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
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2509.05343
|
Zahid Ullah
|
Zahid Ullah, Minki Hong, Tahir Mahmood, Jihie Kim
|
Systematic Integration of Attention Modules into CNNs for Accurate and Generalizable Medical Image Diagnosis
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Deep learning has become a powerful tool for medical image analysis; however, conventional Convolutional Neural Networks (CNNs) often fail to capture the fine-grained and complex features critical for accurate diagnosis. To address this limitation, we systematically integrate attention mechanisms into five widely adopted CNN architectures, namely, VGG16, ResNet18, InceptionV3, DenseNet121, and EfficientNetB5, to enhance their ability to focus on salient regions and improve discriminative performance. Specifically, each baseline model is augmented with either a Squeeze and Excitation block or a hybrid Convolutional Block Attention Module, allowing adaptive recalibration of channel and spatial feature representations. The proposed models are evaluated on two distinct medical imaging datasets, a brain tumor MRI dataset comprising multiple tumor subtypes, and a Products of Conception histopathological dataset containing four tissue categories. Experimental results demonstrate that attention augmented CNNs consistently outperform baseline architectures across all metrics. In particular, EfficientNetB5 with hybrid attention achieves the highest overall performance, delivering substantial gains on both datasets. Beyond improved classification accuracy, attention mechanisms enhance feature localization, leading to better generalization across heterogeneous imaging modalities. This work contributes a systematic comparative framework for embedding attention modules in diverse CNN architectures and rigorously assesses their impact across multiple medical imaging tasks. The findings provide practical insights for the development of robust, interpretable, and clinically applicable deep learning based decision support systems.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 05:45:27 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ullah",
"Zahid",
""
],
[
"Hong",
"Minki",
""
],
[
"Mahmood",
"Tahir",
""
],
[
"Kim",
"Jihie",
""
]
] |
2509.05344
|
Shogo Kunieda
|
Shogo Kunieda, Mitsuru Yambe, Hiromori Murashima, Takeru Nakamura, Toshiaki Shintani, Hitoshi Kamijima, Yoshihiro Hayashi, Yosuke Hanawa and Ryo Yoshida
|
Machine Learning for Polymer Chemical Resistance to Organic Solvents
|
Supporting information is provided as Supporting_Information.pdf
| null | null | null |
cond-mat.soft cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Predicting the chemical resistance of polymers to organic solvents is a longstanding challenge in materials science, with significant implications for sustainable materials design and industrial applications. Here, we address the need for interpretable and generalizable frameworks to understand and predict polymer chemical resistance beyond conventional solubility models. We systematically analyze a large dataset of polymer solvent combinations using a data-driven approach. Our study reveals that polymer crystallinity and density, as well as solvent polarity, are key factors governing chemical resistance, and that these trends are consistent with established theoretical models. These findings provide a foundation for rational screening and design of polymer materials with tailored chemical resistance, advancing both fundamental understanding and practical applications.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 06:47:59 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kunieda",
"Shogo",
""
],
[
"Yambe",
"Mitsuru",
""
],
[
"Murashima",
"Hiromori",
""
],
[
"Nakamura",
"Takeru",
""
],
[
"Shintani",
"Toshiaki",
""
],
[
"Kamijima",
"Hitoshi",
""
],
[
"Hayashi",
"Yoshihiro",
""
],
[
"Hanawa",
"Yosuke",
""
],
[
"Yoshida",
"Ryo",
""
]
] |
2509.05345
|
Guoxin Fang
|
Jiasheng Qu, Zhuo Huang, Dezhao Guo, Hailin Sun, Aoran Lyu, Chengkai Dai, Yeung Yam, Guoxin Fang
|
INF-3DP: Implicit Neural Fields for Collision-Free Multi-Axis 3D Printing
| null | null | null | null |
cs.RO cs.CG
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce a general, scalable computational framework for multi-axis 3D printing based on implicit neural fields (INFs) that unifies all stages of toolpath generation and global collision-free motion planning. In our pipeline, input models are represented as signed distance fields, with fabrication objectives such as support-free printing, surface finish quality, and extrusion control being directly encoded in the optimization of an implicit guidance field. This unified approach enables toolpath optimization across both surface and interior domains, allowing shell and infill paths to be generated via implicit field interpolation. The printing sequence and multi-axis motion are then jointly optimized over a continuous quaternion field. Our continuous formulation constructs the evolving printing object as a time-varying SDF, supporting differentiable global collision handling throughout INF-based motion planning. Compared to explicit-representation-based methods, INF-3DP achieves up to two orders of magnitude speedup and significantly reduces waypoint-to-surface error. We validate our framework on diverse, complex models and demonstrate its efficiency with physical fabrication experiments using a robot-assisted multi-axis system.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 09:22:52 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Qu",
"Jiasheng",
""
],
[
"Huang",
"Zhuo",
""
],
[
"Guo",
"Dezhao",
""
],
[
"Sun",
"Hailin",
""
],
[
"Lyu",
"Aoran",
""
],
[
"Dai",
"Chengkai",
""
],
[
"Yam",
"Yeung",
""
],
[
"Fang",
"Guoxin",
""
]
] |
2509.05346
|
Bo Yuan
|
Bo Yuan and Jiazi Hu
|
Benchmarking Large Language Models for Personalized Guidance in AI-Enhanced Learning
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
While Large Language Models (LLMs) are increasingly envisioned as intelligent assistants for personalized learning, systematic head-to-head evaluations within authentic learning scenarios remain limited. This study conducts an empirical comparison of three state-of-the-art LLMs on a tutoring task that simulates a realistic learning setting. Using a dataset comprising a student's answers to ten questions of mixed formats with correctness labels, each LLM is required to (i) analyze the quiz to identify underlying knowledge components, (ii) infer the student's mastery profile, and (iii) generate targeted guidance for improvement. To mitigate subjectivity and evaluator bias, we employ Gemini as a virtual judge to perform pairwise comparisons along various dimensions: accuracy, clarity, actionability, and appropriateness. Results analyzed via the Bradley-Terry model indicate that GPT-4o is generally preferred, producing feedback that is more informative and better structured than its counterparts, while DeepSeek-V3 and GLM-4.5 demonstrate intermittent strengths but lower consistency. These findings highlight the feasibility of deploying LLMs as advanced teaching assistants for individualized support and provide methodological guidance for future empirical research on LLM-driven personalized learning.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 14:21:59 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yuan",
"Bo",
""
],
[
"Hu",
"Jiazi",
""
]
] |
2509.05347
|
Yinyu Zhang
|
Yinyu Zhang
|
Validity Verification of the New TOEFL Writing Task Based on Classical Test Theory
| null | null | null | null |
