id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.12797 | Alex Keshavarzi | Jim Alexander, Vassilis Anastassopoulos, Grigor Atoian, Rick Baartman,
Stefan Bae{\ss}ler, Franco Bedeschi, John Benante, Martin Berz, Michael
Blaskiewicz, Themis Bowcock, Kevin Brown, Dmitry Budker, Sergey Burdin,
Brendan C. Casey, Gianluigi Casse, Giovanni Cantatore, Timothy Chupp, Hooman
Davoudiasl, Dmitri Denisov, Bhawin Dhital, Milind V. Diwan, Renee Fatemi,
George Fanourakis, Wolfram Fischer, Peter Graham, Frederick Gray, Antonios
Gardikiotis, Claudio Gatti, James Gooding, Boxing Gou, Selcuk Haciomeroglu,
Georg H. Hoffstaetter, Haixin Huang, Marco Incagli, Hoyong Jeong, David
Kaplan, Marin Karuza, David Kawall, Alexander Keshavarzi, On Kim, Younggeun
Kim, Ivan Koop, Valeri Lebedev, Jonathan Lee, Soohyung Lee, Alberto Lusiani,
William J. Marciano, Marios Maroudas, Andrei Matlashov, Francois Meot, James
P. Miller, William M. Morse, James Mott, Zhanibek Omarov, Cenap Ozben,
Giovanni Maria Piacentino, Matthew Poelker, Dinko Pocanic, Boris Podobedov,
Joe Price, Xin Qian, Surjeet Rajendran, Deepak Raparia, Sergio Rescia, B. Lee
Roberts, Yannis K. Semertzidis, Alexander Silenko, Amarjit Soni, Edward
Stephenson, Riad Suleiman, Michael Syphers, Pia Thoerngren, Volodya
Tishchenko, Nicholaos Tsoupas, Spyros Tzamarias, Alessandro Variola, Graziano
Venanzoni, Eva Vilella, Joost Vossebeld, Peter Winter, Bogdan Wojtsekhowski,
Eunil Won, Konstantin Zioutas | Status of the Proton EDM Experiment (pEDM) | Document submitted as part of the US National Input to the European
Strategy for Particle Physics Update 2026
(https://indico.cern.ch/event/1439855/contributions/6461648/). arXiv admin
note: text overlap with arXiv:2205.00830 | null | null | null | hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Proton EDM Experiment (pEDM) is the first direct search for the proton
electric dipole moment (EDM) with the aim of being the first experiment to
probe the Standard Model (SM) prediction of any particle EDM. Phase-I of pEDM
will achieve $10^{-29} e\cdot$cm, improving current indirect limits by four
orders of magnitude. This will establish a new standard of precision in nucleon
EDM searches and offer a unique sensitivity to better understand the Strong CP
problem. The experiment is ideally positioned to explore physics beyond the
Standard Model (BSM), with sensitivity to axionic dark matter via the signal of
an oscillating proton EDM and across a wide mass range of BSM models from
$\mathcal{O}(1\text{GeV})$ to $\mathcal{O}(10^3\text{TeV})$. Utilizing the
frozen-spin technique in a highly symmetric storage ring that leverages
existing infrastructure at Brookhaven National Laboratory (BNL), pEDM builds
upon the technological foundation and experimental expertise of the highly
successful Muon $g$$-$$2$ Experiments. With significant R\&D and prototyping
already underway, pEDM is preparing a conceptual design report (CDR) to offer a
cost-effective, high-impact path to discovering new sources of CP violation and
advancing our understanding of fundamental physics. It will play a vital role
in complementing the physics goals of the next-generation collider while
simultaneously contributing to sustaining particle physics research and
training early-career researchers during gaps between major collider
operations.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 09:59:22 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 11:11:48 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Alexander",
"Jim",
""
],
[
"Anastassopoulos",
"Vassilis",
""
],
[
"Atoian",
"Grigor",
""
],
[
"Baartman",
"Rick",
""
],
[
"Baeßler",
"Stefan",
""
],
[
"Bedeschi",
"Franco",
""
],
[
"Benante",
"John",
""
],
[
"Berz",
"Martin",
""
],
[
"Blaskiewicz",
"Michael",
""
],
[
"Bowcock",
"Themis",
""
],
[
"Brown",
"Kevin",
""
],
[
"Budker",
"Dmitry",
""
],
[
"Burdin",
"Sergey",
""
],
[
"Casey",
"Brendan C.",
""
],
[
"Casse",
"Gianluigi",
""
],
[
"Cantatore",
"Giovanni",
""
],
[
"Chupp",
"Timothy",
""
],
[
"Davoudiasl",
"Hooman",
""
],
[
"Denisov",
"Dmitri",
""
],
[
"Dhital",
"Bhawin",
""
],
[
"Diwan",
"Milind V.",
""
],
[
"Fatemi",
"Renee",
""
],
[
"Fanourakis",
"George",
""
],
[
"Fischer",
"Wolfram",
""
],
[
"Graham",
"Peter",
""
],
[
"Gray",
"Frederick",
""
],
[
"Gardikiotis",
"Antonios",
""
],
[
"Gatti",
"Claudio",
""
],
[
"Gooding",
"James",
""
],
[
"Gou",
"Boxing",
""
],
[
"Haciomeroglu",
"Selcuk",
""
],
[
"Hoffstaetter",
"Georg H.",
""
],
[
"Huang",
"Haixin",
""
],
[
"Incagli",
"Marco",
""
],
[
"Jeong",
"Hoyong",
""
],
[
"Kaplan",
"David",
""
],
[
"Karuza",
"Marin",
""
],
[
"Kawall",
"David",
""
],
[
"Keshavarzi",
"Alexander",
""
],
[
"Kim",
"On",
""
],
[
"Kim",
"Younggeun",
""
],
[
"Koop",
"Ivan",
""
],
[
"Lebedev",
"Valeri",
""
],
[
"Lee",
"Jonathan",
""
],
[
"Lee",
"Soohyung",
""
],
[
"Lusiani",
"Alberto",
""
],
[
"Marciano",
"William J.",
""
],
[
"Maroudas",
"Marios",
""
],
[
"Matlashov",
"Andrei",
""
],
[
"Meot",
"Francois",
""
],
[
"Miller",
"James P.",
""
],
[
"Morse",
"William M.",
""
],
[
"Mott",
"James",
""
],
[
"Omarov",
"Zhanibek",
""
],
[
"Ozben",
"Cenap",
""
],
[
"Piacentino",
"Giovanni Maria",
""
],
[
"Poelker",
"Matthew",
""
],
[
"Pocanic",
"Dinko",
""
],
[
"Podobedov",
"Boris",
""
],
[
"Price",
"Joe",
""
],
[
"Qian",
"Xin",
""
],
[
"Rajendran",
"Surjeet",
""
],
[
"Raparia",
"Deepak",
""
],
[
"Rescia",
"Sergio",
""
],
[
"Roberts",
"B. Lee",
""
],
[
"Semertzidis",
"Yannis K.",
""
],
[
"Silenko",
"Alexander",
""
],
[
"Soni",
"Amarjit",
""
],
[
"Stephenson",
"Edward",
""
],
[
"Suleiman",
"Riad",
""
],
[
"Syphers",
"Michael",
""
],
[
"Thoerngren",
"Pia",
""
],
[
"Tishchenko",
"Volodya",
""
],
[
"Tsoupas",
"Nicholaos",
""
],
[
"Tzamarias",
"Spyros",
""
],
[
"Variola",
"Alessandro",
""
],
[
"Venanzoni",
"Graziano",
""
],
[
"Vilella",
"Eva",
""
],
[
"Vossebeld",
"Joost",
""
],
[
"Winter",
"Peter",
""
],
[
"Wojtsekhowski",
"Bogdan",
""
],
[
"Won",
"Eunil",
""
],
[
"Zioutas",
"Konstantin",
""
]
] |
2504.13024 | Daniel Gonzalez-Alvarado | Daniel Gonzalez-Alvarado, Fabio Schlindwein, Jonas Cassel, Laura
Steingruber, Stefania Petra, Christoph Schn\"orr | Riemannian Patch Assignment Gradient Flows | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper introduces patch assignment flows for metric data labeling on
graphs. Labelings are determined by regularizing initial local labelings
through the dynamic interaction of both labels and label assignments across the
graph, entirely encoded by a dictionary of competing labeled patches and
mediated by patch assignment variables. Maximal consistency of patch
assignments is achieved by geometric numerical integration of a Riemannian
ascent flow, as critical point of a Lagrangian action functional. Experiments
illustrate properties of the approach, including uncertainty quantification of
label assignments.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 15:34:58 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 19:54:20 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Gonzalez-Alvarado",
"Daniel",
""
],
[
"Schlindwein",
"Fabio",
""
],
[
"Cassel",
"Jonas",
""
],
[
"Steingruber",
"Laura",
""
],
[
"Petra",
"Stefania",
""
],
[
"Schnörr",
"Christoph",
""
]
] |
2504.13217 | Jennifer Haase | Jennifer Haase, Finn Klessascheck, Jan Mendling, Sebastian Pokutta | Sustainability via LLM Right-sizing | 17 pages, 2 Figures, 6 Tables | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have become increasingly embedded in
organizational workflows. This has raised concerns over their energy
consumption, financial costs, and data sovereignty. While performance
benchmarks often celebrate cutting-edge models, real-world deployment decisions
require a broader perspective: when is a smaller, locally deployable model
"good enough"? This study offers an empirical answer by evaluating eleven
proprietary and open-weight LLMs across ten everyday occupational tasks,
including summarizing texts, generating schedules, and drafting emails and
proposals. Using a dual-LLM-based evaluation framework, we automated task
execution and standardized evaluation across ten criteria related to output
quality, factual accuracy, and ethical responsibility. Results show that GPT-4o
delivers consistently superior performance but at a significantly higher cost
and environmental footprint. Notably, smaller models like Gemma-3 and Phi-4
achieved strong and reliable results on most tasks, suggesting their viability
in contexts requiring cost-efficiency, local deployment, or privacy. A cluster
analysis revealed three model groups -- premium all-rounders, competent
generalists, and limited but safe performers -- highlighting trade-offs between
quality, control, and sustainability. Significantly, task type influenced model
effectiveness: conceptual tasks challenged most models, while aggregation and
transformation tasks yielded better performances. We argue for a shift from
performance-maximizing benchmarks to task- and context-aware sufficiency
assessments that better reflect organizational priorities. Our approach
contributes a scalable method to evaluate AI models through a sustainability
lens and offers actionable guidance for responsible LLM deployment in practice.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 04:00:40 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 03:43:30 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Haase",
"Jennifer",
""
],
[
"Klessascheck",
"Finn",
""
],
[
"Mendling",
"Jan",
""
],
[
"Pokutta",
"Sebastian",
""
]
] |
2504.13228 | Chiara Th\"oni | Anna C.M. Th\"oni, Yoram Bachrach and Tal Kachman | Modelling Mean-Field Games with Neural Ordinary Differential Equations | null | null | null | null | cs.LG cs.GT | http://creativecommons.org/licenses/by/4.0/ | Mean-field game theory relies on approximating games that would otherwise
have been intractable to model. While the games can be solved analytically via
the associated system of partial derivatives, this approach is not model-free,
can lead to the loss of the existence or uniqueness of solutions and may suffer
from modelling bias. To reduce the dependency between the model and the game,
we combine mean-field game theory with deep learning in the form of neural
ordinary differential equations. The resulting model is data-driven,
lightweight and can learn extensive strategic interactions that are hard to
capture using mean-field theory alone. In addition, the model is based on
automatic differentiation, making it more robust and objective than approaches
based on finite differences. We highlight the efficiency and flexibility of our
approach by solving three mean-field games that vary in their complexity,
observability and the presence of noise. Using these results, we show that the
model is flexible, lightweight and requires few observations to learn the
distribution underlying the data.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 13:38:42 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:56:24 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Thöni",
"Anna C. M.",
""
],
[
"Bachrach",
"Yoram",
""
],
[
"Kachman",
"Tal",
""
]
] |
2504.13230 | Norbert Hegyv\'ari | Norbert Hegyv\'ari | Note on the sumset of squares | This result essentially is proved -- with the same tools -- by
Mei-Chu Chang. See in J. Funct. Anal. 207 (2004), no 2, 444-460. So the
author will withdraw this ArXiv file | null | null | null | math.CO | http://creativecommons.org/publicdomain/zero/1.0/ | It is proved that for any non-empty finite subset $Q$ of the square numbers,
$ |Q+Q|\geq C'|Q|(\log |Q|)^{1/3+o(1)} $.
This result essentially is proved -- with the same tools -- by Mei-Chu Chang.
See in J. Funct. Anal. 207 (2004), no 2, 444-460.
So the author will withdraw this ArXiv file.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 13:45:59 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 19:12:11 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Hegyvári",
"Norbert",
""
]
] |
2504.13497 | Nicolas Menicucci | Tom B. Harris, Takaya Matsuura, Ben Q. Baragiola, Nicolas C. Menicucci | Logical channel for heralded and pure loss with the
Gottesman-Kitaev-Preskill code | (v2) typos corrected, explained odd-numbered heralding result; (v1)
19 pages, 6 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Photon loss is the dominant source of noise in optical quantum systems. The
Gottesman-Kitaev-Preskill (GKP) bosonic code provides significant protection;
however, even low levels of loss can generate uncorrectable errors that another
concatenated code must handle. In this work, we characterize these errors by
deriving analytic expressions for the logical channel that arises from pure
loss acting on approximate GKP qubits. Unlike random displacement noise, we
find that the loss-induced logical channel is not a stochastic Pauli channel.
We also provide analytic expressions for the logical channel for "heralded
loss," when the light scattered out of the signal mode is measured either by
photon number counting -- i.e., photon subtraction -- or heterodyne detection.
These offer a pathway to intentionally inducing non-Pauli channels for, e.g.,
magic-state production.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 06:25:37 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 06:26:26 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Harris",
"Tom B.",
""
],
[
"Matsuura",
"Takaya",
""
],
[
"Baragiola",
"Ben Q.",
""
],
[
"Menicucci",
"Nicolas C.",
""
]
] |
2504.13571 | Tomer Milo | Tomer Milo | On Tightness of the Figiel-Lindenstrauss-Milman inequality | 4 pages | null | null | null | math.MG | http://creativecommons.org/licenses/by/4.0/ | This note aims to explore the tightness of the classical
Figiel-Lindenstrauss-Milman inequality, which states that there exists some
constant $c>0$ such that for any dimension $n$ and any symmetric polytope $P
\subset \mathbb{R}^n$: $\log|V| \cdot \log|\mathcal{F}| \geq cn$, where $V$ and
$\mathcal{F}$ denote the sets of vertices and facets of $P$, respectively. We
show that this inequality is asymptotically tight in almost all possible cases,
except for a special case which was previously studied by Barvinok.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 09:12:35 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 20:01:23 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Milo",
"Tomer",
""
]
] |
2504.13757 | Mark Simkin | Dankrad Feist, Gottfried Herold, Mark Simkin, Benedikt Wagner | Robust Distributed Arrays: Provably Secure Networking for Data
Availability Sampling | null | null | null | null | cs.DS cs.DC | http://creativecommons.org/licenses/by/4.0/ | Data Availability Sampling (DAS), a central component of Ethereum's roadmap,
enables clients to verify data availability without requiring any single client
to download the entire dataset. DAS operates by having clients randomly
retrieve individual symbols of erasure-encoded data from a peer-to-peer
network. While the cryptographic and encoding aspects of DAS have recently
undergone formal analysis, the peer-to-peer networking layer remains
underexplored, with a lack of security definitions and efficient, provably
secure constructions. In this work, we address this gap by introducing a novel
distributed data structure that can serve as the networking layer for DAS,
which we call robust distributed arrays. That is, we rigorously define a
robustness property of a distributed data structure in an open permissionless
network, that mimics a collection of arrays. Then, we give a simple and
efficient construction and formally prove its robustness. Notably, every
individual node is required to store only small portions of the data, and
accessing array positions incurs minimal latency. The robustness of our
construction relies solely on the presence of a minimal absolute number of
honest nodes in the network. In particular, we avoid any honest majority
assumption. Beyond DAS, we anticipate that robust distributed arrays can have
wider applications in distributed systems.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 15:46:38 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 12:23:28 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Feist",
"Dankrad",
""
],
[
"Herold",
"Gottfried",
""
],
[
"Simkin",
"Mark",
""
],
[
"Wagner",
"Benedikt",
""
]
] |
2504.13763 | Ryota Takatsuki | Ryota Takatsuki, Sonia Joseph, Ippei Fujisawa, Ryota Kanai | Decoding Vision Transformers: the Diffusion Steering Lens | 12 pages, 17 figures. Accepted to the CVPR 2025 Workshop on
Mechanistic Interpretability for Vision (MIV) | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Logit Lens is a widely adopted method for mechanistic interpretability of
transformer-based language models, enabling the analysis of how internal
representations evolve across layers by projecting them into the output
vocabulary space. Although applying Logit Lens to Vision Transformers (ViTs) is
technically straightforward, its direct use faces limitations in capturing the
richness of visual representations. Building on the work of Toker et al.
(2024)~\cite{Toker2024-ve}, who introduced Diffusion Lens to visualize
intermediate representations in the text encoders of text-to-image diffusion
models, we demonstrate that while Diffusion Lens can effectively visualize
residual stream representations in image encoders, it fails to capture the
direct contributions of individual submodules. To overcome this limitation, we
propose \textbf{Diffusion Steering Lens} (DSL), a novel, training-free approach
that steers submodule outputs and patches subsequent indirect contributions. We
validate our method through interventional studies, showing that DSL provides
an intuitive and reliable interpretation of the internal processing in ViTs.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 16:00:53 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 10:04:20 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Takatsuki",
"Ryota",
""
],
[
"Joseph",
"Sonia",
""
],
[
"Fujisawa",
"Ippei",
""
],
[
"Kanai",
"Ryota",
""
]
] |
2504.13945 | Zhanglin Wu | Zhanglin Wu, Tengfei Song, Ning Xie, Mengli Zhu, Weidong Zhang, Shuang
Wu, Pengfei Li, Chong Li, Junhao Zhu, Hao Yang, Shiliang Sun | Evaluating Menu OCR and Translation: A Benchmark for Aligning Human and
Automated Evaluations in Large Vision-Language Models | 12 pages, 5 figures, 5 Tables | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The rapid advancement of large vision-language models (LVLMs) has
significantly propelled applications in document understanding, particularly in
optical character recognition (OCR) and multilingual translation. However,
current evaluations of LVLMs, like the widely used OCRBench, mainly focus on
verifying the correctness of their short-text responses and long-text responses
with simple layout, while the evaluation of their ability to understand long
texts with complex layout design is highly significant but largely overlooked.
