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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2505.01862 | Linus Nwankwo | Linus Nwankwo, Bjoern Ellensohn, Ozan \"Ozdenizci, Elmar Rueckert | ReLI: A Language-Agnostic Approach to Human-Robot Interaction | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Adapting autonomous agents to industrial, domestic, and other daily tasks is
currently gaining momentum. However, in the global or cross-lingual application
contexts, ensuring effective interaction with the environment and executing
unrestricted human task-specified instructions in diverse languages remains an
unsolved problem. To address this challenge, we propose ReLI, a
language-agnostic framework designed to enable autonomous agents to converse
naturally, semantically reason about the environment, and to perform downstream
tasks, regardless of the task instruction's linguistic origin. First, we ground
large-scale pre-trained foundation models and transform them into
language-to-action models that can directly provide common-sense reasoning and
high-level robot control through natural, free-flow human-robot conversational
interactions. Further, we perform cross-lingual grounding of the models to
ensure that ReLI generalises across the global languages. To demonstrate the
ReLI's robustness, we conducted extensive simulated and real-world experiments
on various short- and long-horizon tasks, including zero-shot and few-shot
spatial navigation, scene information retrieval, and query-oriented tasks. We
benchmarked the performance on 140 languages involving over 70K multi-turn
conversations. On average, ReLI achieved over 90%$\pm$0.2 accuracy in
cross-lingual instruction parsing and task execution success rates. These
results demonstrate the ReLI's potential to enhance natural human-robot
interaction in the real world while championing linguistic diversity.
Demonstrations and resources will be publicly available at
https://linusnep.github.io/ReLI/.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 16:48:05 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 13:46:20 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Nwankwo",
"Linus",
""
],
[
"Ellensohn",
"Bjoern",
""
],
[
"Özdenizci",
"Ozan",
""
],
[
"Rueckert",
"Elmar",
""
]
] |
2505.01884 | Jaya Sreevalsan-Nair | Siddharth Kothari, Srinivasan Murali, Sankalp Kothari, Ujjwal Verma,
Jaya Sreevalsan-Nair | Adversarial Robustness of Deep Learning Models for Inland Water Body
Segmentation from SAR Images | 21 pages, 15 figures, 2 tables | null | null | null | eess.IV cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Inland water body segmentation from Synthetic Aperture Radar (SAR) images is
an important task needed for several applications, such as flood mapping. While
SAR sensors capture data in all-weather conditions as high-resolution images,
differentiating water and water-like surfaces from SAR images is not
straightforward. Inland water bodies, such as large river basins, have complex
geometry, which adds to the challenge of segmentation. U-Net is a widely used
deep learning model for land-water segmentation of SAR images. In practice,
manual annotation is often used to generate the corresponding water masks as
ground truth. Manual annotation of the images is prone to label noise owing to
data poisoning attacks, especially due to complex geometry. In this work, we
simulate manual errors in the form of adversarial attacks on the U-Net model
and study the robustness of the model to human errors in annotation. Our
results indicate that U-Net can tolerate a certain level of corruption before
its performance drops significantly. This finding highlights the crucial role
that the quality of manual annotations plays in determining the effectiveness
of the segmentation model. The code and the new dataset, along with adversarial
examples for robust training, are publicly available. (GitHub link -
https://github.com/GVCL/IWSeg-SAR-Poison.git)
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 18:18:59 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 17:26:22 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Kothari",
"Siddharth",
""
],
[
"Murali",
"Srinivasan",
""
],
[
"Kothari",
"Sankalp",
""
],
[
"Verma",
"Ujjwal",
""
],
[
"Sreevalsan-Nair",
"Jaya",
""
]
] |
2505.01913 | Ali Khalatpour | Ali Khalatpour, Luke Qi, Martin M. Fejer, Amir Safavi-Naeini | Roughness-Limited Performance in Ultra-Low-Loss Lithium Niobate Cavities | 20 pages, 14 figures | null | null | null | physics.optics physics.app-ph | http://creativecommons.org/licenses/by/4.0/ | Achieving low optical loss is critical for scaling complex photonic systems.
Thin-film lithium niobate (TFLN) offers strong electro-optic and nonlinear
properties in a compact platform, making it ideal for quantum and nonlinear
optics. While $Q$ factors above $10^7$ have been achieved, they remain below
the intrinsic material limit. We present a systematic study of scattering
losses due to roughness in TFLN racetrack cavities, isolating contributions
from sidewall and interface roughness. Quality factors up to $27 \times 10^6$
are demonstrated in waveguides with widths of $2.2\lambda$ ($\sim3.5\,\mu$m),
where interface roughness dominates, and up to $1.2 \times 10^7$ in narrower
waveguides $0.8\lambda$ wide ($\sim1.2\,\mu$m), where sidewall roughness is the
primary limitation. Our modeling framework, based on 3D wave simulations
informed by AFM-measured roughness, is material-independent and broadly
applicable across integrated photonic platforms.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 20:00:18 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 17:58:17 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Khalatpour",
"Ali",
""
],
[
"Qi",
"Luke",
""
],
[
"Fejer",
"Martin M.",
""
],
[
"Safavi-Naeini",
"Amir",
""
]
] |
2505.01921 | Shanyan Lai | Shanyan Lai | Multilayer Perceptron Neural Network Models in Asset Pricing: An
Empirical Study on Large-Cap US Stocks | null | null | 10.5281/zenodo.15333718 | null | q-fin.PR q-fin.CP q-fin.RM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, MLP models with dynamic structure are applied to factor models
for asset pricing tasks. Concretely, the MLP pyramid model structure was
employed on firm-characteristic-sorted portfolio factors for modelling the
large-capital US stocks. It was further developed as a practicable factor
investing strategy based on the predictions. The main findings in this chapter
were evaluated from two angles: model performance and investing performance,
which were compared from the periods with and without COVID-19. The empirical
results indicated that with the restrictions of the data size, the MLP models
no longer perform "deeper, better", while the proposed MLP models with two and
three hidden layers have higher flexibility to model the factors in this case.
This study also verified the idea of previous works that MLP models for factor
investing have more meaning in the downside risk control than in pursuing the
absolute annual returns.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 20:55:40 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 14:40:57 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Lai",
"Shanyan",
""
]
] |
2505.01959 | Yi Ding | Leyi Yan, Linda Wang, Sihang Liu, Yi Ding | EnsembleCI: Ensemble Learning for Carbon Intensity Forecasting | 5 pages, 5 figures, 3 tables, In The 16th ACM International
Conference on Future and Sustainable Energy Systems (E-ENERGY'25) | null | 10.1145/3679240.3734630 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Carbon intensity (CI) measures the average carbon emissions generated per
unit of electricity, making it a crucial metric for quantifying and managing
the environmental impact. Accurate CI predictions are vital for minimizing
carbon footprints, yet the state-of-the-art method (CarbonCast) falls short due
to its inability to address regional variability and lack of adaptability. To
address these limitations, we introduce EnsembleCI, an adaptive, end-to-end
ensemble learning-based approach for CI forecasting. EnsembleCI combines
weighted predictions from multiple sublearners, offering enhanced flexibility
and regional adaptability. In evaluations across 11 regional grids, EnsembleCI
consistently surpasses CarbonCast, achieving the lowest mean absolute
percentage error (MAPE) in almost all grids and improving prediction accuracy
by an average of 19.58%. While performance still varies across grids due to
inherent regional diversity, EnsembleCI reduces variability and exhibits
greater robustness in long-term forecasting compared to CarbonCast and
identifies region-specific key features, underscoring its interpretability and
practical relevance. These findings position EnsembleCI as a more accurate and
reliable solution for CI forecasting. EnsembleCI source code and data used in
this paper are available at https://github.com/emmayly/EnsembleCI.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 01:56:18 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 15:33:51 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Yan",
"Leyi",
""
],
[
"Wang",
"Linda",
""
],
[
"Liu",
"Sihang",
""
],
[
"Ding",
"Yi",
""
]
] |
2505.01998 | Chen Xiaoliang | Xiaoliang Chen (1), Xin Yu (1), Le Chang (1), Yunhe Huang (1),
Jiashuai He (1), Shibo Zhang (1), Jin Li (1), Likai Lin (1), Ziyu Zeng (1),
Xianling Tu (1), Shuyu Zhang (1) ((1) SoundAI Technology Co., Ltd) | A Synergistic Framework of Nonlinear Acoustic Computing and
Reinforcement Learning for Real-World Human-Robot Interaction | 34 pages, 11 figures, 10 tables, and 10 equations | null | null | null | cs.RO cs.AI physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a novel framework integrating nonlinear acoustic
computing and reinforcement learning to enhance advanced human-robot
interaction under complex noise and reverberation. Leveraging physically
informed wave equations (e.g., Westervelt, KZK), the approach captures
higher-order phenomena such as harmonic generation and shock formation. By
embedding these models in a reinforcement learning-driven control loop, the
system adaptively optimizes key parameters (e.g., absorption, beamforming) to
mitigate multipath interference and non-stationary noise. Experimental
evaluations, covering far-field localization, weak signal detection, and
multilingual speech recognition, demonstrate that this hybrid strategy
surpasses traditional linear methods and purely data-driven baselines,
achieving superior noise suppression, minimal latency, and robust accuracy in
demanding real-world scenarios. The proposed system demonstrates broad
application prospects in AI hardware, robot, machine audition, artificial
audition, and brain-machine interfaces.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 06:03:12 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 16:09:59 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Chen",
"Xiaoliang",
"",
"SoundAI Technology Co., Ltd"
],
[
"Yu",
"Xin",
"",
"SoundAI Technology Co., Ltd"
],
[
"Chang",
"Le",
"",
"SoundAI Technology Co., Ltd"
],
[
"Huang",
"Yunhe",
"",
"SoundAI Technology Co., Ltd"
],
[
"He",
"Jiashuai",
"",
"SoundAI Technology Co., Ltd"
],
[
"Zhang",
"Shibo",
"",
"SoundAI Technology Co., Ltd"
],
[
"Li",
"Jin",
"",
"SoundAI Technology Co., Ltd"
],
[
"Lin",
"Likai",
"",
"SoundAI Technology Co., Ltd"
],
[
"Zeng",
"Ziyu",
"",
"SoundAI Technology Co., Ltd"
],
[
"Tu",
"Xianling",
"",
"SoundAI Technology Co., Ltd"
],
[
"Zhang",
"Shuyu",
"",
"SoundAI Technology Co., Ltd"
]
] |
2505.02048 | Sebastian Rassmann | Sebastian Rassmann, David K\"ugler, Christian Ewert, Martin Reuter | Regression is all you need for medical image translation | null | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | The acquisition of information-rich images within a limited time budget is
crucial in medical imaging. Medical image translation (MIT) can help enhance
and supplement existing datasets by generating synthetic images from acquired
data. While Generative Adversarial Nets (GANs) and Diffusion Models (DMs) have
achieved remarkable success in natural image generation, their benefits -
creativity and image realism - do not necessarily transfer to medical
applications where highly accurate anatomical information is required. In fact,
the imitation of acquisition noise or content hallucination hinder clinical
utility. Here, we introduce YODA (You Only Denoise once - or Average), a novel
2.5D diffusion-based framework for volumetric MIT. YODA unites diffusion and
regression paradigms to produce realistic or noise-free outputs. Furthermore,
we propose Expectation-Approximation (ExpA) DM sampling, which draws
inspiration from MRI signal averaging. ExpA-sampling suppresses generated noise
and, thus, eliminates noise from biasing the evaluation of image quality.
Through extensive experiments on four diverse multi-modal datasets - comprising
multi-contrast brain MRI and pelvic MRI-CT - we show that diffusion and
regression sampling yield similar results in practice. As such, the
computational overhead of diffusion sampling does not provide systematic
benefits in medical information translation. Building on these insights, we
demonstrate that YODA outperforms several state-of-the-art GAN and DM methods.
Notably, YODA-generated images are shown to be interchangeable with, or even
superior to, physical acquisitions for several downstream tasks. Our findings
challenge the presumed advantages of DMs in MIT and pave the way for the
practical application of MIT in medical imaging.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 09:57:10 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 05:56:47 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Rassmann",
"Sebastian",
""
],
[
"Kügler",
"David",
""
],
[
"Ewert",
"Christian",
""
],
[
"Reuter",
"Martin",
""
]
] |
2505.02064 | Sicheng Tao | Shuhang Xun, Sicheng Tao, Jungang Li, Yibo Shi, Zhixin Lin, Zhanhui
Zhu, Yibo Yan, Hanqian Li, Linghao Zhang, Shikang Wang, Yixin Liu, Hanbo
Zhang, Ying Ma, Xuming Hu | RTV-Bench: Benchmarking MLLM Continuous Perception, Understanding and
Reasoning through Real-Time Video | 13 pages, 4 figures, 5 tables | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Multimodal Large Language Models (MLLMs) increasingly excel at perception,
understanding, and reasoning. However, current benchmarks inadequately evaluate
their ability to perform these tasks continuously in dynamic, real-world
environments. To bridge this gap, we introduce RTV-Bench, a fine-grained
benchmark for MLLM real-time video analysis. RTV-Bench uses three key
principles: (1) Multi-Timestamp Question Answering (MTQA), where answers evolve
with scene changes; (2) Hierarchical Question Structure, combining basic and
advanced queries; and (3) Multi-dimensional Evaluation, assessing the ability
of continuous perception, understanding, and reasoning. RTV-Bench contains 552
diverse videos (167.2 hours) and 4,631 high-quality QA pairs. We evaluated
leading MLLMs, including proprietary (GPT-4o, Gemini 2.0), open-source offline
(Qwen2.5-VL, VideoLLaMA3), and open-source real-time (VITA-1.5,
InternLM-XComposer2.5-OmniLive) models. Experiment results show open-source
real-time models largely outperform offline ones but still trail top
proprietary models. Our analysis also reveals that larger model size or higher
frame sampling rates do not significantly boost RTV-Bench performance,
sometimes causing slight decreases. This underscores the need for better model
architectures optimized for video stream processing and long sequences to
advance real-time video analysis with MLLMs. Our benchmark toolkit is available
at: https://github.com/LJungang/RTV-Bench.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 10:55:21 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 01:51:03 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Xun",
"Shuhang",
""
],
[
"Tao",
"Sicheng",
""
],
[
"Li",
"Jungang",
""
],
[
"Shi",
"Yibo",
""
],
[
"Lin",
"Zhixin",
""
],
[
"Zhu",
"Zhanhui",
""
],
[
"Yan",
"Yibo",
""
],
[
"Li",
"Hanqian",
""
],
[
"Zhang",
"Linghao",
""
],
[
"Wang",
"Shikang",
""
],
[
"Liu",
"Yixin",
""
],
[
"Zhang",
"Hanbo",
""
],
[
"Ma",
"Ying",
""
],
[
"Hu",
"Xuming",
""
]
] |
2505.02102 | Thomas D. Cohen | Thomas D. Cohen | Gauge invariance and color charge fluctuations | 4 pages | null | null | null | hep-lat hep-ph hep-th nucl-th | http://creativecommons.org/licenses/by/4.0/ | The nature of confinement is connected with color charge. Unfortunately, the
color charge densities in QCD, the Noether charge densities associated with the
global color invariance, are not invariant under local color rotations. This
implies that the expectation values of the net color charge in any region of
any physical state in QCD, states that satisfy the color Gauss law, are
automatically zero for all components of color. In this paper it is shown that
the expectation value of the square of the net color charge in a region, a
measure of the color charge fluctuations, is necessarily nonzero when evaluated
in physical states and the result, while depending on the scheme and scale by
which the theory is regulated, is gauge invariant. This holds despite the
formal lack of gauge invariance of the operator. Moreover, there is a
particular combination of the color charge fluctuations for the vacuum and for
a system describable by a non-trivial density matrix that is independent that
has a well-defined continuum limit
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 13:14:40 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Cohen",
"Thomas D.",
""
]
] |
2505.02118 | Wei Liu | Wei Liu, Zhongyu Niu, Lang Gao, Zhiying Deng, Jun Wang, Haozhao Wang,
Ruixuan Li | Adversarial Cooperative Rationalization: The Risk of Spurious
Correlations in Even Clean Datasets | ICML 2025 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates the self-rationalization framework constructed with a
cooperative game, where a generator initially extracts the most informative
segment from raw input, and a subsequent predictor utilizes the selected subset
for its input. The generator and predictor are trained collaboratively to
maximize prediction accuracy. In this paper, we first uncover a potential
caveat: such a cooperative game could unintentionally introduce a sampling bias
during rationale extraction. Specifically, the generator might inadvertently
create an incorrect correlation between the selected rationale candidate and
the label, even when they are semantically unrelated in the original dataset.
Subsequently, we elucidate the origins of this bias using both detailed
theoretical analysis and empirical evidence. Our findings suggest a direction
for inspecting these correlations through attacks, based on which we further
introduce an instruction to prevent the predictor from learning the
correlations. Through experiments on six text classification datasets and two
graph classification datasets using three network architectures (GRUs, BERT,
and GCN), we show that our method not only significantly outperforms recent
rationalization methods, but also achieves comparable or even better results
than a representative LLM (llama3.1-8b-instruct).
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 14:00:04 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 07:07:13 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Liu",
"Wei",
""
],
[
"Niu",
"Zhongyu",
""
],
[
"Gao",
"Lang",
""
],
[
"Deng",
"Zhiying",
""
],
[
"Wang",
"Jun",
""
],
[
"Wang",
"Haozhao",
""
],
[
"Li",
"Ruixuan",
""
]
] |
2505.02138 | Chenxi Liu | Chenxi Liu, Hao Miao, Qianxiong Xu, Shaowen Zhou, Cheng Long, Yan
Zhao, Ziyue Li, Rui Zhao | Efficient Multivariate Time Series Forecasting via Calibrated Language
Models with Privileged Knowledge Distillation | Accepted by ICDE 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multivariate time series forecasting (MTSF) endeavors to predict future
observations given historical data, playing a crucial role in time series data
management systems. With advancements in large language models (LLMs), recent
studies employ textual prompt tuning to infuse the knowledge of LLMs into MTSF.