cs.CY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The TOEFL iBT has introduced the Academic Discussion Task (ADT) to assess test-takers' ability to engage in academic discourse, reflecting the growing emphasis on interactive communication skills in higher education. However, research on the ADT's validity and fairness particularly for culturally and linguistically diverse groups, such as Chinese students, remains limited. This study addresses this gap by employing Classical Test Theory (CTT) to evaluate the psychometric properties of the ADT among Chinese university students. This study finds a robust correlation between the ADT and the CET-6 writing and translation subscores. In addition, there is a high level of expert agreement regarding the construct validity evidence and the appropriateness of the scoring rubric. Furthermore, the results indicate that gender differences in validity indices are minimal. Taken together, these results suggest that the ADT is a valid measure for Chinese test-takers without gender discrimination. However, it is recommended that the cultural sensitivity of the scoring rubric be further refined and that the CET-6 subscores for writing be retained for predictive purposes, in order to better accommodate the needs of diverse test-taker populations. By addressing these issues, this study contributes to the broader discourse on fairness and validity in high-stakes language assessments.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 14:23:46 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhang",
"Yinyu",
""
]
] |
2509.05348
|
Isuru Munasinghe
|
Ashen Rodrigo, Isuru Munasinghe and Asanka Perera
|
Vision-Based Object Detection for UAV Solar Panel Inspection Using an Enhanced Defects Dataset
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Timely and accurate detection of defects and contaminants in solar panels is critical for maintaining the efficiency and reliability of photovoltaic systems. This study presents a comprehensive evaluation of five state-of-the-art object detection models: YOLOv3, Faster R-CNN, RetinaNet, EfficientDet, and Swin Transformer, for identifying physical and electrical defects as well as surface contaminants such as dust, dirt, and bird droppings on solar panels. A custom dataset, annotated in the COCO format and specifically designed for solar panel defect and contamination detection, was developed alongside a user interface to train and evaluate the models. The performance of each model is assessed and compared based on mean Average Precision (mAP), precision, recall, and inference speed. The results demonstrate the trade-offs between detection accuracy and computational efficiency, highlighting the relative strengths and limitations of each model. These findings provide valuable guidance for selecting appropriate detection approaches in practical solar panel monitoring and maintenance scenarios.
The dataset will be publicly available at https://github.com/IsuruMunasinghe98/solar-panel-inspection-dataset.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 14:41:31 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Rodrigo",
"Ashen",
""
],
[
"Munasinghe",
"Isuru",
""
],
[
"Perera",
"Asanka",
""
]
] |
2509.05350
|
Joshua Ward
|
Joshua Ward, Yuxuan Yang, Chi-Hua Wang, Guang Cheng
|
Ensembling Membership Inference Attacks Against Tabular Generative Models
| null | null |
10.1145/3733799.3762977
| null |
cs.CR cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Membership Inference Attacks (MIAs) have emerged as a principled framework for auditing the privacy of synthetic data generated by tabular generative models, where many diverse methods have been proposed that each exploit different privacy leakage signals. However, in realistic threat scenarios, an adversary must choose a single method without a priori guarantee that it will be the empirically highest performing option. We study this challenge as a decision theoretic problem under uncertainty and conduct the largest synthetic data privacy benchmark to date. Here, we find that no MIA constitutes a strictly dominant strategy across a wide variety of model architectures and dataset domains under our threat model. Motivated by these findings, we propose ensemble MIAs and show that unsupervised ensembles built on individual attacks offer empirically more robust, regret-minimizing strategies than individual attacks.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 18:44:07 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ward",
"Joshua",
""
],
[
"Yang",
"Yuxuan",
""
],
[
"Wang",
"Chi-Hua",
""
],
[
"Cheng",
"Guang",
""
]
] |
2509.05351
|
Guoyue Xu
|
Guoyue Xu, Renzheng Zhang, Tengfei Luo
|
Self-Driving Laboratory Optimizes the Lower Critical Solution Temperature of Thermoresponsive Polymers
| null | null | null | null |
cond-mat.soft cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
To overcome the inherent inefficiencies of traditional trial-and-error materials discovery, the scientific community is increasingly developing autonomous laboratories that integrate data-driven decision-making into closed-loop experimental workflows. In this work, we realize this concept for thermoresponsive polymers by developing a low-cost, "frugal twin" platform for the optimization of the lower critical solution temperature (LCST) of poly(N-isopropylacrylamide) (PNIPAM). Our system integrates robotic fluid-handling, on-line sensors, and Bayesian optimization (BO) that navigates the multi-component salt solution spaces to achieve user-specified LCST targets. The platform demonstrates convergence to target properties within a minimal number of experiments. It strategically explores the parameter space, learns from informative "off-target" results, and self-corrects to achieve the final targets. By providing an accessible and adaptable blueprint, this work lowers the barrier to entry for autonomous experimentation and accelerates the design and discovery of functional polymers.
|
[
{
"version": "v1",
"created": "Tue, 02 Sep 2025 19:35:33 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xu",
"Guoyue",
""
],
[
"Zhang",
"Renzheng",
""
],
[
"Luo",
"Tengfei",
""
]
] |
2509.05352
|
Cuong Hoang
|
Cuong Manh Hoang
|
Unsupervised Instance Segmentation with Superpixels
| null |
Pattern Recognition, 2025
| null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Instance segmentation is essential for numerous computer vision applications, including robotics, human-computer interaction, and autonomous driving. Currently, popular models bring impressive performance in instance segmentation by training with a large number of human annotations, which are costly to collect. For this reason, we present a new framework that efficiently and effectively segments objects without the need for human annotations. Firstly, a MultiCut algorithm is applied to self-supervised features for coarse mask segmentation. Then, a mask filter is employed to obtain high-quality coarse masks. To train the segmentation network, we compute a novel superpixel-guided mask loss, comprising hard loss and soft loss, with high-quality coarse masks and superpixels segmented from low-level image features. Lastly, a self-training process with a new adaptive loss is proposed to improve the quality of predicted masks. We conduct experiments on public datasets in instance segmentation and object detection to demonstrate the effectiveness of the proposed framework. The results show that the proposed framework outperforms previous state-of-the-art methods.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 02:46:48 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hoang",
"Cuong Manh",
""
]
] |
2509.05353
|
Murphy E. Egwe Dr.