In this paper, we propose Menu OCR and Translation Benchmark (MOTBench), a
specialized evaluation framework emphasizing the pivotal role of menu
translation in cross-cultural communication. MOTBench requires LVLMs to
accurately recognize and translate each dish, along with its price and unit
items on a menu, providing a comprehensive assessment of their visual
understanding and language processing capabilities. Our benchmark is comprised
of a collection of Chinese and English menus, characterized by intricate
layouts, a variety of fonts, and culturally specific elements across different
languages, along with precise human annotations. Experiments show that our
automatic evaluation results are highly consistent with professional human
evaluation. We evaluate a range of publicly available state-of-the-art LVLMs,
and through analyzing their output to identify the strengths and weaknesses in
their performance, offering valuable insights to guide future advancements in
LVLM development. MOTBench is available at https://github.com/gitwzl/MOTBench.
| [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 03:08:57 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 02:19:53 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Apr 2025 09:16:55 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Wu",
"Zhanglin",
""
],
[
"Song",
"Tengfei",
""
],
[
"Xie",
"Ning",
""
],
[
"Zhu",
"Mengli",
""
],
[
"Zhang",
"Weidong",
""
],
[
"Wu",
"Shuang",
""
],
[
"Li",
"Pengfei",
""
],
[
"Li",
"Chong",
""
],
[
"Zhu",
"Junhao",
""
],
[
"Yang",
"Hao",
""
],
[
"Sun",
"Shiliang",
""
]
] |
2504.13969 | Nayoung Choi | Nayoung Choi, Peace Cyebukayire, Jinho D. Choi | Tinker Tales: Interactive Storytelling Framework for Early Childhood
Narrative Development and AI Literacy | null | null | null | null | cs.HC cs.AI cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper presents Tinker Tales, an interactive storytelling framework in
the format of a board game, designed to support both narrative development and
AI literacy in early childhood. The framework integrates tangible and
speech-based interactions with AI through NFC chip-attached pawns and tokens,
along with a speaker and microphone. Children select and define key story
elements-such as characters, places, items, and emotions-using the pawns and
tokens, providing further details to the AI and receiving proper assistance,
similar to how adults prompt AI for specific tasks (e.g., writing). For
evaluation, several game sessions were simulated with a child AI agent, and the
quality and safety of the generated stories were assessed from various
perspectives. This work highlights the potential of combining physical and
digital elements in AI literacy, offering a safe and engaging way for children
to learn how to effectively collaborate with AI.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 17:47:55 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 23:59:44 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Choi",
"Nayoung",
""
],
[
"Cyebukayire",
"Peace",
""
],
[
"Choi",
"Jinho D.",
""
]
] |
2504.14051 | Raghavv Goel Mr. | Raghavv Goel, Junyoung Park, Mukul Gagrani, Dalton Jones, Matthew
Morse, Harper Langston, Mingu Lee, Chris Lott | CAOTE: KV Caching through Attention Output Error based Token Eviction | 14 pages, 2 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | While long context support of large language models has extended their
abilities, it also incurs challenges in memory and compute which becomes
crucial bottlenecks in resource-restricted devices. Token eviction, a widely
adopted post-training methodology designed to alleviate the bottlenecks by
evicting less important tokens from the cache, typically uses attention scores
as proxy metrics for token importance. However, one major limitation of
attention score as a token-wise importance metrics is that it lacks the
information about contribution of tokens to the attention output. In this
paper, we propose a simple eviction criterion based on the contribution of
cached tokens to attention outputs. Our method, CAOTE, optimizes for eviction
error due to token eviction, by seamlessly integrating attention scores and
value vectors. This is the first method which uses value vector information on
top of attention-based eviction scores. Additionally, CAOTE can act as a
meta-heuristic method with flexible usage with any token eviction method. We
show that CAOTE, when combined with the state-of-the-art attention score-based
methods, always improves accuracies on the downstream task, indicating the
importance of leveraging information from values during token eviction process.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 19:46:54 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 05:04:58 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Goel",
"Raghavv",
""
],
[
"Park",
"Junyoung",
""
],
[
"Gagrani",
"Mukul",
""
],
[
"Jones",
"Dalton",
""
],
[
"Morse",
"Matthew",
""
],
[
"Langston",
"Harper",
""
],
[
"Lee",
"Mingu",
""
],
[
"Lott",
"Chris",
""
]
] |
2504.14077 | YueMing Shen | Yueming Shen | Asymptotic well-calibration of the posterior predictive $p$-value under
the modified Kolmogorov-Smirnov test | null | null | null | null | math.ST stat.ME stat.TH | http://creativecommons.org/licenses/by/4.0/ | The posterior predictive $p$-value is a widely used tool for Bayesian model
checking. However, under most test statistics, its asymptotic null distribution
is more concentrated around 1/2 than uniform. Consequently, its finite-sample
behavior is difficult to interpret and tends to lack power, which is a
well-known issue among practitioners. A common choice of test statistic is the
Kolmogorov-Smirnov test with plug-in estimators. It provides a global measure
of model-data discrepancy for real-valued observations and is sensitive to
model misspecification. In this work, we establish that under this test
statistic, the posterior predictive $p$-value converges in distribution to
uniform under the null. We further use numerical experiments to demonstrate
that this $p$-value is well-behaved in finite samples and can effectively
detect a wide range of alternative models.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 21:01:17 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 19:42:38 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Shen",
"Yueming",
""
]
] |
2504.14215 | Kang Zhou | Bo Feng, Liang Zhang, Kang Zhou | Hidden Zeros and $2$-split via BCFW Recursion Relation | 28 pages, 8 figures | null | null | null | hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | In this paper, we provide another angle to understand recent discoveries,
i.e., the hidden zeros and corresponding 2-split behavior using the BCFW
recursion relation. For the hidden zeros, we show that although the BCFW
recursion relation is not directly applicable for computing amplitudes of the
non-linear sigma model, we can indeed prove the zeros using the modified BCFW
recursion relation. Our work also indicates that for the 2-split to hold, the
current should be carefully defined.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 07:43:32 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 01:53:40 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Feng",
"Bo",
""
],
[
"Zhang",
"Liang",
""
],
[
"Zhou",
"Kang",
""
]
] |
2504.14270 | Michael Benedikt | Sam Adam-Day, Michael Benedikt, Alberto Larrauri | Convergence Laws for Extensions of First-Order Logic with Averaging | null | null | null | null | cs.LO math.CO | http://creativecommons.org/licenses/by/4.0/ | For many standard models of random structure, first-order logic sentences
exhibit a convergence phenomenon on random inputs. The most well-known example
is for random graphs with constant edge probability, where the probabilities of
first-order sentences converge to 0 or 1. In other cases, such as certain
``sparse random graph'' models, the probabilities of sentences converge,
although not necessarily to 0 or 1. In this work we deal with extensions of
first-order logic with aggregate operators, variations of averaging. These
logics will consist of real-valued terms, and we allow arbitrary Lipschitz
functions to be used as ``connectives''. We show that some of the well-known
convergence laws extend to this setting.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 11:46:37 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 12:27:17 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Apr 2025 07:47:43 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Adam-Day",
"Sam",
""
],
[
"Benedikt",
"Michael",
""
],
[
"Larrauri",
"Alberto",
""
]
] |
2504.14333 | Zhanwang Deng | Zhanwang Deng, Jiang Hu, Kangkang Deng, Zaiwen Wen | An efficient primal dual semismooth Newton method for semidefinite
programming | 35 pages | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present an efficient semismooth Newton method, named SSNCP,
for solving a class of semidefinite programming problems. Our approach is
rooted in an equivalent semismooth system derived from the saddle point problem
induced by the augmented Lagrangian duality. An additional correction step is
incorporated after the semismooth Newton step to ensure that the iterates
eventually reside on a manifold where the semismooth system is locally smooth.
Global convergence is achieved by carefully designing inexact criteria and
leveraging the $\alpha$-averaged property to analyze the error. The correction
steps address challenges related to the lack of smoothness in local convergence
analysis. Leveraging the smoothness established by the correction steps and
assuming a local error bound condition, we establish the local superlinear
convergence rate without requiring the stringent assumptions of nonsingularity
or strict complementarity. Furthermore, we prove that SSNCP converges to an
$\varepsilon$-stationary point with an iteration complexity of
$\widetilde{\mathcal{O}}(\varepsilon^{-3/2})$. Numerical experiments on various
datasets, especially the Mittelmann benchmark, demonstrate the high efficiency
and robustness of SSNCP compared to state-of-the-art solvers.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 15:56:07 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 08:17:34 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Deng",
"Zhanwang",
""
],
[
"Hu",
"Jiang",
""
],
[
"Deng",
"Kangkang",
""
],
[
"Wen",
"Zaiwen",
""
]
] |
2504.14373 | Chen Guo | Chen Guo, Zhuo Su, Jian Wang, Shuang Li, Xu Chang, Zhaohu Li, Yang
Zhao, Guidong Wang, Ruqi Huang | SEGA: Drivable 3D Gaussian Head Avatar from a Single Image | null | null | null | null | cs.GR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Creating photorealistic 3D head avatars from limited input has become
increasingly important for applications in virtual reality, telepresence, and
digital entertainment. While recent advances like neural rendering and 3D
Gaussian splatting have enabled high-quality digital human avatar creation and
animation, most methods rely on multiple images or multi-view inputs, limiting
their practicality for real-world use. In this paper, we propose SEGA, a novel
approach for Single-imagE-based 3D drivable Gaussian head Avatar creation that
combines generalized prior models with a new hierarchical UV-space Gaussian
Splatting framework. SEGA seamlessly combines priors derived from large-scale
2D datasets with 3D priors learned from multi-view, multi-expression, and
multi-ID data, achieving robust generalization to unseen identities while
ensuring 3D consistency across novel viewpoints and expressions. We further
present a hierarchical UV-space Gaussian Splatting framework that leverages
FLAME-based structural priors and employs a dual-branch architecture to
disentangle dynamic and static facial components effectively. The dynamic
branch encodes expression-driven fine details, while the static branch focuses
on expression-invariant regions, enabling efficient parameter inference and
precomputation. This design maximizes the utility of limited 3D data and
achieves real-time performance for animation and rendering. Additionally, SEGA
performs person-specific fine-tuning to further enhance the fidelity and
realism of the generated avatars. Experiments show our method outperforms
state-of-the-art approaches in generalization ability, identity preservation,
and expression realism, advancing one-shot avatar creation for practical
applications.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 18:23:31 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 12:27:03 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Guo",
"Chen",
""
],
[
"Su",
"Zhuo",
""
],
[
"Wang",
"Jian",
""
],
[
"Li",
"Shuang",
""
],
[
"Chang",
"Xu",
""
],
[
"Li",
"Zhaohu",
""
],
[
"Zhao",
"Yang",
""
],
[
"Wang",
"Guidong",
""
],
[
"Huang",
"Ruqi",
""
]
] |
2504.14383 | Heyi Zhu | Daniel C. Isaksen, Hana Jia Kong, Guchuan Li, Yangyang Ruan, Heyi Zhu | Exotic periodic phenomena in the cohomology of the moduli stack of
$1$-dimensional formal group laws | 46 pages, 11 figures. Comments welcome! | null | null | null | math.AT | http://creativecommons.org/licenses/by/4.0/ | We describe some periodic structure in the cohomology of the moduli stack of
1-dimensional formal group laws, also known as the $E_2$-page of the classical
Adams--Novikov spectral sequence. This structure is distinct from the familiar
$v_n$-periodicities, and it displays interesting number-theoretic properties.
Our techniques involve the $\mathbb{C}$-motivic Adams spectral sequence, and we
obtain analogous periodic structure in $\mathbb{C}$-motivic stable homotopy.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 19:13:13 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 21:42:09 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Isaksen",
"Daniel C.",
""
],
[
"Kong",
"Hana Jia",
""
],
[
"Li",
"Guchuan",
""
],
[
"Ruan",
"Yangyang",
""
],
[
"Zhu",
"Heyi",
""
]
] |
2504.14680 | Anoop Bhat | Anoop Bhat and Geordan Gutow and Bhaskar Vundurthy and Zhongqiang Ren
and Sivakumar Rathinam and Howie Choset | A Complete and Bounded-Suboptimal Algorithm for a Moving Target
Traveling Salesman Problem with Obstacles in 3D | Accepted to ICRA 2025 | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The moving target traveling salesman problem with obstacles (MT-TSP-O) seeks
an obstacle-free trajectory for an agent that intercepts a given set of moving
targets, each within specified time windows, and returns to the agent's
starting position. Each target moves with a constant velocity within its time
windows, and the agent has a speed limit no smaller than any target's speed. We
present FMC*-TSP, the first complete and bounded-suboptimal algorithm for the
MT-TSP-O, and results for an agent whose configuration space is $\mathbb{R}^3$.
Our algorithm interleaves a high-level search and a low-level search, where the
high-level search solves a generalized traveling salesman problem with time
windows (GTSP-TW) to find a sequence of targets and corresponding time windows
for the agent to visit. Given such a sequence, the low-level search then finds
an associated agent trajectory. To solve the low-level planning problem, we
develop a new algorithm called FMC*, which finds a shortest path on a graph of
convex sets (GCS) via implicit graph search and pruning techniques specialized
for problems with moving targets. We test FMC*-TSP on 280 problem instances
with up to 40 targets and demonstrate its smaller median runtime than a
baseline based on prior work.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 16:51:29 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 00:08:04 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Bhat",
"Anoop",
""
],
[
"Gutow",
"Geordan",
""
],
[
"Vundurthy",
"Bhaskar",
""
],
[
"Ren",
"Zhongqiang",
""
],
[
"Rathinam",
"Sivakumar",
""
],
[
"Choset",
"Howie",
""
]
] |
2504.14842 | Xiao Ma | Xiao Ma | A Short Proof of Coding Theorems for Reed-Muller Codes Under a Mild
Assumption | 12 pages, 1 figure | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, by treating Reed-Muller (RM) codes as a special class of
low-density parity-check (LDPC) codes and assuming that sub-blocks of the
parity-check matrix are randomly interleaved to each other as Gallager's codes,
we present a short proof that RM codes are entropy-achieving as source coding
for Bernoulli sources and capacity-achieving as channel coding for binary
memoryless symmetric (BMS) channels, also known as memoryless binary-input
output-symmetric (BIOS) channels, in terms of bit error rate (BER) under
maximum-likelihood (ML) decoding.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 03:54:27 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 08:41:12 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ma",
"Xiao",
""
]
] |
2504.14897 | Andong Hu | Andong Hu, Luca Pennati, Ivy Peng and Stefano Markidis | Physics-Aware Compression of Plasma Distribution Functions with
GPU-Accelerated Gaussian Mixture Models | 15 pages, 8 figures | null | null | null | cs.CE | http://creativecommons.org/licenses/by/4.0/ | Data compression is a critical technology for large-scale plasma simulations.
Storing complete particle information requires Terabyte-scale data storage, and
analysis requires ad-hoc scalable post-processing tools. We propose a
physics-aware in-situ compression method using Gaussian Mixture Models (GMMs)
to approximate electron and ion velocity distribution functions with a number
of Gaussian components. This GMM-based method allows us to capture plasma
features such as mean velocity and temperature, and it enables us to identify
heating processes and generate beams. We first construct a histogram to reduce
computational overhead and apply GPU-accelerated, in-situ GMM fitting within
iPIC3D, a large-scale implicit Particle-in-Cell simulator, ensuring real-time
compression. The compressed representation is stored using the ADIOS 2 library,
thus optimizing the I/O process. The GPU and histogramming implementation
provides a significant speed-up with respect to GMM on particles (both in time
and required memory at run-time), enabling real-time compression. Compared to
algorithms like SZ, MGARD, and BLOSC2, our GMM-based method has a physics-based
approach, retaining the physical interpretation of plasma phenomena such as
beam formation, acceleration, and heating mechanisms. Our GMM algorithm
achieves a compression ratio of up to $10^4$, requiring a processing time
comparable to, or even lower than, standard compression engines.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 07:07:49 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 02:47:53 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Hu",
"Andong",
""
],
[
"Pennati",
"Luca",
""
],
[
"Peng",
"Ivy",
""
],
[
"Markidis",
"Stefano",
""
]
] |
2504.14898 | Bert de Vries | Bert de Vries, Wouter Nuijten, Thijs van de Laar, Wouter Kouw, Sepideh
Adamiat, Tim Nisslbeck, Mykola Lukashchuk, Hoang Minh Huu Nguyen, Marco
Hidalgo Araya, Raphael Tresor, Thijs Jenneskens, Ivana Nikoloska, Raaja
Ganapathy Subramanian, Bart van Erp, Dmitry Bagaev and Albert Podusenko | Expected Free Energy-based Planning as Variational Inference | 18 pages | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | We address the problem of planning under uncertainty, where an agent must
choose actions that not only achieve desired outcomes but also reduce
uncertainty. Traditional methods often treat exploration and exploitation as
separate objectives, lacking a unified inferential foundation. Active
inference, grounded in the Free Energy Principle, provides such a foundation by
minimizing Expected Free Energy (EFE), a cost function that combines utility
with epistemic drives, such as ambiguity resolution and novelty seeking.
However, the computational burden of EFE minimization had remained a
significant obstacle to its scalability. In this paper, we show that EFE-based
planning arises naturally from minimizing a variational free energy functional
on a generative model augmented with preference and epistemic priors. This
result reinforces theoretical consistency with the Free Energy Principle by
casting planning under uncertainty itself as a form of variational inference.
Our formulation yields policies that jointly support goal achievement and
information gain, while incorporating a complexity term that accounts for
bounded computational resources. This unifying framework connects and extends
existing methods, enabling scalable, resource-aware implementations of active
inference agents.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 07:09:05 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 12:05:19 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"de Vries",
"Bert",
""
],
[
"Nuijten",
"Wouter",
""
],
[
"van de Laar",
"Thijs",
""
],
[
"Kouw",
"Wouter",
""
],
[
"Adamiat",
"Sepideh",
""
],
[
"Nisslbeck",
"Tim",
""
],
[
"Lukashchuk",
"Mykola",
""
],
[
"Nguyen",
"Hoang Minh Huu",
""
],
[
"Araya",
"Marco Hidalgo",
""
],
[
"Tresor",
"Raphael",
""
],
[
"Jenneskens",
"Thijs",
""
],
[
"Nikoloska",
"Ivana",
""
],
[
"Subramanian",
"Raaja Ganapathy",
""
],
[
"van Erp",
"Bart",
""
],
[
"Bagaev",
"Dmitry",
""
],
[
"Podusenko",
"Albert",
""
]
] |
2504.14910 | Stefano Longhi | Stefano Longhi | Erratic non-Hermitian skin localization | 7 pages, 5 figures, accepted for publication in the Physical Review
Letters | null | null | null | cond-mat.dis-nn physics.optics quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A novel localization phenomenon, termed erratic non-Hermitian skin
localization, has been identified in disordered globally-reciprocal
non-Hermitian lattices. Unlike conventional non-Hermitian skin effect and
Anderson localization, it features macroscopic eigenstate localization at
irregular, disorder-dependent positions with sub-exponential decay. Using the
Hatano-Nelson model with disordered imaginary gauge fields as a case study,
this effect is linked to stochastic interfaces governed by the universal order
statistics of random walks. Finite-size scaling analysis confirms the localized
nature of the eigenstates. This discovery challenges conventional wave
localization paradigms, offering new avenues for understanding and controlling
localization phenomena in non-Hermitian physics.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 07:27:59 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 16:28:54 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Longhi",
"Stefano",
""
]
] |
2504.14921 | Songping Wang | Songping Wang, Hanqing Liu, Yueming Lyu, Xiantao Hu, Ziwen He, Wei
Wang, Caifeng Shan, and Liang Wang | Fast Adversarial Training with Weak-to-Strong Spatial-Temporal
Consistency in the Frequency Domain on Videos | After the submission of the paper, we realized that the study still
has room for expansion. In order to make the research findings more profound
and comprehensive, we have decided to withdraw the paper so that we can
conduct further research and expansion | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adversarial Training (AT) has been shown to significantly enhance adversarial
robustness via a min-max optimization approach. However, its effectiveness in
video recognition tasks is hampered by two main challenges. First, fast
adversarial training for video models remains largely unexplored, which
severely impedes its practical applications. Specifically, most video
adversarial training methods are computationally costly, with long training
times and high expenses. Second, existing methods struggle with the trade-off
between clean accuracy and adversarial robustness. To address these challenges,
we introduce Video Fast Adversarial Training with Weak-to-Strong consistency
(VFAT-WS), the first fast adversarial training method for video data.
Specifically, VFAT-WS incorporates the following key designs: First, it
integrates a straightforward yet effective temporal frequency augmentation
(TF-AUG), and its spatial-temporal enhanced form STF-AUG, along with a
single-step PGD attack to boost training efficiency and robustness. Second, it
devises a weak-to-strong spatial-temporal consistency regularization, which
seamlessly integrates the simpler TF-AUG and the more complex STF-AUG.
Leveraging the consistency regularization, it steers the learning process from
simple to complex augmentations. Both of them work together to achieve a better
trade-off between clean accuracy and robustness. Extensive experiments on
UCF-101 and HMDB-51 with both CNN and Transformer-based models demonstrate that
VFAT-WS achieves great improvements in adversarial robustness and corruption
robustness, while accelerating training by nearly 490%.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 07:40:35 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 13:22:33 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Wang",
"Songping",
""
],
[
"Liu",
"Hanqing",
""
],
[
"Lyu",
"Yueming",
""
],
[
"Hu",
"Xiantao",
""
],
[
"He",
"Ziwen",
""
],
[
"Wang",
"Wei",
""
],
[
"Shan",
"Caifeng",
""
],
[
"Wang",
"Liang",
""
]
] |
2504.14960 | Zhenhuan Liu | Dennis Liu, Zijie Yan, Xin Yao, Tong Liu, Vijay Korthikanti, Evan Wu,
Shiqing Fan, Gao Deng, Hongxiao Bai, Jianbin Chang, Ashwath Aithal, Michael
Andersch, Mohammad Shoeybi, Jiajie Yao, Chandler Zhou, David Wu, Xipeng Li,
June Yang | MoE Parallel Folding: Heterogeneous Parallelism Mappings for Efficient
Large-Scale MoE Model Training with Megatron Core | null | null | null | null | cs.LG cs.DC | http://creativecommons.org/licenses/by/4.0/ | Mixture of Experts (MoE) models enhance neural network scalability by
dynamically selecting relevant experts per input token, enabling larger model
sizes while maintaining manageable computation costs. However, efficient
training of large-scale MoE models across thousands of GPUs presents
significant challenges due to limitations in existing parallelism strategies.