However, the deployment of LLMs often suffers from low efficiency during the
inference phase. To address this problem, we introduce TimeKD, an efficient
MTSF framework that leverages the calibrated language models and privileged
knowledge distillation. TimeKD aims to generate high-quality future
representations from the proposed cross-modality teacher model and cultivate an
effective student model. The cross-modality teacher model adopts calibrated
language models (CLMs) with ground truth prompts, motivated by the paradigm of
Learning Under Privileged Information (LUPI). In addition, we design a
subtractive cross attention (SCA) mechanism to refine these representations. To
cultivate an effective student model, we propose an innovative privileged
knowledge distillation (PKD) mechanism including correlation and feature
distillation. PKD enables the student to replicate the teacher's behavior while
minimizing their output discrepancy. Extensive experiments on real data offer
insight into the effectiveness, efficiency, and scalability of the proposed
TimeKD.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 14:57:42 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 12:35:29 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Liu",
"Chenxi",
""
],
[
"Miao",
"Hao",
""
],
[
"Xu",
"Qianxiong",
""
],
[
"Zhou",
"Shaowen",
""
],
[
"Long",
"Cheng",
""
],
[
"Zhao",
"Yan",
""
],
[
"Li",
"Ziyue",
""
],
[
"Zhao",
"Rui",
""
]
] |
2505.02156 | Minzheng Wang | Minzheng Wang, Yongbin Li, Haobo Wang, Xinghua Zhang, Nan Xu, Bingli
Wu, Fei Huang, Haiyang Yu, Wenji Mao | Think on your Feet: Adaptive Thinking via Reinforcement Learning for
Social Agents | Work in Progress. The code and data are available, see
https://github.com/MozerWang/AMPO | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Effective social intelligence simulation requires language agents to
dynamically adjust reasoning depth, a capability notably absent in current
approaches. While existing methods either lack this kind of reasoning
capability or enforce uniform long chain-of-thought reasoning across all
scenarios, resulting in excessive token usage and inappropriate social
simulation. In this paper, we propose $\textbf{A}$daptive $\textbf{M}$ode
$\textbf{L}$earning ($\textbf{AML}$) that strategically selects from four
thinking modes (intuitive reaction $\rightarrow$ deep contemplation) based on
real-time context. Our framework's core innovation, the $\textbf{A}$daptive
$\textbf{M}$ode $\textbf{P}$olicy $\textbf{O}$ptimization ($\textbf{AMPO}$)
algorithm, introduces three key advancements over existing methods: (1)
Multi-granular thinking mode design, (2) Context-aware mode switching across
social interaction, and (3) Token-efficient reasoning via depth-adaptive
processing. Extensive experiments on social intelligence tasks confirm that AML
achieves 15.6% higher task performance than state-of-the-art methods. Notably,
our method outperforms GRPO by 7.0% with 32.8% shorter reasoning chains. These
results demonstrate that context-sensitive thinking mode selection, as
implemented in AMPO, enables more human-like adaptive reasoning than GRPO's
fixed-depth approach.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 15:39:58 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 09:27:31 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Minzheng",
""
],
[
"Li",
"Yongbin",
""
],
[
"Wang",
"Haobo",
""
],
[
"Zhang",
"Xinghua",
""
],
[
"Xu",
"Nan",
""
],
[
"Wu",
"Bingli",
""
],
[
"Huang",
"Fei",
""
],
[
"Yu",
"Haiyang",
""
],
[
"Mao",
"Wenji",
""
]
] |
2505.02157 | Chi-Thiem Hoang | Thiem Hoang (KASI & UST) | Toward A General Theory of Grain Alignment and Disruption by Radiative
Torques and Magnetic Relaxation | 30 pages, 32 figures. Several typos fixed; To be submitted. Comments
are welcome | null | null | null | astro-ph.GA astro-ph.EP astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | We generalize the magnetically enhanced radiative torque (MRAT) alignment
theory for general astrophysical environments described by a dimensionless
parameter $U/(n_{1}T_{2})$ with $U$ local radiation strength, $n_{1}=n_{\rm
H}/(10{\rm cm}^{-3})$ the hydrogen density, and $T_{2}=T_{\rm gas}/100{\rm K}$
the gas temperature. We first derive the critical magnetic relaxation
$\delta_{\rm mag,cri}$ required to produce high-J attractors for different RAT
models and local conditions and find that $\delta_{\rm mag,cri}$ must be larger
for stronger radiation fields. We then numerically study the grain alignment
and rotational disruption by the MRAT mechanism taking into account gas
collisions and magnetic fluctuations. We find that, for the collision-dominated
(CD) regime ($U/(n_{1}T_{2})\leq 1$), collisional and magnetic excitations can
slowly transport large grains from low-J rotation to high-J attractors, leading
to the perfect {\it slow alignment} within $\sim 10-100$ damping times due to
MRATs. However, for the radiation-dominated (RD) regime ($U/(n_{1}T_{2})>1$),
only a fraction of grains can be fast aligned at high-J attractors by MRATs,
and the majority of grains are trapped at low-J rotation due to strong
radiative torques, a new effect we term {\it radiative torque (RAT) trapping}.
For extreme radiation fields of $U/(n_{1}T_{2})>10^{4}$, the efficiency of
magnetic relaxation on grain alignment is suppressed, and grains only have fast
alignment and disruption purely determined by RATs. We quantified the fraction
of grains with fast alignment at high-J attractors, $f_{\rm high-J}^{\rm
fast}$, for different RAT models, magnetic relaxation, and $U/(n_{1}T_{2})$,
and found that the maximum $f_{\rm high-J}^{\rm fast}$ can reach $45\%$ by
MRATs and $22\%$ by RATs.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 15:40:53 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 12:30:56 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Hoang",
"Thiem",
"",
"KASI & UST"
]
] |
2505.02183 | Aidan Young PhD | Tom Meyerovitch, Aidan Young | Alternating and non-alternating deterministic Markov games | null | null | null | null | cs.IT cs.GT math.DS math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Two variants of a deterministic multi-round, zero-sum, two-player game are
presented: a turn-alternating version and an non-alternating version. The
non-alternating version occurs in the computation of the covering radius of
constrained systems, a quantity of interest in coding theory.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 17:04:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Meyerovitch",
"Tom",
""
],
[
"Young",
"Aidan",
""
]
] |
2505.02289 | Meicheng Xiong | Meicheng Xiong, Di Zhu | Lakeplace: Sensing interactions between lakes and human activities | null | null | null | null | physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | Urban freshwater ecosystems, composed of rivers, ponds, lakes, and other
water bodies, have essential socioeconomic and ecological values for urban
residents. However, research investigating how individuals interact with lakes
remains limited, especially within cities and at fine spatiotemporal
resolutions. To fill this gap, we propose a data-driven analytical framework
that comprehensively senses human-lake interactions and profiles the
social-demographic characteristics of intra-city lakes. The term "lakeplace" is
proposed to depict a place containing lakes and human activities within it. For
each lake, the geographic boundary of its lakeplace refers to the first-order
administrative units, reflecting the neighboring scale of lake socioeconomics.
Utilizing large-scale individual mobile positioning data, we performed
lakeplace sensing on the 2,036 major lakes in the Twin Cities Metropolitan Area
(TCMA), Minnesota, and the people interacting with them. The popularity of each
lakeplace was measured by its temporal visitations and further categorized as
on-lake and around-lake human activities. Popular lakeplaces were investigated
to depict whether the attractiveness of a lake is mostly brought by the lake
itself, or the social-demographic environment around it. The lakeplace sensing
framework offers a practical approach to the spatiotemporal characteristics of
human activities and understanding the social-demographic knowledge related to
human-lake systems. Our work exemplifies the social sensing of
human-environment interactions via geospatial big data, shedding light on
human-oriented sustainable urban planning and urban water resource management.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 23:12:27 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 16:18:46 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Xiong",
"Meicheng",
""
],
[
"Zhu",
"Di",
""
]
] |
2505.02293 | Jasmine Jerry Aloor | Jason J. Choi, Jasmine Jerry Aloor, Jingqi Li, Maria G. Mendoza, Hamsa
Balakrishnan and Claire J. Tomlin | Resolving Conflicting Constraints in Multi-Agent Reinforcement Learning
with Layered Safety | Accepted for publication at the 2025 Robotics: Science and Systems
Conference. 18 pages, 8 figures | null | null | null | cs.RO cs.MA cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | Preventing collisions in multi-robot navigation is crucial for deployment.
This requirement hinders the use of learning-based approaches, such as
multi-agent reinforcement learning (MARL), on their own due to their lack of
safety guarantees. Traditional control methods, such as reachability and
control barrier functions, can provide rigorous safety guarantees when
interactions are limited only to a small number of robots. However, conflicts
between the constraints faced by different agents pose a challenge to safe
multi-agent coordination.
To overcome this challenge, we propose a method that integrates multiple
layers of safety by combining MARL with safety filters. First, MARL is used to
learn strategies that minimize multiple agent interactions, where multiple
indicates more than two. Particularly, we focus on interactions likely to
result in conflicting constraints within the engagement distance. Next, for
agents that enter the engagement distance, we prioritize pairs requiring the
most urgent corrective actions. Finally, a dedicated safety filter provides
tactical corrective actions to resolve these conflicts. Crucially, the design
decisions for all layers of this framework are grounded in reachability
analysis and a control barrier-value function-based filtering mechanism.
We validate our Layered Safe MARL framework in 1) hardware experiments using
Crazyflie drones and 2) high-density advanced aerial mobility (AAM) operation
scenarios, where agents navigate to designated waypoints while avoiding
collisions. The results show that our method significantly reduces conflict
while maintaining safety without sacrificing much efficiency (i.e., shorter
travel time and distance) compared to baselines that do not incorporate layered
safety. The project website is available at
https://dinamo-mit.github.io/Layered-Safe-MARL/
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 23:42:52 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Choi",
"Jason J.",
""
],
[
"Aloor",
"Jasmine Jerry",
""
],
[
"Li",
"Jingqi",
""
],
[
"Mendoza",
"Maria G.",
""
],
[
"Balakrishnan",
"Hamsa",
""
],
[
"Tomlin",
"Claire J.",
""
]
] |
2505.02305 | Shin Hong | Kieun Kim (1), Seongmin Lee (2), Shin Hong (1) ((1) Chungbuk National
University, (2) Max Planck Institute for Security and Privacy) | Refining Fuzzed Crashing Inputs for Better Fault Diagnosis | This paper will be presented in the Posters track at FSE 2025
(https://conf.researchr.org/track/fse-2025/fse-2025-posters) | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | We present DiffMin, a technique that refines a fuzzed crashing input to gain
greater similarities to given passing inputs to help developers analyze the
crashing input to identify the failure-inducing condition and locate buggy code
for debugging. DiffMin iteratively applies edit actions to transform a fuzzed
input while preserving the crash behavior. Our pilot study with the Magma
benchmark demonstrates that DiffMin effectively minimizes the differences
between crashing and passing inputs while enhancing the accuracy of
spectrum-based fault localization, highlighting its potential as a valuable
pre-debugging step after greybox fuzzing.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 00:58:16 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 07:49:50 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Kim",
"Kieun",
""
],
[
"Lee",
"Seongmin",
""
],
[
"Hong",
"Shin",
""
]
] |
2505.02380 | Arnab Sanyal | Arnab Sanyal, Prithwish Mukherjee, Gourav Datta, Sandeep P. Chinchali | EntroLLM: Entropy Encoded Weight Compression for Efficient Large
Language Model Inference on Edge Devices | 6 pages, 1 reference page. Under submission and review at ISLPED 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) demonstrate exceptional performance across
various tasks, but their large storage and computational requirements constrain
their deployment on edge devices. To address this, we propose EntroLLM, a novel
compression framework that integrates mixed quantization with entropy coding to
reduce storage overhead while maintaining model accuracy. Our method applies a
layer-wise mixed quantization scheme - choosing between symmetric and
asymmetric quantization based on individual layer weight distributions - to
optimize compressibility. We then employ Huffman encoding for lossless
compression of the quantized weights, significantly reducing memory bandwidth
requirements. Furthermore, we introduce parallel Huffman decoding, which
enables efficient retrieval of encoded weights during inference, ensuring
minimal latency impact. Our experiments on edge-compatible LLMs, including
smolLM-1.7B-Instruct, phi3-mini-4k-Instruct, and mistral-7B-Instruct,
demonstrate that EntroLLM achieves up to $30\%$ storage reduction compared to
uint8 models and up to $65%$ storage reduction compared to uint4 models, while
preserving perplexity and accuracy, on language benchmark tasks. We further
show that our method enables $31.9\%$ - $146.6\%$ faster inference throughput
on memory-bandwidth-limited edge devices, such as NVIDIA Jetson P3450, by
reducing the required data movement. The proposed approach requires no
additional re-training and is fully compatible with existing post-training
quantization methods, making it a practical solution for edge LLMs.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 05:42:14 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 09:37:57 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Sanyal",
"Arnab",
""
],
[
"Mukherjee",
"Prithwish",
""
],
[
"Datta",
"Gourav",
""
],
[
"Chinchali",
"Sandeep P.",
""
]
] |
2505.02478 | Baptiste Coquinot | Baptiste Coquinot, Mathieu Liz\'ee, Lyd\'eric Bocquet and Nikita
Kavokine | Electrons as intermediate charge carriers in ion transport through
nanoscale channels | null | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | The ionic conductivity of a nanofluidic channel is one of its most basic
properties, which is systematically characterized in experiments. While most
studies to date have focused on the DC conductivity, there has recently been a
surge of interest in the behavior of nanochannels under AC voltages. Here, we
show that in nanochannels with electrically conducting walls, electrons can
contribute to transporting ionic current through the channel. We identify a
critical frequency and lengthscale above which electrons in the channel walls
become the main charge carriers, resulting in a strongly increased
conductivity. Our findings open the way to the engineering of nanofluidic
functionalities based on ion-electron coupling.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 09:00:37 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 07:57:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Coquinot",
"Baptiste",
""
],
[
"Lizée",
"Mathieu",
""
],
[
"Bocquet",
"Lydéric",
""
],
[
"Kavokine",
"Nikita",
""
]
] |
2505.02508 | Yuchun Qian | Yang Lyu, Yuchun Qian, Tan Minh Nguyen, Xin T. Tong | Resolving Memorization in Empirical Diffusion Model for Manifold Data in
High-Dimensional Spaces | null | null | null | null | stat.ML cs.LG math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | Diffusion models is a popular computational tool to generate new data
samples. It utilizes a forward diffusion process that add noise to the data
distribution and then use a reverse process to remove noises to produce samples
from the data distribution. However, when the empirical data distribution
consists of $n$ data point, using the empirical diffusion model will
necessarily produce one of the existing data points. This is often referred to
as the memorization effect, which is usually resolved by sophisticated machine
learning procedures in the current literature. This work shows that the
memorization problem can be resolved by a simple inertia update step at the end
of the empirical diffusion model simulation. Our inertial diffusion model
requires only the empirical diffusion model score function and it does not
require any further training. We show that choosing the inertia diffusion model
sample distribution is an $O\left(n^{-\frac{2}{d+4}}\right)$ Wasserstein-1
approximation of a data distribution lying on a $C^2$ manifold of dimension
$d$. Since this estimate is significant smaller the Wasserstein1 distance
between population and empirical distributions, it rigorously shows the
inertial diffusion model produces new data samples. Remarkably, this upper
bound is completely free of the ambient space dimension, since there is no
training involved. Our analysis utilizes the fact that the inertial diffusion
model samples are approximately distributed as the Gaussian kernel density
estimator on the manifold. This reveals an interesting connection between
diffusion model and manifold learning.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 09:40:41 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 05:10:46 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Lyu",
"Yang",
""
],
[
"Qian",
"Yuchun",
""
],
[
"Nguyen",
"Tan Minh",
""
],
[
"Tong",
"Xin T.",
""
]
] |
2505.02537 | Alberto Sinigaglia | Davide Sartor, Alberto Sinigaglia, Gian Antonio Susto | Advancing Constrained Monotonic Neural Networks: Achieving Universal
Approximation Beyond Bounded Activations | International Conference on Machine Learning | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Conventional techniques for imposing monotonicity in MLPs by construction
involve the use of non-negative weight constraints and bounded activation
functions, which pose well-known optimization challenges. In this work, we
generalize previous theoretical results, showing that MLPs with non-negative
weight constraint and activations that saturate on alternating sides are
universal approximators for monotonic functions. Additionally, we show an
equivalence between the saturation side in the activations and the sign of the
weight constraint. This connection allows us to prove that MLPs with convex
monotone activations and non-positive constrained weights also qualify as
universal approximators, in contrast to their non-negative constrained
counterparts. Our results provide theoretical grounding to the empirical
effectiveness observed in previous works while leading to possible
architectural simplification. Moreover, to further alleviate the optimization
difficulties, we propose an alternative formulation that allows the network to
adjust its activations according to the sign of the weights. This eliminates
the requirement for weight reparameterization, easing initialization and
improving training stability. Experimental evaluation reinforces the validity
of the theoretical results, showing that our novel approach compares favourably
to traditional monotonic architectures.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 10:18:48 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 11:45:55 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Sartor",
"Davide",
""
],
[
"Sinigaglia",
"Alberto",
""
],
[
"Susto",
"Gian Antonio",
""
]
] |
2505.02549 | Yongxiang Li | Yongxiang Li, Yuan Sun, Yang Qin, Dezhong Peng, Xi Peng, Peng Hu | Robust Duality Learning for Unsupervised Visible-Infrared Person
Re-Identification | null | null | 10.1109/TIFS.2025.3536613 | null | cs.CV cs.MM | http://creativecommons.org/licenses/by/4.0/ | Unsupervised visible-infrared person re-identification (UVI-ReID) aims to
retrieve pedestrian images across different modalities without costly
annotations, but faces challenges due to the modality gap and lack of
supervision. Existing methods often adopt self-training with
clustering-generated pseudo-labels but implicitly assume these labels are
always correct. In practice, however, this assumption fails due to inevitable
pseudo-label noise, which hinders model learning. To address this, we introduce
a new learning paradigm that explicitly considers Pseudo-Label Noise (PLN),
characterized by three key challenges: noise overfitting, error accumulation,
and noisy cluster correspondence. To this end, we propose a novel Robust
Duality Learning framework (RoDE) for UVI-ReID to mitigate the effects of noisy
pseudo-labels. First, to combat noise overfitting, a Robust Adaptive Learning
mechanism (RAL) is proposed to dynamically emphasize clean samples while
down-weighting noisy ones. Second, to alleviate error accumulation-where the
model reinforces its own mistakes-RoDE employs dual distinct models that are
alternately trained using pseudo-labels from each other, encouraging diversity
and preventing collapse. However, this dual-model strategy introduces
misalignment between clusters across models and modalities, creating noisy
cluster correspondence. To resolve this, we introduce Cluster Consistency
Matching (CCM), which aligns clusters across models and modalities by measuring
cross-cluster similarity. Extensive experiments on three benchmarks demonstrate
the effectiveness of RoDE.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 10:36:52 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 07:22:39 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Li",
"Yongxiang",
""
],
[
"Sun",
"Yuan",
""
],
[
"Qin",
"Yang",
""
],
[
"Peng",
"Dezhong",
""
],
[
"Peng",
"Xi",
""
],
[
"Hu",
"Peng",
""
]
] |
2505.02560 | Andreas Konstantin Kruff | Andreas Konstantin Kruff, Timo Breuer, Philipp Schaer | Evaluating Contrastive Feedback for Effective User Simulations | null | null | 10.1145/3726302.3730189 | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | The use of Large Language Models (LLMs) for simulating user behavior in the
domain of Interactive Information Retrieval has recently gained significant
popularity. However, their application and capabilities remain highly debated
and understudied. This study explores whether the underlying principles of
contrastive training techniques, which have been effective for fine-tuning
LLMs, can also be applied beneficially in the area of prompt engineering for
user simulations.
Previous research has shown that LLMs possess comprehensive world knowledge,
which can be leveraged to provide accurate estimates of relevant documents.
This study attempts to simulate a knowledge state by enhancing the model with
additional implicit contextual information gained during the simulation. This
approach enables the model to refine the scope of desired documents further.
The primary objective of this study is to analyze how different modalities of
contextual information influence the effectiveness of user simulations.