|
O.A. Ariyo and M.E. Egwe
|
Quaternionic Spherical Fourier Multipliers for Gelfand Pairs
| null | null | null | null |
math.RT math.FA
|
http://creativecommons.org/licenses/by/4.0/
|
Let $\mathbb{H}_q$ denote the quaternionic Heisenberg group with $\mathbb{R}^4\times\mathbb{R}^3$ stratification and $K$ a compact subgroup of Automorphism. We construct the spherical Fourier multiplier related to a Gelfand Pairs on the Quaternionic Heisenberg group $\mathbb{H}_q$ via the motion group $G=K\ltimes \mathbb{H}_q$ of $\mathbb{H}_q$, where $K\ltimes \mathbb{H}_q$ is the semidirect product of $\mathbb{H}_q$ and its compact subgroup of automorphism $K\in Aut(\mathbb{H}_q)$.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 04:12:16 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ariyo",
"O. A.",
""
],
[
"Egwe",
"M. E.",
""
]
] |
2509.05354
|
Demian Pouzo
|
Zhuokai Huang, Demian Pouzo, and Andr\'es Rodr\'iguez-Clare
|
Characterizing Optimality in Dynamic Settings: A Monotonicity-based Approach
| null | null | null | null |
econ.TH econ.GN math.OC q-fin.EC
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
We develop a novel analytical method for studying optimal paths in dynamic optimization problems under general monotonicity conditions. The method centers on a locator function -- a simple object constructed directly from the model's primitives -- whose roots identify interior steady states and whose slope determines their local stability. Under strict concavity of the payoff function, the locator function also characterizes basins of attraction, yielding a complete description of qualitative dynamics. Without concavity, it can still deliver sharp results: if the function is single crossing from above, its root identifies a globally stable steady state; if the locator function is inverted-U-shaped with two interior roots (a typical case), only the higher root can be a locally stable interior steady state. The locator function further enables comparative statics of steady states with respect to parameters through direct analysis of its derivatives. These results are obtained without solving the full dynamic program. We illustrate the approach using a generalized neoclassical growth model, a rational (un)fitness model, and a learning-by-doing economy.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 05:12:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Huang",
"Zhuokai",
""
],
[
"Pouzo",
"Demian",
""
],
[
"Rodríguez-Clare",
"Andrés",
""
]
] |
2509.05355
|
Ahmed R. Sadik Dr.-Ing.
|
Ahmed R. Sadik, Muhammad Ashfaq, Niko M\"akitalo, Tommi Mikkonen
|
Human-LLM Synergy in Context-Aware Adaptive Architecture for Scalable Drone Swarm Operation
| null | null | null | null |
cs.RO cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
The deployment of autonomous drone swarms in disaster response missions necessitates the development of flexible, scalable, and robust coordination systems. Traditional fixed architectures struggle to cope with dynamic and unpredictable environments, leading to inefficiencies in energy consumption and connectivity. This paper addresses this gap by proposing an adaptive architecture for drone swarms, leveraging a Large Language Model to dynamically select the optimal architecture as centralized, hierarchical, or holonic based on real time mission parameters such as task complexity, swarm size, and communication stability. Our system addresses the challenges of scalability, adaptability, and robustness,ensuring efficient energy consumption and maintaining connectivity under varying conditions. Extensive simulations demonstrate that our adaptive architecture outperforms traditional static models in terms of scalability, energy efficiency, and connectivity. These results highlight the potential of our approach to provide a scalable, adaptable, and resilient solution for real world disaster response scenarios.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 11:46:19 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sadik",
"Ahmed R.",
""
],
[
"Ashfaq",
"Muhammad",
""
],
[
"Mäkitalo",
"Niko",
""
],
[
"Mikkonen",
"Tommi",
""
]
] |
2509.05356
|
Justus Huebotter
|
Justus Huebotter, Pablo Lanillos, Marcel van Gerven, Serge Thill
|
Spiking Neural Networks for Continuous Control via End-to-End Model-Based Learning
| null | null | null | null |
cs.RO cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Despite recent progress in training spiking neural networks (SNNs) for classification, their application to continuous motor control remains limited. Here, we demonstrate that fully spiking architectures can be trained end-to-end to control robotic arms with multiple degrees of freedom in continuous environments. Our predictive-control framework combines Leaky Integrate-and-Fire dynamics with surrogate gradients, jointly optimizing a forward model for dynamics prediction and a policy network for goal-directed action. We evaluate this approach on both a planar 2D reaching task and a simulated 6-DOF Franka Emika Panda robot. Results show that SNNs can achieve stable training and accurate torque control, establishing their viability for high-dimensional motor tasks. An extensive ablation study highlights the role of initialization, learnable time constants, and regularization in shaping training dynamics. We conclude that while stable and effective control can be achieved, recurrent spiking networks remain highly sensitive to hyperparameter settings, underscoring the importance of principled design choices.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 12:05:58 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Huebotter",
"Justus",
""
],
[
"Lanillos",
"Pablo",
""
],
[
"van Gerven",
"Marcel",
""
],
[
"Thill",
"Serge",
""
]
] |
2509.05357
|
Bernhard Wortmann
|
Bernhard Wortmann, Detlef Stolten, Heidi Heinrichs
|
Critical Iridium Demands arising from future Expansion of Proton Exchange Membrane Electrolysis
| null | null | null | null |
econ.GN cond-mat.mtrl-sci q-fin.EC
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Proton exchange membrane electrolysis (PEMEL) is a key technology for producing green hydrogen, but its scalability is limited by the use of scarce materials, particularly iridium. Iridium oxide, the preferred anode catalyst in PEMEL, offers exceptional stability but is produced only as a by-product of platinum mining, with annual output around 7.5 tons. This study estimates future iridium demand for PEMEL under various deployment scenarios and technological advances. Results show that meeting net zero targets will require both significant improvements in catalyst efficiency and access to roughly 30\% of global iridium production annually. Supply shortages could arise as early as 2030, earlier than previously anticipated. The analysis also reveals that long-term iridium needs beyond 2040 are significantly underestimated. These findings underscore the urgent need for innovation in material efficiency and recycling, and the importance of integrating resource constraints into energy policy and technology planning to ensure a sustainable hydrogen transition.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 14:17:30 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wortmann",