We introduce an end-to-end training framework for large-scale MoE models that
utilizes five-dimensional hybrid parallelism: Tensor Parallelism, Expert
Parallelism, Context Parallelism, Data Parallelism, and Pipeline Parallelism.
Central to our approach is MoE Parallel Folding, a novel strategy that
decouples the parallelization of attention and MoE layers in Transformer
models, allowing each layer type to adopt optimal parallel configurations.
Additionally, we develop a flexible token-level dispatcher that supports both
token-dropping and token-dropless MoE training across all five dimensions of
parallelism. This dispatcher accommodates dynamic tensor shapes and coordinates
different parallelism schemes for Attention and MoE layers, facilitating
complex parallelism implementations. Our experiments demonstrate significant
improvements in training efficiency and scalability. We achieve up to 49.3%
Model Flops Utilization (MFU) for the Mixtral 8x22B model and 39.0% MFU for the
Qwen2-57B-A14B model on H100 GPUs, outperforming existing methods. The
framework scales efficiently up to 1,024 GPUs and maintains high performance
with sequence lengths up to 128K tokens, validating its effectiveness for
large-scale MoE model training. The code is available in Megatron-Core.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 08:39:47 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 05:44:18 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Liu",
"Dennis",
""
],
[
"Yan",
"Zijie",
""
],
[
"Yao",
"Xin",
""
],
[
"Liu",
"Tong",
""
],
[
"Korthikanti",
"Vijay",
""
],
[
"Wu",
"Evan",
""
],
[
"Fan",
"Shiqing",
""
],
[
"Deng",
"Gao",
""
],
[
"Bai",
"Hongxiao",
""
],
[
"Chang",
"Jianbin",
""
],
[
"Aithal",
"Ashwath",
""
],
[
"Andersch",
"Michael",
""
],
[
"Shoeybi",
"Mohammad",
""
],
[
"Yao",
"Jiajie",
""
],
[
"Zhou",
"Chandler",
""
],
[
"Wu",
"David",
""
],
[
"Li",
"Xipeng",
""
],
[
"Yang",
"June",
""
]
] |
2504.14962 | Wai-Ho Yeung | Wai-Ho Yeung | A Spectral Splitting Theorem for the $N$-Bakry \'Emery Ricci tensor | 20 pages. this article draws heavily from arXiv:2412.12707. Change:
Correct typos, mistakes in eq(3.28) and latter estimate in [v1]. Adding a
remark in the end of section 1 | null | null | null | math.DG math.AP | http://creativecommons.org/licenses/by/4.0/ | We extend the spectral generalization of the Cheeger-Gromoll splitting
theorem to smooth metric measure space. We show that if a complete non-compact
weighted Riemannian manifold $(M,g,e^{-f}\,dvolg)$ of dimension $n\ge 2$ has at
least two ends where $f$ is smooth and bounded. If there is some $N\in
(0,\infty)$ and $\gamma<\left(\frac{1}{(n-1)\left(1 + \frac{n-1}{N}\right)} +
\frac{n-1}{4}\right)^{-1}$ such that $$\lambda_1(-\gamma
\Delta_f+\operatorname{Ric}^N_f)\ge 0$$then $M$ splits isometrically as
$\mathbb{R}\times X$ for some complete Riemannian manifold $X$ with
$(\operatorname{Ric}_X)^N_f\ge 0$. The estimate can recover the spectral
splitting result and its sharp constant $\frac{4}{n-1}$ in
Antonelli-Pozzetta-Xu and and Catino--Mari--Mastrolia--Roncoroni.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 08:44:32 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 08:31:16 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Apr 2025 03:59:05 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Yeung",
"Wai-Ho",
""
]
] |
2504.14985 | Fatih Deniz | Fatih Deniz, Dorde Popovic, Yazan Boshmaf, Euisuh Jeong, Minhaj Ahmad,
Sanjay Chawla, Issa Khalil | aiXamine: Simplified LLM Safety and Security | null | null | null | null | cs.CR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evaluating Large Language Models (LLMs) for safety and security remains a
complex task, often requiring users to navigate a fragmented landscape of ad
hoc benchmarks, datasets, metrics, and reporting formats. To address this
challenge, we present aiXamine, a comprehensive black-box evaluation platform
for LLM safety and security. aiXamine integrates over 40 tests (i.e.,
benchmarks) organized into eight key services targeting specific dimensions of
safety and security: adversarial robustness, code security, fairness and bias,
hallucination, model and data privacy, out-of-distribution (OOD) robustness,
over-refusal, and safety alignment. The platform aggregates the evaluation
results into a single detailed report per model, providing a detailed breakdown
of model performance, test examples, and rich visualizations. We used aiXamine
to assess over 50 publicly available and proprietary LLMs, conducting over 2K
examinations. Our findings reveal notable vulnerabilities in leading models,
including susceptibility to adversarial attacks in OpenAI's GPT-4o, biased
outputs in xAI's Grok-3, and privacy weaknesses in Google's Gemini 2.0.
Additionally, we observe that open-source models can match or exceed
proprietary models in specific services such as safety alignment, fairness and
bias, and OOD robustness. Finally, we identify trade-offs between distillation
strategies, model size, training methods, and architectural choices.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 09:26:05 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 16:52:54 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Deniz",
"Fatih",
""
],
[
"Popovic",
"Dorde",
""
],
[
"Boshmaf",
"Yazan",
""
],
[
"Jeong",
"Euisuh",
""
],
[
"Ahmad",
"Minhaj",
""
],
[
"Chawla",
"Sanjay",
""
],
[
"Khalil",
"Issa",
""
]
] |
2504.15019 | Partha Sarathi Mohapatra | Partha Sarathi Mohapatra, Puduru Viswanadha Reddy, Georges Zaccour | Feedback Stackelberg-Nash equilibria in difference games with
quasi-hierarchical interactions and inequality constraints | null | null | null | null | math.OC cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study a class of two-player deterministic finite-horizon
difference games with coupled inequality constraints, where each player has two
types of decision variables: one involving sequential interactions and the
other simultaneous interactions. We refer to this class of games as
quasi-hierarchical dynamic games and define a solution concept called the
feedback Stackelberg-Nash (FSN) equilibrium. Under separability assumption on
cost functions, we provide a recursive formulation of the FSN solution using
dynamic programming. We show that the FSN solution can be derived from the
parametric feedback Stackelberg solution of an associated unconstrained game
involving only sequential interactions, with a specific choice of the
parameters that satisfy certain implicit complementarity conditions. For the
linear-quadratic case, we show that an FSN solution is obtained by
reformulating these complementarity conditions as a single large-scale linear
complementarity problem. Finally, we illustrate our results using a dynamic
duopoly game with production constraints.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 11:05:03 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 09:46:22 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Apr 2025 11:09:58 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Mohapatra",
"Partha Sarathi",
""
],
[
"Reddy",
"Puduru Viswanadha",
""
],
[
"Zaccour",
"Georges",
""
]
] |
2504.15166 | Frederick Truman-Williams | Frederick Truman-Williams and Giorgos Minas | Simulating biochemical reactions: The Linear Noise Approximation can
capture non-linear dynamics | 44 pages, 16 figures | null | null | null | q-bio.QM math.PR physics.chem-ph q-bio.MN | http://creativecommons.org/licenses/by-nc-nd/4.0/ | There is a plethora of highly stochastic non-linear dynamical systems in
fields such as molecular biology, chemistry, epidemiology, and ecology. Yet,
none of the currently available stochastic models are both accurate and
computationally efficient for long-term predictions of large systems. The
Linear Noise Approximation (LNA) model for biochemical reaction networks is
analytically tractable, which makes it computationally efficient for
simulation, analysis, and inference. However, it is only accurate for linear
systems and short-time transitions. Other methods can achieve greater accuracy
across a wider range of systems, including non-linear ones, but lack analytical
tractability. This paper seeks to challenge the prevailing view by
demonstrating that the Linear Noise Approximation can indeed capture non-linear
dynamics after certain modifications. We introduce a new framework that
utilises centre manifold theory allowing us to identify simple interventions to
the LNA that do not significantly compromise its computational efficiency. We
develop specific algorithms for systems that exhibit oscillations or
bi-stability and demonstrate their accuracy and computational efficiency across
multiple examples.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 15:16:30 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 16:31:05 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Truman-Williams",
"Frederick",
""
],
[
"Minas",
"Giorgos",
""
]
] |
2504.15305 | Abhishek Tyagi | Abhishek Tyagi, Charu Gaur | SLAM-Based Navigation and Fault Resilience in a Surveillance Quadcopter
with Embedded Vision Systems | 18 pages, 21 figures, 15 tables. Onboard processing using Raspberry
Pi 4 and Arduino Nano. Includes ORB-SLAM3-based navigation, LQR control,
rotor fault recovery, object detection, and PCA face recognition. Real-world
and simulation tests included. Designed for GPS-denied autonomous UAV
surveillance | null | null | null | cs.RO cs.CV cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | We present an autonomous aerial surveillance platform, Veg, designed as a
fault-tolerant quadcopter system that integrates visual SLAM for
GPS-independent navigation, advanced control architecture for dynamic
stability, and embedded vision modules for real-time object and face
recognition. The platform features a cascaded control design with an LQR
inner-loop and PD outer-loop trajectory control. It leverages ORB-SLAM3 for
6-DoF localization and loop closure, and supports waypoint-based navigation
through Dijkstra path planning over SLAM-derived maps. A real-time Failure
Detection and Identification (FDI) system detects rotor faults and executes
emergency landing through re-routing. The embedded vision system, based on a
lightweight CNN and PCA, enables onboard object detection and face recognition
with high precision. The drone operates fully onboard using a Raspberry Pi 4
and Arduino Nano, validated through simulations and real-world testing. This
work consolidates real-time localization, fault recovery, and embedded AI on a
single platform suitable for constrained environments.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 17:10:26 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 03:17:04 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Tyagi",
"Abhishek",
""
],
[
"Gaur",
"Charu",
""
]
] |
2504.15308 | Andre Souza | Andre N. Souza, Simone Silvestri, Katherine Deck, Tobias Bischoff,
Raffaele Ferrari, and Glenn R. Flierl | Surface to Seafloor: A Generative AI Framework for Decoding the Ocean
Interior State | null | null | null | null | physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | Understanding subsurface ocean dynamics is essential for quantifying oceanic
heat and mass transport, but direct observations at depth remain sparse due to
logistical and technological constraints. In contrast, satellite missions
provide rich surface datasets-such as sea surface height, temperature, and
salinity-that offer indirect but potentially powerful constraints on the ocean
interior. Here, we present a probabilistic framework based on score-based
diffusion models to reconstruct three-dimensional subsurface velocity and
buoyancy fields, including the energetic ocean eddy field, from surface
observations. Using a 15-level primitive equation simulation of an idealized
double-gyre system, we evaluate the skill of the model in inferring the mean
circulation and the mesoscale variability at depth under varying levels of
surface information. We find that the generative model successfully recovers
key dynamical structures and provides physically meaningful uncertainty
estimates, with predictive skill diminishing systematically as the surface
resolution decreases or the inference depth increases. These results
demonstrate the potential of generative approaches for ocean state estimation
and uncertainty quantification, particularly in regimes where traditional
deterministic methods are underconstrained or ill-posed.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 21:39:08 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 13:36:33 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Souza",
"Andre N.",
""
],
[
"Silvestri",
"Simone",
""
],
[
"Deck",
"Katherine",
""
],
[
"Bischoff",
"Tobias",
""
],
[
"Ferrari",
"Raffaele",
""
],
[
"Flierl",
"Glenn R.",
""
]
] |
2504.15327 | Tianliang Yao | Tianliang Yao, Bo Lu, Markus Kowarschik, Yixuan Yuan, Hubin Zhao,
Sebastien Ourselin, Kaspar Althoefer, Junbo Ge, Peng Qi | Advancing Embodied Intelligence in Robotic-Assisted Endovascular
Procedures: A Systematic Review of AI Solutions | 41 pages, 7 figures | null | null | null | cs.RO cs.LG | http://creativecommons.org/licenses/by/4.0/ | Endovascular procedures have revolutionized the treatment of vascular
diseases thanks to minimally invasive solutions that significantly reduce
patient recovery time and enhance clinical outcomes. However, the precision and
dexterity required during these procedures poses considerable challenges for
interventionists. Robotic systems have emerged offering transformative
solutions, addressing issues such as operator fatigue, radiation exposure, and
the inherent limitations of human precision. The integration of Embodied
Intelligence (EI) into these systems signifies a paradigm shift, enabling
robots to navigate complex vascular networks and adapt to dynamic physiological
conditions. Data-driven approaches, advanced computer vision, medical image
analysis, and machine learning techniques, are at the forefront of this
evolution. These methods augment procedural intelligence by facilitating
real-time vessel segmentation, device tracking, and anatomical landmark
detection. Reinforcement learning and imitation learning further refine
navigation strategies and replicate experts' techniques. This review
systematically examines the integration of EI principles into robotic
technologies, in relation to endovascular procedures. We discuss recent
advancements in intelligent perception and data-driven control, and their
practical applications in robot-assisted endovascular procedures. By critically
evaluating current limitations and emerging opportunities, this review
establishes a framework for future developments, emphasizing the potential for
greater autonomy and improved clinical outcomes. Emerging trends and specific
areas of research, such as federated learning for medical data sharing,
explainable AI for clinical decision support, and advanced human-robot
collaboration paradigms, are also explored, offering insights into the future
direction of this rapidly evolving field.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 13:49:30 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 15:44:20 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Yao",
"Tianliang",
""
],
[
"Lu",
"Bo",
""
],
[
"Kowarschik",
"Markus",
""
],
[
"Yuan",
"Yixuan",
""
],
[
"Zhao",
"Hubin",
""
],
[
"Ourselin",
"Sebastien",
""
],
[
"Althoefer",
"Kaspar",
""
],
[
"Ge",
"Junbo",
""
],
[
"Qi",
"Peng",
""
]
] |
2504.15382 | Michael Sekatchev | Michael Sekatchev, Xunyu Liang, Fereshteh Majidi, Ben Scully, Ludovic
Van Waerbeke, Ariel Zhitnitsky | The Glow of Axion Quark Nugget Dark Matter: (III) The Mysteries of the
Milky Way UV Background | 25 pages, 7 figures, prepared for submission to JCAP | null | null | null | astro-ph.CO astro-ph.GA astro-ph.HE hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Axion quark nuggets (AQNs) are hypothetical objects with nuclear density that
would have formed during the quark-hadron transition and could make up most of
the dark matter today. These objects have a mass greater than a few grams and
are sub-micrometer in size. They would also help explain the matter-antimatter
asymmetry and the similarity between visible and dark components of the
universe, i.e. $\Omega_{\text{DM}} \sim \Omega_{\text{visible}}$. These
composite objects behave as cold dark matter, interacting with ordinary matter
and producing pervasive electromagnetic radiation. This work aims to calculate
the FUV electromagnetic signature in a 1 kpc region surrounding the solar
system, resulting from the interaction between antimatter AQNs and baryons. To
this end, we use the high-resolution hydrodynamic simulation of the Milky Way,
FIRE-2 Latter suite, to select solar system-like regions. From the simulated
gas and dark matter distributions in these regions, we calculate the FUV
background radiation generated by the AQN model. We find that the results are
consistent with the FUV excess recently confirmed by the Alice spectrograph
aboard New Horizons, which corroborated the FUV excess initially discovered by
GALEX a decade ago. We also discuss the potential cosmological implications of
our work, which suggest the existence of a new source of FUV radiation in
galaxies, linked to the interaction between dark matter and baryons.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 18:48:11 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Sekatchev",
"Michael",
""
],
[
"Liang",
"Xunyu",
""
],
[
"Majidi",
"Fereshteh",
""
],
[
"Scully",
"Ben",
""
],
[
"Van Waerbeke",
"Ludovic",
""
],
[
"Zhitnitsky",
"Ariel",
""
]
] |
2504.15410 | Matteo Antonio Inajetovic | Matteo Inajetovic and Petros Wallden and Anna Pappa | Verifiable End-to-End Delegated Variational Quantum Algorithms | 9 pages, 1 figure | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Variational quantum algorithms (VQAs) have emerged as promising candidates
for solving complex optimization and machine learning tasks on near-term
quantum hardware. However, executing quantum operations remains challenging for
small-scale users because of several hardware constraints, making it desirable
to delegate parts of the computation to more powerful quantum devices. In this
work, we introduce a framework for delegated variational quantum algorithms
(DVQAs), where a client with limited quantum capabilities delegates the
execution of a VQA to a more powerful quantum server. In particular, we
introduce a protocol that enables a client to delegate a variational quantum
algorithm to a server while ensuring that the input, the output and also the
computation itself remain secret. Additionally, if the protocol does not abort,
the client can be certain that the computation outcome is indeed correct. This
work builds on the general verification protocol introduced by Fitzimons and
Kashefi (2017), tailoring it to VQAs. Our approach first proposes a verifiable
Protocol for delegating the quantum computation required at each optimization
step of a VQA, and then combines the iterative steps into an error-resilient
optimization process that offers end-to-end verifiable algorithm execution. We
also simulate the performance of our protocol tackling the Transverse Field
Ising Model. Our results demonstrate that secure delegation of variational
quantum algorithms is a realistic solution for near-term quantum networks,
paving the way for practical quantum cloud computing applications.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 19:36:42 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 05:27:37 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Inajetovic",
"Matteo",
""
],
[
"Wallden",
"Petros",
""
],
[
"Pappa",
"Anna",
""
]
] |
2504.15437 | Ryan Landvater | Ryan Erik Landvater, Ulysses Balis | Iris: A Next Generation Digital Pathology Rendering Engine | 11 pages, 8 figures | Journal of Pathology Informatics, 16, 100414 (2025) | 10.1016/j.jpi.2024.100414 | null | cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Digital pathology is a tool of rapidly evolving importance within the
discipline of pathology. Whole slide imaging promises numerous advantages;
however, adoption is limited by challenges in ease of use and speed of
high-quality image rendering relative to the simplicity and visual quality of
glass slides. We introduce Iris, a new high-performance digital pathology
rendering system. Specifically, we outline and detail the performance metrics
of Iris Core, the core rendering engine technology. Iris Core comprises machine
code modules written from the ground up in C++ and using Vulkan, a low-level
and low-overhead cross-platform graphical processing unit application program
interface, and our novel rapid tile buffering algorithms. We provide a detailed
explanation of Iris Core's system architecture, including the stateless
isolation of core processes, interprocess communication paradigms, and explicit
synchronization paradigms that provide powerful control over the graphical
processing unit. Iris Core achieves slide rendering at the sustained maximum
frame rate on all tested platforms and buffers an entire new slide field of,
view without overlapping pixels, in 10 ms with enhanced detail in 30 ms. It is
able to buffer and compute high-fidelity reduction-enhancements for viewing
low-power cytology with increased visual quality at a rate of 100-160 us per
slide tile, and with a cumulative median buffering rate of 1.36 GB of
decompressed image data per second. This buffering rate allows for an entirely
new field of view to be fully buffered and rendered in less than a single
monitor refresh on a standard display, and high detail features within 2-3
monitor refresh frames. These metrics far exceed previously published
specifications, beyond an order of magnitude in some contexts. The system shows
no slowing with high use loads, but rather increases performance due to cache
mechanisms.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 21:01:57 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 15:02:16 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Landvater",
"Ryan Erik",
""
],
[
"Balis",
"Ulysses",
""
]
] |
2504.15501 | Michael Reitz | Piper Fowler-Wright, Michael Reitz, Joel Yuen-Zhou | Mapping molecular polariton transport via pump-probe microscopy | 8 pages, 4 figures. Includes supplementary movie file showing
pump-probe dynamics | null | null | null | quant-ph cond-mat.mes-hall physics.chem-ph physics.optics | http://creativecommons.org/licenses/by/4.0/ | We demonstrate how the transport properties of molecular polaritons in
optical cavities can be extracted from a microscopic modeling of pump-probe
spectroscopy. Our approach combines a mean-field treatment of the light-matter
Hamiltonian with a perturbative expansion of both light and matter components,
along with spatial coarse-graining. This approach extends semiclassical cavity
spectroscopy to multimode light-matter interactions, providing full access to
spatially resolved transient spectra. By simulating a microscopy experiment
with counter-propagating pump and probe pulses, we compute the differential
transmission and show how molecular dephasing and persistent dark exciton
populations drive sub-group-velocity transport of the root-mean-square
displacement. We analyze transport across the polariton dispersion, showing how
velocity renormalization correlates with excitonic weight, consistent with
experimental observations, and further its dependence on the rate of molecular
dephasing. Our results highlight the need to consider measured spectroscopic
observables when characterizing transport in polaritonic systems.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 00:46:24 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 01:20:57 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Fowler-Wright",
"Piper",
""
],
[
"Reitz",
"Michael",
""
],
[
"Yuen-Zhou",
"Joel",
""
]
] |
2504.15507 | Shaila Sharmin | Shaila Sharmin, Anwar Hossain Zahid, Subhankar Bhattacharjee, Chiamaka
Igwilo, Miryung Kim and Wei Le | Automatically Detecting Numerical Instability in Machine Learning
Applications via Soft Assertions | 22 pages, 5 figures. Accepted at FSE 2025 | null | 10.1145/3729394 | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Machine learning (ML) applications have become an integral part of our lives.
