, there is a growing
need for reliable models that fully explain representation content. We propose
a mutual information (MI)-based method that decomposes neural network
representations into three exhaustive components: total mutual information,
decision-related information, and redundant information. This theoretically
complete framework captures the entire input-representation relationship,
surpassing partial explanations like those from Grad-CAM. Using two lightweight
modules integrated into architectures such as CNNs and Transformers,we estimate
these components and demonstrate their interpretive power through
visualizations on ResNet and prototype network applied to image classification
and few-shot learning tasks. Our approach is distinguished by three key
features: 1. Rooted in mutual information theory, it delivers a thorough and
theoretically grounded interpretation, surpassing the scope of existing
interpretability methods. 2. Unlike conventional methods that focus on
explaining decisions, our approach centers on interpreting representations. 3.
It seamlessly integrates into pre-existing network architectures, requiring
only fine-tuning of the inserted modules.
| [
{
"version": "v1",
"created": "Sun, 28 Mar 2021 12:26:56 GMT"
},
{
"version": "v2",
"created": "Sat, 19 Apr 2025 12:58:16 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Gu",
"Lifeng",
""
]
] |
2104.02461 | Sanjeev Saxena | Waseem Akram and Sanjeev Saxena | Sorted Range Reporting and Range Minima Queries | null | In Applied Algorithm: ICAA 2025, Lecture Notes in Computer
Science, vol 15505 | 10.1007/978-3-031-84543-7_2 | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given an array A[1: n] of n elements drawn from an ordered set, the sorted
range selection problem is to build a data structure that can be used to answer
the following type of queries efficiently: Given a pair of indices i, j $ (1\le
i\le j \le n)$, and a positive integer k, report the k smallest elements from
the sub-array A[i: j] in order. Brodal et al. (Brodal, G.S., Fagerberg, R.,
Greve, M., and L{\'o}pez-Ortiz, A., Online sorted range reporting. Algorithms
and Computation (2009) pp. 173--182) introduced the problem and gave an optimal
solution. After O(n log n) time for preprocessing, the query time is O(k). The
space used is O(n).
In this paper, we propose the only other possible optimal trade-off for the
problem. We present a linear space solution to the problem that takes O(k log
k) time to answer a range selection query. The preprocessing time is O(n).
Moreover, the proposed algorithm reports the output elements one by one in
non-decreasing order. Our solution is simple and practical.
We also describe an extremely simple method for range minima queries (most of
whose parts are known) which takes al most (but not exactly) linear time. We
believe that this method may be, in practice, faster and easier to implement in
most cases.
| [
{
"version": "v1",
"created": "Tue, 6 Apr 2021 12:39:28 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Dec 2022 07:17:24 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Aug 2023 08:27:22 GMT"
},
{
"version": "v4",
"created": "Tue, 19 Sep 2023 09:55:20 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Akram",
"Waseem",
""
],
[
"Saxena",
"Sanjeev",
""
]
] |
2105.04917 | Maksim Zhukovskii | Mikhail Isaev, Igor Rodionov, Rui-Ray Zhang, Maksim Zhukovskii | Extremal independence in discrete random systems | null | null | null | null | math.PR math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $\mathbf{X}(n) \in \mathbb{R}^d$ be a sequence of random vectors, where
$n\in\mathbb{N}$ and $d = d(n)$. Under certain weakly dependence conditions, we
prove that the distribution of the maximal component of $\mathbf{X}$ and the
distribution of the maximum of their independent copies are asymptotically
equivalent. Our result on extremal independence relies on new lower and upper
bounds for the probability that none of a given finite set of events occurs. As
applications, we obtain the distribution of various extremal characteristics of
random discrete structures such as maximum codegree in binomial random
hypergraphs and the maximum number of cliques sharing a given vertex in
binomial random graphs. We also generalise Berman-type conditions for a
sequence of Gaussian random vectors to possess the extremal independence
property.
| [
{
"version": "v1",
"created": "Tue, 11 May 2021 10:06:51 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Oct 2022 07:50:00 GMT"
},
{
"version": "v3",
"created": "Sat, 19 Apr 2025 19:08:15 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Isaev",
"Mikhail",
""
],
[
"Rodionov",
"Igor",
""
],
[
"Zhang",
"Rui-Ray",
""
],
[
"Zhukovskii",
"Maksim",
""
]
] |
2107.02017 | Yueheng Huang | Yueheng Huang, Nong Xiang, Jiale Chen and Zong Xu | Chaotic diffusion in multi-scale turbulence | 8 pages, 3 figures | null | null | null | physics.plasm-ph nlin.CD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates chaotic diffusion in multi-scale turbulence driven by
nonlinear wave-particle resonance coupling. Turbulent waves with distinct
characteristic wavelengths across scales coherently interact with charged
particles when their phase velocities match the particles' velocities. A
multi-wavenumber mapping framework is developed to model chaotic transport
under multi-scale turbulence. By analytically deriving velocity correlation
functions, we quantify the diffusion coefficient under conditions of
cross-scale wave intensity parity. A critical analysis reveals that chaotic
dynamics at smaller scales prove insufficient to completely erase phase-space
correlations established by large-scale turbulent components. The largest-scale
turbulence components dominate deviations from quasi-linear (QL) theory
predictions, establishing a scale-dependent hierarchy in chaotic transport.
Mere reduction of inter-wave phase velocity spacing for small-scale components
cannot recover QL diffusion at finite wave amplitudes in multi-scale
turbulence. Incorporating a larger-scale component into a small-scale-driven
strong chaotic system can induce non-QL diffusion. Specifically, for two-scale
turbulence, the QL approximation systematically underestimates transport.
Increasing the number of smaller-scale components with strong overlap
parameters drives convergence toward the QL approximation. This framework
provides a methodology for analyzing resonance-driven turbulence in laboratory
and astrophysical plasmas.
| [
{
"version": "v1",
"created": "Thu, 1 Jul 2021 02:46:38 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Apr 2025 11:18:11 GMT"
},
{
"version": "v3",
"created": "Mon, 21 Apr 2025 06:26:47 GMT"
}
] | 2025-04-22T00:00:00 | [
[
"Huang",
"Yueheng",
""
],
[
"Xiang",
"Nong",
""
],
[
"Chen",
"Jiale",
""
],
[
"Xu",
"Zong",
""
]
] |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.