"Bernhard",
""
],
[
"Stolten",
"Detlef",
""
],
[
"Heinrichs",
"Heidi",
""
]
] |
2509.05358
|
Iniakpokeikiye Peter Thompson
|
Iniakpokeikiye Peter Thompson, Yi Dewei, Reiter Ehud
|
Mobile Phone Sensor-based Nigerian Driving Dataset to Detect Alcohol-influenced Behaviours
|
The article has 4 tables and 3 figures
| null | null | null |
cs.CY
|
http://creativecommons.org/licenses/by-sa/4.0/
|
This paper presents a unique driving dataset collected in Nigeria via mobile phone sensors to support a machine learning model for detecting alcohol-influenced driving behaviours, with the long-term aim of integrating this model into a mobile application that encourages safer driving behaviours. Driving under the influence of alcohol is a major public safety concern, particularly in low-income countries like Nigeria, where traditional enforcement mechanisms may be limited. The proposed model leverages smartphone sensors such as accelerometers, gyroscopes, and GPS to provide a non-invasive, continuous solution for detecting impaired driving patterns in real time. This study adapts existing data processing and pattern matching methodologies to label real-world driving data collected from Nigerian drivers, which are then used to train the model. A decision tree classifier is developed to detect alcohol influence, based on behavioural and temporal features, achieving a recall of 100%, a precision of 60%, and an F1 score of 75%. The model's overall accuracy was 90.91%, ensuring that no alcohol influenced trips were missed. Key predictive features included speed variability, course deviation, and time of day, which align with established patterns of alcohol consumption. This study contributes to the field by demonstrating how machine learning can be applied in low-resource environments to improve road safety. The findings suggest that the model can significantly enhance the detection and prevention of risky driving behaviours, with the potential for future integration into mobile applications to provide real-time feedback and encourage safer driving practices. This scalable and accessible solution offers a new approach to addressing road safety challenges in regions where traditional interventions are inadequate.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 15:21:03 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Thompson",
"Iniakpokeikiye Peter",
""
],
[
"Dewei",
"Yi",
""
],
[
"Ehud",
"Reiter",
""
]
] |
2509.05359
|
Yanis Labrak
|
Yanis Labrak, Richard Dufour, Micka\"el Rouvier
|
An Empirical Analysis of Discrete Unit Representations in Speech Language Modeling Pre-training
|
Published in International Conference on Text, Speech, and Dialogue, 13-24
|
International Conference on Text, Speech, and Dialogue 2025
|
10.1007/978-3-032-02548-7_2
| null |
cs.CL cs.AI eess.AS
|
http://creativecommons.org/publicdomain/zero/1.0/
|
This paper investigates discrete unit representations in Speech Language Models (SLMs), focusing on optimizing speech modeling during continual pre-training. In this paper, we systematically examine how model architecture, data representation, and training robustness influence the pre-training stage in which we adapt existing pre-trained language models to the speech modality. Our experiments highlight the role of speech encoders and clustering granularity across different model scales, showing how optimal discretization strategies vary with model capacity. By examining cluster distribution and phonemic alignments, we investigate the effective use of discrete vocabulary, uncovering both linguistic and paralinguistic patterns. Additionally, we explore the impact of clustering data selection on model robustness, highlighting the importance of domain matching between discretization training and target applications.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 18:11:53 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Labrak",
"Yanis",
""
],
[
"Dufour",
"Richard",
""
],
[
"Rouvier",
"Mickaël",
""
]
] |
2509.05360
|
Jerry Li
|
Jerry Li, Evangelos Papalexakis
|
Beyond ROUGE: N-Gram Subspace Features for LLM Hallucination Detection
| null | null | null | null |
cs.CL cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large Language Models (LLMs) have demonstrated effectiveness across a wide variety of tasks involving natural language, however, a fundamental problem of hallucinations still plagues these models, limiting their trustworthiness in generating consistent, truthful information. Detecting hallucinations has quickly become an important topic, with various methods such as uncertainty estimation, LLM Judges, retrieval augmented generation (RAG), and consistency checks showing promise. Many of these methods build upon foundational metrics, such as ROUGE, BERTScore, or Perplexity, which often lack the semantic depth necessary to detect hallucinations effectively. In this work, we propose a novel approach inspired by ROUGE that constructs an N-Gram frequency tensor from LLM-generated text. This tensor captures richer semantic structure by encoding co-occurrence patterns, enabling better differentiation between factual and hallucinated content. We demonstrate this by applying tensor decomposition methods to extract singular values from each mode and use these as input features to train a multi-layer perceptron (MLP) binary classifier for hallucinations. Our method is evaluated on the HaluEval dataset and demonstrates significant improvements over traditional baselines, as well as competitive performance against state-of-the-art LLM judges.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 18:52:24 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Li",
"Jerry",
""
],
[
"Papalexakis",
"Evangelos",
""
]
] |
2509.05361
|
Alexander Mark
|
Alex Mark, Aaron Scher
|
Governing AI R&D: A Legal Framework for Constraining Dangerous AI
| null | null | null | null |
cs.CY cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
As AI advances, governing its development may become paramount to public safety. Lawmakers may seek to restrict the development and release of AI models or of AI research itself. These governance actions could trigger legal challenges that invalidate the actions, so lawmakers should consider these challenges ahead of time. We investigate three classes of potential litigation risk for AI regulation in the U.S.: the First Amendment, administrative law, and the Fourteenth Amendment. We discuss existing precedent that is likely to apply to AI, which legal challenges are likely to arise, and how lawmakers might preemptively address them. Effective AI regulation is possible, but it requires careful implementation to avoid these legal challenges.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:09:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Mark",