ML applications extensively use floating-point computation and involve very
large/small numbers; thus, maintaining the numerical stability of such complex
computations remains an important challenge. Numerical bugs can lead to system
crashes, incorrect output, and wasted computing resources. In this paper, we
introduce a novel idea, namely soft assertions (SA), to encode safety/error
conditions for the places where numerical instability can occur. A soft
assertion is an ML model automatically trained using the dataset obtained
during unit testing of unstable functions. Given the values at the unstable
function in an ML application, a soft assertion reports how to change these
values in order to trigger the instability. We then use the output of soft
assertions as signals to effectively mutate inputs to trigger numerical
instability in ML applications. In the evaluation, we used the GRIST benchmark,
a total of 79 programs, as well as 15 real-world ML applications from GitHub.
We compared our tool with 5 state-of-the-art (SOTA) fuzzers. We found all the
GRIST bugs and outperformed the baselines. We found 13 numerical bugs in
real-world code, one of which had already been confirmed by the GitHub
developers. While the baselines mostly found the bugs that report NaN and INF,
our tool \tool found numerical bugs with incorrect output. We showed one case
where the Tumor Detection Model, trained on Brain MRI images, should have
predicted "tumor", but instead, it incorrectly predicted "no tumor" due to the
numerical bugs. Our replication package is located at
https://figshare.com/s/6528d21ccd28bea94c32.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 00:55:33 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:46:21 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Sharmin",
"Shaila",
""
],
[
"Zahid",
"Anwar Hossain",
""
],
[
"Bhattacharjee",
"Subhankar",
""
],
[
"Igwilo",
"Chiamaka",
""
],
[
"Kim",
"Miryung",
""
],
[
"Le",
"Wei",
""
]
] |
2504.15515 | Wuchen Li | Wuchen Li | Transport f divergences | Comments are welcome | null | null | null | math.ST cs.AI cs.IT math.IT stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We define a class of divergences to measure differences between probability
density functions in one-dimensional sample space. The construction is based on
the convex function with the Jacobi operator of mapping function that
pushforwards one density to the other. We call these information measures
transport f-divergences. We present several properties of transport
$f$-divergences, including invariances, convexities, variational formulations,
and Taylor expansions in terms of mapping functions. Examples of transport
f-divergences in generative models are provided.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 01:25:41 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 01:18:37 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Li",
"Wuchen",
""
]
] |
2504.15610 | MD Millat Hosen | Md Millat Hosen | A LoRA-Based Approach to Fine-Tuning LLMs for Educational Guidance in
Resource-Constrained Settings | 18 pages, 6 figures (3 graphs + 3 flowchart/architecture diagrams),
submitted as a preprint for review consideration in AI for Education or
Machine Learning applications in low-resource settings. Includes detailed
experiments with LoRA and quantization methods for efficient LLM fine-tuning | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | The current study describes a cost-effective method for adapting large
language models (LLMs) for academic advising with study-abroad contexts in mind
and for application in low-resource methods for acculturation. With the
Mistral-7B-Instruct model applied with a Low-Rank Adaptation (LoRA) method and
a 4-bit quantization method, the model underwent training in two distinct
stages related to this study's purpose to enhance domain specificity while
maintaining computational efficiency. In Phase 1, the model was conditioned
with a synthetic dataset via the Gemini Pro API, and in Phase 2, it was trained
with manually curated datasets from the StudyAbroadGPT project to achieve
enhanced, contextualized responses. Technical innovations entailed
memory-efficient quantization, parameter-efficient adaptation, and continuous
training analytics via Weights & Biases. After training, this study
demonstrated a reduction in training loss by 52.7%, 92% accuracy in
domain-specific recommendations, achieved 95% markdown-based formatting
support, and a median run-rate of 100 samples per second on off-the-shelf GPU
equipment. These findings support the effective application of
instruction-tuned LLMs within educational advisers, especially in low-resource
institutional scenarios. Limitations included decreased generalizability and
the application of a synthetically generated dataset, but this framework is
scalable for adding new multilingual-augmented and real-time academic advising
processes. Future directions may include plans for the integration of
retrieval-augmented generation, applying dynamic quantization routines, and
connecting to real-time academic databases to increase adaptability and
accuracy.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 06:08:13 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 04:59:47 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Hosen",
"Md Millat",
""
]
] |
2504.15621 | Taichi Katayama | Taichi Katayama | Regularization of elliptic multiple zeta values | 20 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we show that regularized elliptic multiple zeta values are
given by polynomials in elliptic multiple zeta values with admissible indices
and special ones whose indices consist of 0 and 1.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 06:24:48 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 15:12:27 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Katayama",
"Taichi",
""
]
] |
2504.15774 | Boris Bellalta Dr. | Boris Bellalta and Francesc Wilhelmi and Lorenzo Galati-Giordano and
Giovanni Geraci | Performance Analysis of IEEE 802.11bn Non-Primary Channel Access | null | null | null | null | cs.NI cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a performance analysis of the Non-Primary Channel Access
(NPCA) mechanism, a new feature introduced in IEEE 802.11bn to enhance spectrum
utilization in Wi-Fi networks. NPCA enables devices to contend for and transmit
on the secondary channel when the primary channel is occupied by transmissions
from an Overlapping Basic Service Set (OBSS). We develop a Continuous-Time
Markov Chain (CTMC) model that captures the interactions among OBSSs in dense
WLAN environments when NPCA is enabled, incorporating new NPCA-specific states
and transitions. In addition to the analytical insights offered by the model,
we conduct numerical evaluations and simulations to quantify NPCA's impact on
throughput and channel access delay across various scenarios. Our results show
that NPCA can significantly improve throughput and reduce access delays in
favorable conditions for BSSs that support the mechanism. Moreover, NPCA helps
mitigate the OBSS performance anomaly, where low-rate OBSS transmissions
degrade network performance for all nearby devices. However, we also observe
trade-offs: NPCA may increase contention on secondary channels, potentially
reducing transmission opportunities for BSSs operating there.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 10:30:15 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Bellalta",
"Boris",
""
],
[
"Wilhelmi",
"Francesc",
""
],
[
"Galati-Giordano",
"Lorenzo",
""
],
[
"Geraci",
"Giovanni",
""
]
] |
2504.15800 | Jihoon Kwon | Chanyeol Choi, Jihoon Kwon, Jaeseon Ha, Hojun Choi, Chaewoon Kim,
Yongjae Lee, Jy-yong Sohn, Alejandro Lopez-Lira | FinDER: Financial Dataset for Question Answering and Evaluating
Retrieval-Augmented Generation | 10 pages, 3 figures, ICLR 2025 Workshop Advances in Financial AI | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | In the fast-paced financial domain, accurate and up-to-date information is
critical to addressing ever-evolving market conditions. Retrieving this
information correctly is essential in financial Question-Answering (QA), since
many language models struggle with factual accuracy in this domain. We present
FinDER, an expert-generated dataset tailored for Retrieval-Augmented Generation
(RAG) in finance. Unlike existing QA datasets that provide predefined contexts
and rely on relatively clear and straightforward queries, FinDER focuses on
annotating search-relevant evidence by domain experts, offering 5,703
query-evidence-answer triplets derived from real-world financial inquiries.
These queries frequently include abbreviations, acronyms, and concise
expressions, capturing the brevity and ambiguity common in the realistic search
behavior of professionals. By challenging models to retrieve relevant
information from large corpora rather than relying on readily determined
contexts, FinDER offers a more realistic benchmark for evaluating RAG systems.
We further present a comprehensive evaluation of multiple state-of-the-art
retrieval models and Large Language Models, showcasing challenges derived from
a realistic benchmark to drive future research on truthful and precise RAG in
the financial domain.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 11:30:13 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:49:10 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Choi",
"Chanyeol",
""
],
[
"Kwon",
"Jihoon",
""
],
[
"Ha",
"Jaeseon",
""
],
[
"Choi",
"Hojun",
""
],
[
"Kim",
"Chaewoon",
""
],
[
"Lee",
"Yongjae",
""
],
[
"Sohn",
"Jy-yong",
""
],
[
"Lopez-Lira",
"Alejandro",
""
]
] |
2504.15806 | Kai Luo | Kai Luo, Juan Tang, Mingchao Cai, Xiaoqing Zeng, Manqi Xie, Ming Yan | DAE-KAN: A Kolmogorov-Arnold Network Model for High-Index
Differential-Algebraic Equations | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Kolmogorov-Arnold Networks (KANs) have emerged as a promising alternative to
Multi-layer Perceptrons (MLPs) due to their superior function-fitting abilities
in data-driven modeling. In this paper, we propose a novel framework, DAE-KAN,
for solving high-index differential-algebraic equations (DAEs) by integrating
KANs with Physics-Informed Neural Networks (PINNs). This framework not only
preserves the ability of traditional PINNs to model complex systems governed by
physical laws but also enhances their performance by leveraging the
function-fitting strengths of KANs. Numerical experiments demonstrate that for
DAE systems ranging from index-1 to index-3, DAE-KAN reduces the absolute
errors of both differential and algebraic variables by 1 to 2 orders of
magnitude compared to traditional PINNs. To assess the effectiveness of this
approach, we analyze the drift-off error and find that both PINNs and DAE-KAN
outperform classical numerical methods in controlling this phenomenon. Our
results highlight the potential of neural network methods, particularly
DAE-KAN, in solving high-index DAEs with substantial computational accuracy and
generalization, offering a promising solution for challenging partial
differential-algebraic equations.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 11:42:02 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 06:21:23 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Luo",
"Kai",
""
],
[
"Tang",
"Juan",
""
],
[
"Cai",
"Mingchao",
""
],
[
"Zeng",
"Xiaoqing",
""
],
[
"Xie",
"Manqi",
""
],
[
"Yan",
"Ming",
""
]
] |
2504.15811 | Markus Braden | D. Gorkov, M. M\"uller, G. Waldherr, A. Gr\"unwald, J. Stein, S.
Giemsa, A. C. Komarek, P. B\"oni and M. Braden | KOMPASS: the new cold neutron triple-axis-spectrometer specialized for
polarization analysis | 16 pages, 10 Figures | null | null | null | physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | KOMPASS is a polarized triple-axis cold neutron spectrometer recently
installed at the FRM II neutron source. The instrument is designed to operate
exclusively with polarized neutrons and is specialized in longitudinal
polarization analysis using Helmholtz coils and spherical zero-field neutron
polarimetry using a Cryopad device. The advanced guide system polarizes and
focuses flexibly in the scattering plane. A first, fixed parabolic focusing
part contains a series of three polarizing supermirror V-cavities that produce
a highly polarized beam. By exchanging straight and parabolic front-end guide
sections, the resolution of the instrument can be optimized to meet
experimental requirements. Large, double-focusing monochromator and analyzer
units with pyrolytic graphite crystals enable efficient and adaptive energy
selection, and a compact supermirror cavity analyzes the final neutron
polarization. Alternatively, or in combination with this cavity, a Heusler
polarization analyzer can be used. KOMPASS offers full- and half-polarized
configurations with or without secondary energy analysis and provides a wide
range of polarization options. Therefore, KOMPASS is well suited for various
studies of static and dynamic magnetic correlations with energy transfers on
the neutron energy loss side up to ~12 meV.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 11:50:57 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:49:44 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Gorkov",
"D.",
""
],
[
"Müller",
"M.",
""
],
[
"Waldherr",
"G.",
""
],
[
"Grünwald",
"A.",
""
],
[
"Stein",
"J.",
""
],
[
"Giemsa",
"S.",
""
],
[
"Komarek",
"A. C.",
""
],
[
"Böni",
"P.",
""
],
[
"Braden",
"M.",
""
]
] |
2504.15814 | Tobias Weinzierl | Timothy Stokes, Tobias Weinzierl, Han Zhang, Baojiu Li | Fast Higher-Order Interpolation and Restriction in ExaHyPE Avoiding
Non-physical Reflections | null | null | null | null | cs.CE cs.MS gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Wave equations help us to understand phenomena ranging from earthquakes to
tsunamis. These phenomena materialise over very large scales. It would be
computationally infeasible to track them over a regular mesh. Yet, since the
phenomena are localised, adaptive mesh refinement (AMR) can be used to
construct meshes with a higher resolution close to the regions of interest.
ExaHyPE is a software engine created to solve wave problems using AMR, and we
use it as baseline to construct our numerical relativity application called
ExaGRyPE. To advance the mesh in time, we have to interpolate and restrict
along resolution transitions in each and every time step. ExaHyPE's vanilla
code version uses a d-linear tensor-product approach. In benchmarks of a
stationary black hole this performs slowly and leads to errors in conserved
quantities near AMR boundaries. We therefore introduce a set of higher-order
interpolation schemes where the derivatives are calculated at each coarse grid
cell to approximate the enclosed fine cells. The resulting methods run faster
than the tensor-product approach. Most importantly, when running the stationary
black hole simulation using the higher order methods the errors near the AMR
boundaries are removed.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 11:52:58 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 08:43:40 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Stokes",
"Timothy",
""
],
[
"Weinzierl",
"Tobias",
""
],
[
"Zhang",
"Han",
""
],
[
"Li",
"Baojiu",
""
]
] |
2504.15844 | Zafer Esen | Zafer Esen, Philipp R\"ummer, Tjark Weber | Sound and Complete Invariant-Based Heap Encodings (Technical Report) | Added acknowledgements | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Verification of programs operating on mutable, heap-allocated data structures
poses significant challenges due to potentially unbounded structures like
linked lists and trees. In this paper, we present a novel relational heap
encoding leveraging uninterpreted predicates and prophecy variables, reducing
heap verification tasks to satisfiability checks over integers in constrained
Horn clauses (CHCs). To the best of our knowledge, our approach is the first
invariant-based method that achieves both soundness and completeness for
heap-manipulating programs. We provide formal proofs establishing the
correctness of our encodings. Through an experimental evaluation we demonstrate
that our method significantly extends the capability of existing CHC-based
verification tools, allowing automatic verification of programs with heap
previously unreachable by state-of-the-art tools.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 12:40:21 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 12:34:58 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Esen",
"Zafer",
""
],
[
"Rümmer",
"Philipp",
""
],
[
"Weber",
"Tjark",
""
]
] |
2504.15865 | Lotfi Abdelkrim Mecharbat | Lotfi Abdelkrim Mecharbat, Ibrahim Almakky, Martin Takac, Mohammad
Yaqub | MedNNS: Supernet-based Medical Task-Adaptive Neural Network Search | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Deep learning (DL) has achieved remarkable progress in the field of medical
imaging. However, adapting DL models to medical tasks remains a significant
challenge, primarily due to two key factors: (1) architecture selection, as
different tasks necessitate specialized model designs, and (2) weight
initialization, which directly impacts the convergence speed and final
performance of the models. Although transfer learning from ImageNet is a widely
adopted strategy, its effectiveness is constrained by the substantial
differences between natural and medical images. To address these challenges, we
introduce Medical Neural Network Search (MedNNS), the first Neural Network
Search framework for medical imaging applications. MedNNS jointly optimizes
architecture selection and weight initialization by constructing a meta-space
that encodes datasets and models based on how well they perform together. We
build this space using a Supernetwork-based approach, expanding the model zoo
size by 51x times over previous state-of-the-art (SOTA) methods. Moreover, we
introduce rank loss and Fr\'echet Inception Distance (FID) loss into the
construction of the space to capture inter-model and inter-dataset
relationships, thereby achieving more accurate alignment in the meta-space.
Experimental results across multiple datasets demonstrate that MedNNS
significantly outperforms both ImageNet pre-trained DL models and SOTA Neural
Architecture Search (NAS) methods, achieving an average accuracy improvement of
1.7% across datasets while converging substantially faster. The code and the
processed meta-space is available at https://github.com/BioMedIA-MBZUAI/MedNNS.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 13:04:40 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 05:28:18 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Mecharbat",
"Lotfi Abdelkrim",
""
],
[
"Almakky",
"Ibrahim",
""
],
[
"Takac",
"Martin",
""
],
[
"Yaqub",
"Mohammad",
""
]
] |
2504.15876 | Lei Chen | Qizhen Wu, Lei Chen, Kexin Liu, and Jinhu L\"u | Bidirectional Task-Motion Planning Based on Hierarchical Reinforcement
Learning for Strategic Confrontation | null | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | In swarm robotics, confrontation scenarios, including strategic
confrontations, require efficient decision-making that integrates discrete
commands and continuous actions. Traditional task and motion planning methods
separate decision-making into two layers, but their unidirectional structure
fails to capture the interdependence between these layers, limiting
adaptability in dynamic environments. Here, we propose a novel bidirectional
approach based on hierarchical reinforcement learning, enabling dynamic
interaction between the layers. This method effectively maps commands to task
allocation and actions to path planning, while leveraging cross-training
techniques to enhance learning across the hierarchical framework. Furthermore,
we introduce a trajectory prediction model that bridges abstract task
representations with actionable planning goals. In our experiments, it achieves
over 80% in confrontation win rate and under 0.01 seconds in decision time,
outperforming existing approaches. Demonstrations through large-scale tests and
real-world robot experiments further emphasize the generalization capabilities
and practical applicability of our method.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 13:22:58 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 15:00:10 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Wu",
"Qizhen",
""
],
[
"Chen",
"Lei",
""
],
[
"Liu",
"Kexin",
""
],
[
"Lü",
"Jinhu",
""
]
] |
2504.15918 | Chang Zong | Chang Zong, Bin Li, Shoujun Zhou, Jian Wan, Lei Zhang | Ask2Loc: Learning to Locate Instructional Visual Answers by Asking
Questions | 16 pages, 8 figures | null | null | null | cs.CV cs.AI cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Locating specific segments within an instructional video is an efficient way
to acquire guiding knowledge. Generally, the task of obtaining video segments
for both verbal explanations and visual demonstrations is known as visual
answer localization (VAL). However, users often need multiple interactions to
obtain answers that align with their expectations when using the system. During
these interactions, humans deepen their understanding of the video content by
asking themselves questions, thereby accurately identifying the location.