Various user configurations were tested, where models are provided with
summaries of already judged relevant, irrelevant, or both types of documents in
a contrastive manner. The focus of this study is the assessment of the impact
of the prompting techniques on the simulated user agent performance. We hereby
lay the foundations for leveraging LLMs as part of more realistic simulated
users.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 11:02:31 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 11:44:32 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Kruff",
"Andreas Konstantin",
""
],
[
"Breuer",
"Timo",
""
],
[
"Schaer",
"Philipp",
""
]
] |
2505.02572 | Sergei Alexandrov | Sergei Alexandrov | Mock modularity at work, or black holes in a forest | invited review for "Entropy" | null | null | null | hep-th gr-qc math-ph math.AG math.MP math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mock modular forms, first invented by Ramanujan, provide a beautiful
generalization of the usual modular forms. In recent years, it was found that
they capture generating functions of the number of microstates of BPS black
holes appearing in compactifications of string theory with 8 and 16
supercharges. This review describes these results and their applications which
range from the actual computation of these generating functions for both
compact and non-compact compactification manifolds (encoding, respectively,
Donaldson-Thomas and Vafa-Witten topological invariants) to the construction of
new non-commutative structures on moduli spaces of Calabi-Yau threefolds.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 11:22:03 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Alexandrov",
"Sergei",
""
]
] |
2505.02579 | Lingxiao Kong | Lingxiao Kong, Cong Yang, Susanne Neufang, Oya Deniz Beyan, Zeyd
Boukhers | EMORL: Ensemble Multi-Objective Reinforcement Learning for Efficient and
Flexible LLM Fine-Tuning | 13 pages, 9 figures, submitted to SIGDIAL 2025 conference | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Recent advances in reinforcement learning (RL) for large language model (LLM)
fine-tuning show promise in addressing multi-objective tasks but still face
significant challenges, including complex objective balancing, low training
efficiency, poor scalability, and limited explainability. Leveraging ensemble
learning principles, we introduce an Ensemble Multi-Objective RL (EMORL)
framework that fine-tunes multiple models with individual objectives while
optimizing their aggregation after the training to improve efficiency and
flexibility. Our method is the first to aggregate the last hidden states of
individual models, incorporating contextual information from multiple
objectives. This approach is supported by a hierarchical grid search algorithm
that identifies optimal weighted combinations. We evaluate EMORL on counselor
reflection generation tasks, using text-scoring LLMs to evaluate the
generations and provide rewards during RL fine-tuning. Through comprehensive
experiments on the PAIR and Psych8k datasets, we demonstrate the advantages of
EMORL against existing baselines: significantly lower and more stable training
consumption ($17,529\pm 1,650$ data points and $6,573\pm 147.43$ seconds),
improved scalability and explainability, and comparable performance across
multiple objectives.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 11:30:46 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 06:26:11 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Kong",
"Lingxiao",
""
],
[
"Yang",
"Cong",
""
],
[
"Neufang",
"Susanne",
""
],
[
"Beyan",
"Oya Deniz",
""
],
[
"Boukhers",
"Zeyd",
""
]
] |
2505.02592 | I-Chao Shen | Yuki Tatsukawa, Anran Qi, I-Chao Shen, Takeo Igarashi | GarmentImage: Raster Encoding of Garment Sewing Patterns with Diverse
Topologies | 10 pages, SIGGRAPH 2025 (Conference Track) | null | null | null | cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Garment sewing patterns are the design language behind clothing, yet their
current vector-based digital representations weren't built with machine
learning in mind. Vector-based representation encodes a sewing pattern as a
discrete set of panels, each defined as a sequence of lines and curves,
stitching information between panels and the placement of each panel around a
body. However, this representation causes two major challenges for neural
networks: discontinuity in latent space between patterns with different
topologies and limited generalization to garments with unseen topologies in the
training data. In this work, we introduce GarmentImage, a unified raster-based
sewing pattern representation. GarmentImage encodes a garment sewing pattern's
geometry, topology and placement into multi-channel regular grids. Machine
learning models trained on GarmentImage achieve seamless transitions between
patterns with different topologies and show better generalization capabilities
compared to models trained on vector-based representation. We demonstrate the
effectiveness of GarmentImage across three applications: pattern exploration in
latent space, text-based pattern editing, and image-to-pattern prediction. The
results show that GarmentImage achieves superior performance on these
applications using only simple convolutional networks.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 11:59:20 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 08:39:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Tatsukawa",
"Yuki",
""
],
[
"Qi",
"Anran",
""
],
[
"Shen",
"I-Chao",
""
],
[
"Igarashi",
"Takeo",
""
]
] |
2505.02619 | Ayda Gholamhosseinian | A. Gholamhosseinian, R. W. Corkery, I. Brevik, Mathias Bostrom | Can An Uncertainty Relation Generate A Plasma? | null | null | null | null | physics.gen-ph | http://creativecommons.org/licenses/by/4.0/ | We explore the fundamental idea that there may be a role for the Casimir
effect, via an uncertainty relation, in the generation of electron-positron and
quark-gluon plasmas. We investigate this concept, reviewing the possible
contribution of semi-classical electrodynamics to nuclear interactions,
specifically focusing on the Casimir effect at sub-Fermi length scales. The
main result is a temperature distance relation, derived from the time-energy
uncertainty relation, which can have observable consequences at these extreme
scales. From a more general perspective, since the energy-time uncertainty
relation appears to be a significant physical quantity, we also provide a brief
overview of recent developments in this direction in Sec. 3.2.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 12:48:06 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Gholamhosseinian",
"A.",
""
],
[
"Corkery",
"R. W.",
""
],
[
"Brevik",
"I.",
""
],
[
"Bostrom",
"Mathias",
""
]
] |
2505.02641 | William McDonough | William F McDonough | Earths composition: origin, evolution and energy budget | 47 pages, 9 figures and 8 tables; update has the correction for the
wrong input for H in bulk silicate Earth in Table 3 and for H in the bulk
Earth in Table 8 | null | null | null | physics.geo-ph astro-ph.EP astro-ph.SR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | One in every two atoms in the Earth, Mars, and the Moon is oxygen; it is the
third most abundant element in the solar system. The oxygen isotopic
compositions of the terrestrial planets are different from those of the Sun and
demonstrate that these planets are not direct compositional analogs of the
solar photosphere. Likewise, the Suns O/Fe, Fe/Mg and Mg/Si values are distinct
from those of inner solar system chondrites and terrestrial planets. These four
elements (O, Fe, Mg, Si) make up about 94% by mass of the rocky planets and
their abundances are determined uniquely using geophysical, geochemical and
cosmochemical constraints. The rocky planets grew rapidly from planetesimals,
most of which were differentiated, having a core and a mantle, before being
accreted. Planetary growth in the early stages of protoplanetary disk evolution
was rapid and was only partially recorded by the meteoritic record. The
noncarbonaceous meteorites (NC) provide insights into the early history of the
inner solar system and are used to construct a framework for how the rocky
planets were assembled. NC chondrites have chondrule ages that are two to three
million years younger than t_zero (the age of calcium-aluminum inclusions,
CAI), documenting that chondrites are middle- to late-stage products of solar
system evolution. The composition of the Earth, its current form of mantle
convection, and the amount of radiogenic power that drives its engine remain
controversial topics. Earths dynamics are driven by primordial and radiogenic
heat sources. Measurement of the Earths geoneutrino flux defines its radiogenic
power and restricts its bulk composition. Using the latest data from the
KamLAND and Borexino geoneutrino experiments affirms that the Earth has 20 TW
of radiogenic power and sets the proportions of refractory lithophile elements
in the bulk silicate Earth at 2.5 to 2.7 times that in CI chondrites.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 13:36:50 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 06:08:27 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"McDonough",
"William F",
""
]
] |
2505.02648 | Mingcheng Li | Mingcheng Li, Xiaolu Hou, Ziyang Liu, Dingkang Yang, Ziyun Qian,
Jiawei Chen, Jinjie Wei, Yue Jiang, Qingyao Xu, Lihua Zhang | MCCD: Multi-Agent Collaboration-based Compositional Diffusion for
Complex Text-to-Image Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models have shown excellent performance in text-to-image
generation. Nevertheless, existing methods often suffer from performance
bottlenecks when handling complex prompts that involve multiple objects,
characteristics, and relations. Therefore, we propose a Multi-agent
Collaboration-based Compositional Diffusion (MCCD) for text-to-image generation
for complex scenes. Specifically, we design a multi-agent collaboration-based
scene parsing module that generates an agent system comprising multiple agents
with distinct tasks, utilizing MLLMs to extract various scene elements
effectively. In addition, Hierarchical Compositional diffusion utilizes a
Gaussian mask and filtering to refine bounding box regions and enhance objects
through region enhancement, resulting in the accurate and high-fidelity
generation of complex scenes. Comprehensive experiments demonstrate that our
MCCD significantly improves the performance of the baseline models in a
training-free manner, providing a substantial advantage in complex scene
generation.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 13:50:03 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 15:18:25 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Li",
"Mingcheng",
""
],
[
"Hou",
"Xiaolu",
""
],
[
"Liu",
"Ziyang",
""
],
[
"Yang",
"Dingkang",
""
],
[
"Qian",
"Ziyun",
""
],
[
"Chen",
"Jiawei",
""
],
[
"Wei",
"Jinjie",
""
],
[
"Jiang",
"Yue",
""
],
[
"Xu",
"Qingyao",
""
],
[
"Zhang",
"Lihua",
""
]
] |
2505.02659 | Andrey Sidorenko | Andrey Sidorenko | A Note on Statistically Accurate Tabular Data Generation Using Large
Language Models | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have shown promise in synthetic tabular data
generation, yet existing methods struggle to preserve complex feature
dependencies, particularly among categorical variables. This work introduces a
probability-driven prompting approach that leverages LLMs to estimate
conditional distributions, enabling more accurate and scalable data synthesis.
The results highlight the potential of prompting probability distributions to
enhance the statistical fidelity of LLM-generated tabular data.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 14:05:15 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 08:34:46 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Sidorenko",
"Andrey",
""
]
] |
2505.02704 | Bojin Wu | Bojin Wu, Jing Chen | VGLD: Visually-Guided Linguistic Disambiguation for Monocular Depth
Scale Recovery | 21 pages, conference | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a robust method for monocular depth scale recovery. Monocular
depth estimation can be divided into two main directions: (1) relative depth
estimation, which provides normalized or inverse depth without scale
information, and (2) metric depth estimation, which involves recovering depth
with absolute scale. To obtain absolute scale information for practical
downstream tasks, utilizing textual information to recover the scale of a
relative depth map is a highly promising approach. However, since a single
image can have multiple descriptions from different perspectives or with
varying styles, it has been shown that different textual descriptions can
significantly affect the scale recovery process. To address this issue, our
method, VGLD, stabilizes the influence of textual information by incorporating
high-level semantic information from the corresponding image alongside the
textual description. This approach resolves textual ambiguities and robustly
outputs a set of linear transformation parameters (scalars) that can be
globally applied to the relative depth map, ultimately generating depth
predictions with metric-scale accuracy. We validate our method across several
popular relative depth models(MiDas, DepthAnything), using both indoor scenes
(NYUv2) and outdoor scenes (KITTI). Our results demonstrate that VGLD functions
as a universal alignment module when trained on multiple datasets, achieving
strong performance even in zero-shot scenarios. Code is available at:
https://github.com/pakinwu/VGLD.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 14:57:16 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 03:06:28 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wu",
"Bojin",
""
],
[
"Chen",
"Jing",
""
]
] |
2505.02737 | Gerard Pons | Gerard Pons, Besim Bilalli and Anna Queralt | Knowledge Graphs for Enhancing Large Language Models in Entity
Disambiguation | Pre-print submitted to ISWC 2024 | Proc. 23rd Int. Semantic Web Conf. (ISWC 2024), LNCS, Springer,
2024 | 10.1007/978-3-031-77844-5_9 | null | cs.LG cs.AI cs.DB | http://creativecommons.org/licenses/by/4.0/ | Recent advances in Large Language Models (LLMs) have positioned them as a
prominent solution for Natural Language Processing tasks. Notably, they can
approach these problems in a zero or few-shot manner, thereby eliminating the
need for training or fine-tuning task-specific models. However, LLMs face some
challenges, including hallucination and the presence of outdated knowledge or
missing information from specific domains in the training data. These problems
cannot be easily solved by retraining the models with new data as it is a
time-consuming and expensive process. To mitigate these issues, Knowledge
Graphs (KGs) have been proposed as a structured external source of information
to enrich LLMs. With this idea, in this work we use KGs to enhance LLMs for
zero-shot Entity Disambiguation (ED). For that purpose, we leverage the
hierarchical representation of the entities' classes in a KG to gradually prune
the candidate space as well as the entities' descriptions to enrich the input
prompt with additional factual knowledge. Our evaluation on popular ED datasets
shows that the proposed method outperforms non-enhanced and description-only
enhanced LLMs, and has a higher degree of adaptability than task-specific
models. Furthermore, we conduct an error analysis and discuss the impact of the
leveraged KG's semantic expressivity on the ED performance.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 15:40:24 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 06:44:35 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Pons",
"Gerard",
""
],
[
"Bilalli",
"Besim",
""
],
[
"Queralt",
"Anna",
""
]
] |
2505.02741 | Yihang Yuan Mr | Yihang Yuan, Ali Aghdaei, Zhuo Feng | dyGRASS: Dynamic Spectral Graph Sparsification via Localized Random
Walks on GPUs | null | null | null | null | cs.SI cs.NA math.NA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This work presents dyGRASS, an efficient dynamic algorithm for spectral
sparsification of large undirected graphs that undergo streaming edge
insertions and deletions. At its core, dyGRASS employs a random-walk-based
method to efficiently estimate node-to-node distances in both the original
graph (for decremental update) and its sparsifier (for incremental update). For
incremental updates, dyGRASS enables the identification of spectrally critical
edges among the updates to capture the latest structural changes. For
decremental updates, dyGRASS facilitates the recovery of important edges from
the original graph back into the sparsifier. To further enhance computational
efficiency, dyGRASS employs a GPU-based non-backtracking random walk scheme
that allows multiple walkers to operate simultaneously across various target
updates. This parallelization significantly improves both the performance and
scalability of the proposed dyGRASS framework. Our comprehensive experimental
evaluations reveal that dyGRASS achieves approximately a 10x speedup compared
to the state-of-the-art incremental sparsification (inGRASS) algorithm while
eliminating the setup overhead and improving solution quality in incremental
spectral sparsification tasks. Moreover, dyGRASS delivers high efficiency and
superior solution quality for fully dynamic graph sparsification, accommodating
both edge insertions and deletions across a diverse range of graph instances
originating from integrated circuit simulations, finite element analysis, and
social networks.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 15:48:44 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 16:45:21 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Yuan",
"Yihang",
""
],
[
"Aghdaei",
"Ali",
""
],
[
"Feng",
"Zhuo",
""
]
] |
2505.02774 | Arash Ahmadi | Arash Ahmadi, Ya\u{g}{\i}z Murat, Pei-Chen Kuan, Mustafa
G\"undo\u{g}an and Markus Krutzik | Controlled Displacement of Stored Light at Room Temperature | 5 pages, 3 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We report the demonstration of spatially translating a stored optical pulse
at room temperature over distances exceeding one optical wavelength. By
implementing an interferometric scheme, we further measure the average speed of
this linear translation, thus harnessing a stopped-light experiment for a
sensing application. This work extends the use of quantum memories beyond
quantum communication and information contexts, opening a pathway to novel
methods of velocity measurements with high sensitivity.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 16:37:18 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 11:47:40 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ahmadi",
"Arash",
""
],
[
"Murat",
"Yağız",
""
],
[
"Kuan",
"Pei-Chen",
""
],
[
"Gündoğan",
"Mustafa",
""
],
[
"Krutzik",
"Markus",
""
]
] |
2505.02825 | Otto Brookes | Alex Hoi Hang Chan, Otto Brookes, Urs Waldmann, Hemal Naik, Iain D.
Couzin, Majid Mirmehdi, No\"el Adiko Houa, Emmanuelle Normand, Christophe
Boesch, Lukas Boesch, Mimi Arandjelovic, Hjalmar K\"uhl, Tilo Burghardt,
Fumihiro Kano | Towards Application-Specific Evaluation of Vision Models: Case Studies
in Ecology and Biology | Accepted at CVPR Workshop, CV4Animals 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Computer vision methods have demonstrated considerable potential to
streamline ecological and biological workflows, with a growing number of
datasets and models becoming available to the research community. However,
these resources focus predominantly on evaluation using machine learning
metrics, with relatively little emphasis on how their application impacts
downstream analysis. We argue that models should be evaluated using
application-specific metrics that directly represent model performance in the
context of its final use case. To support this argument, we present two
disparate case studies: (1) estimating chimpanzee abundance and density with
camera trap distance sampling when using a video-based behaviour classifier and
(2) estimating head rotation in pigeons using a 3D posture estimator. We show
that even models with strong machine learning performance (e.g., 87% mAP) can
yield data that leads to discrepancies in abundance estimates compared to
expert-derived data. Similarly, the highest-performing models for posture
estimation do not produce the most accurate inferences of gaze direction in
pigeons. Motivated by these findings, we call for researchers to integrate
application-specific metrics in ecological/biological datasets, allowing for
models to be benchmarked in the context of their downstream application and to
facilitate better integration of models into application workflows.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 17:51:56 GMT"
},
{
"version": "v2",
"created": "Tue, 6 May 2025 10:17:58 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Chan",
"Alex Hoi Hang",
""
],
[
"Brookes",
"Otto",
""
],
[
"Waldmann",
"Urs",
""
],
[
"Naik",
"Hemal",
""
],
[
"Couzin",
"Iain D.",
""
],
[
"Mirmehdi",
"Majid",
""
],
[
"Houa",
"Noël Adiko",
""
],
[
"Normand",
"Emmanuelle",
""
],
[
"Boesch",
"Christophe",
""
],
[
"Boesch",
"Lukas",
""
],
[
"Arandjelovic",
"Mimi",
""
],
[
"Kühl",
"Hjalmar",
""
],
[
"Burghardt",
"Tilo",
""
],
[
"Kano",
"Fumihiro",
""
]
] |
2505.02838 | K Mahesh Krishna | K. Mahesh Krishna | p-adic Heisenberg-Robertson-Schrodinger and p-adic Maccone-Pati
Uncertainty Principles | 6 Pages, 0 Figures | null | null | null | math.FA math.OA | http://creativecommons.org/licenses/by/4.0/ | Let $\mathcal{X}$ be a p-adic Hilbert space. Let $A:\mathcal{D}(A)\subseteq
\mathcal{X}\to \mathcal{X}$ and $B: \mathcal{D}(B)\subseteq \mathcal{X}\to
\mathcal{X}$ be possibly unbounded linear operators. For $x \in \mathcal{D}(A)$
with $\langle x, x \rangle =1$, define $ \Delta_x(A):= \|Ax- \langle Ax, x
\rangle x \|.$ Then for all $x \in \mathcal{D}(AB)\cap \mathcal{D}(BA)$ with
$\langle x, x \rangle =1$, we show that \begin{align*} (1) \quad \quad \quad
\max\{\Delta_x(A), \Delta_x(B)\}\geq \frac{\sqrt{\bigg|\big\langle [A,B]x, x
\big\rangle ^2+\big(\langle \{A,B\}x, x \rangle -2\langle Ax, x \rangle\langle
Bx, x \rangle\big)^2\bigg|}}{\sqrt{|2|}} \end{align*} and \begin{align*} (2)
\quad \quad \quad \max\{\Delta_x(A), \Delta_x(B)\} \geq |\langle (A+B)x, y
\rangle |, \quad \forall y \in \mathcal{X} \text{ satisfying } \|y\|\leq 1,
\langle x, y \rangle =0. \end{align*} We call Inequality (1) as p-adic
Heisenberg-Robertson-Schrodinger uncertainty principle and Inequality (2) as
p-adic Maccone-Pati uncertainty principle.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 03:30:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Krishna",
"K. Mahesh",
""
]
] |
2505.02839 | Fabio Bagarello Dr. | Fabio Bagarello | A raised hand effect as a decision making process | In press in "Quantum Economics and Finance", for the special issue
"Mathematical Aspects in Non-Equilibrium Systems: from Micro to Macro". arXiv
admin note: text overlap with arXiv:2303.05278 | null | null | null | physics.soc-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we will analyse a group of agents and their attitude to follow,
or not, some rules. The model is based on some quantum-like ideas, and in
particular on an Hamiltonian operator $H$ describing the dynamics of the
agents, assuming they are driven by some mutual interactions and that they are
subjected to an external source of "information" used by the agents to decide
whether to obey or not these rules. We will discuss how the relative strengths
of the parameters of $H$ determine this attitude and we will discuss in
particular the role of the external information. We will also apply our general
idea to a specific situation, involving drivers and pedestrians trying to cross
a road.
| [
{
"version": "v1",
"created": "Thu, 17 Apr 2025 19:04:44 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Bagarello",
"Fabio",
""
]
] |
2505.02841 | Ivan Monlineris | Marco Masera, Alessandro Leone, Johannes K\"oster and Ivan Molineris | Snakemaker: Seamlessly transforming ad-hoc analyses into sustainable
Snakemake workflows with generative AI | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Reproducibility and sustainability present significant challenges in
bioinformatics software development, where rapidly evolving tools and complex
workflows often result in short-lived or difficult-to-adapt pipelines. This
paper introduces Snakemaker, a tool that leverages generative AI to facilitate
researchers build sustainable data analysis pipelines by converting
unstructured code into well-defined Snakemake workflows. Snakemaker
non-invasively tracks the work performed in the terminal by the researcher,
analyzes execution patterns, and generates Snakemake workflows that can be
integrated into existing pipelines. Snakemaker also supports the transformation
of monolithic Ipython Notebooks into modular Snakemake pipelines, resolving the
global state of the notebook into discrete, file-based interactions between
rules. An integrated chat assistant provides users with fine-grained control
through natural language instructions. Snakemaker generates high-quality
Snakemake workflows by adhering to the best practices, including Conda
environment tracking, generic rule generation and loop unrolling. By lowering
the barrier between prototype and production-quality code, Snakemaker addresses
a critical gap in computational reproducibility for bioinformatics research.
| [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 06:00:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Masera",
"Marco",
""
],
[
"Leone",
"Alessandro",
""
],
[
"Köster",
"Johannes",
""
],
[
"Molineris",
"Ivan",
""
]
] |
2505.02842 | Olga Mironenko | Olga Mironenko, Hadi Banaee, Amy Loutfi | Evaluation of Coordination Strategies for Underground Automated Vehicle
Fleets in Mixed Traffic | Accepted for publication in the Proceedings of the 2025 IEEE
Intelligent Vehicles Symposium (IV 2025) | null | null | null | physics.soc-ph cs.HC cs.MA cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates the efficiency and safety outcomes of implementing
different adaptive coordination models for automated vehicle (AV) fleets,
managed by a centralized coordinator that dynamically responds to
human-controlled vehicle behavior. The simulated scenarios replicate an
underground mining environment characterized by narrow tunnels with limited
connectivity. To address the unique challenges of such settings, we propose a
novel metric - Path Overlap Density (POD) - to predict efficiency and
potentially the safety performance of AV fleets. The study also explores the
impact of map features on AV fleets performance. The results demonstrate that
both AV fleet coordination strategies and underground tunnel network
characteristics significantly influence overall system performance. While map
features are critical for optimizing efficiency, adaptive coordination
strategies are essential for ensuring safe operations.
| [
{
"version": "v1",
"created": "Sun, 27 Apr 2025 16:22:55 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Mironenko",
"Olga",
""
],
[
"Banaee",
"Hadi",
""
],
[
"Loutfi",
"Amy",
""
]
] |
2505.02843 | Miriam Cobo MSc | Miriam Cobo, David Corral Fontecha, Wilson Silva, Lara Lloret Iglesias | Physical foundations for trustworthy medical imaging: a review for
artificial intelligence researchers | 17 pages, 2 figures | null | null | null | eess.IV cs.AI cs.CV physics.med-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Artificial intelligence in medical imaging has seen unprecedented growth in
the last years, due to rapid advances in deep learning and computing resources.