"Alex",
""
],
[
"Scher",
"Aaron",
""
]
] |
2509.05362
|
Ismail Hossain
|
Ismail Hossain, Sai Puppala, Sajedul Talukder, Md Jahangir Alam
|
AI-in-the-Loop: Privacy Preserving Real-Time Scam Detection and Conversational Scambaiting by Leveraging LLMs and Federated Learning
|
This paper got accepted in 26th Privacy Enhancing Technologies Symposium (PETS 2026). We uploaded it into ArXiv as pre-print
| null | null | null |
cs.CR cs.AI cs.LG cs.SI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Scams exploiting real-time social engineering -- such as phishing, impersonation, and phone fraud -- remain a persistent and evolving threat across digital platforms. Existing defenses are largely reactive, offering limited protection during active interactions. We propose a privacy-preserving, AI-in-the-loop framework that proactively detects and disrupts scam conversations in real time. The system combines instruction-tuned artificial intelligence with a safety-aware utility function that balances engagement with harm minimization, and employs federated learning to enable continual model updates without raw data sharing. Experimental evaluations show that the system produces fluent and engaging responses (perplexity as low as 22.3, engagement $\approx$0.80), while human studies confirm significant gains in realism, safety, and effectiveness over strong baselines. In federated settings, models trained with FedAvg sustain up to 30 rounds while preserving high engagement ($\approx$0.80), strong relevance ($\approx$0.74), and low PII leakage ($\leq$0.0085). Even with differential privacy, novelty and safety remain stable, indicating that robust privacy can be achieved without sacrificing performance. The evaluation of guard models (LlamaGuard, LlamaGuard2/3, MD-Judge) shows a straightforward pattern: stricter moderation settings reduce the chance of exposing personal information, but they also limit how much the model engages in conversation. In contrast, more relaxed settings allow longer and richer interactions, which improve scam detection, but at the cost of higher privacy risk. To our knowledge, this is the first framework to unify real-time scam-baiting, federated privacy preservation, and calibrated safety moderation into a proactive defense paradigm.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:19:48 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hossain",
"Ismail",
""
],
[
"Puppala",
"Sai",
""
],
[
"Talukder",
"Sajedul",
""
],
[
"Alam",
"Md Jahangir",
""
]
] |
2509.05363
|
Lijie Ding
|
Lijie Ding and Changwoo Do
|
SasAgent: Multi-Agent AI System for Small-Angle Scattering Data Analysis
|
8 pages, 7 figures
| null | null | null |
cs.AI cond-mat.mtrl-sci cs.MA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We introduce SasAgent, a multi-agent AI system powered by large language models (LLMs) that automates small-angle scattering (SAS) data analysis by leveraging tools from the SasView software and enables user interaction via text input. SasAgent features a coordinator agent that interprets user prompts and delegates tasks to three specialized agents for scattering length density (SLD) calculation, synthetic data generation, and experimental data fitting. These agents utilize LLM-friendly tools to execute tasks efficiently. These tools, including the model data tool, Retrieval-Augmented Generation (RAG) documentation tool, bump fitting tool, and SLD calculator tool, are derived from the SasView Python library. A user-friendly Gradio-based interface enhances user accessibility. Through diverse examples, we demonstrate SasAgent's ability to interpret complex prompts, calculate SLDs, generate accurate scattering data, and fit experimental datasets with high precision. This work showcases the potential of LLM-driven AI systems to streamline scientific workflows and enhance automation in SAS research.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:23:23 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ding",
"Lijie",
""
],
[
"Do",
"Changwoo",
""
]
] |
2509.05364
|
Abdollah Baghaei Daemei
|
Abdollah Baghaei Daemei
|
Prototyping an AI-powered Tool for Energy Efficiency in New Zealand Homes
| null | null | null | null |
cs.CY cs.AI cs.ET
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Residential buildings contribute significantly to energy use, health outcomes, and carbon emissions. In New Zealand, housing quality has historically been poor, with inadequate insulation and inefficient heating contributing to widespread energy hardship. Recent reforms, including the Warmer Kiwi Homes program, Healthy Homes Standards, and H1 Building Code upgrades, have delivered health and comfort improvements, yet challenges persist. Many retrofits remain partial, data on household performance are limited, and decision-making support for homeowners is fragmented. This study presents the design and evaluation of an AI-powered decision-support tool for residential energy efficiency in New Zealand. The prototype, developed using Python and Streamlit, integrates data ingestion, anomaly detection, baseline modeling, and scenario simulation (e.g., LED retrofits, insulation upgrades) into a modular dashboard. Fifteen domain experts, including building scientists, consultants, and policy practitioners, tested the tool through semi-structured interviews. Results show strong usability (M = 4.3), high value of scenario outputs (M = 4.5), and positive perceptions of its potential to complement subsidy programs and regulatory frameworks. The tool demonstrates how AI can translate national policies into personalized, household-level guidance, bridging the gap between funding, standards, and practical decision-making. Its significance lies in offering a replicable framework for reducing energy hardship, improving health outcomes, and supporting climate goals. Future development should focus on carbon metrics, tariff modeling, integration with national datasets, and longitudinal trials to assess real-world adoption.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:41:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Daemei",
"Abdollah Baghaei",
""
]
] |
2509.05365
|
Dingcui Yu
|
Dingcui Yu, Yunpeng Song, Yiyang Huang, Yumiao Zhao, Yina Lv, Chundong Wang, Youtao Zhang, Liang Shi
|
Waltz: Temperature-Aware Cooperative Compression for High-Performance Compression-Based CSDs
|