Therefore, we propose a new task, named In-VAL, to simulate the multiple
interactions between humans and videos in the procedure of obtaining visual
answers. The In-VAL task requires interactively addressing several semantic gap
issues, including 1) the ambiguity of user intent in the input questions, 2)
the incompleteness of language in video subtitles, and 3) the fragmentation of
content in video segments. To address these issues, we propose Ask2Loc, a
framework for resolving In-VAL by asking questions. It includes three key
modules: 1) a chatting module to refine initial questions and uncover clear
intentions, 2) a rewriting module to generate fluent language and create
complete descriptions, and 3) a searching module to broaden local context and
provide integrated content. We conduct extensive experiments on three
reconstructed In-VAL datasets. Compared to traditional end-to-end and two-stage
methods, our proposed Ask2Loc can improve performance by up to 14.91 (mIoU) on
the In-VAL task. Our code and datasets can be accessed at
https://github.com/changzong/Ask2Loc.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 14:03:16 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 03:01:06 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zong",
"Chang",
""
],
[
"Li",
"Bin",
""
],
[
"Zhou",
"Shoujun",
""
],
[
"Wan",
"Jian",
""
],
[
"Zhang",
"Lei",
""
]
] |
2504.15922 | Waleed Abdeen | Waleed Abdeen, Michael Unterkalmsteiner, Krzysztof Wnuk, Alessio
Ferrari, and Panagiota Chatzipetrou | Language Models to Support Multi-Label Classification of Industrial Data | Accepted at SANER Conference 2025. Awaiting publication by IEEE | null | null | null | cs.SE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Multi-label requirements classification is a challenging task, especially
when dealing with numerous classes at varying levels of abstraction. The
difficulties increases when a limited number of requirements is available to
train a supervised classifier. Zero-shot learning (ZSL) does not require
training data and can potentially address this problem. This paper investigates
the performance of zero-shot classifiers (ZSCs) on a multi-label industrial
dataset. We focuse on classifying requirements according to a taxonomy designed
to support requirements tracing. We compare multiple variants of ZSCs using
different embeddings, including 9 language models (LMs) with a reduced number
of parameters (up to 3B), e.g., BERT, and 5 large LMs (LLMs) with a large
number of parameters (up to 70B), e.g., Llama. Our ground truth includes 377
requirements and 1968 labels from 6 output spaces. For the evaluation, we adopt
traditional metrics, i.e., precision, recall, F1, and $F_\beta$, as well as a
novel label distance metric Dn. This aims to better capture the
classification's hierarchical nature and provides a more nuanced evaluation of
how far the results are from the ground truth. 1) The top-performing model on 5
out of 6 output spaces is T5-xl, with maximum $F_\beta$ = 0.78 and Dn = 0.04,
while BERT base outperformed the other models in one case, with maximum
$F_\beta$ = 0.83 and Dn = 0.04. 2) LMs with smaller parameter size produce the
best classification results compared to LLMs. Thus, addressing the problem in
practice is feasible as limited computing power is needed. 3) The model
architecture (autoencoding, autoregression, and sentence-to-sentence)
significantly affects the classifier's performance. We conclude that using ZSL
for multi-label requirements classification offers promising results. We also
present a novel metric that can be used to select the top-performing model for
this problem
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 14:06:02 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:24:40 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Abdeen",
"Waleed",
""
],
[
"Unterkalmsteiner",
"Michael",
""
],
[
"Wnuk",
"Krzysztof",
""
],
[
"Ferrari",
"Alessio",
""
],
[
"Chatzipetrou",
"Panagiota",
""
]
] |
2504.15939 | Alexander Paarmann | Niclas S. Mueller, Alexander P. Fellows, Ben John, Andrew E. Naclerio,
Christian Carbogno, Katayoun Gharagozloo-Hubmann, Dami\'an Bal\'a\v{z}, Ryan
A. Kowalski, Hendrik H. Heenen, Christoph Scheurer, Karsten Reuter, Joshua D.
Caldwell, Martin Wolf, Piran R. Kidambi, Martin Th\"amer, Alexander Paarmann | Full Crystallographic Imaging of Hexagonal Boron Nitride Monolayers with
Phonon-Enhanced Sum-Frequency Microscopy | null | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall physics.optics | http://creativecommons.org/licenses/by/4.0/ | Hexagonal boron nitride (hBN) is an important 2D material for van der Waals
heterostructures, single photon emitters, and infrared nanophotonics. The
optical characterization of mono- and few-layer samples of hBN however remains
a challenge as the material is almost invisible optically. Here we introduce
phase-resolved sum-frequency microscopy as a technique for imaging monolayers
of hBN grown by chemical vapor deposition (CVD) and visualize their crystal
orientation. A combination of femtosecond mid-infrared (IR) and visible laser
pulses is used for sum-frequency generation (SFG), which is imaged in a
wide-field optical microscope. The IR laser resonantly excites a phonon of hBN
that leads to an ~800-fold enhancement of the SFG intensity, making it possible
to image large 100x100 {\mu}m2 sample areas in less than 1 s. Implementing
heterodyne detection in combination with azimuthal rotation of the sample
further provides full crystallographic information. Through combined knowledge
of topography and crystal orientation, we find that triangular domains of
CVD-grown monolayer hBN have nitrogen-terminated zigzag edges. Overall, SFG
microscopy can be used as an ultra-sensitive tool to image crystal structure,
strain, stacking sequences, and twist angles, and is applicable to the wide
range of van der Waals structures, where location and identification of
monolayer regions and interfaces with broken inversion symmetry is of paramount
importance.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 14:28:30 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 07:46:06 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Mueller",
"Niclas S.",
""
],
[
"Fellows",
"Alexander P.",
""
],
[
"John",
"Ben",
""
],
[
"Naclerio",
"Andrew E.",
""
],
[
"Carbogno",
"Christian",
""
],
[
"Gharagozloo-Hubmann",
"Katayoun",
""
],
[
"Baláž",
"Damián",
""
],
[
"Kowalski",
"Ryan A.",
""
],
[
"Heenen",
"Hendrik H.",
""
],
[
"Scheurer",
"Christoph",
""
],
[
"Reuter",
"Karsten",
""
],
[
"Caldwell",
"Joshua D.",
""
],
[
"Wolf",
"Martin",
""
],
[
"Kidambi",
"Piran R.",
""
],
[
"Thämer",
"Martin",
""
],
[
"Paarmann",
"Alexander",
""
]
] |
2504.15965 | Yaxiong Wu | Yaxiong Wu, Sheng Liang, Chen Zhang, Yichao Wang, Yongyue Zhang,
Huifeng Guo, Ruiming Tang, Yong Liu | From Human Memory to AI Memory: A Survey on Memory Mechanisms in the Era
of LLMs | 26 pages, 1 figure, 3 tables | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Memory is the process of encoding, storing, and retrieving information,
allowing humans to retain experiences, knowledge, skills, and facts over time,
and serving as the foundation for growth and effective interaction with the
world. It plays a crucial role in shaping our identity, making decisions,
learning from past experiences, building relationships, and adapting to
changes. In the era of large language models (LLMs), memory refers to the
ability of an AI system to retain, recall, and use information from past
interactions to improve future responses and interactions. Although previous
research and reviews have provided detailed descriptions of memory mechanisms,
there is still a lack of a systematic review that summarizes and analyzes the
relationship between the memory of LLM-driven AI systems and human memory, as
well as how we can be inspired by human memory to construct more powerful
memory systems. To achieve this, in this paper, we propose a comprehensive
survey on the memory of LLM-driven AI systems. In particular, we first conduct
a detailed analysis of the categories of human memory and relate them to the
memory of AI systems. Second, we systematically organize existing
memory-related work and propose a categorization method based on three
dimensions (object, form, and time) and eight quadrants. Finally, we illustrate
some open problems regarding the memory of current AI systems and outline
possible future directions for memory in the era of large language models.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 15:05:04 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 13:47:27 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Wu",
"Yaxiong",
""
],
[
"Liang",
"Sheng",
""
],
[
"Zhang",
"Chen",
""
],
[
"Wang",
"Yichao",
""
],
[
"Zhang",
"Yongyue",
""
],
[
"Guo",
"Huifeng",
""
],
[
"Tang",
"Ruiming",
""
],
[
"Liu",
"Yong",
""
]
] |
2504.15973 | Hyungrok Kim | Leron Borsten, Michael J. Duff, Dimitri Kanakaris, Hyungrok Kim | Duality Anomalies in Linearized Gravity | 8 pages. Added acknowledgments | null | null | null | hep-th gr-qc math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Classical linearized gravity admits a dual formulation in terms of a
higher-rank tensor field. Proposing a prescription for the instanton sectors of
linearized gravity and its dual, we show that they may be quantum inequivalent
in even dimensions. The duality anomaly is obtained by resolving the dual
graviton theories into vector-valued $p$-form electrodynamics and is controlled
by the Reidemeister torsion, Ray-Singer torsion and Euler characteristic of the
cotangent bundle. Under the proposed instanton prescription the duality anomaly
vanishes for an odd number of spacetime dimensions as a consequence of the
celebrated Cheeger-M\"uller theorem. In the presence of a gravitational
$\theta$-term, the partition function is a modular form in direct analogy to
Abelian S-duality for Maxwell theory.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 15:18:47 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 15:50:37 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Borsten",
"Leron",
""
],
[
"Duff",
"Michael J.",
""
],
[
"Kanakaris",
"Dimitri",
""
],
[
"Kim",
"Hyungrok",
""
]
] |
2504.16008 | Hang Ren | Hang Ren, Yipei Zhang, Wendy M. Billings, Rebecca Tomann, Nikolay V.
Tkachenko, Martin Head-Gordon, K. Birgitta Whaley | An Error Mitigated Non-Orthogonal Quantum Eigensolver via Shadow
Tomography | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a shadow-tomography-enhanced Non-Orthogonal Quantum Eigensolver
(NOQE) for more efficient and accurate electronic structure calculations on
near-term quantum devices. By integrating shadow tomography into the NOQE, the
measurement cost scales linearly rather than quadratically with the number of
reference states, while also reducing the required qubits and circuit depth by
half. This approach enables extraction of all matrix elements via randomized
measurements and classical postprocessing. We analyze its sample complexity and
show that, for small systems, it remains constant in the high-precision regime,
while for larger systems, it scales linearly with the system size. We further
apply shadow-based error mitigation to suppress noise-induced bias without
increasing quantum resources. Demonstrations on the hydrogen molecule in the
strongly correlated regime achieve chemical accuracy under realistic noise,
showing that our method is both resource-efficient and noise-resilient for
practical quantum chemistry simulations in the near term.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:16:09 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 16:39:28 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ren",
"Hang",
""
],
[
"Zhang",
"Yipei",
""
],
[
"Billings",
"Wendy M.",
""
],
[
"Tomann",
"Rebecca",
""
],
[
"Tkachenko",
"Nikolay V.",
""
],
[
"Head-Gordon",
"Martin",
""
],
[
"Whaley",
"K. Birgitta",
""
]
] |
2504.16020 | Soham Sane | Soham Sane | AlphaGrad: Non-Linear Gradient Normalization Optimizer | null | null | null | null | cs.LG cs.AI cs.NE stat.ML | http://creativecommons.org/licenses/by/4.0/ | We introduce AlphaGrad, a memory-efficient, conditionally stateless optimizer
addressing the memory overhead and hyperparameter complexity of adaptive
methods like Adam. AlphaGrad enforces scale invariance via tensor-wise L2
gradient normalization followed by a smooth hyperbolic tangent transformation,
$g' = \tanh(\alpha \cdot \tilde{g})$, controlled by a single steepness
parameter $\alpha$. Our contributions include: (1) the AlphaGrad algorithm
formulation; (2) a formal non-convex convergence analysis guaranteeing
stationarity; (3) extensive empirical evaluation on diverse RL benchmarks (DQN,
TD3, PPO). Compared to Adam, AlphaGrad demonstrates a highly context-dependent
performance profile. While exhibiting instability in off-policy DQN, it
provides enhanced training stability with competitive results in TD3 (requiring
careful $\alpha$ tuning) and achieves substantially superior performance in
on-policy PPO. These results underscore the critical importance of empirical
$\alpha$ selection, revealing strong interactions between the optimizer's
dynamics and the underlying RL algorithm. AlphaGrad presents a compelling
alternative optimizer for memory-constrained scenarios and shows significant
promise for on-policy learning regimes where its stability and efficiency
advantages can be particularly impactful.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 16:33:14 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 01:25:32 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Sane",
"Soham",
""
]
] |
2504.16046 | Jingyu Zhang | Jingyu Zhang, Jiacan Yu, Marc Marone, Benjamin Van Durme, Daniel
Khashabi | Certified Mitigation of Worst-Case LLM Copyright Infringement | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The exposure of large language models (LLMs) to copyrighted material during
pre-training raises concerns about unintentional copyright infringement post
deployment. This has driven the development of "copyright takedown" methods,
post-training approaches aimed at preventing models from generating content
substantially similar to copyrighted ones. While current mitigation approaches
are somewhat effective for average-case risks, we demonstrate that they
overlook worst-case copyright risks exhibits by the existence of long, verbatim
quotes from copyrighted sources. We propose BloomScrub, a remarkably simple yet
highly effective inference-time approach that provides certified copyright
takedown. Our method repeatedly interleaves quote detection with rewriting
techniques to transform potentially infringing segments. By leveraging
efficient data sketches (Bloom filters), our approach enables scalable
copyright screening even for large-scale real-world corpora. When quotes beyond
a length threshold cannot be removed, the system can abstain from responding,
offering certified risk reduction. Experimental results show that BloomScrub
reduces infringement risk, preserves utility, and accommodates different levels
of enforcement stringency with adaptive abstention. Our results suggest that
lightweight, inference-time methods can be surprisingly effective for copyright
prevention.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:16:53 GMT"
},
{
"version": "v2",
"created": "Wed, 23 Apr 2025 16:13:08 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zhang",
"Jingyu",
""
],
[
"Yu",
"Jiacan",
""
],
[
"Marone",
"Marc",
""
],
[
"Van Durme",
"Benjamin",
""
],
[
"Khashabi",
"Daniel",
""
]
] |
2504.16077 | Yuanpeng Qu | Yuanpeng Qu and Hajime Nobuhara | Intent-aware Diffusion with Contrastive Learning for Sequential
Recommendation | Accepted at SIGIR 2025. 10 pages, 6 figures, 3 tables | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Contrastive learning has proven effective in training sequential
recommendation models by incorporating self-supervised signals from augmented
views. Most existing methods generate multiple views from the same interaction
sequence through stochastic data augmentation, aiming to align their
representations in the embedding space. However, users typically have specific
intents when purchasing items (e.g., buying clothes as gifts or cosmetics for
beauty). Random data augmentation used in existing methods may introduce noise,
disrupting the latent intent information implicit in the original interaction
sequence. Moreover, using noisy augmented sequences in contrastive learning may
mislead the model to focus on irrelevant features, distorting the embedding
space and failing to capture users' true behavior patterns and intents. To
address these issues, we propose Intent-aware Diffusion with contrastive
learning for sequential Recommendation (InDiRec). The core idea is to generate
item sequences aligned with users' purchasing intents, thus providing more
reliable augmented views for contrastive learning. Specifically, InDiRec first
performs intent clustering on sequence representations using K-means to build
intent-guided signals. Next, it retrieves the intent representation of the
target interaction sequence to guide a conditional diffusion model, generating
positive views that share the same underlying intent. Finally, contrastive
learning is applied to maximize representation consistency between these
intent-aligned views and the original sequence. Extensive experiments on five
public datasets demonstrate that InDiRec achieves superior performance compared
to existing baselines, learning more robust representations even under noisy
and sparse data conditions.
| [
{
"version": "v1",
"created": "Tue, 22 Apr 2025 17:55:56 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Qu",
"Yuanpeng",
""
],
[
"Nobuhara",
"Hajime",
""
]
] |
2504.16085 | Yun-Cheng Tsai | Yun-Cheng Tsai | Blockchain-Driven Solutions for Carbon Credit Trading: A Decentralized
Platform for SMEs | 9 pages, 3 figures | null | null | null | cs.CR cs.CY | http://creativecommons.org/licenses/by/4.0/ | The increasing demand for sustainability and compliance with global carbon
regulations has posed significant challenges for small and medium-sized
enterprises (SMEs). This paper proposes a blockchain-based decentralized carbon
credit trading platform tailored for SMEs in Taiwan, aiming to simplify the
complex carbon trading process and lower market entry barriers. Drawing upon
the Diffusion of Innovations theory and transaction cost economics, we
illustrate how blockchain technology can reduce informational asymmetry and
intermediary costs in carbon markets. By integrating Ethereum-based smart
contracts, the platform automates transactions, enhances transparency, and
reduces administrative burdens - addressing key obstacles such as technical
complexity and market risks. A controlled experimental design was conducted to
compare the proposed system with a conventional centralized carbon trading
platform. Statistical analysis confirms its effectiveness in minimizing time
and expenses while ensuring compliance with the Carbon Border Adjustment
Mechanism (CBAM) and the Clean Competition Act (CCA). User satisfaction was
measured using the Kano model, with the results identifying essential features
and prioritizing future enhancements. This study contributes a more
comprehensive solution for SMEs seeking to achieve carbon neutrality,
underscoring the transformative potential of blockchain technology in global
carbon markets.
| [
{
"version": "v1",
"created": "Fri, 27 Dec 2024 13:28:54 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Tsai",
"Yun-Cheng",
""
]
] |
2504.16086 | Guanzhou Ji | Guanzhou Ji, Azadeh O. Sawyer, Srinivasa G. Narasimhan | Digital Kitchen Remodeling: Editing and Relighting Intricate Indoor
Scenes from a Single Panorama | Submitted to IES25 - The Lighting Conference, Anaheim, California,
August 21 - 23, 2025 | null | null | null | cs.GR cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel virtual staging application for kitchen remodeling from a
single panorama. To ensure the realism of the virtual rendered scene, we
capture real-world High Dynamic Range (HDR) panoramas and recover the absolute
scene radiance for high-quality scene relighting. Our application pipeline
consists of three key components: (1) HDR photography for capturing paired
indoor and outdoor panoramas, (2) automatic kitchen layout generation with new
kitchen components, and (3) an editable rendering pipeline that flexibly edits
scene materials and relights the new virtual scene with global illumination.
Additionally, we contribute a novel Pano-Pano HDR dataset with 141 paired
indoor and outdoor panoramas and present a low-cost photometric calibration
method for panoramic HDR photography.
| [
{
"version": "v1",
"created": "Wed, 5 Feb 2025 02:44:58 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ji",
"Guanzhou",
""
],
[
"Sawyer",
"Azadeh O.",
""
],
[
"Narasimhan",
"Srinivasa G.",
""
]
] |
2504.16087 | Lukas Daniel Klausner | Eva-Maria Maier, Leonie Maria Tanczer, Lukas Daniel Klausner | Surveillance Disguised as Protection: A Comparative Analysis of
Sideloaded and In-Store Parental Control Apps | 18 pages, 2 figures | Proc. Priv. Enh. Technol. 2025 (2), 2025, 107-124 | 10.56553/popets-2025-0052 | null | cs.CR cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Parental control applications, software tools designed to manage and monitor
children's online activities, serve as essential safeguards for parents in the
digital age. However, their usage has sparked concerns about security and
privacy violations inherent in various child monitoring products. Sideloaded
software (i. e. apps installed outside official app stores) poses an increased
risk, as it is not bound by the regulations of trusted platforms. Despite this,
the market of sideloaded parental control software has remained widely
unexplored by the research community. This paper examines 20 sideloaded
parental control apps and compares them to 20 apps available on the Google Play
Store. We base our analysis on privacy policies, Android package kit (APK)
files, application behaviour, network traffic and application functionalities.
Our findings reveal that sideloaded parental control apps fall short compared
to their in-store counterparts, lacking specialised parental control features
and safeguards against misuse while concealing themselves on the user's device.