Applications cover the full range of existing medical imaging modalities, with
unique characteristics driven by the physics of each technique. Yet, artificial
intelligence professionals entering the field, and even experienced developers,
often lack a comprehensive understanding of the physical principles underlying
medical image acquisition, which hinders their ability to fully leverage its
potential. The integration of physics knowledge into artificial intelligence
algorithms enhances their trustworthiness and robustness in medical imaging,
especially in scenarios with limited data availability. In this work, we review
the fundamentals of physics in medical images and their impact on the latest
advances in artificial intelligence, particularly, in generative models and
reconstruction algorithms. Finally, we explore the integration of physics
knowledge into physics-inspired machine learning models, which leverage
physics-based constraints to enhance the learning of medical imaging features.
| [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 09:35:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Cobo",
"Miriam",
""
],
[
"Fontecha",
"David Corral",
""
],
[
"Silva",
"Wilson",
""
],
[
"Iglesias",
"Lara Lloret",
""
]
] |
2505.02844 | Zhikai Wang | Zhikai Wang, Yanyan Shen, Zibin Zhang, Kangyi Lin | Feature Staleness Aware Incremental Learning for CTR Prediction | The code can be found in https://github.com/cloudcatcher888/FeSAIL | null | 10.24963/ijcai.2023/261 | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Click-through Rate (CTR) prediction in real-world recommender systems often
deals with billions of user interactions every day. To improve the training
efficiency, it is common to update the CTR prediction model incrementally using
the new incremental data and a subset of historical data. However, the feature
embeddings of a CTR prediction model often get stale when the corresponding
features do not appear in current incremental data. In the next period, the
model would have a performance degradation on samples containing stale
features, which we call the feature staleness problem. To mitigate this
problem, we propose a Feature Staleness Aware Incremental Learning method for
CTR prediction (FeSAIL) which adaptively replays samples containing stale
features. We first introduce a staleness aware sampling algorithm (SAS) to
sample a fixed number of stale samples with high sampling efficiency. We then
introduce a staleness aware regularization mechanism (SAR) for a fine-grained
control of the feature embedding updating. We instantiate FeSAIL with a general
deep learning-based CTR prediction model and the experimental results
demonstrate FeSAIL outperforms various state-of-the-art methods on four
benchmark datasets.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 09:05:47 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Zhikai",
""
],
[
"Shen",
"Yanyan",
""
],
[
"Zhang",
"Zibin",
""
],
[
"Lin",
"Kangyi",
""
]
] |
2505.02845 | Jeremias Gerner | Jeremias Gerner, Klaus Bogenberger, Stefanie Schmidtner | Floating Car Observers in Intelligent Transportation Systems: Detection
Modeling and Temporal Insights | null | null | null | null | physics.soc-ph cs.CV cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Floating Car Observers (FCOs) extend traditional Floating Car Data (FCD) by
integrating onboard sensors to detect and localize other traffic participants,
providing richer and more detailed traffic data. In this work, we explore
various modeling approaches for FCO detections within microscopic traffic
simulations to evaluate their potential for Intelligent Transportation System
(ITS) applications. These approaches range from 2D raytracing to high-fidelity
co-simulations that emulate real-world sensors and integrate 3D object
detection algorithms to closely replicate FCO detections. Additionally, we
introduce a neural network-based emulation technique that effectively
approximates the results of high-fidelity co-simulations. This approach
captures the unique characteristics of FCO detections while offering a fast and
scalable solution for modeling. Using this emulation method, we investigate the
impact of FCO data in a digital twin of a traffic network modeled in SUMO.
Results demonstrate that even at a 20% penetration rate, FCOs using LiDAR-based
detections can identify 65% of vehicles across various intersections and
traffic demand scenarios. Further potential emerges when temporal insights are
integrated, enabling the recovery of previously detected but currently unseen
vehicles. By employing data-driven methods, we recover over 80% of these
vehicles with minimal positional deviations. These findings underscore the
potential of FCOs for ITS, particularly in enhancing traffic state estimation
and monitoring under varying penetration rates and traffic conditions.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 19:38:13 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Gerner",
"Jeremias",
""
],
[
"Bogenberger",
"Klaus",
""
],
[
"Schmidtner",
"Stefanie",
""
]
] |
2505.02846 | Kim Kaivanto | Kim Kaivanto | The Precautionary Principle and the Innovation Principle: Incompatible
Guides for AI Innovation Governance? | 47 pages | null | null | null | cs.CY cs.AI econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In policy debates concerning the governance and regulation of Artificial
Intelligence (AI), both the Precautionary Principle (PP) and the Innovation
Principle (IP) are advocated by their respective interest groups. Do these
principles offer wholly incompatible and contradictory guidance? Does one
necessarily negate the other? I argue here that provided attention is
restricted to weak-form PP and IP, the answer to both of these questions is
"No." The essence of these weak formulations is the requirement to fully
account for type-I error costs arising from erroneously preventing the
innovation's diffusion through society (i.e. mistaken regulatory red-lighting)
as well as the type-II error costs arising from erroneously allowing the
innovation to diffuse through society (i.e. mistaken regulatory
green-lighting). Within the Signal Detection Theory (SDT) model developed here,
weak-PP red-light (weak-IP green-light) determinations are optimal for
sufficiently small (large) ratios of expected type-I to type-II error costs.
For intermediate expected cost ratios, an amber-light 'wait-and-monitor' policy
is optimal. Regulatory sandbox instruments allow AI testing and experimentation
to take place within a structured environment of limited duration and societal
scale, whereby the expected cost ratio falls within the 'wait-and-monitor'
range. Through sandboxing regulators and innovating firms learn more about the
expected cost ratio, and what respective adaptations -- of regulation, of
technical solution, of business model, or combination thereof, if any -- are
needed to keep the ratio out of the weak-PP red-light zone.
| [
{
"version": "v1",
"created": "Thu, 1 May 2025 17:48:18 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Kaivanto",
"Kim",
""
]
] |
2505.02847 | Zhaopeng Tu | Bang Zhang, Ruotian Ma, Qingxuan Jiang, Peisong Wang, Jiaqi Chen,
Zheng Xie, Xingyu Chen, Yue Wang, Fanghua Ye, Jian Li, Yifan Yang, Zhaopeng
Tu, Xiaolong Li | Sentient Agent as a Judge: Evaluating Higher-Order Social Cognition in
Large Language Models | null | null | null | null | cs.CL cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | Assessing how well a large language model (LLM) understands human, rather
than merely text, remains an open challenge. To bridge the gap, we introduce
Sentient Agent as a Judge (SAGE), an automated evaluation framework that
measures an LLM's higher-order social cognition. SAGE instantiates a Sentient
Agent that simulates human-like emotional changes and inner thoughts during
interaction, providing a more realistic evaluation of the tested model in
multi-turn conversations. At every turn, the agent reasons about (i) how its
emotion changes, (ii) how it feels, and (iii) how it should reply, yielding a
numerical emotion trajectory and interpretable inner thoughts. Experiments on
100 supportive-dialogue scenarios show that the final Sentient emotion score
correlates strongly with Barrett-Lennard Relationship Inventory (BLRI) ratings
and utterance-level empathy metrics, validating psychological fidelity. We also
build a public Sentient Leaderboard covering 18 commercial and open-source
models that uncovers substantial gaps (up to 4x) between frontier systems
(GPT-4o-Latest, Gemini2.5-Pro) and earlier baselines, gaps not reflected in
conventional leaderboards (e.g., Arena). SAGE thus provides a principled,
scalable and interpretable tool for tracking progress toward genuinely
empathetic and socially adept language agents.
| [
{
"version": "v1",
"created": "Thu, 1 May 2025 19:06:10 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Zhang",
"Bang",
""
],
[
"Ma",
"Ruotian",
""
],
[
"Jiang",
"Qingxuan",
""
],
[
"Wang",
"Peisong",
""
],
[
"Chen",
"Jiaqi",
""
],
[
"Xie",
"Zheng",
""
],
[
"Chen",
"Xingyu",
""
],
[
"Wang",
"Yue",
""
],
[
"Ye",
"Fanghua",
""
],
[
"Li",
"Jian",
""
],
[
"Yang",
"Yifan",
""
],
[
"Tu",
"Zhaopeng",
""
],
[
"Li",
"Xiaolong",
""
]
] |
2505.02848 | Kexin Ding | Kexin Ding, Mu Zhou, Akshay Chaudhari, Shaoting Zhang, Dimitris N.
Metaxas | Aligning Large Language Models with Healthcare Stakeholders: A Pathway
to Trustworthy AI Integration | null | null | null | null | cs.CY cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The wide exploration of large language models (LLMs) raises the awareness of
alignment between healthcare stakeholder preferences and model outputs. This
alignment becomes a crucial foundation to empower the healthcare workflow
effectively, safely, and responsibly. Yet the varying behaviors of LLMs may not
always match with healthcare stakeholders' knowledge, demands, and values. To
enable a human-AI alignment, healthcare stakeholders will need to perform
essential roles in guiding and enhancing the performance of LLMs. Human
professionals must participate in the entire life cycle of adopting LLM in
healthcare, including training data curation, model training, and inference. In
this review, we discuss the approaches, tools, and applications of alignments
between healthcare stakeholders and LLMs. We demonstrate that LLMs can better
follow human values by properly enhancing healthcare knowledge integration,
task understanding, and human guidance. We provide outlooks on enhancing the
alignment between humans and LLMs to build trustworthy real-world healthcare
applications.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 00:59:49 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ding",
"Kexin",
""
],
[
"Zhou",
"Mu",
""
],
[
"Chaudhari",
"Akshay",
""
],
[
"Zhang",
"Shaoting",
""
],
[
"Metaxas",
"Dimitris N.",
""
]
] |
2505.02849 | Wenli Yang | Mohsen Balavar, Wenli Yang, David Herbert and Soonja Yeom | Enhancing tutoring systems by leveraging tailored promptings and domain
knowledge with Large Language Models | null | null | null | null | cs.CY cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in artificial intelligence (AI) and machine learning have
reignited interest in their impact on Computer-based Learning (CBL). AI-driven
tools like ChatGPT and Intelligent Tutoring Systems (ITS) have enhanced
learning experiences through personalisation and flexibility. ITSs can adapt to
individual learning needs and provide customised feedback based on a student's
performance, cognitive state, and learning path. Despite these advances,
challenges remain in accommodating diverse learning styles and delivering
real-time, context-aware feedback. Our research aims to address these gaps by
integrating skill-aligned feedback via Retrieval Augmented Generation (RAG)
into prompt engineering for Large Language Models (LLMs) and developing an
application to enhance learning through personalised tutoring in a computer
science programming context. The pilot study evaluated a proposed system using
three quantitative metrics: readability score, response time, and feedback
depth, across three programming tasks of varying complexity. The system
successfully sorted simulated students into three skill-level categories and
provided context-aware feedback. This targeted approach demonstrated better
effectiveness and adaptability compared to general methods.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 02:30:39 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Balavar",
"Mohsen",
""
],
[
"Yang",
"Wenli",
""
],
[
"Herbert",
"David",
""
],
[
"Yeom",
"Soonja",
""
]
] |
2505.02850 | Nicy Scaria | Nicy Scaria, Silvester John Joseph Kennedy, Diksha Seth, Ananya
Thakur, Deepak Subramani | Harnessing Structured Knowledge: A Concept Map-Based Approach for
High-Quality Multiple Choice Question Generation with Effective Distractors | null | null | null | null | cs.CL cs.AI cs.CY cs.DB | http://creativecommons.org/licenses/by-sa/4.0/ | Generating high-quality MCQs, especially those targeting diverse cognitive
levels and incorporating common misconceptions into distractor design, is
time-consuming and expertise-intensive, making manual creation impractical at
scale. Current automated approaches typically generate questions at lower
cognitive levels and fail to incorporate domain-specific misconceptions. This
paper presents a hierarchical concept map-based framework that provides
structured knowledge to guide LLMs in generating MCQs with distractors. We
chose high-school physics as our test domain and began by developing a
hierarchical concept map covering major Physics topics and their
interconnections with an efficient database design. Next, through an automated
pipeline, topic-relevant sections of these concept maps are retrieved to serve
as a structured context for the LLM to generate questions and distractors that
specifically target common misconceptions. Lastly, an automated validation is
completed to ensure that the generated MCQs meet the requirements provided. We
evaluate our framework against two baseline approaches: a base LLM and a
RAG-based generation. We conducted expert evaluations and student assessments
of the generated MCQs. Expert evaluation shows that our method significantly
outperforms the baseline approaches, achieving a success rate of 75.20% in
meeting all quality criteria compared to approximately 37% for both baseline
methods. Student assessment data reveal that our concept map-driven approach
achieved a significantly lower guess success rate of 28.05% compared to 37.10%
for the baselines, indicating a more effective assessment of conceptual
understanding. The results demonstrate that our concept map-based approach
enables robust assessment across cognitive levels and instant identification of
conceptual gaps, facilitating faster feedback loops and targeted interventions
at scale.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 06:36:06 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Scaria",
"Nicy",
""
],
[
"Kennedy",
"Silvester John Joseph",
""
],
[
"Seth",
"Diksha",
""
],
[
"Thakur",
"Ananya",
""
],
[
"Subramani",
"Deepak",
""
]
] |
2505.02851 | Franklin Zhang | Franklin Zhang, Sonya Zhang, Alon Halevy | 30DayGen: Leveraging LLMs to Create a Content Corpus for Habit Formation | 8 pages (main content), 4 figures. Submitted to ACL BEA2025 | null | null | null | cs.CL cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present 30 Day Me, a habit formation application that
leverages Large Language Models (LLMs) to help users break down their goals
into manageable, actionable steps and track their progress. Central to the app
is the 30DAYGEN system, which generates 3,531 unique 30-day challenges sourced
from over 15K webpages, and enables runtime search of challenge ideas aligned
with user-defined goals. We showcase how LLMs can be harnessed to rapidly
construct domain specific content corpora for behavioral and educational
purposes, and propose a practical pipeline that incorporates effective LLM
enhanced approaches for content generation and semantic deduplication.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 08:53:27 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Zhang",
"Franklin",
""
],
[
"Zhang",
"Sonya",
""
],
[
"Halevy",
"Alon",
""
]
] |
2505.02852 | Vinh Truong Dr | Vinh Truong (RMIT University) | Examining gender and cultural influences on customer emotions | null | null | null | null | econ.GN q-fin.EC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding consumer emotional experiences on e-commerce platforms is
essential for businesses striving to enhance customer engagement and
personalisation. Recent research has demonstrated that these experiences are
more intricate and diverse than previously examined, encompassing a wider range
of discrete emotions and spanning multiple-dimensional scales. This study
examines how gender and cultural differences shape these complex emotional
responses, revealing significant variations between male and female consumers
across all sentiment, valence, arousal, and dominance scores. Additionally,
clear cultural distinctions emerge, with Western and Eastern consumers
displaying markedly different emotional behaviours across the larger spectrum
of emotions, including admiration, amusement, approval, caring, curiosity,
desire, disappointment, optimism, and pride. Furthermore, the study uncovers a
critical interaction between gender and culture in shaping consumer emotions.