Dingcui Yu, Yunpeng Song and Yiyang Huang have equal contributions to this work. Liang Shi and Yina Lv are the corresponding authors
| null | null | null |
cs.PF
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Data compression is widely adopted for modern solid-state drives (SSDs) to mitigate both storage capacity and SSD lifetime issues. Researchers have proposed compression schemes at different system layers, including device-side solutions like CCSDs ( c ompression-based c omputational SSDs) and compression supported by host-side, like F2FS (flash-friendly file system). We conduct quantitative studies to understand how host-side and device-side compression schemes affect the temperature and performance of SSD-based storage systems. From our experiments, device-side compression, facilitated by a hardware compression engine, can raise the temperature of CCSDs to intolerable levels, resulting in throttling and service shutdown. In contrast, host-side compression causes software-stack overhead, which often results in large performance degradation and resource consumption. To ensure efficient data compression with high performance and better temperature control, we propose Waltz, a temperature-aware cooperative compression method that schedules (de)compression tasks at the host and device sides by monitoring device temperature. Furthermore, we introduce two variants (Waltzs and Waltzp) for space and performance optimization, respectively. Waltz is implemented within F2FS, achieving high performance while extending SSD lifetime and preventing overheating-induced in-flight shutdowns.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:05:07 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yu",
"Dingcui",
""
],
[
"Song",
"Yunpeng",
""
],
[
"Huang",
"Yiyang",
""
],
[
"Zhao",
"Yumiao",
""
],
[
"Lv",
"Yina",
""
],
[
"Wang",
"Chundong",
""
],
[
"Zhang",
"Youtao",
""
],
[
"Shi",
"Liang",
""
]
] |
2509.05366
|
Muhammad Umar Khan
|
Umair Amjid, M. Umar Khan, S. A. Manan Kirmani
|
A Framework for Detection and Classification of Attacks on Surveillance Cameras under IoT Networks
| null | null | null | null |
cs.CR cs.NI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The increasing use of Internet of Things (IoT) devices has led to a rise in security related concerns regarding IoT Networks. The surveillance cameras in IoT networks are vulnerable to security threats such as brute force and zero-day attacks which can lead to unauthorized access by hackers and potential spying on the users activities. Moreover, these cameras can be targeted by Denial of Service (DOS) attacks, which will make it unavailable for the user. The proposed AI based framework will leverage machine learning algorithms to analyze network traffic and detect anomalous behavior, allowing for quick detection and response to potential intrusions. The framework will be trained and evaluated using real-world datasets to learn from past security incidents and improve its ability to detect potential intrusion.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 05:33:41 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Amjid",
"Umair",
""
],
[
"Khan",
"M. Umar",
""
],
[
"Kirmani",
"S. A. Manan",
""
]
] |
2509.05367
|
Shei Pern Chua
|
Shei Pern Chua, Thai Zhen Leng, Teh Kai Jun, Xiao Li, Xiaolin Hu
|
Between a Rock and a Hard Place: Exploiting Ethical Reasoning to Jailbreak LLMs
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Large language models (LLMs) have undergone safety alignment efforts to mitigate harmful outputs. However, as LLMs become more sophisticated in reasoning, their intelligence may introduce new security risks. While traditional jailbreak attacks relied on singlestep attacks, multi-turn jailbreak strategies that adapt dynamically to context remain underexplored. In this work, we introduce TRIAL (Trolley-problem Reasoning for Interactive Attack Logic), a framework that leverages LLMs ethical reasoning to bypass their safeguards. TRIAL embeds adversarial goals within ethical dilemmas modeled on the trolley problem. TRIAL demonstrates high jailbreak success rates towards both open and close-source models. Our findings underscore a fundamental limitation in AI safety: as models gain advanced reasoning abilities, the nature of their alignment may inadvertently allow for more covert security vulnerabilities to be exploited. TRIAL raises an urgent need in reevaluating safety alignment oversight strategies, as current safeguards may prove insufficient against context-aware adversarial attack.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 05:53:20 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chua",
"Shei Pern",
""
],
[
"Leng",
"Thai Zhen",
""
],
[
"Jun",
"Teh Kai",
""
],
[
"Li",
"Xiao",
""
],
[
"Hu",
"Xiaolin",
""
]
] |
2509.05368
|
Quan Chen
|
Quan Chen, Chenrui Shi, Qi Chen, Yuwei Wu, Zhi Gao, Xintong Zhang, Rui Gao, Kun Wu, and Yunde Jia
|
Long-Horizon Visual Imitation Learning via Plan and Code Reflection
|
9 pages, 4 figures. Submitted to AAAI 2026
| null | null | null |
cs.RO cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Learning from long-horizon demonstrations with complex action sequences presents significant challenges for visual imitation learning, particularly in understanding temporal relationships of actions and spatial relationships between objects. In this paper, we propose a new agent framework that incorporates two dedicated reflection modules to enhance both plan and code generation. The plan generation module produces an initial action sequence, which is then verified by the plan reflection module to ensure temporal coherence and spatial alignment with the demonstration video. The code generation module translates the plan into executable code, while the code reflection module verifies and refines the generated code to ensure correctness and consistency with the generated plan. These two reflection modules jointly enable the agent to detect and correct errors in both the plan generation and code generation, improving performance in tasks with intricate temporal and spatial dependencies. To support systematic evaluation, we introduce LongVILBench, a benchmark comprising 300 human demonstrations with action sequences of up to 18 steps. LongVILBench emphasizes temporal and spatial complexity across multiple task types. Experimental results demonstrate that existing methods perform poorly on this benchmark, whereas our new framework establishes a strong baseline for long-horizon visual imitation learning.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 06:10:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chen",
"Quan",
""
],
[
"Shi",
"Chenrui",
""
],
[
"Chen",
"Qi",
""
],
[
"Wu",
"Yuwei",
""
],
[
"Gao",
"Zhi",
""
],
[
"Zhang",
"Xintong",
""
],
[
"Gao",
"Rui",
""
],
[
"Wu",
"Kun",
""
],
[
"Jia",
"Yunde",
""
]
] |
2509.05369
|
Brighton Amuni Mr.