Alarmingly, three apps transmitted sensitive data unencrypted, half lacked a
privacy policy and 8 out of 20 were flagged for potential stalkerware
indicators of compromise (IOC).
| [
{
"version": "v1",
"created": "Sat, 8 Mar 2025 00:22:20 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Maier",
"Eva-Maria",
""
],
[
"Tanczer",
"Leonie Maria",
""
],
[
"Klausner",
"Lukas Daniel",
""
]
] |
2504.16088 | William Boebert | William Earl Boebert | Paths Not Taken: A Secure Computing Tutorial | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | This paper is a tutorial on the proven but currently under-appreciated
security mechanisms associated with "tagged" or "descriptor" architectures. The
tutorial shows how the principles behind such architectures can be applied to
mitigate or eliminate vulnerabilities. The tutorial incorporates systems
engineering practices by presenting the mechanisms in an informal model of an
integrated artifact in its operational environment. The artifact is a
special-purpose hardware/software system called a "Guard" which robustly hosts
defensive software. It is hoped that this tutorial may encourage teachers to
include significant past work in their curricula and students who are
self-teaching to add that work to their exploration of secure computing.
| [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 18:17:47 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Boebert",
"William Earl",
""
]
] |
2504.16089 | Hina Binte Haq | Hina Binte Haq, Syed Taha Ali, Asad Salman, Patrick McCorry and Siamak
F. Shahandashti | Carbyne: An Ultra-Lightweight DoS-Resilient Mempool for Bitcoin | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The increasing adoption of cryptocurrencies has significantly amplified the
resource requirements for operating full nodes, creating substantial barriers
to entry. Unlike miners, who are financially incentivized through block rewards
and transaction fees, full nodes lack direct economic compensation for their
critical role in maintaining the network. A key resource burden is the
transaction pool, which is particularly memory-intensive as it temporarily
stores unconfirmed transactions awaiting verification and propagation across
the network. We present Neonpool, a novel optimization for transaction pool
leveraging bloom filter variants to drastically reduce memory consumption by up
to 200 (e.g., 400 MB to 2 MB) while maintaining over 99.99% transaction
processing accuracy. Implemented in C++ and evaluated on unique Bitcoin and
Ethereum datasets, Neonpool enables efficient operation on lightweight clients,
such as smartphones, IoT devices, and systems-on-a-chip, without requiring a
hard fork. By lowering the cost of node participation, Neonpool enhances
decentralization and strengthens the overall security and robustness of
cryptocurrency networks.
| [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 11:03:13 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Haq",
"Hina Binte",
""
],
[
"Ali",
"Syed Taha",
""
],
[
"Salman",
"Asad",
""
],
[
"McCorry",
"Patrick",
""
],
[
"Shahandashti",
"Siamak F.",
""
]
] |
2504.16090 | Yue Yu | Yue Yu, Xinyue Hu, Sivaraman Rajaganapathy, Jingna Feng, Ahmed
Abdelhameed, Xiaodi Li, Jianfu Li, Ken Liu, Liu Yang, Nilufer Taner, Phil
Fiero, Soulmaz Boroumand, Richard Larsen, Maneesh Goyal, Clark Otley, Nansu
Zong, John Halamka, Cui Tao | Launching Insights: A Pilot Study on Leveraging Real-World Observational
Data from the Mayo Clinic Platform to Advance Clinical Research | 12 pages, 3 figures, 2 tables | null | null | null | cs.CY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Backgrounds: Artificial intelligence (AI) is transforming healthcare, yet
translating AI models from theoretical frameworks to real-world clinical
applications remains challenging. The Mayo Clinic Platform (MCP) was
established to address these challenges by providing a scalable ecosystem that
integrates real-world multiple modalities data from multiple institutions,
advanced analytical tools, and secure computing environments to support
clinical research and AI development. Methods: In this study, we conducted four
research projects leveraging MCP's data infrastructure and analytical
capabilities to demonstrate its potential in facilitating real-world evidence
generation and AI-driven clinical insights. Utilizing MCP's tools and
environment, we facilitated efficient cohort identification, data extraction,
and subsequent statistical or AI-powered analyses. Results: The results
underscore MCP's role in accelerating translational research by offering
de-identified, standardized real-world data and facilitating AI model
validation across diverse healthcare settings. Compared to Mayo's internal
Electronic Health Record (EHR) data, MCP provides broader accessibility,
enhanced data standardization, and multi-institutional integration, making it a
valuable resource for both internal and external researchers. Conclusion:
Looking ahead, MCP is well-positioned to transform clinical research through
its scalable ecosystem, effectively bridging the divide between AI innovation
and clinical deployment. Future investigations will build upon this foundation,
further exploring MCP's capacity to advance precision medicine and enhance
patient outcomes.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 16:06:21 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Yu",
"Yue",
""
],
[
"Hu",
"Xinyue",
""
],
[
"Rajaganapathy",
"Sivaraman",
""
],
[
"Feng",
"Jingna",
""
],
[
"Abdelhameed",
"Ahmed",
""
],
[
"Li",
"Xiaodi",
""
],
[
"Li",
"Jianfu",
""
],
[
"Liu",
"Ken",
""
],
[
"Yang",
"Liu",
""
],
[
"Taner",
"Nilufer",
""
],
[
"Fiero",
"Phil",
""
],
[
"Boroumand",
"Soulmaz",
""
],
[
"Larsen",
"Richard",
""
],
[
"Goyal",
"Maneesh",
""
],
[
"Otley",
"Clark",
""
],
[
"Zong",
"Nansu",
""
],
[
"Halamka",
"John",
""
],
[
"Tao",
"Cui",
""
]
] |
2504.16091 | Siddhartha Bhoi | Siddhartha Siddhiprada Bhoi, Arathi Arakala, Amy Beth Corman, and Asha
Rao | Post-Quantum Homomorphic Encryption: A Case for Code-Based Alternatives | null | null | null | null | cs.CR math.HO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Homomorphic Encryption (HE) allows secure and privacy-protected computation
on encrypted data without the need to decrypt it. Since Shor's algorithm
rendered prime factorisation and discrete logarithm-based ciphers insecure with
quantum computations, researchers have been working on building post-quantum
homomorphic encryption (PQHE) algorithms. Most of the current PQHE algorithms
are secured by Lattice-based problems and there have been limited attempts to
build ciphers based on error-correcting code-based problems. This review
presents an overview of the current approaches to building PQHE schemes and
justifies code-based encryption as a novel way to diversify post-quantum
algorithms. We present the mathematical underpinnings of existing code-based
cryptographic frameworks and their security and efficiency guarantees. We
compare lattice-based and code-based homomorphic encryption solutions
identifying challenges that have inhibited the progress of code-based schemes.
We finally propose five new research directions to advance post-quantum
code-based homomorphic encryption.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 06:49:22 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Bhoi",
"Siddhartha Siddhiprada",
""
],
[
"Arakala",
"Arathi",
""
],
[
"Corman",
"Amy Beth",
""
],
[
"Rao",
"Asha",
""
]
] |
2504.16092 | Mahrad Almotahari | Mahrad Almotahari | Cooperative Speech, Semantic Competence, and AI | 25 pages | null | null | null | cs.CY cs.CL cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Cooperative speech is purposive. From the speaker's perspective, one crucial
purpose is the transmission of knowledge. Cooperative speakers care about
getting things right for their conversational partners. This attitude is a kind
of respect. Cooperative speech is an ideal form of communication because
participants have respect for each other. And having respect within a
cooperative enterprise is sufficient for a particular kind of moral standing:
we ought to respect those who have respect for us. Respect demands reciprocity.
I maintain that large language models aren't owed the kind of respect that
partly constitutes a cooperative conversation. This implies that they aren't
cooperative interlocutors, otherwise we would be obliged to reciprocate the
attitude. Leveraging this conclusion, I argue that present-day LLMs are
incapable of assertion and that this raises an overlooked doubt about their
semantic competence. One upshot of this argument is that knowledge of meaning
isn't just a subject for the cognitive psychologist. It's also a subject for
the moral psychologist.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 20:34:41 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Almotahari",
"Mahrad",
""
]
] |
2504.16093 | Tom Chou | Yurun Ge, Lucas B\"ottcher, Tom Chou, Maria R. D'Orsogna | Efficient Portfolio Selection through Preference Aggregation with
Quicksort and the Bradley--Terry Model | 15pp, 4 figs | null | null | null | q-fin.PM cs.AI math.PR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | How to allocate limited resources to projects that will yield the greatest
long-term benefits is a problem that often arises in decision-making under
uncertainty. For example, organizations may need to evaluate and select
innovation projects with risky returns. Similarly, when allocating resources to
research projects, funding agencies are tasked with identifying the most
promising proposals based on idiosyncratic criteria. Finally, in participatory
budgeting, a local community may need to select a subset of public projects to
fund. Regardless of context, agents must estimate the uncertain values of a
potentially large number of projects. Developing parsimonious methods to
compare these projects, and aggregating agent evaluations so that the overall
benefit is maximized, are critical in assembling the best project portfolio.
Unlike in standard sorting algorithms, evaluating projects on the basis of
uncertain long-term benefits introduces additional complexities. We propose
comparison rules based on Quicksort and the Bradley--Terry model, which
connects rankings to pairwise "win" probabilities. In our model, each agent
determines win probabilities of a pair of projects based on his or her specific
evaluation of the projects' long-term benefit. The win probabilities are then
appropriately aggregated and used to rank projects. Several of the methods we
propose perform better than the two most effective aggregation methods
currently available. Additionally, our methods can be combined with sampling
techniques to significantly reduce the number of pairwise comparisons. We also
discuss how the Bradley--Terry portfolio selection approach can be implemented
in practice.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 23:16:30 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ge",
"Yurun",
""
],
[
"Böttcher",
"Lucas",
""
],
[
"Chou",
"Tom",
""
],
[
"D'Orsogna",
"Maria R.",
""
]
] |
2504.16094 | Jingzhou Shen | Jingzhou Shen, Tianya Zhao, Yanzhao Wu, Xuyu Wang | NeRF-APT: A New NeRF Framework for Wireless Channel Prediction | Accepted by IEEE INFOCOM WKSHPS: DeepWireless 2025: Deep Learning for
Wireless Communications, Sensing, and Security, to appear. 6 pages, 6
figures, 2 tables | null | null | null | cs.NI | http://creativecommons.org/licenses/by/4.0/ | Neural radiance fields (NeRFs) have recently attracted significant attention
in the field of wireless channel prediction, primarily due to their capability
for high-fidelity reconstruction of complex wireless measurement environments.
However, the ray-tracing component of NeRF-based methods faces challenges in
realistically representing wireless scenarios, mainly due to the limited
expressive power of multilayer perceptrons (MLPs). To overcome this issue, in
this paper, we propose NeRF-APT, an encoder-decoder architecture integrated
within a NeRF-based wireless channel prediction framework. Our architecture
leverages the strengths of NeRF-like models in learning environmental features
and exploits encoder-decoder modules' capabilities for critical information
extraction. Additionally, we incorporate an attention mechanism within the skip
connections between encoder and decoder layers, significantly enhancing
contextual understanding across layers. Extensive experimental evaluations
conducted on several realistic and synthetic datasets demonstrate that our
proposed method outperforms existing state-of-the-art approaches in wireless
channel prediction.
| [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 16:52:00 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Shen",
"Jingzhou",
""
],
[
"Zhao",
"Tianya",
""
],
[
"Wu",
"Yanzhao",
""
],
[
"Wang",
"Xuyu",
""
]
] |
2504.16095 | Jonathan Gl\"ockle | Jonathan Gl\"ockle | Initial data rigidity implies spacetime rigidity | Comments welcome! | null | null | null | math.DG gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article, we revisit the initial data rigidity theorem of Eichmair,
Galloway and Mendes (arxiv:2009.09527). The goal is to strengthen their result
by showing that the initial data sets concerned carry a vector field that is
lightlike and parallel in an ambient sense. This will be used in a second step
to show that among the spacetimes satisfying the dominant energy condition
there exists locally essentially one spacetime extending these initial data
sets. This local uniqueness theorem also applies in the context of other
initial data rigidity theorems. Notably, the one in the spin case due the
author (arxiv:2304.02331) and a recent study of the mass zero case in the
positive energy theorem due to Hirsch and Zhang (arxiv:2403.15984).
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:44:59 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Glöckle",
"Jonathan",
""
]
] |
2504.16096 | Jiaxing Xu | Jiaxing Xu, Kai He, Yue Tang, Wei Li, Mengcheng Lan, Xia Dong, Yiping
Ke and Mengling Feng | BrainPrompt: Multi-Level Brain Prompt Enhancement for Neurological
Condition Identification | null | null | null | null | q-bio.NC cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Neurological conditions, such as Alzheimer's Disease, are challenging to
diagnose, particularly in the early stages where symptoms closely resemble
healthy controls. Existing brain network analysis methods primarily focus on
graph-based models that rely solely on imaging data, which may overlook
important non-imaging factors and limit the model's predictive power and
interpretability. In this paper, we present BrainPrompt, an innovative
framework that enhances Graph Neural Networks (GNNs) by integrating Large
Language Models (LLMs) with knowledge-driven prompts, enabling more effective
capture of complex, non-imaging information and external knowledge for
neurological disease identification. BrainPrompt integrates three types of
knowledge-driven prompts: (1) ROI-level prompts to encode the identity and
function of each brain region, (2) subject-level prompts that incorporate
demographic information, and (3) disease-level prompts to capture the temporal
progression of disease. By leveraging these multi-level prompts, BrainPrompt
effectively harnesses knowledge-enhanced multi-modal information from LLMs,
enhancing the model's capability to predict neurological disease stages and
meanwhile offers more interpretable results. We evaluate BrainPrompt on two
resting-state functional Magnetic Resonance Imaging (fMRI) datasets from
neurological disorders, showing its superiority over state-of-the-art methods.
Additionally, a biomarker study demonstrates the framework's ability to extract
valuable and interpretable information aligned with domain knowledge in
neuroscience.
| [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 06:45:16 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Xu",
"Jiaxing",
""
],
[
"He",
"Kai",
""
],
[
"Tang",
"Yue",
""
],
[
"Li",
"Wei",
""
],
[
"Lan",
"Mengcheng",
""
],
[
"Dong",
"Xia",
""
],
[
"Ke",
"Yiping",
""
],
[
"Feng",
"Mengling",
""
]
] |
2504.16097 | Pedro Bento | Arthur Buzelin, Pedro Robles Dutenhefner, Turi Rezende, Luisa G.
Porfirio, Pedro Bento, Yan Aquino, Jose Fernandes, Caio Santana, Gabriela
Miana, Gisele L. Pappa, Antonio Ribeiro, Wagner Meira Jr | A CNN-based Local-Global Self-Attention via Averaged Window Embeddings
for Hierarchical ECG Analysis | null | null | null | null | eess.SP cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Cardiovascular diseases remain the leading cause of global mortality,
emphasizing the critical need for efficient diagnostic tools such as
electrocardiograms (ECGs). Recent advancements in deep learning, particularly
transformers, have revolutionized ECG analysis by capturing detailed waveform
features as well as global rhythm patterns. However, traditional transformers
struggle to effectively capture local morphological features that are critical
for accurate ECG interpretation. We propose a novel Local-Global Attention ECG
model (LGA-ECG) to address this limitation, integrating convolutional inductive
biases with global self-attention mechanisms. Our approach extracts queries by
averaging embeddings obtained from overlapping convolutional windows, enabling
fine-grained morphological analysis, while simultaneously modeling global
context through attention to keys and values derived from the entire sequence.
Experiments conducted on the CODE-15 dataset demonstrate that LGA-ECG
outperforms state-of-the-art models and ablation studies validate the
effectiveness of the local-global attention strategy. By capturing the
hierarchical temporal dependencies and morphological patterns in ECG signals,
this new design showcases its potential for clinical deployment with robust
automated ECG classification.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 01:21:18 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Buzelin",
"Arthur",
""
],
[
"Dutenhefner",
"Pedro Robles",
""
],
[
"Rezende",
"Turi",
""
],
[
"Porfirio",
"Luisa G.",
""
],
[
"Bento",
"Pedro",
""
],
[
"Aquino",
"Yan",
""
],
[
"Fernandes",
"Jose",
""
],
[
"Santana",
"Caio",
""
],
[
"Miana",
"Gabriela",
""
],
[
"Pappa",
"Gisele L.",
""
],
[
"Ribeiro",
"Antonio",
""
],
[
"Meira",
"Wagner",
"Jr"
]
] |
2504.16099 | Luyuan Zhang | Luyuan Zhang, Xidong Mu, An Liu and Yuanwei Liu | Two-Timescale Joint Transmit and Pinching Beamforming for
Pinching-Antenna Systems | 5 pages, 4 figures, letter | null | null | null | eess.SP cs.AI cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pinching antenna systems (PASS) have been proposed as a revolutionary
flexible antenna technology which facilitates line-of-sight links via numerous
low-cost pinching antennas with adjustable activation positions over
waveguides. This letter proposes a two-timescale joint transmit and pinching
beamforming design for the maximization of sum rate of a PASS-based downlink
multi-user multiple input single output system. A primal dual decomposition
method is developed to decouple the two-timescale problem into two
sub-problems: 1) A Karush-Kuhn-Tucker-guided dual learning-based approach is
proposed to solve the short-term transmit beamforming design sub-problem; 2)
The long-term pinching beamforming design sub-problem is tackled by adopting a
stochastic successive convex approximation method. Simulation results
demonstrate that the proposed two-timescale algorithm achieves a significant
performance gain compared to other baselines.
| [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 16:58:35 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zhang",
"Luyuan",
""
],
[
"Mu",
"Xidong",
""
],
[
"Liu",
"An",
""
],
[
"Liu",
"Yuanwei",
""
]
] |
2504.16100 | Eloi Lindas | Eloi Lindas, Yannig Goude, Philippe Ciais | Towards Accurate Forecasting of Renewable Energy : Building Datasets and
Benchmarking Machine Learning Models for Solar and Wind Power in France | 24 pages, 4 tables, 18 figures | null | null | null | eess.SP cs.AI cs.LG stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Accurate prediction of non-dispatchable renewable energy sources is essential
for grid stability and price prediction. Regional power supply forecasts are
usually indirect through a bottom-up approach of plant-level forecasts,
incorporate lagged power values, and do not use the potential of spatially
resolved data. This study presents a comprehensive methodology for predicting
solar and wind power production at country scale in France using machine
learning models trained with spatially explicit weather data combined with
spatial information about production sites capacity. A dataset is built
spanning from 2012 to 2023, using daily power production data from RTE (the
national grid operator) as the target variable, with daily weather data from
ERA5, production sites capacity and location, and electricity prices as input
features. Three modeling approaches are explored to handle spatially resolved
weather data: spatial averaging over the country, dimension reduction through
principal component analysis, and a computer vision architecture to exploit
complex spatial relationships. The study benchmarks state-of-the-art machine
learning models as well as hyperparameter tuning approaches based on
cross-validation methods on daily power production data. Results indicate that
cross-validation tailored to time series is best suited to reach low error. We
found that neural networks tend to outperform traditional tree-based models,
which face challenges in extrapolation due to the increasing renewable capacity
over time. Model performance ranges from 4% to 10% in nRMSE for midterm
horizon, achieving similar error metrics to local models established at a
single-plant level, highlighting the potential of these methods for regional
power supply forecasting.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 15:30:54 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Lindas",
"Eloi",
""
],
[
"Goude",
"Yannig",
""
],
[
"Ciais",
"Philippe",
""
]
] |
2504.16101 | Lei Kang | Lei Kang, Xuanshuo Fu, Javier Vazquez-Corral, Ernest Valveny,
Dimosthenis Karatzas | xLSTM-ECG: Multi-label ECG Classification via Feature Fusion with xLSTM | null | null | null | null | eess.SP cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Cardiovascular diseases (CVDs) remain the leading cause of mortality
worldwide, highlighting the critical need for efficient and accurate diagnostic
tools. Electrocardiograms (ECGs) are indispensable in diagnosing various heart
conditions; however, their manual interpretation is time-consuming and
error-prone. In this paper, we propose xLSTM-ECG, a novel approach that
leverages an extended Long Short-Term Memory (xLSTM) network for multi-label
classification of ECG signals, using the PTB-XL dataset. To the best of our
knowledge, this work represents the first design and application of xLSTM
modules specifically adapted for multi-label ECG classification. Our method
employs a Short-Time Fourier Transform (STFT) to convert time-series ECG
waveforms into the frequency domain, thereby enhancing feature extraction. The
xLSTM architecture is specifically tailored to address the complexities of
12-lead ECG recordings by capturing both local and global signal features.
Comprehensive experiments on the PTB-XL dataset reveal that our model achieves
strong multi-label classification performance, while additional tests on the
Georgia 12-Lead dataset underscore its robustness and efficiency. This approach
significantly improves ECG classification accuracy, thereby advancing clinical
diagnostics and patient care. The code will be publicly available upon
acceptance.
| [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 16:12:46 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Kang",
"Lei",
""
],
[
"Fu",
"Xuanshuo",
""
],
[
"Vazquez-Corral",
"Javier",
""
],
[
"Valveny",
"Ernest",
""
],
[
"Karatzas",
"Dimosthenis",
""
]
] |
2504.16102 | Xiwen Li | Xiwen Li, Ross Whitaker, Tolga Tasdizen | Audio and Multiscale Visual Cues Driven Cross-modal Transformer for
Idling Vehicle Detection | null | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Idling vehicle detection (IVD) supports real-time systems that reduce
pollution and emissions by dynamically messaging drivers to curb excess idling
behavior. In computer vision, IVD has become an emerging task that leverages
video from surveillance cameras and audio from remote microphones to localize
and classify vehicles in each frame as moving, idling, or engine-off. As with
other cross-modal tasks, the key challenge lies in modeling the correspondence
between audio and visual modalities, which differ in representation but provide
complementary cues -- video offers spatial and motion context, while audio
conveys engine activity beyond the visual field. The previous end-to-end model,
which uses a basic attention mechanism, struggles to align these modalities
effectively, often missing vehicle detections. To address this issue, we
propose AVIVDNetv2, a transformer-based end-to-end detection network. It
incorporates a cross-modal transformer with global patch-level learning, a
multiscale visual feature fusion module, and decoupled detection heads.