Notably, gender-based emotional disparities are more pronounced in Western
cultures than in Eastern ones, an aspect that has been largely overlooked in
previous research. From a theoretical perspective, this study advances the
understanding of gender and cultural variations in online consumer behaviour by
integrating insights from neuroscience theories and Hofstede cultural dimension
model. Practically, it offers valuable guidance for businesses, equipping them
with the tools to more accurately interpret customer feedback, refine sentiment
and emotional analysis models, and develop personalised marketing strategies.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 09:06:05 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Truong",
"Vinh",
"",
"RMIT University"
]
] |
2505.02853 | Francesco Balzan Dr | Francesco Balzan, Pedro P. Santos, Maurizio Gabbrielli, Mahault
Albarracin, Manuel Lopes | A Computational Model of Inclusive Pedagogy: From Understanding to
Application | This is a preprint version of a manuscript intended for submission to
the International Journal of Artificial Intelligence in Education (IJAIED) | null | null | null | cs.CY cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Human education transcends mere knowledge transfer, it relies on
co-adaptation dynamics -- the mutual adjustment of teaching and learning
strategies between agents. Despite its centrality, computational models of
co-adaptive teacher-student interactions (T-SI) remain underdeveloped. We argue
that this gap impedes Educational Science in testing and scaling contextual
insights across diverse settings, and limits the potential of Machine Learning
systems, which struggle to emulate and adaptively support human learning
processes. To address this, we present a computational T-SI model that
integrates contextual insights on human education into a testable framework. We
use the model to evaluate diverse T-SI strategies in a realistic synthetic
classroom setting, simulating student groups with unequal access to sensory
information. Results show that strategies incorporating co-adaptation
principles (e.g., bidirectional agency) outperform unilateral approaches (i.e.,
where only the teacher or the student is active), improving the learning
outcomes for all learning types. Beyond the testing and scaling of
context-dependent educational insights, our model enables hypothesis generation
in controlled yet adaptable environments. This work bridges non-computational
theories of human education with scalable, inclusive AI in Education systems,
providing a foundation for equitable technologies that dynamically adapt to
learner needs.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 12:26:31 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Balzan",
"Francesco",
""
],
[
"Santos",
"Pedro P.",
""
],
[
"Gabbrielli",
"Maurizio",
""
],
[
"Albarracin",
"Mahault",
""
],
[
"Lopes",
"Manuel",
""
]
] |
2505.02854 | Masumi Morishige | Masumi Morishige, Ryo Koshihara | Ensuring Reproducibility in Generative AI Systems for General Use Cases:
A Framework for Regression Testing and Open Datasets | 15 pages, 10 figures | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Reproducibility and reliability remain pressing challenges for generative AI
systems whose behavior can drift with each model update or prompt revision. We
introduce GPR-bench, a lightweight, extensible benchmark that operationalizes
regression testing for general purpose use cases. GPR-bench couples an open,
bilingual (English and Japanese) dataset covering eight task categories (e.g.,
text generation, code generation, and information retrieval) and 10 scenarios
in each task categories (80 total test cases for each language) with an
automated evaluation pipeline that employs "LLM-as-a-Judge" scoring of
correctness and conciseness. Experiments across three recent model versions -
gpt-4o-mini, o3-mini, and o4-mini - and two prompt configurations (default
versus concise-writing instruction) reveal heterogeneous quality. Our results
show that newer models generally improve correctness, but the differences are
modest and not statistically significant, suggesting that GPR-bench may not be
sufficiently challenging to differentiate between recent model versions. In
contrast, the concise-writing instruction significantly enhances conciseness
(+12.37 pp, Mann-Whitney U test: p < 0.001, effect size r = 0.2995) with
minimal degradations on accuracy (-1.7 pp), demonstrating the effectiveness of
prompt engineering. Released under the MIT License, GPR- bench lowers the
barrier to initiating reproducibility monitoring and provides a foundation for
community-driven extensions, while also raising important considerations about
benchmark design for rapidly evolving language models.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 12:31:43 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Morishige",
"Masumi",
""
],
[
"Koshihara",
"Ryo",
""
]
] |
2505.02855 | Antoine Derimay | Antoine Derimay | Poisson boundaries of building lattices and rigidity with
hyperbolic-like targets | 48 pages. Comments welcome! arXiv admin note: text overlap with
arXiv:2005.07477 by other authors | null | null | null | math.GR math.GT math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove that a Poisson boundary of any regular thick Euclidean building, as
well as lattices thereof is the space of chambers at infinity of the building
with the harmonic measure class. We then use this result to generalize rigidity
results of Guirardel-Horbez-L\'ecureux on morphisms and cocycles from lattices
in buildings to groups with negative curvature properties.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 14:04:43 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Derimay",
"Antoine",
""
]
] |
2505.02856 | Mahir Akgun Dr | Mahir Akgun, Hadi Hosseini | AI Education in a Mirror: Challenges Faced by Academic and Industry
Experts | To appear in AIED 2025 | null | null | null | cs.CY cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As Artificial Intelligence (AI) technologies continue to evolve, the gap
between academic AI education and real-world industry challenges remains an
important area of investigation. This study provides preliminary insights into
challenges AI professionals encounter in both academia and industry, based on
semi-structured interviews with 14 AI experts - eight from industry and six
from academia. We identify key challenges related to data quality and
availability, model scalability, practical constraints, user behavior, and
explainability. While both groups experience data and model adaptation
difficulties, industry professionals more frequently highlight deployment
constraints, resource limitations, and external dependencies, whereas academics
emphasize theoretical adaptation and standardization issues. These exploratory
findings suggest that AI curricula could better integrate real-world
complexities, software engineering principles, and interdisciplinary learning,
while recognizing the broader educational goals of building foundational and
ethical reasoning skills.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 16:52:49 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Akgun",
"Mahir",
""
],
[
"Hosseini",
"Hadi",
""
]
] |
2505.02857 | Larissa Rocha | Angelica Pereira Souza and Anderson Uch\^oa and Edna Dias Canedo and
Juliana Alves Pereira and Claudia Pinto Pereira and Larissa Rocha | Overcoming Obstacles: Challenges of Gender Inequality in Undergraduate
ICT Programs | 8 pages. To be published in the Proceedings of the 6th ACM/IEEE
Workshop on Gender Equality, Diversity, and Inclusion in Software Engineering
(GE), 2025 in Ottawa, Ontario, Canada | null | null | null | cs.CY cs.SE | http://creativecommons.org/licenses/by/4.0/ | Context: Gender inequality is a widely discussed issue across various
sectors, including Information Technology and Communication (ICT). In Brazil,
women represent less than 18% of ICT students in higher education. Prior
studies highlight gender-related barriers that discourage women from staying in
ICT. However, they provide limited insights into their perceptions as
undergraduate students and the factors influencing their participation and
confidence. Goal: This study explores the perceptions of women undergraduate
students in ICT regarding gender inequality. Method: A survey of 402 women from
18 Brazilian states enrolled in ICT courses was conducted using a mixed-method
approach, combining quantitative and qualitative analyses. Results: Women
students reported experiencing discriminatory practices from peers and
professors, both inside and outside the classroom. Gender stereotypes were
found to undermine their self-confidence and self-esteem, occasionally leading
to course discontinuation. Conclusions: Factors such as lack of representation,
inappropriate jokes, isolation, mistrust, and difficulty being heard contribute
to harmful outcomes, including reduced participation and reluctance to take
leadership roles. Addressing these issues is essential to creating a safe and
respectful learning environment for all students.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 18:28:48 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Souza",
"Angelica Pereira",
""
],
[
"Uchôa",
"Anderson",
""
],
[
"Canedo",
"Edna Dias",
""
],
[
"Pereira",
"Juliana Alves",
""
],
[
"Pereira",
"Claudia Pinto",
""
],
[
"Rocha",
"Larissa",
""
]
] |
2505.02858 | Henry Tari | Henry Tari, Nojus Sereiva, Rishabh Kaushal, Thales Bertaglia, Adriana
Iamnitchi | Towards High-Fidelity Synthetic Multi-platform Social Media Datasets via
Large Language Models | arXiv admin note: text overlap with arXiv:2407.08323 | null | null | null | cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social media datasets are essential for research on a variety of topics, such
as disinformation, influence operations, hate speech detection, or influencer
marketing practices. However, access to social media datasets is often
constrained due to costs and platform restrictions. Acquiring datasets that
span multiple platforms, which is crucial for understanding the digital
ecosystem, is particularly challenging. This paper explores the potential of
large language models to create lexically and semantically relevant social
media datasets across multiple platforms, aiming to match the quality of real
data. We propose multi-platform topic-based prompting and employ various
language models to generate synthetic data from two real datasets, each
consisting of posts from three different social media platforms. We assess the
lexical and semantic properties of the synthetic data and compare them with
those of the real data. Our empirical findings show that using large language
models to generate synthetic multi-platform social media data is promising,
different language models perform differently in terms of fidelity, and a
post-processing approach might be needed for generating high-fidelity synthetic
datasets for research. In addition to the empirical evaluation of three state
of the art large language models, our contributions include new fidelity
metrics specific to multi-platform social media datasets.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 18:56:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Tari",
"Henry",
""
],
[
"Sereiva",
"Nojus",
""
],
[
"Kaushal",
"Rishabh",
""
],
[
"Bertaglia",
"Thales",
""
],
[
"Iamnitchi",
"Adriana",
""
]
] |
2505.02859 | Johannes Schneider | Jonas Bokstaller and Julia Altheimer and Julian Dormehl and Alina Buss
and Jasper Wiltfang and Johannes Schneider and Maximilian R\"oglinger | Enhancing ML Model Interpretability: Leveraging Fine-Tuned Large
Language Models for Better Understanding of AI | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Across various sectors applications of eXplainableAI (XAI) gained momentum as
the increasing black-boxedness of prevailing Machine Learning (ML) models
became apparent. In parallel, Large Language Models (LLMs) significantly
developed in their abilities to understand human language and complex patterns.
By combining both, this paper presents a novel reference architecture for the
interpretation of XAI through an interactive chatbot powered by a fine-tuned
LLM. We instantiate the reference architecture in the context of
State-of-Health (SoH) prediction for batteries and validate its design in
multiple evaluation and demonstration rounds. The evaluation indicates that the
implemented prototype enhances the human interpretability of ML, especially for
users with less experience with XAI.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 20:57:55 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Bokstaller",
"Jonas",
""
],
[
"Altheimer",
"Julia",
""
],
[
"Dormehl",
"Julian",
""
],
[
"Buss",
"Alina",
""
],
[
"Wiltfang",
"Jasper",
""
],
[
"Schneider",
"Johannes",
""
],
[
"Röglinger",
"Maximilian",
""
]
] |
2505.02860 | Keith Paarporn | Keith Paarporn, Adel Aghajan, Jason R. Marden | Allocation of Heterogeneous Resources in General Lotto Games | Submitted for journal publication | null | null | null | econ.TH cs.CR cs.GT cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | The allocation of resources plays an important role in the completion of
system objectives and tasks, especially in the presence of strategic
adversaries. Optimal allocation strategies are becoming increasingly more
complex, given that multiple heterogeneous types of resources are at a system
planner's disposal. In this paper, we focus on deriving optimal strategies for
the allocation of heterogeneous resources in a well-known competitive resource
allocation model known as the General Lotto game. In standard formulations,
outcomes are determined solely by the players' allocation strategies of a
common, single type of resource across multiple contests. In particular, a
player wins a contest if it sends more resources than the opponent. Here, we
propose a multi-resource extension where the winner of a contest is now
determined not only by the amount of resources allocated, but also by the
composition of resource types that are allocated. We completely characterize
the equilibrium payoffs and strategies for two distinct formulations. The first
consists of a weakest-link/best-shot winning rule, and the second considers a
winning rule based on a weighted linear combination of the allocated resources.
We then consider a scenario where the resource types are costly to purchase,
and derive the players' equilibrium investments in each of the resource types.
| [
{
"version": "v1",
"created": "Fri, 2 May 2025 23:26:10 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Paarporn",
"Keith",
""
],
[
"Aghajan",
"Adel",
""
],
[
"Marden",
"Jason R.",
""
]
] |
2505.02861 | Kushagra Agrawal | Kushagra Agrawal, Nisharg Nargund | Neural Orchestration for Multi-Agent Systems: A Deep Learning Framework
for Optimal Agent Selection in Multi-Domain Task Environments | null | null | null | null | cs.MA cs.AI cs.NE | http://creativecommons.org/licenses/by/4.0/ | Multi-agent systems (MAS) are foundational in simulating complex real-world
scenarios involving autonomous, interacting entities. However, traditional MAS
architectures often suffer from rigid coordination mechanisms and difficulty
adapting to dynamic tasks. We propose MetaOrch, a neural orchestration
framework for optimal agent selection in multi-domain task environments. Our
system implements a supervised learning approach that models task context,
agent histories, and expected response quality to select the most appropriate
agent for each task. A novel fuzzy evaluation module scores agent responses
along completeness, relevance, and confidence dimensions, generating soft
supervision labels for training the orchestrator. Unlike previous methods that
hard-code agent-task mappings, MetaOrch dynamically predicts the most suitable
agent while estimating selection confidence. Experiments in simulated
environments with heterogeneous agents demonstrate that our approach achieves
86.3% selection accuracy, significantly outperforming baseline strategies
including random selection and round-robin scheduling. The modular architecture
emphasizes extensibility, allowing agents to be registered, updated, and
queried independently. Results suggest that neural orchestration offers a
powerful approach to enhancing the autonomy, interpretability, and adaptability
of multi-agent systems across diverse task domains.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 02:58:25 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Agrawal",
"Kushagra",
""
],
[
"Nargund",
"Nisharg",
""
]
] |
2505.02862 | Haoming Yang | Haoming Yang, Ke Ma, Xiaojun Jia, Yingfei Sun, Qianqian Xu, Qingming
Huang | Cannot See the Forest for the Trees: Invoking Heuristics and Biases to
Elicit Irrational Choices of LLMs | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Despite the remarkable performance of Large Language Models (LLMs), they
remain vulnerable to jailbreak attacks, which can compromise their safety
mechanisms. Existing studies often rely on brute-force optimization or manual
design, failing to uncover potential risks in real-world scenarios. To address
this, we propose a novel jailbreak attack framework, ICRT, inspired by
heuristics and biases in human cognition. Leveraging the simplicity effect, we
employ cognitive decomposition to reduce the complexity of malicious prompts.
Simultaneously, relevance bias is utilized to reorganize prompts, enhancing
semantic alignment and inducing harmful outputs effectively. Furthermore, we
introduce a ranking-based harmfulness evaluation metric that surpasses the
traditional binary success-or-failure paradigm by employing ranking aggregation
methods such as Elo, HodgeRank, and Rank Centrality to comprehensively quantify
the harmfulness of generated content. Experimental results show that our
approach consistently bypasses mainstream LLMs' safety mechanisms and generates
high-risk content, providing insights into jailbreak attack risks and
contributing to stronger defense strategies.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 05:28:11 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Yang",
"Haoming",
""
],
[
"Ma",
"Ke",
""
],
[
"Jia",
"Xiaojun",
""
],
[
"Sun",
"Yingfei",
""
],
[
"Xu",
"Qianqian",
""
],
[
"Huang",
"Qingming",
""
]
] |
2505.02863 | Xiaoming Zhai | Newnew Deng and Edward Jiusi Liu and Xiaoming Zhai | Understanding University Students' Use of Generative AI: The Roles of
Demographics and Personality Traits | null | null | null | null | cs.CY cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The use of generative AI (GAI) among university students is rapidly
increasing, yet empirical research on students' GAI use and the factors
influencing it remains limited. To address this gap, we surveyed 363
undergraduate and graduate students in the United States, examining their GAI
usage and how it relates to demographic variables and personality traits based
on the Big Five model (i.e., extraversion, agreeableness, conscientiousness,
and emotional stability, and intellect/imagination). Our findings reveal: (a)
Students in higher academic years are more inclined to use GAI and prefer it
over traditional resources. (b) Non-native English speakers use and adopt GAI
more readily than native speakers. (c) Compared to White, Asian students report
higher GAI usage, perceive greater academic benefits, and express a stronger
preference for it. Similarly, Black students report a more positive impact of
GAI on their academic performance. Personality traits also play a significant
role in shaping perceptions and usage of GAI. After controlling demographic
factors, we found that personality still significantly predicts GAI use and
attitudes: (a) Students with higher conscientiousness use GAI less. (b)
Students who are higher in agreeableness perceive a less positive impact of GAI
on academic performance and express more ethical concerns about using it for
academic work. (c) Students with higher emotional stability report a more
positive impact of GAI on learning and fewer concerns about its academic use.
(d) Students with higher extraversion show a stronger preference for GAI over
traditional resources. (e) Students with higher intellect/imagination tend to
prefer traditional resources. These insights highlight the need for
universities to provide personalized guidance to ensure students use GAI
effectively, ethically, and equitably in their academic pursuits.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 06:58:29 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Deng",
"Newnew",
""
],
[
"Liu",
"Edward Jiusi",
""
],
[
"Zhai",
"Xiaoming",
""
]
] |
2505.02864 | Kaidi Wang | Kaidi Wang, Zhiguo Ding, George K. Karagiannidis | Antenna Activation and Resource Allocation in Multi-Waveguide
Pinching-Antenna Systems | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pinching antennas, as a novel flexible-antenna technology capable of
establishing line of sight (LoS) connections and effectively mitigating
large-scale path loss, have recently attracted considerable research interests.
However, the implementation of ideal pinching-antenna systems involves
determining and adjusting pinching antennas to an arbitrary position on
waveguides, which presents challenges to both practical deployment and related
optimization. This paper investigates a practical pinching-antennas system in
multi-waveguide scenarios, where pinching antennas are installed at
pre-configured discrete positions to serve downlink users with non-orthogonal
multiple access (NOMA). To improve system throughput, a sophisticated
optimization problem is formulated by jointly considering waveguide assignment,
antenna activation, successive interference cancellation (SIC) decoding order
design, and power allocation. By treating waveguide assignment and antenna
activation as two coalition-formation games, a novel game-theoretic algorithm
is developed, in which the optimal decoding order is derived and incorporated.
For power allocation, monotonic optimization and successive convex
approximation (SCA) are employed to construct global optimal and low-complexity
solutions, respectively. Simulation results demonstrate that the NOMA-based
pinching-antenna system exhibits superior performance compared to the
considered benchmark systems, and the proposed solutions provide significant
improvement in terms of sum rate and outage probability.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 11:05:17 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Kaidi",
""
],
[
"Ding",
"Zhiguo",
""
],
[
"Karagiannidis",
"George K.",
""
]
] |
2505.02865 | Zhihai Wang | Zhihai Wang, Jie Wang, Jilai Pan, Xilin Xia, Huiling Zhen, Mingxuan
Yuan, Jianye Hao, Feng Wu | Accelerating Large Language Model Reasoning via Speculative Search | Accepted by ICML2025 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tree-search-based reasoning methods have significantly enhanced the reasoning
capability of large language models (LLMs) by facilitating the exploration of
multiple intermediate reasoning steps, i.e., thoughts. However, these methods
suffer from substantial inference latency, as they have to generate numerous
reasoning thoughts, severely limiting LLM applicability. To address this
challenge, we propose a novel Speculative Search (SpecSearch) framework that
significantly accelerates LLM reasoning by optimizing thought generation.
Specifically, SpecSearch utilizes a small model to strategically collaborate
with a large model at both thought and token levels, efficiently generating
high-quality reasoning thoughts. The major pillar of SpecSearch is a novel
quality-preserving rejection mechanism, which effectively filters out thoughts
whose quality falls below that of the large model's outputs. Moreover, we show
that SpecSearch preserves comparable reasoning quality to the large model.
Experiments on both the Qwen and Llama models demonstrate that SpecSearch
significantly outperforms state-of-the-art approaches, achieving up to
2.12$\times$ speedup with comparable reasoning quality.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 12:14:08 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Zhihai",
""
],
[
"Wang",
"Jie",
""
],
[
"Pan",
"Jilai",
""
],
[
"Xia",
"Xilin",
""
],
[
"Zhen",
"Huiling",
""
],
[
"Yuan",
"Mingxuan",
""
],
[
"Hao",
"Jianye",
""
],
[
"Wu",
"Feng",
""
]
] |
2505.02866 | Ximeng Li | Ximeng Li, Jing Ren, Xi-Li Zhang | Probing Boson Clouds with Supermassive Black Hole Binaries | 27 pages, 12 figures, 1 table | null | null | null | hep-ph astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Rotating black holes can generate boson clouds via superradiance when the
boson's Compton wavelength is comparable to the black hole's size. In binary
systems, these clouds can produce distinctive observational imprints. Recent
studies accounting for nonlinearities induced by orbital backreaction suggest
that if the binary forms at a large separation, resonance transitions can
significantly deplete the cloud, minimizing later observational consequences
except for very specific orbital inclinations. In this paper, we extend this
framework to supermassive black hole binaries (SMBHBs), considering the
influence of their astrophysical evolutionary histories. We find that, before
entering the gravitational wave (GW) radiation stage, the additional energy
loss channels can accelerate orbital evolution. This acceleration makes
hyperfine resonant transitions inefficient, allowing a sufficient portion of
the cloud to remain for later direct observations. We further discuss the
ionization effects and cloud depletion occurring at this stage. Based on these
theoretical insights, we explore how multi-messenger observations for SMBHBs
can be utilized to detect the ionization effects of boson clouds by examining
changes in the orbital period decay rate via electromagnetic measurements and
variations in GW strain over a wide frequency band. Our findings reveal a
complex dependence on the binary's total mass, mass ratio, and boson mass,
emphasizing the significant role of astrophysical evolution histories in
detecting boson clouds within binaries.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 14:14:44 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Li",
"Ximeng",
""
],
[
"Ren",
"Jing",
""
],
[
"Zhang",
"Xi-Li",
""
]
] |
2505.02867 | Ruiqi Wang | Ruiqi Wang and Hao Zhang | RESAnything: Attribute Prompting for Arbitrary Referring Segmentation | 42 pages, 31 figures. For more details:
https://suikei-wang.github.io/RESAnything/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present an open-vocabulary and zero-shot method for arbitrary referring
expression segmentation (RES), targeting input expressions that are more
general than what prior works were designed to handle. Specifically, our inputs
encompass both object- and part-level labels as well as implicit references
pointing to properties or qualities of object/part function, design, style,
material, etc. Our model, coined RESAnything, leverages Chain-of-Thoughts (CoT)
reasoning, where the key idea is attribute prompting. We generate detailed
descriptions of object/part attributes including shape, color, and location for
potential segment proposals through systematic prompting of a large language
model (LLM), where the proposals are produced by a foundational image
segmentation model. Our approach encourages deep reasoning about object or part
attributes related to function, style, design, etc., enabling the system to
handle implicit queries without any part annotations for training or
fine-tuning. As the first zero-shot and LLM-based RES method, RESAnything
achieves clearly superior performance among zero-shot methods on traditional
RES benchmarks and significantly outperforms existing methods on challenging
scenarios involving implicit queries and complex part-level relations. Finally,
we contribute a new benchmark dataset to offer ~3K carefully curated RES
instances to assess part-level, arbitrary RES solutions.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 15:19:20 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Ruiqi",
""
],
[
"Zhang",
"Hao",
""
]
] |
2505.02868 | Anurag K S V | Shubham Chouhan, Anurag K.S.V., G. Raghavan, Kanaka Raju P | FPGA-based Toeplitz Strong Extractor for Quantum Random Number
Generators | 5 pages, 5 figures | 2024 IEEE 5th India Council International Subsections Conference
(INDISCON) | 10.1109/INDISCON62179.2024.10744392 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum Random Number Generators (QRNGs) serve as high-entropy sources for
Quantum Key Distribution (QKD) systems. However, the raw data from these
quantum sources require post-processing to achieve a nearly uniform
distribution. This work presents a state-of-the-art implementation of the
Toeplitz Strong Extractor on an FPGA, achieving a benchmark extraction speed of
26.57 Gbps. A detailed implementation flow of the post-processing on the FPGA
is provided, along with the execution speeds obtained for different randomness
extraction ratios. Raw data from an in-house phase noise-based QRNG is
processed on the FPGA using this implementation, and the output is validated
using the NIST STS 2.1.2 statistical randomness test suite.