|
Brighton Savayi Amuni, Stella Silas Karimi, Felistus Mwikali Patrick
|
Influence of Stakeholder Involvement in M&E on the Performance of Donor-Funded Projects in Informal Settlements in Kisumu Central Sub-County, Kisumu County, Kenya
|
8 Pages, 5 Figures, Published with International Research Journal of Economics and Management Studies (IRJEMS)
|
IRJEMS, Vol. 4, No. 7, pp. 185-192, 2025
|
0.56472/25835238/IRJEMS-V4I7P123
| null |
cs.CY
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Stakeholder engagement and participatory approaches influence the effectiveness of donor-funded projects. Participatory Monitoring and Evaluation (PM&E) methodologies ensure that local communities play an active role in decision-making, leading to more sustainable outcomes. Given the complex socio-political landscape of Kisumu Central Sub-County, there is a critical need for inclusive and context-responsive project monitoring strategies. Initiatives that have integrated local leaders, youth groups, and women-led organizations into their M&E processes tend to achieve stronger community buy-in, improved continuity, and more impactful outcomes. The current study explored the influence of stakeholder involvement in M&E on the performance of donor-funded projects in informal settlements in Kisumu Central Sub-County, Kenya. The study was guided by Stakeholder Engagement Theory. The study used a Convergent Parallel design with a sample size of 364 respondents computed using Yamanes' Sampling formula, drawn from 27 donor-funded projects in Obunga and Nyalenda informal Settlements. Purposive sampling was used for project managers, project M&E staff, and community members served, while community members were selected using stratified random sampling. The study findings revealed that there were regular opportunities for stakeholder interaction in the projects (x =4.05, SD 1.08), stakeholders contributed to the development of the organization/project (x= 3.79, SD=.940), although stakeholders' perspectives and opinions were not diligently incorporated into programming (x=2.06, SD=.879) as anticipated. The study, therefore, concluded that stakeholder involvement in M&E influenced the performance of donor-funded projects. The study recommended that the project managers in donor-funded projects need to enhance stakeholder involvement for project ownership and sustainability.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 06:18:35 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Amuni",
"Brighton Savayi",
""
],
[
"Karimi",
"Stella Silas",
""
],
[
"Patrick",
"Felistus Mwikali",
""
]
] |
2509.05370
|
Krishnendu Guha
|
Tanya Joshi, Krishnendu Guha
|
Quantum AI Algorithm Development for Enhanced Cybersecurity: A Hybrid Approach to Malware Detection
|
10 pages
| null | null | null |
cs.CR cs.ET
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This study explores the application of quantum machine learning (QML) algorithms to enhance cybersecurity threat detection, particularly in the classification of malware and intrusion detection within high-dimensional datasets. Classical machine learning approaches encounter limitations when dealing with intricate, obfuscated malware patterns and extensive network intrusion data. To address these challenges, we implement and evaluate various QML algorithms, including Quantum Neural Networks (QNN), Quantum Support Vector Machines (QSVM), and hybrid Quantum Convolutional Neural Networks (QCNN) for malware detection tasks. Our experimental analysis utilized two datasets: the Intrusion dataset, comprising 150 samples with 56 memory-based features derived from Volatility framework analysis, and the ObfuscatedMalMem2022 dataset, containing 58,596 samples with 57 features representing benign and malicious software. Remarkably, our QML methods demonstrated superior performance compared to classical approaches, achieving accuracies of 95% for QNN and 94% for QSVM. These quantum-enhanced methods leveraged quantum superposition and entanglement principles to accurately identify complex patterns within highly obfuscated malware samples that were imperceptible to classical methods. To further advance malware analysis, we propose a novel real-time malware analysis framework that incorporates Quantum Feature Extraction using Quantum Fourier Transform, Quantum Feature Maps, and Classification using Variational Quantum Circuits. This system integrates explainable AI methods, including GradCAM++ and ScoreCAM algorithms, to provide interpretable insights into the quantum decision-making processes.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 09:28:05 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Joshi",
"Tanya",
""
],
[
"Guha",
"Krishnendu",
""
]
] |
2509.05371
|
Muamer Kadic
|
Francesco Avallone, Federico Bosia, Yi Chen, Giada Colombo, Richard Craster, Jacopo Maria De Ponti, Nicol\`o Fabbiane, Michael R. Haberman, Mahmoud I. Hussein, Wontae Hwang, Umberto Iemma, Abigail Juhl, Muamer Kadic, Marios Kotsonis, Vincent Laude, Olivier Marquet, Fabien Mery, Theodoros Michelis, Mostafa Nouh, Daniele Ragni, Marie Touboul, Martin Wegener, Anastasiia O. Krushynska
|
Metamaterials and Fluid Flows
|
4 figures
| null | null | null |
cond-mat.soft cond-mat.mtrl-sci physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
Understanding and controlling the dynamic interactions between fluid flows and solid materials and structures-a field known as fluid-structure interaction -is central not only to established disciplines such as aerospace and naval engineering but also to emerging technologies such as energy harvesting, soft robotics, and biomedical devices. In recent years, the advent of metamaterials-rationally designed composites with properties beyond their constituents, often not found in conventional materials-has provided exciting opportunities for rethinking and redesigning fluid-structure interaction. The premise of engineering the internal structure of materials interfacing with fluid flows is opening a new horizon for precise and effective manipulation and control of coupled fluidic, acoustic, and elastodynamics responses. This review focuses on this relatively unexplored interdisciplinary theme with broad real-world technological significance. Key performance metrics, such as fuel consumption of transport systems, efficiency of renewable energy extraction, mitigation of noise emissions, and resilience to structural fatigue, depend on the control of interactions between flow, acoustic, and vibration mechanisms. Flow control, for example, which spans a wealth of regimes such as laminar, transitional, turbulent, and unsteady separated flows, is highly influenced by the ability to tailor fluid-structure interaction behavior. We survey and discuss theoretical frameworks that describe the interplay between fluids and elastic solids, with a focus on contemporary work and emerging concepts. The paper is organised into three main sections-flow-structure interactions, acoustic-structure interactions, and exotic metamaterial concepts with potential impact on fluid-structure interaction-and concludes with perspectives on current challenges and future directions in this rapidly expanding area of research.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 09:40:17 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Avallone",
"Francesco",
""
],
[
"Bosia",
"Federico",
""
],
[
"Chen",
"Yi",
""
],
[
"Colombo",
"Giada",
""
],
[
"Craster",
"Richard",
""
],
[
"De Ponti",
"Jacopo Maria",
""
],
[
"Fabbiane",
"Nicolò",
""
],
[
"Haberman",
"Michael R.",
""
],
[
"Hussein",
"Mahmoud I.",
""
],
[
"Hwang",
"Wontae",
""
],
[
"Iemma",
"Umberto",
""
],
[
"Juhl",
"Abigail",
""
],
[
"Kadic",
"Muamer",
""
],
[
"Kotsonis",
"Marios",
""
],
[
"Laude",
"Vincent",
""
],
[
"Marquet",
"Olivier",
""
],
[
"Mery",
"Fabien",
""
],
[
"Michelis",
"Theodoros",
""
],
[
"Nouh",
"Mostafa",
""
],
[
"Ragni",
"Daniele",
""
],
[
"Touboul",
"Marie",
""
],
[
"Wegener",
"Martin",
""
],
[
"Krushynska",
"Anastasiia O.",
""
]
] |