Extensive experiments show that AVIVDNetv2 improves mAP by 7.66 over the
disjoint baseline and 9.42 over the E2E baseline, with consistent AP gains
across all vehicle categories. Furthermore, AVIVDNetv2 outperforms the
state-of-the-art method for sounding object localization, establishing a new
performance benchmark on the AVIVD dataset.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 21:10:17 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Li",
"Xiwen",
""
],
[
"Whitaker",
"Ross",
""
],
[
"Tasdizen",
"Tolga",
""
]
] |
2504.16103 | Oussema Dhaouadi | Oussema Dhaouadi, Johannes Meier, Jacques Kaiser, Daniel Cremers | Shape Your Ground: Refining Road Surfaces Beyond Planar Representations | null | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Road surface reconstruction from aerial images is fundamental for autonomous
driving, urban planning, and virtual simulation, where smoothness, compactness,
and accuracy are critical quality factors. Existing reconstruction methods
often produce artifacts and inconsistencies that limit usability, while
downstream tasks have a tendency to represent roads as planes for simplicity
but at the cost of accuracy. We introduce FlexRoad, the first framework to
directly address road surface smoothing by fitting Non-Uniform Rational
B-Splines (NURBS) surfaces to 3D road points obtained from photogrammetric
reconstructions or geodata providers. Our method at its core utilizes the
Elevation-Constrained Spatial Road Clustering (ECSRC) algorithm for robust
anomaly correction, significantly reducing surface roughness and fitting
errors. To facilitate quantitative comparison between road surface
reconstruction methods, we present GeoRoad Dataset (GeRoD), a diverse
collection of road surface and terrain profiles derived from openly accessible
geodata. Experiments on GeRoD and the photogrammetry-based DeepScenario Open 3D
Dataset (DSC3D) demonstrate that FlexRoad considerably surpasses commonly used
road surface representations across various metrics while being insensitive to
various input sources, terrains, and noise types. By performing ablation
studies, we identify the key role of each component towards high-quality
reconstruction performance, making FlexRoad a generic method for realistic road
surface modeling.
| [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 21:20:44 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Dhaouadi",
"Oussema",
""
],
[
"Meier",
"Johannes",
""
],
[
"Kaiser",
"Jacques",
""
],
[
"Cremers",
"Daniel",
""
]
] |
2504.16104 | Aditya Gulati | Aditya Gulati, Moreno D'Inc\`a, Nicu Sebe, Bruno Lepri, Nuria Oliver | Uncovering an Attractiveness Bias in Multimodal Large Language Models: A
Case Study with LLaVA | null | null | null | null | cs.CY | http://creativecommons.org/licenses/by/4.0/ | Physical attractiveness matters. It has been shown to influence human
perception and decision-making, often leading to biased judgments that favor
those deemed attractive in what is referred to as "the attractiveness halo
effect". While extensively studied in human judgments in a broad set of
domains, including hiring, judicial sentencing or credit granting, the role
that attractiveness plays in the assessments and decisions made by multimodal
large language models (MLLMs) is unknown. To address this gap, we conduct an
empirical study using 91 socially relevant scenarios and a diverse dataset of
924 face images, corresponding to 462 individuals both with and without beauty
filters applied to them, evaluated on LLaVA, a state-of-the-art, open source
MLLM. Our analysis reveals that attractiveness impacts the decisions made by
the MLLM in over 80% of the scenarios, demonstrating substantial bias in model
behavior in what we refer to as an attractiveness bias. Similarly to humans, we
find empirical evidence of the existence of the attractiveness halo effect,
such that more attractive individuals are more likely to be attributed positive
traits, such as intelligence or confidence, by the MLLM. Furthermore, we
uncover a gender, age and race bias in 83%, 73% and 57% of the scenarios,
respectively, which is impacted by attractiveness, particularly in the case of
gender, highlighting the intersectional nature of the attractiveness bias. Our
findings suggest that societal stereotypes and cultural norms intersect with
perceptions of attractiveness, amplifying or mitigating this bias in multimodal
generative AI models in a complex way. Our work emphasizes the need to account
for intersectionality in algorithmic bias detection and mitigation efforts and
underscores the challenges of addressing bias in modern multimodal large
language models.
| [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 16:02:55 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Gulati",
"Aditya",
""
],
[
"D'Incà",
"Moreno",
""
],
[
"Sebe",
"Nicu",
""
],
[
"Lepri",
"Bruno",
""
],
[
"Oliver",
"Nuria",
""
]
] |
2504.16105 | Duvier Suarez Fontanella | D. Su\'arez-Fontanella, A. Cabo Montes de Oca | About the Upstream Contamination | 7 pages, 5 figures. This article has been published in the Revista
Cubana de F\'isica | Revista Cubana de F\'isica, 36(1) 2019 | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This study investigates the upward transport of waterborne pollutants from a
lower container to an upper container through vertically falling water streams.
While previous analyses have primarily focused on inclined channels, we extend
the theoretical framework to consider vertical configurations. Two distinct
cases are examined: (i) a vertical flow within a cylindrical tube, and (ii) a
free-falling water jet. For the first case, we derive an analytical expression
for the critical water flux required to prevent the upward migration of
particles. In the second case, we establish a relationship among water flux,
particle size, and vertical position along the stream, which determines the
feasibility of upward particle transport. Our findings reveal a fundamental
difference between the two configurations. In the tubular flow case, surface
tension has negligible influence on particle motion. In contrast, in the
free-fall scenario, upward particle transport is only possible in the presence
of surface tension. Moreover, we demonstrate that for any given water flux,
there exists a threshold height difference beyond which contamination of the
upper container is not possible. Increasing the water flux further inhibits any
upward transport of pollutants
| [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 17:25:48 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Suárez-Fontanella",
"D.",
""
],
[
"de Oca",
"A. Cabo Montes",
""
]
] |
2504.16106 | Marc-Emmanuel Coupvent Des Graviers | Marc-Emmanuel Coupvent des Graviers (1), Lotfi Kobrosly (1 and 2),
Christophe Guettier (1), Tristan Cazenave (2) ((1) Safran Electronics and
Defense, France, (2) LAMSADE, Universit\'e Paris Dauphine-PSL, Place du
Mar\'echal de Lattre de Tassigny, Paris, France) | Updating Lower and Upper Bounds for the Job-Shop Scheduling Problem Test
Instances | null | null | null | null | cs.DS cs.DM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The Job-Shop Scheduling Problem (JSSP) and its variant, the Flexible Job-Shop
Scheduling Problem (FJSSP), are combinatorial optimization problems studied
thoroughly in the literature. Generally, the aim is to reduce the makespan of a
scheduling solution corresponding to a problem instance. Thus, finding upper
and lower bounds for an optimal makespan enables the assessment of performances
for multiple approaches addressed so far. We use OR-Tools, a solver portfolio,
to compute new bounds for some open benchmark instances, in order to reduce the
gap between upper and lower bounds. We find new numerical lower bounds for
multiple benchmark instances, up to closing the Taillard's ta33 instance. We
also improve upper bounds for four instances, namely Taillard's ta26 & ta45 and
Dauzere's 05a & 06a. Additionally we share an optimal solution for Taillard's
ta45 as well as Hurink-edata's car5.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 10:09:19 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Graviers",
"Marc-Emmanuel Coupvent des",
"",
"1 and 2"
],
[
"Kobrosly",
"Lotfi",
"",
"1 and 2"
],
[
"Guettier",
"Christophe",
""
],
[
"Cazenave",
"Tristan",
""
]
] |
2504.16107 | Shiyuan Li | Shiyuan Li, Yuyue Zhou, Chi Zhang, Liang Kong, Kebin Liu, Yihan Xie,
Chong He | Phased Array Calibration based on Rotating-Element Harmonic
Electric-Field Vector with Time Modulation | null | null | null | null | eess.SP cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Calibration is crucial for ensuring the performance of phased array since
amplitude-phase imbalance between elements results in significant performance
degradation. While amplitude-only calibration methods offer advantages when
phase measurements are impractical, conventional approaches face two key
challenges: they typically require high-resolution phase shifters and remain
susceptible to phase errors inherent in these components. To overcome these
limitations, we propose a Rotating element Harmonic Electric-field Vector
(RHEV) strategy, which enables precise calibration through time modulation
principles. The proposed technique functions as follows. Two 1-bit phase
shifters are periodically phase-switched at the same frequency, each generating
corresponding harmonics. By adjusting the relative delay between their
modulation timings, the phase difference between the $+1$st harmonics produced
by the two elements can be precisely controlled, utilizing the time-shift
property of the Fourier transform. Furthermore, the +1st harmonic generated by
sequential modulation of individual elements exhibits a linear relationship
with the amplitude of the modulated element, enabling amplitude ambiguity
resolution. The proposed RHEV-based calibration method generates phase shifts
through relative timing delays rather than physical phase shifter adjustments,
rendering it less susceptible to phase shift errors. Additionally, since the
calibration process exclusively utilizes the $+1$st harmonic, which is produced
solely by the modulated unit, the method demonstrates consistent performance
regardless of array size. Extensive numerical simulations, practical in-channel
and over-the-air (OTA) calibration experiments demonstrate the effectiveness
and distinct advantages of the proposed method.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 14:34:55 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Li",
"Shiyuan",
""
],
[
"Zhou",
"Yuyue",
""
],
[
"Zhang",
"Chi",
""
],
[
"Kong",
"Liang",
""
],
[
"Liu",
"Kebin",
""
],
[
"Xie",
"Yihan",
""
],
[
"He",
"Chong",
""
]
] |
2504.16108 | Sebastian Barros | Sebastian Barros | Trusted Identities for AI Agents: Leveraging Telco-Hosted eSIM
Infrastructure | 25 pages, 1 diagram, 2 tables | null | null | null | cs.CR cs.NI | http://creativecommons.org/licenses/by-sa/4.0/ | The rise of autonomous AI agents in enterprise and industrial environments
introduces a critical challenge: how to securely assign, verify, and manage
their identities across distributed systems. Existing identity frameworks based
on API keys, certificates, or application-layer credentials lack the
infrastructure-grade trust, lifecycle control, and interoperability needed to
manage agents operating independently in sensitive contexts.
In this paper, we propose a conceptual architecture that leverages
telecom-grade eSIM infrastructure, specifically hosted by mobile network
operators (MNOs), to serve as a root of trust for AI agents. Rather than
embedding SIM credentials in hardware devices, we envision a model where telcos
host secure, certified hardware modules (eUICC or HSM) that store and manage
agent-specific eSIM profiles. Agents authenticate remotely via cryptographic
APIs or identity gateways, enabling scalable and auditable access to enterprise
networks and services.
We explore use cases such as onboarding enterprise automation agents,
securing AI-driven financial systems, and enabling trust in inter-agent
communications. We identify current limitations in GSMA and 3GPP standards,
particularly their device centric assumptions, and propose extensions to
support non-physical, software-based agents within trusted execution
environments. This paper is intended as a conceptual framework to open
discussion around standardization, security architecture, and the role of
telecom infrastructure in the evolving agent economy.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 15:36:26 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Barros",
"Sebastian",
""
]
] |
2504.16109 | Jun-Peng Jiang | Jun-Peng Jiang, Si-Yang Liu, Hao-Run Cai, Qile Zhou, Han-Jia Ye | Representation Learning for Tabular Data: A Comprehensive Survey | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Tabular data, structured as rows and columns, is among the most prevalent
data types in machine learning classification and regression applications.
Models for learning from tabular data have continuously evolved, with Deep
Neural Networks (DNNs) recently demonstrating promising results through their
capability of representation learning. In this survey, we systematically
introduce the field of tabular representation learning, covering the
background, challenges, and benchmarks, along with the pros and cons of using
DNNs. We organize existing methods into three main categories according to
their generalization capabilities: specialized, transferable, and general
models. Specialized models focus on tasks where training and evaluation occur
within the same data distribution. We introduce a hierarchical taxonomy for
specialized models based on the key aspects of tabular data -- features,
samples, and objectives -- and delve into detailed strategies for obtaining
high-quality feature- and sample-level representations. Transferable models are
pre-trained on one or more datasets and subsequently fine-tuned on downstream
tasks, leveraging knowledge acquired from homogeneous or heterogeneous sources,
or even cross-modalities such as vision and language. General models, also
known as tabular foundation models, extend this concept further, allowing
direct application to downstream tasks without fine-tuning. We group these
general models based on the strategies used to adapt across heterogeneous
datasets. Additionally, we explore ensemble methods, which integrate the
strengths of multiple tabular models. Finally, we discuss representative
extensions of tabular learning, including open-environment tabular machine
learning, multimodal learning with tabular data, and tabular understanding.
More information can be found in the following repository:
https://github.com/LAMDA-Tabular/Tabular-Survey.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 17:58:23 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Jiang",
"Jun-Peng",
""
],
[
"Liu",
"Si-Yang",
""
],
[
"Cai",
"Hao-Run",
""
],
[
"Zhou",
"Qile",
""
],
[
"Ye",
"Han-Jia",
""
]
] |
2504.16110 | Krti Tallam | Krti Tallam | Security-First AI: Foundations for Robust and Trustworthy Systems | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | The conversation around artificial intelligence (AI) often focuses on safety,
transparency, accountability, alignment, and responsibility. However, AI
security (i.e., the safeguarding of data, models, and pipelines from
adversarial manipulation) underpins all of these efforts. This manuscript
posits that AI security must be prioritized as a foundational layer. We present
a hierarchical view of AI challenges, distinguishing security from safety, and
argue for a security-first approach to enable trustworthy and resilient AI
systems. We discuss core threat models, key attack vectors, and emerging
defense mechanisms, concluding that a metric-driven approach to AI security is
essential for robust AI safety, transparency, and accountability.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 22:53:01 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Tallam",
"Krti",
""
]
] |
2504.16111 | Simon Santschi | Peter Jipsen, Simon Santschi | Residuated lattices do not have the amalgamation property | 2 pages, 1 figure | null | null | null | math.RA math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that the variety of residuated lattices does not have the
amalgamation property.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 23:09:52 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Jipsen",
"Peter",
""
],
[
"Santschi",
"Simon",
""
]
] |
2504.16112 | Myunghyun Rhee | Myunghyun Rhee, Joonseop Sim, Taeyoung Ahn, Seungyong Lee, Daegun
Yoon, Euiseok Kim, Kyoung Park, Youngpyo Joo, Hosik Kim | HPU: High-Bandwidth Processing Unit for Scalable, Cost-effective LLM
Inference via GPU Co-processing | 6 pages | null | null | null | cs.AR cs.AI cs.CL cs.DC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The attention layer, a core component of Transformer-based LLMs, brings out
inefficiencies in current GPU systems due to its low operational intensity and
the substantial memory requirements of KV caches. We propose a High-bandwidth
Processing Unit (HPU), a memoryintensive co-processor that enhances GPU
resource utilization during large-batched LLM inference. By offloading
memory-bound operations, the HPU allows the GPU to focus on compute-intensive
tasks, increasing overall efficiency. Also, the HPU, as an add-on card, scales
out to accommodate surging memory demands driven by large batch sizes and
extended sequence lengths. In this paper, we show the HPU prototype implemented
with PCIe-based FPGA cards mounted on a GPU system. Our novel GPU-HPU
heterogeneous system demonstrates up to 4.1x performance gains and 4.6x energy
efficiency improvements over a GPUonly system, providing scalability without
increasing the number of GPUs.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 03:31:08 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Rhee",
"Myunghyun",
""
],
[
"Sim",
"Joonseop",
""
],
[
"Ahn",
"Taeyoung",
""
],
[
"Lee",
"Seungyong",
""
],
[
"Yoon",
"Daegun",
""
],
[
"Kim",
"Euiseok",
""
],
[
"Park",
"Kyoung",
""
],
[
"Joo",
"Youngpyo",
""
],
[
"Kim",
"Hosik",
""
]
] |
2504.16114 | Martin Uray | Johannes Ferner and Stefan Huber and Saverio Messineo and Angel Pop
and Martin Uray | Persistence-based Hough Transform for Line Detection | Accepted at iDSC'25, Salzburg, Austria | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | The Hough transform is a popular and classical technique in computer vision
for the detection of lines (or more general objects). It maps a pixel into a
dual space -- the Hough space: each pixel is mapped to the set of lines through
this pixel, which forms a curve in Hough space. The detection of lines then
becomes a voting process to find those lines that received many votes by
pixels. However, this voting is done by thresholding, which is susceptible to
noise and other artifacts.
In this work, we present an alternative voting technique to detect peaks in
the Hough space based on persistent homology, which very naturally addresses
limitations of simple thresholding. Experiments on synthetic data show that our
method significantly outperforms the original method, while also demonstrating
enhanced robustness.
This work seeks to inspire future research in two key directions. First, we
highlight the untapped potential of Topological Data Analysis techniques and
advocate for their broader integration into existing methods, including
well-established ones. Secondly, we initiate a discussion on the mathematical
stability of the Hough transform, encouraging exploration of mathematically
grounded improvements to enhance its robustness.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 10:00:08 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ferner",
"Johannes",
""
],
[
"Huber",
"Stefan",
""
],
[
"Messineo",
"Saverio",
""
],
[
"Pop",
"Angel",
""
],
[
"Uray",
"Martin",
""
]
] |
2504.16115 | Yibo Zhang | Yibo Jacky Zhang, Sanmi Koyejo | A Framework for Objective-Driven Dynamical Stochastic Fields | null | null | null | null | cs.AI cs.LG cs.MA nlin.AO | http://creativecommons.org/licenses/by/4.0/ | Fields offer a versatile approach for describing complex systems composed of
interacting and dynamic components. In particular, some of these dynamical and
stochastic systems may exhibit goal-directed behaviors aimed at achieving
specific objectives, which we refer to as $\textit{intelligent fields}$.
However, due to their inherent complexity, it remains challenging to develop a
formal theoretical description of such systems and to effectively translate
these descriptions into practical applications. In this paper, we propose three
fundamental principles -- complete configuration, locality, and purposefulness
-- to establish a theoretical framework for understanding intelligent fields.
Moreover, we explore methodologies for designing such fields from the
perspective of artificial intelligence applications. This initial investigation
aims to lay the groundwork for future theoretical developments and practical
advances in understanding and harnessing the potential of such objective-driven
dynamical stochastic fields.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 15:46:33 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zhang",
"Yibo Jacky",
""
],
[
"Koyejo",
"Sanmi",
""
]
] |
2504.16116 | Yihui Yang | Miracle Master, Rainy Sun, Anya Reese, Joey Ouyang, Alex Chen, Winter
Dong, Frank Li, James Yi, Garry Zhao, Tony Ling, Hobert Wong and Lowes Yang | DMind Benchmark: The First Comprehensive Benchmark for LLM Evaluation in
the Web3 Domain | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Recent advances in Large Language Models (LLMs) have led to significant
progress on a wide range of natural language processing tasks. However, their
effectiveness in specialized and rapidly evolving domains such as Web3 remains
underexplored. In this paper, we introduce DMind Benchmark, a novel framework
that systematically tests LLMs across nine key categories encompassing
blockchain fundamentals, infrastructure, smart contract analysis, decentralized
finance (DeFi), decentralized autonomous organizations (DAOs), non-fungible
tokens (NFTs), token economics, meme concepts, and security vulnerabilities.
DMind Benchmark goes beyond conventional multiple-choice questions by
incorporating domain-specific subjective tasks (e.g., smart contract code
auditing and repair, numeric reasoning on on-chain data, and fill-in
assessments), thereby capturing real-world complexities and stress-testing
model adaptability. We evaluate fifteen popular LLMs (from ChatGPT, DeepSeek,
Claude, and Gemini series) on DMind Benchmark, uncovering performance gaps in
Web3-specific reasoning and application, particularly in emerging areas like
token economics and meme concepts. Even the strongest models face significant
challenges in identifying subtle security vulnerabilities and analyzing complex
DeFi mechanisms. To foster progress in this area, we publicly release our
benchmark dataset, evaluation pipeline, and annotated results at
http://www.dmind.ai, offering a valuable resource for advancing specialized
domain adaptation and the development of more robust Web3-enabled LLMs.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 16:40:39 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Master",
"Miracle",
""
],
[
"Sun",
"Rainy",
""
],
[
"Reese",
"Anya",
""
],
[
"Ouyang",
"Joey",
""
],
[
"Chen",
"Alex",
""
],
[
"Dong",
"Winter",
""
],
[
"Li",
"Frank",
""
],
[
"Yi",
"James",
""
],
[
"Zhao",
"Garry",
""
],
[
"Ling",
"Tony",
""
],
[
"Wong",
"Hobert",
""
],
[
"Yang",
"Lowes",
""
]
] |
2504.16117 | Sridevi Polavaram Dr | Sridevi Polavaram, Xin Zhou, Meenu Ravi, Mohammad Zarei, Anmol
Srivastava | Context-Awareness and Interpretability of Rare Occurrences for Discovery
and Formalization of Critical Failure Modes | Accepted to IEEE Conference for Artificial Intelligence, 2025 | null | null | null | cs.CV cs.AI cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Vision systems are increasingly deployed in critical domains such as
surveillance, law enforcement, and transportation. However, their
vulnerabilities to rare or unforeseen scenarios pose significant safety risks.