| [
{
"version": "v1",
"created": "Sat, 3 May 2025 18:25:30 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Chouhan",
"Shubham",
""
],
[
"V.",
"Anurag K. S.",
""
],
[
"Raghavan",
"G.",
""
],
[
"P",
"Kanaka Raju",
""
]
] |
2505.02869 | Abdul Khaliq | Abdul Khaliq, Syafruddin Karimi, Werry Darta Taifur, Endrizal Ridwan | The quest for explosive bubbles in the Indonesian Rupiah/US exchange
rate: Does the uncertainty trinity matter? | 12 pages | null | 10.5267/j.dsl.2024.1.005 | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | The Generalized Supremum Augmented Dickey-Fuller (GSADF) technique is
performed to resolve whether the Indonesian Rupiah/US exchange rate has
experienced multiple explosive bubbles. The GSADF uncovers that the Indonesian
Rupiah/US exchange rate deviates from the fundamental values by six times from
January 1985 to September 2023, periodically indicating the presence of
numerous explosive behaviors. Once the full-sample period separates into the
managed-floating regime and the free-floating regime, the GSADF still detects
multiple bubbles. Of particular curiosity on uncertainty trinity, this study
underlines that global geopolitical risk negatively drives explosive actions in
the ratio of exchange rates for non-traded and traded goods. The global
economic policy uncertainty negatively affects speculative bubbles in the
exchange rate and the ratio of exchange rates for non-traded. The country's
geopolitical risks negatively strike only speculative bubbles in the exchange
rate. Further, we find heterogeneity in our results by examining different
exchange rate systems. The robustness checks further firmly ascertain across
baseline empirical findings.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 00:46:12 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Khaliq",
"Abdul",
""
],
[
"Karimi",
"Syafruddin",
""
],
[
"Taifur",
"Werry Darta",
""
],
[
"Ridwan",
"Endrizal",
""
]
] |
2505.02870 | Duje Bonacci | Duje Bonacci | Variational Resolution of the Abraham-Lorentz-Dirac Equation Pathologies | PRL style paper, recently submitted. 5 pages of the core derivations
and discussion and 3 pages of appendices. Al comments welcome | null | null | null | physics.class-ph physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | We propose a structural variational resolution of the Abraham-Lorentz-Dirac
(ALD) pathologies. By deriving the Variational Kinematic Constraint (VKC) and
the Variational Dynamics Constraint (VDC) from the particle's proper-time
perspective, we show that self-induced variations are forbidden and dynamics
arise solely from first-order proper-time variations of external fields.
Consequently, self-force terms are excluded at the variational level,
eliminating runaway solutions and non-causal behavior without regularization.
Our framework further provides a first-principles derivation of minimal
coupling and reveals gauge invariance as a necessary consequence of
proper-time-based variational structure.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 08:34:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Bonacci",
"Duje",
""
]
] |
2505.02871 | Luis Herrera | L. Herrera, A. Di Prisco, J.Ospino | The birth of a ghost star | 30 pages Latex,3 figures. Invited contribution published in Entropy | Entropy, 27, 412, (2025) | null | null | gr-qc | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a model of an evolving spherically symmetric dissipative
self-gravitating fluid distribution which tends asymptotically to a ghost star,
meaning that the end state of such a system corresponds to a static fluid
distribution with vanishing total mass, and energy-density distribution which
is negative in some regions of the fluid. The model is inspired in a solution
representing a fluid evolving quasi-homologously and with vanishing complexity
factor. However in order to satisfy the asymptotic behavior mentioned above,
the starting solution has to be modified, as a consequence of which the
resulting model only satisfies the two previously mentioned conditions,
asymptotically. Additionally a condition on the variation of the infinitesimal
proper radial distance between two neighboring points per unit of proper time
is imposed, which implies the presence of a cavity surrounding the center.
Putting together all these conditions we are able to obtain an analytical model
depicting the emergence of a ghost star. Some potential observational
consequences of this phenomenon are briefly discussed at the last section.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 09:00:04 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Herrera",
"L.",
""
],
[
"Di Prisco",
"A.",
""
],
[
"Ospino",
"J.",
""
]
] |
2505.02872 | Omer Shubi | Cfir Avraham Hadar, Omer Shubi, Yoav Meiri, Yevgeni Berzak | Decoding Open-Ended Information Seeking Goals from Eye Movements in
Reading | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | When reading, we often have specific information that interests us in a text.
For example, you might be reading this paper because you are curious about LLMs
for eye movements in reading, the experimental design, or perhaps you only care
about the question ``but does it work?''. More broadly, in daily life, people
approach texts with any number of text-specific goals that guide their reading
behavior. In this work, we ask, for the first time, whether open-ended reading
goals can be automatically decoded from eye movements in reading. To address
this question, we introduce goal classification and goal reconstruction tasks
and evaluation frameworks, and use large-scale eye tracking for reading data in
English with hundreds of text-specific information seeking tasks. We develop
and compare several discriminative and generative multimodal LLMs that combine
eye movements and text for goal classification and goal reconstruction. Our
experiments show considerable success on both tasks, suggesting that LLMs can
extract valuable information about the readers' text-specific goals from eye
movements.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 13:23:48 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Hadar",
"Cfir Avraham",
""
],
[
"Shubi",
"Omer",
""
],
[
"Meiri",
"Yoav",
""
],
[
"Berzak",
"Yevgeni",
""
]
] |
2505.02873 | EPTCS | Marsha Chechik (University of Toronto), Arianna Fedeli (Gran Sasso
Science Institute), Gianluca Filippone (Gran Sasso Science Institute),
Federico Formica (McMaster University), Mirgita Frasheri (Aarhus University),
Nico Hochgeschwender (University of Bremen), Lina Marsso (Polytechnique
Montreal) | Proceedings of the First International Workshop on Autonomous Systems
Quality Assurance and Prediction with Digital Twins | null | EPTCS 418, 2025 | 10.4204/EPTCS.418 | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | This volume contains the proceedings of the First International Workshop on
Autonomous Systems Quality Assurance and Prediction with Digital Twins (ASQAP
2025), which was held in Hamilton, Canada, on May 4th, 2025, as a satellite
event of ETAPS 2025. The aim of ASQAP 2025 is to gather experts from academia
and industry to explore the potential of digital twin technology in supporting
quality assurance in autonomous systems, including concepts such as
specification, verification, validation, testing, analysis, and many others.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 16:53:30 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Chechik",
"Marsha",
"",
"University of Toronto"
],
[
"Fedeli",
"Arianna",
"",
"Gran Sasso\n Science Institute"
],
[
"Filippone",
"Gianluca",
"",
"Gran Sasso Science Institute"
],
[
"Formica",
"Federico",
"",
"McMaster University"
],
[
"Frasheri",
"Mirgita",
"",
"Aarhus University"
],
[
"Hochgeschwender",
"Nico",
"",
"University of Bremen"
],
[
"Marsso",
"Lina",
"",
"Polytechnique\n Montreal"
]
] |
2505.02874 | Leopoldo Juli\'an Lechuga L\'opez | L. Juli\'an Lechuga L\'opez, Shaza Elsharief, Dhiyaa Al Jorf, Firas
Darwish, Congbo Ma, and Farah E. Shamout | Uncertainty Quantification for Machine Learning in Healthcare: A Survey | 46 pages, 3 figures, 2 tables, AHLI Conference on Health, Inference,
and Learning (CHIL) | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Uncertainty Quantification (UQ) is pivotal in enhancing the robustness,
reliability, and interpretability of Machine Learning (ML) systems for
healthcare, optimizing resources and improving patient care. Despite the
emergence of ML-based clinical decision support tools, the lack of principled
quantification of uncertainty in ML models remains a major challenge. Current
reviews have a narrow focus on analyzing the state-of-the-art UQ in specific
healthcare domains without systematically evaluating method efficacy across
different stages of model development, and despite a growing body of research,
its implementation in healthcare applications remains limited. Therefore, in
this survey, we provide a comprehensive analysis of current UQ in healthcare,
offering an informed framework that highlights how different methods can be
integrated into each stage of the ML pipeline including data processing,
training and evaluation. We also highlight the most popular methods used in
healthcare and novel approaches from other domains that hold potential for
future adoption in the medical context. We expect this study will provide a
clear overview of the challenges and opportunities of implementing UQ in the ML
pipeline for healthcare, guiding researchers and practitioners in selecting
suitable techniques to enhance the reliability, safety and trust from patients
and clinicians on ML-driven healthcare solutions.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 16:56:22 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"López",
"L. Julián Lechuga",
""
],
[
"Elsharief",
"Shaza",
""
],
[
"Jorf",
"Dhiyaa Al",
""
],
[
"Darwish",
"Firas",
""
],
[
"Ma",
"Congbo",
""
],
[
"Shamout",
"Farah E.",
""
]
] |
2505.02875 | Blessed Ngwenya | Blessed Arthur Ngwenya | The Casimir Effect in Non-Abelian Gauge Theories on the Lattice | PhD Thesis, University of Cape Town (June, 2024), 175 pages | null | null | null | hep-lat | http://creativecommons.org/licenses/by/4.0/ | We present non-perturbative results of the Casimir potential in non-abelian
gauge theories in (2+1)D and (3+1)D for SU(2) and SU(3) in the confined and
deconfined phases. For the first time, geometries beyond parallel plates in
(3+1)D SU($N_c)$ are explored, and we show that the Casimir effect for the
symmetrical and asymmetrical tube and box is attractive. The result for the
tube is contrary to the weakly coupled, massless non-interacting scalar field
theory result where a repulsive Casimir force, described by a negative slope of
the potential, is measured in this geometry. We propose various techniques that
can be used to account for the energy contributions from creating the
boundaries in the geometry of a tube and box where the size of the faces
forming the walls of the geometries changes with separation distance. We show
that increasing the temperature from a confined to a deconfined phase does not
alter the measured potential and we motivate for this observation by showing
that the masses of the particles in the Casimir interactions are lower than the
lowest glueball mass, $M_{0^{++}}$ in the pure gauge ground state, thus
suggesting that the region inside the geometries is a boundary-induced
deconfined phase. Through the measured expectation value of the Polyakov loop,
we propose that the Casimir effect for the asymmetrical tube in the large
separation distance limit $R\to \infty$ should be equivalent to that of the
parallel plate geometry at smallest separation distance, while the asymmetrical
box in the same limit should be equivalent to the symmetrical tube at the
smallest separation distance.
| [
{
"version": "v1",
"created": "Sun, 4 May 2025 22:27:32 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ngwenya",
"Blessed Arthur",
""
]
] |
2505.02876 | Wentao Wu | Xiaoying Wang, Wentao Wu, Vivek Narasayya, Surajit Chaudhuri | Esc: An Early-stopping Checker for Budget-aware Index Tuning | This is the extended version of a paper published at VLDB 2025 | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Index tuning is a time-consuming process. One major performance bottleneck in
existing index tuning systems is the large amount of "what-if" query optimizer
calls that estimate the cost of a given pair of query and index configuration
without materializing the indexes. There has been recent work on budget-aware
index tuning that limits the amount of what-if calls allowed in index tuning.
Existing budget-aware index tuning algorithms, however, typically make fast
progress early on in terms of the best configuration found but slow down when
more and more what-if calls are allocated. This observation of "diminishing
return" on index quality leads us to introduce early stopping for budget-aware
index tuning, where user specifies a threshold on the tolerable loss of index
quality and we stop index tuning if the projected loss with the remaining
budget is below the threshold. We further propose Esc, a low-overhead
early-stopping checker that realizes this new functionality. Experimental
evaluation on top of both industrial benchmarks and real customer workloads
demonstrate that Esc can significantly reduce the number of what-if calls made
during budget-aware index tuning while incur little or zero improvement loss
and little extra computational overhead compared to the overall index tuning
time.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 00:57:57 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wang",
"Xiaoying",
""
],
[
"Wu",
"Wentao",
""
],
[
"Narasayya",
"Vivek",
""
],
[
"Chaudhuri",
"Surajit",
""
]
] |
2505.02877 | Haohua Que | Hele Zhu, Xinyi Huang, Haojia Gao, Mengfei Jiang, Haohua Que and Lei
Mu | A Wireless Collaborated Inference Acceleration Framework for Plant
Disease Recognition | null | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Plant disease is a critical factor affecting agricultural production.
Traditional manual recognition methods face significant drawbacks, including
low accuracy, high costs, and inefficiency. Deep learning techniques have
demonstrated significant benefits in identifying plant diseases, but they still
face challenges such as inference delays and high energy consumption. Deep
learning algorithms are difficult to run on resource-limited embedded devices.
Offloading these models to cloud servers is confronted with the restriction of
communication bandwidth, and all of these factors will influence the
inference's efficiency. We propose a collaborative inference framework for
recognizing plant diseases between edge devices and cloud servers to enhance
inference speed. The DNN model for plant disease recognition is pruned through
deep reinforcement learning to improve the inference speed and reduce energy
consumption. Then the optimal split point is determined by a greedy strategy to
achieve the best collaborated inference acceleration. Finally, the system for
collaborative inference acceleration in plant disease recognition has been
implemented using Gradio to facilitate friendly human-machine interaction.
Experiments indicate that the proposed collaborative inference framework
significantly increases inference speed while maintaining acceptable
recognition accuracy, offering a novel solution for rapidly diagnosing and
preventing plant diseases.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 03:17:32 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Zhu",
"Hele",
""
],
[
"Huang",
"Xinyi",
""
],
[
"Gao",
"Haojia",
""
],
[
"Jiang",
"Mengfei",
""
],
[
"Que",
"Haohua",
""
],
[
"Mu",
"Lei",
""
]
] |
2505.02878 | Wren Yamada | Wren Yamada, Osamu Morimatsu, Toru Sato, Koichi Yazaki | Pole-Expansion of Two-Hadron Imaginary-Time Correlation Function -a new
method of analysis for unstable states in lattice QCD- | 16 pages, 8 figures | null | null | KEK-TH-2717, J-PARC-TH-0316, RIKEN-iTHEMS-Report-25 | hep-lat nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze the pole expansion of the two-hadron imaginary-time correlation
function. We first explain the general idea that the imaginary-time correlation
function is expressed as a sum of the pole terms, the Mittag-Leffler expansion,
in terms of the uniformization variable, which makes the S-matrix
single-valued. We then derive explicit expressions of the pole expansion for
the single-channel ($\rho$ meson) and two-channel ($\Lambda(1405)$) examples
and demonstrate that the pole expansion actually holds employing
phenomenological models, the vector-dominance model for the $\rho$ meson and
the chiral unitary model for $\Lambda(1405)$. From this observation we propose
the pole expansion as a method to extract information of unstable states such
as masses and widths from the two-hadron imaginary-time correlation functions
obtained by lattice QCD simulations.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 04:53:13 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Yamada",
"Wren",
""
],
[
"Morimatsu",
"Osamu",
""
],
[
"Sato",
"Toru",
""
],
[
"Yazaki",
"Koichi",
""
]
] |
2505.02879 | Linh Huynh | Jackson George, Zachariah Yusaf, Stephanie Zoltick, Linh Huynh | Phase transitions in AI-human interaction networks: statistics,
computation, and probabilistic modeling | null | null | null | null | physics.soc-ph math.PR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In recent years, Large Language Models (LLMs) have revolutionized Natural
Language Processing with their ability to generate human-like texts. However, a
fundamental challenge remains in understanding the underlying mechanisms
driving their emergent behaviors, particularly the randomness in their outputs.
This paper investigates the application of spin glass theory as a mathematical
framework to quantify the uncertainty of LLMs. Moreover, we analyze how the
interaction between the noise in LLMs and from social networks shape emergent
collective behaviors of the system. By making connections between LLMs and spin
glass models, we gain insights into the high-dimensional optimization
landscapes of LLMs, the uncertainty in their outputs, and the role of noise in
their learning process. We focus on LLMs' ability to replicate human-written
flitzes, a form of flirtatious poems unique to Dartmouth College, used to
invite peers or a potentially romantic partner to social events. Given flitzes'
playful tone, personal references, and role in complex social networks, they
represent a uniquely creative form of language, making them ideal for exploring
how the temperature parameter in LLMs affects the creativity and verisimilitude
of AI-generated content. To better understand where temperature affects model
behavior, we look for temperature-based phase transitions through statistical
analysis, computational methods, and simulation of our spin glass model. Our
findings demonstrate that temperature not only governs randomness in LLM
output, but also mediates deeper transitions in linguistic structure, perceived
quality, and human-machine alignment. By connecting statistical physics with
language generation, we provide a novel framework for understanding emergent
behavior in LLMs and their interaction with complex social networks.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 06:32:17 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"George",
"Jackson",
""
],
[
"Yusaf",
"Zachariah",
""
],
[
"Zoltick",
"Stephanie",
""
],
[
"Huynh",
"Linh",
""
]
] |
2505.02880 | Zian Liu | Zian Liu, Renjun Jia | LLM4FTS: Enhancing Large Language Models for Financial Time Series
Prediction | 12 pages, 9figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Predicting financial time series presents significant challenges due to
inherent low signal-to-noise ratios and intricate temporal patterns.