To address these challenges, we introduce Context-Awareness and
Interpretability of Rare Occurrences (CAIRO), an ontology-based human-assistive
discovery framework for failure cases (or CP - Critical Phenomena) detection
and formalization. CAIRO by design incentivizes human-in-the-loop for testing
and evaluation of criticality that arises from misdetections, adversarial
attacks, and hallucinations in AI black-box models. Our robust analysis of
object detection model(s) failures in automated driving systems (ADS) showcases
scalable and interpretable ways of formalizing the observed gaps between camera
perception and real-world contexts, resulting in test cases stored as explicit
knowledge graphs (in OWL/XML format) amenable for sharing, downstream analysis,
logical reasoning, and accountability.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 17:12:37 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Polavaram",
"Sridevi",
""
],
[
"Zhou",
"Xin",
""
],
[
"Ravi",
"Meenu",
""
],
[
"Zarei",
"Mohammad",
""
],
[
"Srivastava",
"Anmol",
""
]
] |
2504.16118 | Milad Rahmati | Milad Rahmati | Towards Explainable and Lightweight AI for Real-Time Cyber Threat
Hunting in Edge Networks | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | As cyber threats continue to evolve, securing edge networks has become
increasingly challenging due to their distributed nature and resource
limitations. Many AI-driven threat detection systems rely on complex deep
learning models, which, despite their high accuracy, suffer from two major
drawbacks: lack of interpretability and high computational cost. Black-box AI
models make it difficult for security analysts to understand the reasoning
behind their predictions, limiting their practical deployment. Moreover,
conventional deep learning techniques demand significant computational
resources, rendering them unsuitable for edge devices with limited processing
power. To address these issues, this study introduces an Explainable and
Lightweight AI (ELAI) framework designed for real-time cyber threat detection
in edge networks. Our approach integrates interpretable machine learning
algorithms with optimized lightweight deep learning techniques, ensuring both
transparency and computational efficiency. The proposed system leverages
decision trees, attention-based deep learning, and federated learning to
enhance detection accuracy while maintaining explainability. We evaluate ELAI
using benchmark cybersecurity datasets, such as CICIDS and UNSW-NB15, assessing
its performance across diverse cyberattack scenarios. Experimental results
demonstrate that the proposed framework achieves high detection rates with
minimal false positives, all while significantly reducing computational demands
compared to traditional deep learning methods. The key contributions of this
work include: (1) a novel interpretable AI-based cybersecurity model tailored
for edge computing environments, (2) an optimized lightweight deep learning
approach for real-time cyber threat detection, and (3) a comprehensive analysis
of explainability techniques in AI-driven cybersecurity applications.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 23:45:39 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Rahmati",
"Milad",
""
]
] |
2504.16119 | Bo-Han Wu | Bo-Han Wu, Shi-Yuan Ma, Sri Krishna Vadlamani, Hyeongrak Choi, Dirk
Englund | Micro-Ring Perceptron Sensor for High-Speed, Low-Power Radio-Frequency
Signal | null | null | null | null | eess.SP physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radio-frequency (RF) sensing enables long-range, high-resolution detection
for applications such as radar and wireless communication. RF photonic sensing
mitigates the bandwidth limitations and high transmission losses of electronic
systems by transducing the detected RF signals into broadband optical carriers.
However, these sensing systems remain limited by detector noise and Nyquist
rate sampling with analog-to-digital converters, particularly under low-power
and high-data rate conditions. To overcome these limitations, we introduce the
micro-ring perceptron (MiRP) sensor, a physics-inspired AI framework that
integrates the micro-ring (MiR) dynamics-based analog processor with a
machine-learning-driven digital backend. By embedding the nonlinear optical
dynamics of MiRs into an end-to-end architecture, MiRP sensing maps the input
signal into a learned feature space for the subsequent digital neural network.
The trick is to encode the entire temporal structure of the incoming signal
into each output sample in order to enable effectively sub-Nyquist sampling
without loss of task-relevant information. Evaluations of three target
classification datasets demonstrate the performance advantages of MiRP sensing.
For example, on MNIST, MiRP detection achieves $94\pm0.1$\% accuracy at $1/49$
the Nyquist rate at the input RF signal of $1$~ pW, compared to $11\pm0.4$\%
for the conventional RF detection method. Thus, our sensor framework provides a
robust and efficient solution for the detection of low-power and high-speed
signals in real-world sensing applications.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 02:31:32 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Wu",
"Bo-Han",
""
],
[
"Ma",
"Shi-Yuan",
""
],
[
"Vadlamani",
"Sri Krishna",
""
],
[
"Choi",
"Hyeongrak",
""
],
[
"Englund",
"Dirk",
""
]
] |
2504.16120 | Ha\"ifa Nakouri | Chaima Njeh, Ha\"ifa Nakouri and Fehmi Jaafar | A Data-Centric Approach for Safe and Secure Large Language Models
against Threatening and Toxic Content | This paper is under revision in the International Journal of
Information Security | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLM) have made remarkable progress, but concerns about
potential biases and harmful content persist. To address these apprehensions,
we introduce a practical solution for ensuring LLM's safe and ethical use. Our
novel approach focuses on a post-generation correction mechanism, the
BART-Corrective Model, which adjusts generated content to ensure safety and
security. Unlike relying solely on model fine-tuning or prompt engineering, our
method provides a robust data-centric alternative for mitigating harmful
content. We demonstrate the effectiveness of our approach through experiments
on multiple toxic datasets, which show a significant reduction in mean toxicity
and jail-breaking scores after integration. Specifically, our results show a
reduction of 15% and 21% in mean toxicity and jail-breaking scores with GPT-4,
a substantial reduction of 28% and 5% with PaLM2, a reduction of approximately
26% and 23% with Mistral-7B, and a reduction of 11.1% and 19% with Gemma-2b-it.
These results demonstrate the potential of our approach to improve the safety
and security of LLM, making them more suitable for real-world applications.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 04:57:05 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Njeh",
"Chaima",
""
],
[
"Nakouri",
"Haïfa",
""
],
[
"Jaafar",
"Fehmi",
""
]
] |
2504.16121 | Muhammad Rafsan Kabir | Muhammad Rafsan Kabir, Rafeed Mohammad Sultan, Fuad Rahman, Mohammad
Ruhul Amin, Sifat Momen, Nabeel Mohammed, Shafin Rahman | LegalRAG: A Hybrid RAG System for Multilingual Legal Information
Retrieval | Accepted at IJCNN 2025 | null | null | null | cs.IR cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Natural Language Processing (NLP) and computational linguistic techniques are
increasingly being applied across various domains, yet their use in legal and
regulatory tasks remains limited. To address this gap, we develop an efficient
bilingual question-answering framework for regulatory documents, specifically
the Bangladesh Police Gazettes, which contain both English and Bangla text. Our
approach employs modern Retrieval Augmented Generation (RAG) pipelines to
enhance information retrieval and response generation. In addition to
conventional RAG pipelines, we propose an advanced RAG-based approach that
improves retrieval performance, leading to more precise answers. This system
enables efficient searching for specific government legal notices, making legal
information more accessible. We evaluate both our proposed and conventional RAG
systems on a diverse test set on Bangladesh Police Gazettes, demonstrating that
our approach consistently outperforms existing methods across all evaluation
metrics.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 06:09:54 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Kabir",
"Muhammad Rafsan",
""
],
[
"Sultan",
"Rafeed Mohammad",
""
],
[
"Rahman",
"Fuad",
""
],
[
"Amin",
"Mohammad Ruhul",
""
],
[
"Momen",
"Sifat",
""
],
[
"Mohammed",
"Nabeel",
""
],
[
"Rahman",
"Shafin",
""
]
] |
2504.16122 | Xuhui Zhou | Xuhui Zhou, Zhe Su, Sophie Feng, Jiaxu Zhou, Jen-tse Huang, Hsien-Te
Kao, Spencer Lynch, Svitlana Volkova, Tongshuang Sherry Wu, Anita Woolley,
Hao Zhu, Maarten Sap | SOTOPIA-S4: a user-friendly system for flexible, customizable, and
large-scale social simulation | The first author and the second author contributed equally | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | Social simulation through large language model (LLM) agents is a promising
approach to explore and validate hypotheses related to social science questions
and LLM agents behavior. We present SOTOPIA-S4, a fast, flexible, and scalable
social simulation system that addresses the technical barriers of current
frameworks while enabling practitioners to generate multi-turn and multi-party
LLM-based interactions with customizable evaluation metrics for hypothesis
testing. SOTOPIA-S4 comes as a pip package that contains a simulation engine,
an API server with flexible RESTful APIs for simulation management, and a web
interface that enables both technical and non-technical users to design, run,
and analyze simulations without programming. We demonstrate the usefulness of
SOTOPIA-S4 with two use cases involving dyadic hiring negotiation and
multi-party planning scenarios.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 20:02:59 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zhou",
"Xuhui",
""
],
[
"Su",
"Zhe",
""
],
[
"Feng",
"Sophie",
""
],
[
"Zhou",
"Jiaxu",
""
],
[
"Huang",
"Jen-tse",
""
],
[
"Kao",
"Hsien-Te",
""
],
[
"Lynch",
"Spencer",
""
],
[
"Volkova",
"Svitlana",
""
],
[
"Wu",
"Tongshuang Sherry",
""
],
[
"Woolley",
"Anita",
""
],
[
"Zhu",
"Hao",
""
],
[
"Sap",
"Maarten",
""
]
] |
2504.16123 | Clara Rojas | Alexis Garz\'on, Ricardo Fern\'andez, Vicente A. Ar\'evalo, David
Laroze, Laura M. P\'erez, Benjam\'in de Zayas, and Clara Rojas | The DKP Equation in Presence of a Woods-Saxon Potential: Transmission
Resonances and Bound States | 40 pages, 20 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article, we solve the Duffin-Kemmer-Petiau (DKP) equation in the
presence of the Woods-Saxon potential barrier and well for spin-one particles.
We derive the scattering solution in terms of the Gaussian hypergeometric
function ${}_2F_1(a, b; c; z)$, the regularized Gaussian hypergeometric
function ${}_2\tilde{F}_1(a, b; c; z)$, and the Gamma function $\Gamma(x)$. Our
analysis reveals the presence of transmission resonances. To observe these
resonances, we calculate and plot the transmission $T$ and reflection $R$
coefficients for various parameters of the Woods-Saxon potential barrier. Our
results are compared with those obtained for the square potential barrier and
the cusp potential barrier, which represent limiting cases of the Woods-Saxon
potential barrier. Furthermore, we investigate the bound state solutions,
determining the critical turning point $V_{cr}$ to study particle-antiparticle
creation and compute the norm $N$. Finally, we also compare our results with
those obtained for the square potential well and the cusp potential well,
confirming that pair creation occurs in both potential wells.
| [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 21:42:38 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Garzón",
"Alexis",
""
],
[
"Fernández",
"Ricardo",
""
],
[
"Arévalo",
"Vicente A.",
""
],
[
"Laroze",
"David",
""
],
[
"Pérez",
"Laura M.",
""
],
[
"de Zayas",
"Benjamín",
""
],
[
"Rojas",
"Clara",
""
]
] |
2504.16124 | Arani Mukhopadhyay | Ilias Papailias, Arani Mukhopadhyay, Anish Pal, Shahriar Namvar,
Constantine M. Megaridis | A Scalable Methodology for Reinstating the Superhydrophilicity of
Ambient-Contaminant Compromised Surfaces | Corresponding author: [email protected] | null | 10.1021/acsami.5c00724 | null | physics.flu-dyn cond-mat.mtrl-sci cond-mat.soft physics.chem-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The degradation of the hemi-wicking property of superhydrophilic high-energy
surfaces due to contaminant adsorption from the ambient atmosphere is well
documented. This degradation compromises the performance of such surfaces, thus
affecting their efficacy in real-world applications where hemi-wicking is
critical. In this work, the role of surface micro/nanostructure morphology of
laser-textured metallic surfaces on superhydrophilicity degradation is studied.
We explore intrinsic contact angle variations of superhydrophilic surfaces via
adsorption of organics from the surroundings, which brings about the associated
changes in surface chemistry. Furthermore, we explore condensation from humid
air as a scalable and environment friendly methodology that can reinstate
surface superhydrophilicity to a considerable extent (64% recovery of intrinsic
wettability after three hours of condensation) due to the efficient removal of
physisorbed contaminants from the surface texture features. The present results
strengthen the argument that contact-line movement at fine scales can be used
for de-pinning and removal of adsorbed organic molecules from contaminated
surfaces.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 02:22:47 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Papailias",
"Ilias",
""
],
[
"Mukhopadhyay",
"Arani",
""
],
[
"Pal",
"Anish",
""
],
[
"Namvar",
"Shahriar",
""
],
[
"Megaridis",
"Constantine M.",
""
]
] |
2504.16125 | Hao Di | Xiangyu Chang, Guang Dai, Hao Di, Haishan Ye | Breaking the Prompt Wall (I): A Real-World Case Study of Attacking
ChatGPT via Lightweight Prompt Injection | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This report presents a real-world case study demonstrating how prompt
injection can attack large language model platforms such as ChatGPT according
to a proposed injection framework. By providing three real-world examples, we
show how adversarial prompts can be injected via user inputs, web-based
retrieval, and system-level agent instructions. These attacks, though
lightweight and low-cost, can cause persistent and misleading behaviors in LLM
outputs. Our case study reveals that even commercial-grade LLMs remain
vulnerable to subtle manipulations that bypass safety filters and influence
user decisions. \textbf{More importantly, we stress that this report is not
intended as an attack guide, but as a technical alert. As ethical researchers,
we aim to raise awareness and call upon developers, especially those at OpenAI,
to treat prompt-level security as a critical design priority.
| [
{
"version": "v1",
"created": "Sun, 20 Apr 2025 05:59:00 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Chang",
"Xiangyu",
""
],
[
"Dai",
"Guang",
""
],
[
"Di",
"Hao",
""
],
[
"Ye",
"Haishan",
""
]
] |
2504.16126 | Hua Wang | Cong Chen and Hua Wang | Estimates for generalized fractional integrals associated with operators
on Morrey--Campanato spaces | 25 pages | null | null | null | math.CA | http://creativecommons.org/licenses/by/4.0/ | Let $\mathcal{L}$ be the infinitesimal generator of an analytic semigroup
$\big\{e^{-t\mathcal L}\big\}_{t>0}$ satisfying the Gaussian upper bounds. For
given $0<\alpha<n$, let $\mathcal L^{-\alpha/2}$ be the generalized fractional
integral associated with $\mathcal{L}$, which is defined as \begin{equation*}
\mathcal L^{-\alpha/2}(f)(x):=\frac{1}{\Gamma(\alpha/2)}\int_0^{+\infty}
e^{-t\mathcal L}(f)(x)t^{\alpha/2-1}dt, \end{equation*} where $\Gamma(\cdot)$
is the usual gamma function. For a locally integrable function $b(x)$ defined
on $\mathbb R^n$, the related commutator operator $\big[b,\mathcal
L^{-\alpha/2}\big]$ generated by $b$ and $\mathcal{L}^{-\alpha/2}$ is defined
by \begin{equation*} \big[b,\mathcal
L^{-\alpha/2}\big](f)(x):=b(x)\cdot\mathcal{L}^{-\alpha/2}(f)(x)-\mathcal{L}^{-\alpha/2}(bf)(x).
\end{equation*} A new class of Morrey--Campanato spaces associated with
$\mathcal{L}$ is introduced in this paper. The authors establish some new
estimates for the commutators $\big[b,\mathcal L^{-\alpha/2}\big]$ on
Morrey--Campanato spaces. The corresponding results for higher-order
commutators$\big[b,\mathcal L^{-\alpha/2}\big]^m$($m\in \mathbb{N}$) are also
discussed.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 04:23:15 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Chen",
"Cong",
""
],
[
"Wang",
"Hua",
""
]
] |
2504.16127 | Xingxing Zuo | Xingxing Zuo, Nikhil Ranganathan, Connor Lee, Georgia Gkioxari,
Soon-Jo Chung | MonoTher-Depth: Enhancing Thermal Depth Estimation via Confidence-Aware
Distillation | 8 Pages; The code will be available at
https://github.com/ZuoJiaxing/monother_depth | IEEE Robotics and Automation Letters (RA-L), 2025 | 10.1109/LRA.2025.3536855 | null | cs.CV cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Monocular depth estimation (MDE) from thermal images is a crucial technology
for robotic systems operating in challenging conditions such as fog, smoke, and
low light. The limited availability of labeled thermal data constrains the
generalization capabilities of thermal MDE models compared to foundational RGB
MDE models, which benefit from datasets of millions of images across diverse
scenarios. To address this challenge, we introduce a novel pipeline that
enhances thermal MDE through knowledge distillation from a versatile RGB MDE
model. Our approach features a confidence-aware distillation method that
utilizes the predicted confidence of the RGB MDE to selectively strengthen the
thermal MDE model, capitalizing on the strengths of the RGB model while
mitigating its weaknesses. Our method significantly improves the accuracy of
the thermal MDE, independent of the availability of labeled depth supervision,
and greatly expands its applicability to new scenarios. In our experiments on
new scenarios without labeled depth, the proposed confidence-aware distillation
method reduces the absolute relative error of thermal MDE by 22.88\% compared
to the baseline without distillation.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 05:13:49 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Zuo",
"Xingxing",
""
],
[
"Ranganathan",
"Nikhil",
""
],
[
"Lee",
"Connor",
""
],
[
"Gkioxari",
"Georgia",
""
],
[
"Chung",
"Soon-Jo",
""
]
] |
2504.16128 | Stanley Mugisha | Stanley Mugisha, Rashid Kisitu, Florence Tushabe | Hybrid Knowledge Transfer through Attention and Logit Distillation for
On-Device Vision Systems in Agricultural IoT | 12 pages and 4 figures | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Integrating deep learning applications into agricultural IoT systems faces a
serious challenge of balancing the high accuracy of Vision Transformers (ViTs)
with the efficiency demands of resource-constrained edge devices. Large
transformer models like the Swin Transformers excel in plant disease
classification by capturing global-local dependencies. However, their
computational complexity (34.1 GFLOPs) limits applications and renders them
impractical for real-time on-device inference. Lightweight models such as
MobileNetV3 and TinyML would be suitable for on-device inference but lack the
required spatial reasoning for fine-grained disease detection. To bridge this
gap, we propose a hybrid knowledge distillation framework that synergistically
transfers logit and attention knowledge from a Swin Transformer teacher to a
MobileNetV3 student model. Our method includes the introduction of adaptive
attention alignment to resolve cross-architecture mismatch (resolution,
channels) and a dual-loss function optimizing both class probabilities and
spatial focus. On the lantVillage-Tomato dataset (18,160 images), the distilled
MobileNetV3 attains 92.4% accuracy relative to 95.9% for Swin-L but at an 95%
reduction on PC and < 82% in inference latency on IoT devices. (23ms on PC CPU
and 86ms/image on smartphone CPUs). Key innovations include IoT-centric
validation metrics (13 MB memory, 0.22 GFLOPs) and dynamic resolution-matching
attention maps. Comparative experiments show significant improvements over
standalone CNNs and prior distillation methods, with a 3.5% accuracy gain over
MobileNetV3 baselines. Significantly, this work advances real-time,
energy-efficient crop monitoring in precision agriculture and demonstrates how
we can attain ViT-level diagnostic precision on edge devices. Code and models
will be made available for replication after acceptance.
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 06:56:41 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Mugisha",
"Stanley",
""
],
[
"Kisitu",
"Rashid",
""
],
[
"Tushabe",
"Florence",
""
]
] |
2504.16130 | Ren Pengju | Pengju Ren, Ri-gui Zhou, Yaochong Li | A Self-supervised Learning Method for Raman Spectroscopy based on Masked
Autoencoders | 15 pages, 10 figures | null | null | null | eess.SP cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Raman spectroscopy serves as a powerful and reliable tool for analyzing the
chemical information of substances. The integration of Raman spectroscopy with
deep learning methods enables rapid qualitative and quantitative analysis of
materials. Most existing approaches adopt supervised learning methods. Although
supervised learning has achieved satisfactory accuracy in spectral analysis, it
is still constrained by costly and limited well-annotated spectral datasets for
training. When spectral annotation is challenging or the amount of annotated
data is insufficient, the performance of supervised learning in spectral
material identification declines. In order to address the challenge of feature
extraction from unannotated spectra, we propose a self-supervised learning
paradigm for Raman Spectroscopy based on a Masked AutoEncoder, termed SMAE.
SMAE does not require any spectral annotations during pre-training. By randomly
masking and then reconstructing the spectral information, the model learns
essential spectral features. The reconstructed spectra exhibit certain
denoising properties, improving the signal-to-noise ratio (SNR) by more than
twofold. Utilizing the network weights obtained from masked pre-training, SMAE
achieves clustering accuracy of over 80% for 30 classes of isolated bacteria in
a pathogenic bacterial dataset, demonstrating significant improvements compared
to classical unsupervised methods and other state-of-the-art deep clustering
methods. After fine-tuning the network with a limited amount of annotated data,
SMAE achieves an identification accuracy of 83.90% on the test set, presenting
competitive performance against the supervised ResNet (83.40%).
| [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 10:44:06 GMT"
}
] | 2025-04-24T00:00:00 | [
[
"Ren",
"Pengju",
""
],
[
"Zhou",
"Ri-gui",
""
],
[
"Li",
"Yaochong",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.