Traditional machine learning models exhibit limitations in this forecasting
task constrained by their restricted model capacity. Recent advances in large
language models (LLMs), with their greatly expanded parameter spaces,
demonstrate promising potential for modeling complex dependencies in temporal
sequences. However, existing LLM-based approaches typically focus on
fixed-length patch analysis due to the Transformer architecture, ignoring
market data's multi-scale pattern characteristics. In this study, we propose
$LLM4FTS$, a novel framework that enhances LLM capabilities for temporal
sequence modeling through learnable patch segmentation and dynamic wavelet
convolution modules. Specifically,we first employ K-means++ clustering based on
DTW distance to identify scale-invariant patterns in market data. Building upon
pattern recognition results, we introduce adaptive patch segmentation that
partitions temporal sequences while preserving maximal pattern integrity. To
accommodate time-varying frequency characteristics, we devise a dynamic wavelet
convolution module that emulates discrete wavelet transformation with enhanced
flexibility in capturing time-frequency features. These three modules work
together to improve large language model's ability to handle scale-invariant
patterns in financial time series. Extensive experiments on real-world
financial datasets substantiate the framework's efficacy, demonstrating
superior performance in capturing complex market patterns and achieving
state-of-the-art results in stock return prediction. The successful deployment
in practical trading systems confirms its real-world applicability,
representing a significant advancement in LLM applications for financial
forecasting.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 06:48:34 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Liu",
"Zian",
""
],
[
"Jia",
"Renjun",
""
]
] |
2505.02881 | Kazuki Fujii | Kazuki Fujii, Yukito Tajima, Sakae Mizuki, Hinari Shimada, Taihei
Shiotani, Koshiro Saito, Masanari Ohi, Masaki Kawamura, Taishi Nakamura,
Takumi Okamoto, Shigeki Ishida, Kakeru Hattori, Youmi Ma, Hiroya Takamura,
Rio Yokota, Naoaki Okazaki | Rewriting Pre-Training Data Boosts LLM Performance in Math and Code | 27pages(including appendix), 10 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The performance of large language models (LLMs) in program synthesis and
mathematical reasoning is fundamentally limited by the quality of their
pre-training corpora. We introduce two openly licensed datasets, released under
the Llama 3.3 Community License, that significantly enhance LLM performance by
systematically rewriting public data. SwallowCode (approximately 16.1 billion
tokens) refines Python snippets from The-Stack-v2 through a novel four-stage
pipeline: syntax validation, pylint-based style filtering, and a two-stage LLM
rewriting process that enforces style conformity and transforms snippets into
self-contained, algorithmically efficient examples. Unlike prior methods that
rely on exclusionary filtering or limited transformations, our
transform-and-retain approach upgrades low-quality code, maximizing data
utility. SwallowMath (approximately 2.3 billion tokens) enhances Finemath-4+ by
removing boilerplate, restoring context, and reformatting solutions into
concise, step-by-step explanations. Within a fixed 50 billion token training
budget, continual pre-training of Llama-3.1-8B with SwallowCode boosts pass@1
by +17.0 on HumanEval and +17.7 on HumanEval+ compared to Stack-Edu, surpassing
the baseline model's code generation capabilities. Similarly, substituting
SwallowMath yields +12.4 accuracy on GSM8K and +7.6 on MATH. Ablation studies
confirm that each pipeline stage contributes incrementally, with rewriting
delivering the largest gains. All datasets, prompts, and checkpoints are
publicly available, enabling reproducible research and advancing LLM
pre-training for specialized domains.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 07:38:43 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Fujii",
"Kazuki",
""
],
[
"Tajima",
"Yukito",
""
],
[
"Mizuki",
"Sakae",
""
],
[
"Shimada",
"Hinari",
""
],
[
"Shiotani",
"Taihei",
""
],
[
"Saito",
"Koshiro",
""
],
[
"Ohi",
"Masanari",
""
],
[
"Kawamura",
"Masaki",
""
],
[
"Nakamura",
"Taishi",
""
],
[
"Okamoto",
"Takumi",
""
],
[
"Ishida",
"Shigeki",
""
],
[
"Hattori",
"Kakeru",
""
],
[
"Ma",
"Youmi",
""
],
[
"Takamura",
"Hiroya",
""
],
[
"Yokota",
"Rio",
""
],
[
"Okazaki",
"Naoaki",
""
]
] |
2505.02882 | Qingzheng Lyu | D.D. Su, B.F. Shen, and Q.Z. Lv | Controlling Schwinger tunneling via engineering of virtual particle
phases in vacuum | 10 pages, 8 ffigures | null | null | null | quant-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | An investigation into Schwinger pair production mechanisms is presented,
demonstrating that vacuum tunneling processes can be effectively controlled
through electromagnetic potential modulation while maintaining the strong
ffelds in the interaction region. This challenges the conventional paradigm
that attributes exclusive governance of Schwinger processes to localized ffeld
intensities. Through comprehensive analysis of particle number, momentum
spectra, and spatial distribution of created pairs, we establish that the
observed modulation effects originate from electromagnetic potential - induced
modiffcations to the quantum phase structure of virtual particles. This
phenomenon reveals a profound connection between Schwinger tunneling dynamics
and the geometric phase properties of the quantum vacuum state - a vacuum
analogue to the Aharonov-Bohm effect in charged particle systems. This
discovery not only advances our understanding of electromagnetic interactions
in quantum vacuum but also opens up new experimental opportunities for
realizing Schwinger tunneling processes with existing facilities.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 09:51:47 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Su",
"D. D.",
""
],
[
"Shen",
"B. F.",
""
],
[
"Lv",
"Q. Z.",
""
]
] |
2505.02884 | Guangzhi Sun | Guangzhi Sun, Potsawee Manakul, Xiao Zhan, Mark Gales | Unlearning vs. Obfuscation: Are We Truly Removing Knowledge? | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Unlearning has emerged as a critical capability for large language models
(LLMs) to support data privacy, regulatory compliance, and ethical AI
deployment. Recent techniques often rely on obfuscation by injecting incorrect
or irrelevant information to suppress knowledge. Such methods effectively
constitute knowledge addition rather than true removal, often leaving models
vulnerable to probing. In this paper, we formally distinguish unlearning from
obfuscation and introduce a probing-based evaluation framework to assess
whether existing approaches genuinely remove targeted information. Moreover, we
propose DF-MCQ, a novel unlearning method that flattens the model predictive
distribution over automatically generated multiple-choice questions using
KL-divergence, effectively removing knowledge about target individuals and
triggering appropriate refusal behaviour. Experimental results demonstrate that
DF-MCQ achieves unlearning with over 90% refusal rate and a random choice-level
uncertainty that is much higher than obfuscation on probing questions.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 14:21:08 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Sun",
"Guangzhi",
""
],
[
"Manakul",
"Potsawee",
""
],
[
"Zhan",
"Xiao",
""
],
[
"Gales",
"Mark",
""
]
] |
2505.02885 | Axel Brandenburg | Axel Brandenburg, Evangelia Ntormousi | Magnetic field amplification during a turbulent collapse | 13 pages, 16 figures, 1 table, submitted to ApJ | null | null | NORDITA-2025-023 | astro-ph.GA physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | The question of whether a dynamo can be triggered by gravitational collapse
is of great interest, especially for the early Universe. Here, we employ
supercomoving coordinates to study the magnetic field amplification from
decaying turbulence during gravitational collapse. We perform three-dimensional
simulations and show that for large magnetic Reynolds numbers, there can be
exponential growth of the comoving magnetic field with conformal time before
the decay of turbulence impedes further amplification. The collapse dynamics
only affects the nonlinear feedback from the Lorentz force, which diminishes
more rapidly for shorter collapse times, allowing nearly kinematic continued
growth. We also confirm that helical turbulence is more efficient in driving
dynamo action than nonhelical turbulence, but this difference decreases for
larger collapse times. We also show that for nearly irrotational flows, dynamo
amplification is still possible, but it is always associated with a growth of
vorticity -- even if it still remains very small. In nonmagnetic runs, the
growth of vorticity is associated with viscosity and grows with the Mach
number. In the presence of magnetic fields, vorticity emerges from the curl of
the Lorentz force.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 14:35:36 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Brandenburg",
"Axel",
""
],
[
"Ntormousi",
"Evangelia",
""
]
] |
2505.02886 | David Silver | David H. Silver | Taskmaster Deconstructed: A Quantitative Look at Tension, Volatility,
and Viewer Ratings | 29 pages, includes 5 figures and 18 supplementary visualizations.
Submitted as a preprint. Code and data available at github dot com slash
silverdavi slash taskmaster-stats | null | null | null | physics.soc-ph cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Taskmaster is a British television show that combines comedic performance
with a formal scoring system. Despite the appearance of structured competition,
it remains unclear whether scoring dynamics contribute meaningfully to audience
engagement. We conducted a statistical analysis of 162 episodes across 18
series, using fifteen episode-level metrics to quantify rank volatility, point
spread, lead changes, and winner dominance. None of these metrics showed a
significant association with IMDb ratings, even after controlling for series
effects. Long-term trends suggest that average points have increased over time,
while volatility has slightly declined and rank spread has remained stable.
These patterns indicate an attempt to enhance competitive visibility without
altering the show's structural equilibrium. We also analyzed contestant rank
trajectories and identified five recurring archetypes describing performance
styles. These patterns suggest that viewer interest is shaped more by
contestant behavior than by game mechanics.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 15:46:32 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Silver",
"David H.",
""
]
] |
2505.02887 | Cheng Ge | Cheng Ge, Han-Shen Tae, Zhenqiang Zhang, Lu Lu, Zhijie Huang, Yilin
Wang, Tao Jiang, Wenqing Cai, Shan Chang, David J. Adams, and Rilei Yu | CreoPep: A Universal Deep Learning Framework for Target-Specific Peptide
Design and Optimization | 16 pages, 6 figures | null | null | null | q-bio.BM cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Target-specific peptides, such as conotoxins, exhibit exceptional binding
affinity and selectivity toward ion channels and receptors. However, their
therapeutic potential remains underutilized due to the limited diversity of
natural variants and the labor-intensive nature of traditional optimization
strategies. Here, we present CreoPep, a deep learning-based conditional
generative framework that integrates masked language modeling with a
progressive masking scheme to design high-affinity peptide mutants while
uncovering novel structural motifs. CreoPep employs an integrative augmentation
pipeline, combining FoldX-based energy screening with temperature-controlled
multinomial sampling, to generate structurally and functionally diverse
peptides that retain key pharmacological properties. We validate this approach
by designing conotoxin inhibitors targeting the $\alpha$7 nicotinic
acetylcholine receptor, achieving submicromolar potency in electrophysiological
assays. Structural analysis reveals that CreoPep-generated variants engage in
both conserved and novel binding modes, including disulfide-deficient forms,
thus expanding beyond conventional design paradigms. Overall, CreoPep offers a
robust and generalizable platform that bridges computational peptide design
with experimental validation, accelerating the discovery of next-generation
peptide therapeutics.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 15:56:39 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ge",
"Cheng",
""
],
[
"Tae",
"Han-Shen",
""
],
[
"Zhang",
"Zhenqiang",
""
],
[
"Lu",
"Lu",
""
],
[
"Huang",
"Zhijie",
""
],
[
"Wang",
"Yilin",
""
],
[
"Jiang",
"Tao",
""
],
[
"Cai",
"Wenqing",
""
],
[
"Chang",
"Shan",
""
],
[
"Adams",
"David J.",
""
],
[
"Yu",
"Rilei",
""
]
] |
2505.02888 | Rintaro Ando | Rintaro Ando | When Your Own Output Becomes Your Training Data: Noise-to-Meaning Loops
and a Formal RSI Trigger | 20 pages, 4 figures, 3 tables. Code:
github.com/rintaro-ando-tech/n2m-rsi-demo (v1.0) | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present Noise-to-Meaning Recursive Self-Improvement (N2M-RSI), a minimal
formal model showing that once an AI agent feeds its own outputs back as inputs
and crosses an explicit information-integration threshold, its internal
complexity will grow without bound under our assumptions. The framework unifies
earlier ideas on self-prompting large language models, G\"odelian
self-reference, and AutoML, yet remains implementation-agnostic. The model
furthermore scales naturally to interacting swarms of agents, hinting at
super-linear effects once communication among instances is permitted. For
safety reasons, we omit system-specific implementation details and release only
a brief, model-agnostic toy prototype in Appendix C.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 17:03:07 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ando",
"Rintaro",
""
]
] |
2505.02889 | Oyindolapo Komolafe | Oyindolapo O. Komolafe, Zhimin Mei, David Morales Zarate, Gregory
William Spangenberg | Early Prediction of Sepsis: Feature-Aligned Transfer Learning | A project implemented for MACHINE LEARNING IN HEALTH AND BIOMEDICAL
SCIENCE | null | null | null | cs.LG cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | Sepsis is a life threatening medical condition that occurs when the body has
an extreme response to infection, leading to widespread inflammation, organ
failure, and potentially death. Because sepsis can worsen rapidly, early
detection is critical to saving lives. However, current diagnostic methods
often identify sepsis only after significant damage has already occurred. Our
project aims to address this challenge by developing a machine learning based
system to predict sepsis in its early stages, giving healthcare providers more
time to intervene.
A major problem with existing models is the wide variability in the patient
information or features they use, such as heart rate, temperature, and lab
results. This inconsistency makes models difficult to compare and limits their
ability to work across different hospitals and settings. To solve this, we
propose a method called Feature Aligned Transfer Learning (FATL), which
identifies and focuses on the most important and commonly reported features
across multiple studies, ensuring the model remains consistent and clinically
relevant.
Most existing models are trained on narrow patient groups, leading to
population bias. FATL addresses this by combining knowledge from models trained
on diverse populations, using a weighted approach that reflects each models
contribution. This makes the system more generalizable and effective across
different patient demographics and clinical environments. FATL offers a
practical and scalable solution for early sepsis detection, particularly in
hospitals with limited resources, and has the potential to improve patient
outcomes, reduce healthcare costs, and support more equitable healthcare
delivery.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 17:59:34 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Komolafe",
"Oyindolapo O.",
""
],
[
"Mei",
"Zhimin",
""
],
[
"Zarate",
"David Morales",
""
],
[
"Spangenberg",
"Gregory William",
""
]
] |
2505.02890 | Beena Meena | Beena Meena, Elena Sabbi, Peter Zeidler, Bruce G. Elmegreen, Jan J.
Eldridge, Varun Bajaj, Mario Gennaro, Anna Pasquali, Debra M. Elmegreen, Ralf
S. Klessen, Linda J. Smith, Luciana Bianchi, Aida Wofford, Pietro Facchini,
John S. Gallagher III, Daniela Calzetti, Eva K. Grebel, Angela Adamo | GULP II: Hierarchical Distribution and Evolution of Young Stellar
Structures in NGC 4449 | 23 Pages, 11 Figures, 2 Tables, Accepted for publication in ApJ | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We investigate the hierarchical distribution and evolution of young stellar
structures in the dwarf starburst galaxy NGC 4449 using data from the GULP
survey. By analyzing the spatial distribution of field stars younger than 100
Myr, we identify large-scale stellar complexes and substructures using HDBSCAN
-- a density-based clustering algorithm -- and trace their evolution over time.
While comparing these stellar structures in different regions of the galaxy, we
find that the central bar-like region shows a clear expansion of the structures
within the first $\sim$ 60 Myrs, while the arm-like structure in the NE shows
no discernible trend, possibly due to external perturbations from tidal
interactions with a neighboring galaxy. An age-dependent two point correlation
function (TPCF) analysis shows that young stars exhibit a strong hierarchical
distribution, with clustering strength decreasing over time. The power-law
slope of the TPCF, which starts at $\alpha \sim 0.65$ for stars younger than 5
Myr, shows a slight decline to $\alpha \sim 0.4$ for stars older than 50 Myr,
though it does not reach a completely flat (random) distribution. This trend
indicates a subtle weakening of structural hierarchy among young ($<$100 Myr)
stars, which is primarily driven by internal stellar motions. Future work will
extend this analysis to the remaining 26 galaxies in the GULP survey to better
constrain the role of the galactic environment in shaping the hierarchical
evolution of young stellar populations.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Meena",
"Beena",
""
],
[
"Sabbi",
"Elena",
""
],
[
"Zeidler",
"Peter",
""
],
[
"Elmegreen",
"Bruce G.",
""
],
[
"Eldridge",
"Jan J.",
""
],
[
"Bajaj",
"Varun",
""
],
[
"Gennaro",
"Mario",
""
],
[
"Pasquali",
"Anna",
""
],
[
"Elmegreen",
"Debra M.",
""
],
[
"Klessen",
"Ralf S.",
""
],
[
"Smith",
"Linda J.",
""
],
[
"Bianchi",
"Luciana",
""
],
[
"Wofford",
"Aida",
""
],
[
"Facchini",
"Pietro",
""
],
[
"Gallagher",
"John S.",
"III"
],
[
"Calzetti",
"Daniela",
""
],
[
"Grebel",
"Eva K.",
""
],
[
"Adamo",
"Angela",
""
]
] |
2505.02891 | David Vall\'es-P\'erez | David Vall\'es-P\'erez, Susana Planelles, Vicent Quilis | The eventful life journey of galaxy clusters. I. Impact of DM halo and
ICM properties on their full assembly histories | Accepted for publication in Astronomy & Astrophysics. 19 pages, 10
figures, 2 tables + appendices (3 pages, 4 figures, 2 tables) | null | null | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Context. Galaxy clusters assemble over gigayears in a very anisotropic
environment, which causes a remarkable diversity in their mass assembly
histories (MAH).
Aims. In this work, we have aimed to understand how the present-day
properties of the dark matter halo and the intracluster medium are related to
the whole evolution of these structures.
Methods. To this end, we analysed a $\Lambda$CDM hydrodynamical+$N$-Body
simulation of a $(100 \, h^{-1} \mathrm{Mpc})^3$ volume, containing over 30
clusters and 300 groups, and looked at the individual and the stacked MAHs
(determined from complete merger trees) in relation to properties of the DM
haloes and the ICM at fixed cosmic time (indicators of assembly state).
Results. The ensemble MAHs are well separated when stacked in bins of these
indicators, yielding clear dependencies of evolutionary properties (such as
formation redshift) on fix-time halo properties. Additionally, we find that
different indicators are informative about distinct epochs of accretion.
Finally, by summarising the complex MAH diversity with two parameters, we
describe how different indicators bring complementary information in different
directions of this biparametric space. Overall, halo spin and a combined
indicator appear to be the ones encoding the most information about the MAH.
Conclusions. The results shown here add up to the idea that the dynamical
state of cosmic structures is a multifaceted concept, and warn that single
indicators are incapable of capturing the whole complexity of the process. This
work sheds light on the nature of this characterisation by untangling precisely
when and how several indicators are informative about. In turn, this can
provide clues to better constrain the MAH of observed structures.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Vallés-Pérez",
"David",
""
],
[
"Planelles",
"Susana",
""
],
[
"Quilis",
"Vicent",
""
]
] |
2505.02892 | Federico Roccati | Federico Roccati | Many-Body Quantum Optics in a Bose-Hubbard Waveguide | 5+25 pages; 4 figures | null | null | null | quant-ph cond-mat.mes-hall cond-mat.quant-gas | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Waveguide quantum electrodynamics (QED) studies the interaction between
quantum emitters and guided photons in one-dimension. When the waveguide hosts
interacting photons, it becomes a platform to explore many-body quantum optics.
However, the influence of photonic correlations on emitter dynamics remains
poorly understood. In this work, we study the collective decay and coherent
interactions of quantum emitters coupled to a one-dimensional Bose-Hubbard
waveguide, an array of coupled photonic modes with repulsive on-site
interactions that supports superfluid and Mott insulating phases. We show that
photon-photon interactions alone can trigger a superradiant burst, independent
of emitter spacing and transition frequency. In the off-resonant regime,
emitters exhibit two distinct types of mediated interactions: delocalized
superfluid excitations yield distance-independent couplings, while
Mott-insulator quasiparticles generate short-range interactions mediated by
doublons and holons. Our work bridges many-body physics and waveguide QED,
revealing how photonic many-body states shape emitter dynamics.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Roccati",
"Federico",
""
]
] |
2505.02893 | Zhengyan Shi Mr. | Zhengyan Darius Shi, Carolyn Zhang, T. Senthil | Doping lattice non-abelian quantum Hall states | 32 pages, 2 figures, 14 pages of appendices | null | null | null | cond-mat.str-el cond-mat.supr-con hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study quantum phases of a fluid of mobile charged non-abelian anyons,
which arise upon doping the lattice Moore-Read quantum Hall state at lattice
filling $\nu = 1/2$ and its generalizations to the Read-Rezayi
($\mathrm{RR}_k$) sequence at $\nu = k/(k+2)$. In contrast to their abelian
counterparts, non-abelian anyons present unique challenges due to their
non-invertible fusion rules and non-abelian braiding structures. We address
these challenges using a Chern-Simons-Ginzburg-Landau (CSGL) framework that
incorporates the crucial effect of energy splitting between different anyon
fusion channels at nonzero dopant density. For the Moore-Read state, we show
that doping the charge $e/4$ non-abelion naturally leads to a fully gapped
charge-$2$ superconductor without any coexisting topological order. The chiral
central charge of the superconductor depends on details of the interactions
determining the splitting of anyon fusion channels. For general $\mathrm{RR}_k$
states, our analysis of states obtained by doping the basic non-abelion $a_0$
with charge $e/(k+2)$ reveals a striking even/odd pattern in the Read-Rezayi
index $k$. We develop a general physical picture for anyon-driven
superconductivity based on charge-flux unbinding, and show how it relates to
the CSGL description of doped abelian quantum Hall states. Finally, as a bonus,
we use the CSGL formalism to describe transitions between the $\mathrm{RR}_k$
state and a trivial period-$(k+2)$ CDW insulator at fixed filling, driven by
the gap closure of the fundamental non-abelian anyon $a_0$. Notably, for $k=2$,
this predicts a period-4 CDW neighboring the Moore-Read state at half-filling,
offering a potential explanation of recent numerical observations in models of
twisted MoTe$_2$.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Shi",
"Zhengyan Darius",
""
],
[
"Zhang",
"Carolyn",
""
],
[
"Senthil",
"T.",
""
]
] |
2505.02894 | Nikolay Gnezdilov | Nikolay V. Gnezdilov and Rufus Boyack | Upper bound on $T_c$ in a strongly coupled electron-boson superconductor | 7 pages, 1 figure; supplemental material: 5 pages, 3 figures | null | null | null | cond-mat.str-el cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Migdal-Eliashberg theory (METh) of boson-mediated superconductivity contains
a $\sqrt{\lambda}$ divergence in the critical temperature $T_c$ at strong
electron-boson coupling $\lambda$. In conventional METh, the strong-coupling
regime can be accessed only in the limit that $\lambda_E = \lambda \,
\omega_D/\varepsilon_F\ll1$, where $\omega_D$ is the Debye frequency and
$\varepsilon_F$ is the Fermi energy. Here we go beyond this restriction in the
context of the two-dimensional Yukawa-SYK (Y-SYK) model, which is solvable for
arbitrary values of $\lambda_E$. We find that $T_c\approx 0.18 \,\omega_D
\sqrt{\lambda}$ for large $\lambda$, provided $\lambda_E$ remains small, and
crosses over to a universal value of $T_c \approx 0.04\, \varepsilon_F$ for
large $\lambda_E$. The saturation of $T_c$ is due to a self-consistent account
of the boson dynamics for large $\lambda_E$ and is arguably qualitatively valid
beyond the Y-SYK framework. These results demonstrate how the $\sqrt{\lambda}$
growth of $T_c$ in METh saturates to a universal value independent of $\lambda$
and $\omega_D$, providing an upper bound on the critical temperature at strong
electron-boson coupling.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Gnezdilov",
"Nikolay V.",
""
],
[
"Boyack",
"Rufus",
""
]
] |
2505.02895 | Junyu Zhang | Junyu Zhang, Eiichi Egami, Fengwu Sun, Xiaojing Lin, Jianwei Lyu,
Yongda Zhu, Pierluigi Rinaldi, Yang Sun, Andrew J. Bunker, Rachana
Bhatawdekar, Jakob M. Helton, Roberto Maiolino, Zheng Ma, Brant Robertson,
Sandro Tacchella, Giacomo Venturi, Christina C. Williams, Chris Willott | Abundant Population of Broad H$\alpha$ Emitters in the GOODS-N Field
Revealed by CONGRESS, FRESCO, and JADES | Submitted to ApJ. 23 pages, 16 figures, 3 tables | null | null | null | astro-ph.GA astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We present a spectroscopic search for broad H$\alpha$ emitters at z~3.7-6.5
in the GOODS-N field, utilizing JWST/NIRCam slitless spectroscopy from FRESCO
and CONGRESS, complemented by JADES imaging. We identify 19 broad H$\alpha$
emitters with FWHM > 1000 km/s at z~4-5.5, including 9 new sources. The black
hole masses and AGN bolometric luminosities, inferred from the broad H$\alpha$
components, indicate that most sources are accreting at ~10% of the Eddington
limit. We derive their host stellar masses via SED fitting and find higher
$M_{BH}/M_{*}$ ratios relative to the local $M_{BH}$-$M_{*}$ relations,
consistent with previous studies. We find that 42% of the sample do not satisfy
the widely-used color selection criteria for Little Red Dots (LRDs), with the
majority of these sources lacking the characteristic steep red slope. A
comparison of the average SEDs between our sample and LRDs selected in the same
field reveals that the steep red slopes observed in some LRDs are likely due to
line-boosting effects as previously suggested. Furthermore, we find that 68% of
color-selected LRDs with H$\alpha$ detections in the NIRCam/Grism spectra do
not exhibit broad-line features. While the limited sensitivity of the grism
spectra may hinder the detection of broad-line components in faint sources, our
findings still highlight the enigmatic nature of the LRD population.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Zhang",
"Junyu",
""
],
[
"Egami",
"Eiichi",
""
],
[
"Sun",
"Fengwu",
""
],
[
"Lin",
"Xiaojing",
""
],
[
"Lyu",
"Jianwei",
""
],
[
"Zhu",
"Yongda",
""
],
[
"Rinaldi",
"Pierluigi",
""
],
[
"Sun",
"Yang",
""
],
[
"Bunker",
"Andrew J.",
""
],
[
"Bhatawdekar",
"Rachana",
""
],
[
"Helton",
"Jakob M.",
""
],
[
"Maiolino",
"Roberto",
""
],
[
"Ma",
"Zheng",
""
],
[
"Robertson",
"Brant",
""
],
[
"Tacchella",
"Sandro",
""
],
[
"Venturi",
"Giacomo",
""
],
[
"Williams",
"Christina C.",
""
],
[
"Willott",
"Chris",
""
]
] |
2505.02896 | Xiaojing Lin | Xiaojing Lin, Xiaohui Fan, Fengwu Sun, Junyu Zhang, Eiichi Egami,
Jakob M. Helton, Feige Wang, Haowen Zhang, Andrew J. Bunker, Zheng Cai,
Zhiyuan Ji, Xiangyu Jin, Roberto Maiolino, Maria Anne Pudoka, Pierluigi
Rinaldi, Brant Robertson, Sandro Tacchella, Wei Leong Tee, Yang Sun,
Christopher N. A. Willmer, Chris Willott, and Yongda Zhu | The Large-scale Environments of Low-luminosity AGNs at $3.9 < z < 6$ and
Implications for Their Host Dark Matter Halos from a Complete NIRCam Grism
Redshift Survey | Main text 15 pages; 9 figures; submitted to the ApJ; comments are
welcome | null | null | null | astro-ph.GA astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We study the large-scale environments and clustering properties of 28
low-luminosity AGNs at $z=3.9-6$ in the GOODS-N field. Our sample, identified
from the JWST NIRCam Imaging and WFSS data in CONGRESS and FRESCO surveys with
either broad H$\alpha$ emission lines or V-shape continua, are compared to 782
H$\alpha$ emitters (HAEs) selected from the same data. These AGNs are located
in diverse large-scale environments and do not preferentially reside in denser
environments compared to HAEs. Their overdensity field, $\delta$, averaged over
(15 $h^{-1}$cMpc)$^3$, ranges from $-0.56$ to 10.56, and shows no clear
correlation with broad-line luminosity, black hole (BH) masses, or the AGN
fraction. It suggests that $> 10$ cMpc structures do not significantly
influence BH growth. We measure the two-point cross-correlation function of
AGNs with HAEs, finding a comparable amplitude to that of the HAE
auto-correlation. This indicates similar bias parameters and host dark matter
halo masses for AGNs and HAEs. The correlation length of field AGNs is 4.26
$h^{-1}$cMpc, and 7.66 $h^{-1}$cMpc at $3.9 < z < 5$ and $5 < z < 6$,
respectively. We infer a median host dark matter halo mass of $\log
(M_h/M_\odot)\approx 11.0-11.2$ and host stellar masses of $\log
(M_\star/M_\odot) \approx 8.4-8.6$ by comparing with the UniverseMachine
simulation. Our clustering analysis suggests that low-luminosity AGNs at high
redshift reside in normal star-forming galaxies with overmassive BHs. They
represent an intrinsically distinct population from luminous quasars and could
be a common phase in galaxy evolution.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:00 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Lin",
"Xiaojing",
""
],
[
"Fan",
"Xiaohui",
""
],
[
"Sun",
"Fengwu",
""
],
[
"Zhang",
"Junyu",
""
],
[
"Egami",
"Eiichi",
""
],
[
"Helton",
"Jakob M.",
""
],
[
"Wang",
"Feige",
""
],
[
"Zhang",
"Haowen",
""
],
[
"Bunker",
"Andrew J.",
""
],
[
"Cai",
"Zheng",
""
],
[
"Ji",
"Zhiyuan",
""
],
[
"Jin",
"Xiangyu",
""
],
[
"Maiolino",
"Roberto",
""
],
[
"Pudoka",
"Maria Anne",
""
],
[
"Rinaldi",
"Pierluigi",
""
],
[
"Robertson",
"Brant",
""
],
[
"Tacchella",
"Sandro",
""
],
[
"Tee",
"Wei Leong",
""
],
[
"Sun",
"Yang",
""
],
[
"Willmer",
"Christopher N. A.",
""
],
[
"Willott",
"Chris",
""
],
[
"Zhu",
"Yongda",
""
]
] |
2505.02897 | Jiaxin Qiao | Sridip Pal, Jiaxin Qiao, Balt C. van Rees | Universality of the microcanonical entropy at large spin | 30 pages, 2 figures | null | null | null | hep-th cond-mat.stat-mech math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider rigorous consequences of modular invariance for two-dimensional
unitary non-rational CFTs with $c > 1$. Simple estimates for the torus
partition function can lead to remarkably strong results. We show in particular
that the spectral density of spin-$J$ operators must grow like $\exp\left( \pi
\sqrt{\frac{2}{3}(c-1) J} \right)/\sqrt{2J}$ in any twist interval at or above
$(c-1)/12$, with a known twist-dependent prefactor. This proves that the large
$J$ spectrum becomes dense even without averaging over spins. For twists below
$(c-1)/12$ we establish that the growth must be strictly slower. Finally, we
estimate how fast the maximal gap between two spin-$J$ operators must go to
zero as $J$ becomes large.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Pal",
"Sridip",
""
],
[
"Qiao",
"Jiaxin",
""
],
[
"van Rees",
"Balt C.",
""
]
] |
2505.02898 | Parsa Ghafour | Parsa Ghafour, Saeed Tavasoli | GrAviPaSt's Lens to the Past: Unveiling the Evolution of Filamentary
Structures | 13 pages, 5 figures, 2 tables, prepared for submission | null | null | null | astro-ph.CO astro-ph.GA astro-ph.IM physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper examines the evolution of cosmic filaments across redshifts 1,
0.5, and 0 using the IllustrisTNG100-1 magneto-hydrodynamical simulation. To
achieve this, we introduce GrAviPaSt, a simple and efficient filament
identification method that leverages gravitational potential, an A*-like
path-finding algorithm, and spanning trees. Applying this method to galaxy
distributions at different redshifts allows us to analyze various filament
properties, including their length, thickness, mass density contrast, and
radial profile. Additionally, we investigate dynamic characteristics such as
the mean distance of filament galaxies from the skeleton, their weighted mean
velocity, and velocity trends normalized by their positions within the
filaments. Our findings reveal the evolution of cosmic filaments from redshift
1 to 0, highlighting key differences across classifications. In particular, we
examine the mass density contrast radial profile of filaments connecting two
galaxy groups and those linking two galaxy clusters, identifying distinct
differences in profile shape between these categories. Furthermore, in the
context of weighted mean velocity, we analyze cosmic filaments exhibiting
either negative or positive weighted mean velocity, demonstrating their
differing evolutionary trends in terms of the mean distance of filament
galaxies from the skeleton.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Ghafour",
"Parsa",
""
],
[
"Tavasoli",
"Saeed",
""
]
] |
2505.02899 | Sounak Pal Mr. | Arpan Bhattacharyya, Saptaswa Ghosh, Sounak Pal | The Sky Remembers everything: Celestial amplitude, Shadow and OPE in
quadratic EFT of gravity | 39 Pages, 5 Figures | null | null | null | hep-th gr-qc math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we compute the celestial amplitude arising from higher
curvature corrections to Einstein gravity, incorporating phase dressing. The
inclusion of such corrections leads to effective modifications of the theory's
ultraviolet (UV) behaviour. In the eikonal limit, we find that, in contrast to
Einstein's gravity, where the $u$ and $s$-channel contributions cancel, these
contributions remain non-vanishing in the presence of higher curvature terms.
We examine the analytic structure of the resulting amplitude and derive a
dispersion relation for the phase-dressed eikonal amplitude in quadratic
gravity. Furthermore, we investigate the celestial conformal block expansion of
the Mellin-transformed conformal shadow amplitude within the framework of
celestial conformal field theory (CCFT). As a consequence, we compute the
corresponding operator product expansion (OPE) coefficients using the
Burchnall-Chaundy expansion. In addition, we evaluate the OPE via the Euclidean
OPE inversion formula across various kinematic channels and comment on its
applicability and implications. Finally, we briefly explore the Carrollian
amplitude associated with the corresponding quadratic EFT.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Bhattacharyya",
"Arpan",
""
],
[
"Ghosh",
"Saptaswa",
""
],
[
"Pal",
"Sounak",
""
]
] |
2505.02900 | Yuxuan Zhang | Yuxuan Zhang, Timothy H. Hsieh, Yong Baek Kim, and Yijian Zou | Probing mixed-state phases on a quantum computer via Renyi correlators
and variational decoding | null | null | null | null | quant-ph cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Recent advances have defined nontrivial phases of matter in open quantum
systems, such as many-body quantum states subject to environmental noise. In
this work, we experimentally probe and characterize mixed-state phases on
Quantinuum's H1 quantum computer using two measures: Renyi correlators and the
coding performance of a quantum error-correcting code associated with the
phase. As a concrete example, we probe the low-energy states of the critical
transverse field Ising model under different dephasing noise channels. First,
we employ shadow tomography to observe a newly proposed Renyi correlator in two
distinct phases: one exhibiting power-law decay and the other long-ranged.
Second, we investigate the decoding fidelity of the associated quantum
error-correcting code using a variational quantum circuit, and we find that a
shallow circuit is sufficient to distinguish the above-mentioned two
mixed-state phases through the decoding performance quantified by entanglement
fidelity. Our work is a proof of concept for the quantum simulation and
characterization of mixed-state phases.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Zhang",
"Yuxuan",
""
],
[
"Hsieh",
"Timothy H.",
""
],
[
"Kim",
"Yong Baek",
""
],
[
"Zou",
"Yijian",
""
]
] |
2505.02901 | Alexander Wietek Ph.D. | Alexander Wietek, Luke Staszewski, Martin Ulaga, Paul L. Ebert, Hannes
Karlsson, Siddhartha Sarkar, Henry Shackleton, Aritra Sinha, Rafael D. Soares | XDiag: Exact Diagonalization for quantum many-body systems | 29 pages, 4 figures | null | null | null | cond-mat.str-el physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Exact diagonalization (ED) is a cornerstone technique in quantum many-body
physics, enabling precise solutions to the Schr\"odinger equation for
interacting quantum systems. Despite its utility in studying ground states,
excited states, and dynamical behaviors, the exponential growth of the Hilbert
space with system size presents significant computational challenges. We
introduce XDiag, an open-source software package designed to combine advanced
and efficient algorithms for ED with and without symmetry-adapted bases with
user-friendly interfaces. Implemented in C++ for computational efficiency and
wrapped in Julia for ease of use, XDiag provides a comprehensive toolkit for ED
calculations. Key features of XDiag include the first publicly accessible
implementation of sublattice coding algorithms for large-scale spin system
diagonalizations, efficient Lin table algorithms for symmetry lookups, and
random-hashing techniques for distributed memory parallelization. The library
supports various Hilbert space types (e.g., spin-1/2, electron, and t-J
models), facilitates symmetry-adapted block calculations, and automates
symmetry considerations. The package is complemented by extensive
documentation, a user guide, reproducible benchmarks demonstrating near-linear
scaling on thousands of CPU cores, and over 20 examples covering ground-state
calculations, spectral functions, time evolution, and thermal states. By
integrating high-performance computing with accessible scripting capabilities,
XDiag allows researchers to perform state-of-the-art ED simulations and explore
quantum many-body phenomena with unprecedented flexibility and efficiency.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:01 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Wietek",
"Alexander",
""
],
[
"Staszewski",
"Luke",
""
],
[
"Ulaga",
"Martin",
""
],
[
"Ebert",
"Paul L.",
""
],
[
"Karlsson",
"Hannes",
""
],
[
"Sarkar",
"Siddhartha",
""
],
[
"Shackleton",
"Henry",
""
],
[
"Sinha",
"Aritra",
""
],
[
"Soares",
"Rafael D.",
""
]
] |
2505.02902 | Luca Marzola | M. Fabbrichesi, R. Floreanini, and L. Marzola | Local vs. nonlocal entanglement in top-quark pairs at the LHC | 6 pages, 3 figures | null | null | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | We show that the entanglement observed in top-antitop quark spin states at
the LHC is local in the energy region close to the production threshold. In
contrast, nonlocal entanglement is observed in the central boosted region
defined by a top-quark pair invariant mass $m_{t\bar t} > 800$ GeV and
scattering angles $\Theta$ satisfying $|\cos \Theta |<0.2$. This makes
top-quark pairs a unique laboratory for studying the interplay between
entanglement and Bell locality. The locality of entanglement near the
production threshold is further supported by a recent CMS analysis, which
reports a significance of more than $5\sigma$.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:02 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Fabbrichesi",
"M.",
""
],
[
"Floreanini",
"R.",
""
],
[
"Marzola",
"L.",
""
]
] |
2505.02903 | Paolo Suin | Paolo Suin, Doris Arzoumanian, Annie Zavagno, and Patrick Hennebelle | The role of magnetic field and stellar feedback in the evolution of
filamentary structures in collapsing star-forming clouds | 21 pages, 2 appendices, 16 figures, accepted for publication to A&A | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Context: Filaments are common features in molecular clouds and they play a
key role in star formation (SF). Studying their life cycle is essential to
fully understand the SF process.
Aims: We aim to characterise the impact of magnetic field ($B$) and stellar
feedback on the evolution of filamentary structures in star-forming clouds.
Methods: We performed two numerical simulations of a collapsing
$10^4\,$M$_\odot$ cloud with different mass-to-flux ratios ($\mu=2$ and
$\mu=8$), including early stellar feedback (jets and HII regions). Using
DisPerSE, we extracted the three-dimensional filamentary network and analysed
its properties as it evolves throughout the SF event.
Results: We observed that the filamentary network in the simulations follow
two distinct evolutionary pathways. In the strongly magnetised case, the cloud
maintains a sparser filamentary network, and the arising filaments are
predominantly perpendicular to $B$ lines. With a weak field, the cloud develops
a single central hub, with converging filaments favouring a parallel alignment
relative to $B$. Furthermore, while always accreting, filaments exhibit faster
flows towards the hub relative to the surrounding gas. In the weakly magnetised
run, the central hub dominates the dynamics, and filaments exhibit faster flows
as they approach the central hub. Finally, once the expanding HII region
impacts the filaments, they align to $B$ independently of the initial
configuration.
Conclusions: Magnetic fields play a critical role in shaping the structure
and dynamics of molecular clouds. Stronger magnetic fields slow the cloud's
evolution and inhibit the formation of central hubs, promoting a broader
filamentary network instead.
However, ionising feedback dominates the late-stage evolution, overriding the
initial differences and dictating the final filament configuration.
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:02 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Suin",
"Paolo",
""
],
[
"Arzoumanian",
"Doris",
""
],
[
"Zavagno",
"Annie",
""
],
[
"Hennebelle",
"Patrick",
""
]
] |
2505.02904 | Martine Lokken | Martine Lokken, Andreu Font-Ribera, Patrick McDonald | Weighted FFT estimators for 1D and 3D correlations of the Lyman-$\alpha$
forest | 22 pages, 7 figures, to be submitted to JCAP, comments welcome | null | null | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Correlations in the Lyman-$\alpha$ (Ly$\alpha$) forest, both as a function of
line of sight separation (1D) and 3D separation, provide a unique window to the
distribution of matter at redshifts not accessible by current galaxy surveys.
While optimal quadratic estimators have been used to measure 1D correlations,
they are computationally expensive and difficult to extend to 3D analyses. On
the other hand, estimators based on the Fast Fourier Transform (FFT) are
significantly faster, but are affected by missing data in the spectra (masked
pixels) and so far have not used pixel weights to reduce the uncertainties in
the measurement. In this publication we describe how to compute the window
matrix that enables forward-modeling the impact of masked pixels and weights on
the FFT-based estimators. We use Gaussian and hydrodynamical simulations with
artificially masked pixels to validate the method on the measurement of 1D
correlations. Finally, we show that the formalism can be extended to model the
impact on 3D correlations, in particular on the cross-spectrum, the correlation
of 1D Fourier modes as a function of transverse separation. This work will
enable more precise clustering measurements with the Ly$\alpha$ forest dataset
recently collected by the Dark Energy Spectroscopic Instrument (DESI).
| [
{
"version": "v1",
"created": "Mon, 5 May 2025 18:00:02 GMT"
}
] | 2025-05-07T00:00:00 | [
[
"Lokken",
"Martine",
""
],
[
"Font-Ribera",
"Andreu",
""
],
[
"McDonald",
"Patrick",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.