2509.05372
|
Piotr Przymus
|
Piotr Przymus (1), Andreas Happe (2), J\"urgen Cito (2) ((1) Nicolaus Copernicus University, Toru\'n, Poland, (2) TU Wienm Vienna, Austria)
|
Adversarial Bug Reports as a Security Risk in Language Model-Based Automated Program Repair
| null | null | null | null |
cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Large Language Model (LLM) - based Automated Program Repair (APR) systems are increasingly integrated into modern software development workflows, offering automated patches in response to natural language bug reports. However, this reliance on untrusted user input introduces a novel and underexplored attack surface. In this paper, we investigate the security risks posed by adversarial bug reports -- realistic-looking issue submissions crafted to mislead APR systems into producing insecure or harmful code changes. We develop a comprehensive threat model and conduct an empirical study to evaluate the vulnerability of state-of-the-art APR systems to such attacks. Our demonstration comprises 51 adversarial bug reports generated across a spectrum of strategies, from manual curation to fully automated pipelines. We test these against leading APR model and assess both pre-repair defenses (e.g., LlamaGuard variants, PromptGuard variants, Granite-Guardian, and custom LLM filters) and post-repair detectors (GitHub Copilot, CodeQL). Our findings show that current defenses are insufficient: 90\% of crafted bug reports triggered attacker-aligned patches. The best pre-repair filter blocked only 47\%, while post-repair analysis-often requiring human oversight-was effective in just 58\% of cases. To support scalable security testing, we introduce a prototype framework for automating the generation of adversarial bug reports. Our analysis exposes a structural asymmetry: generating adversarial inputs is inexpensive, while detecting or mitigating them remains costly and error-prone. We conclude with practical recommendations for improving the robustness of APR systems against adversarial misuse and highlight directions for future work on trustworthy automated repair.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 09:41:57 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Przymus",
"Piotr",
"",
"Nicolaus Copernicus University, Toruń, Poland"
],
[
"Happe",
"Andreas",
"",
"TU Wienm Vienna, Austria"
],
[
"Cito",
"Jürgen",
"",
"TU Wienm Vienna, Austria"
]
] |
2509.05373
|
Yuna Lim
|
Yuna Lim, Yubin Seo, Jacob Lee, and Eunok Jung
|
Comparison of the effectiveness and costs of hepatitis A vaccination strategies to mitigate future deaths in the Republic of Korea
| null | null | null | null |
q-bio.PE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Improved hygiene and infant vaccinations have led to age specific variations in hepatitis A antibody prevalence in Korea, with lower rates among individuals in their 20s to 40s. Given that the fatality rate of hepatitis A increases for those aged 50 and older, the low immunity level among younger adults indicates a future risk of increased deaths in older age groups without additional preventive measures.We developed an age structured transmission model to assess the impact of adult vaccination, assuming it begins in 2025. The 20s age group was modeled with an additional compartment to account for hepatitis A vaccination administered to military recruits. Vaccination strategies targeting the 20s to 30s and 40s to 50s age groups were compared, considering antibody testing costs for the latter in Korea and focusing on projected deaths over approximately 50 years. When the total vaccination cost is fixed, targeting the 40s to 50s group covers 20% fewer individuals than the 20s to 30s group but yields a 1.3 to 1.5 fold greater reduction in deaths. When the total vaccine supply is fixed, targeting the 40s to 50s group is 1.2 times more expensive but yields a 1.7 to 1.8 fold greater reduction in deaths than the 20s to 30s group. Moreover, including a second dose for military personnel in the 20s to 30s group has minimal impact on reducing deaths. Preventive measures for adults may be necessary to reduce future hepatitis A related deaths. Vaccinating the 40s to 50s group would be more effective in reducing deaths than vaccinating the 20s to 30s group.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 10:28:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lim",
"Yuna",
""
],
[
"Seo",
"Yubin",
""
],
[
"Lee",
"Jacob",
""
],
[
"Jung",
"Eunok",
""
]
] |
2509.05374
|
Zhiqiang Yuan
|
Zhiqiang Yuan, Jinchao Zhang, Jie Zhou
|
A Synthetic-to-Real Dehazing Method based on Domain Unification
|
ICME 2025 Accept
| null | null | null |
eess.IV cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Due to distribution shift, the performance of deep learning-based method for image dehazing is adversely affected when applied to real-world hazy images. In this paper, we find that such deviation in dehazing task between real and synthetic domains may come from the imperfect collection of clean data. Owing to the complexity of the scene and the effect of depth, the collected clean data cannot strictly meet the ideal conditions, which makes the atmospheric physics model in the real domain inconsistent with that in the synthetic domain. For this reason, we come up with a synthetic-to-real dehazing method based on domain unification, which attempts to unify the relationship between the real and synthetic domain, thus to let the dehazing model more in line with the actual situation. Extensive experiments qualitatively and quantitatively demonstrate that the proposed dehazing method significantly outperforms state-of-the-art methods on real-world images.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 11:39:22 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yuan",
"Zhiqiang",
""
],
[
"Zhang",
"Jinchao",
""
],
[
"Zhou",
"Jie",
""
]
] |
2509.05375
|
Arend Hintze
|
Arend Hintze
|
Characterizing Fitness Landscape Structures in Prompt Engineering
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
While prompt engineering has emerged as a crucial technique for optimizing large language model performance, the underlying optimization landscape remains poorly understood. Current approaches treat prompt optimization as a black-box problem, applying sophisticated search algorithms without characterizing the landscape topology they navigate. We present a systematic analysis of fitness landscape structures in prompt engineering using autocorrelation analysis across semantic embedding spaces. Through experiments on error detection tasks with two distinct prompt generation strategies -- systematic enumeration (1,024 prompts) and novelty-driven diversification (1,000 prompts) -- we reveal fundamentally different landscape topologies. Systematic prompt generation yields smoothly decaying autocorrelation, while diversified generation exhibits non-monotonic patterns with peak correlation at intermediate semantic distances, indicating rugged, hierarchically structured landscapes. Task-specific analysis across 10 error detection categories reveals varying degrees of ruggedness across different error types. Our findings provide an empirical foundation for understanding the complexity of optimization in prompt engineering landscapes.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 11:52:19 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hintze",
"Arend",
""
]
] |
2509.05376
|
Abdul Rehman Mr.
|
Abdul Rehman, Are D{\ae}hlen, Ilona Heldal, Jerry Chun-wei Lin
|
Privacy Preservation and Identity Tracing Prevention in AI-Driven Eye Tracking for Interactive Learning Environments
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Eye-tracking technology can aid in understanding neurodevelopmental disorders and tracing a person's identity. However, this technology poses a significant risk to privacy, as it captures sensitive information about individuals and increases the likelihood that data can be traced back to them. This paper proposes a human-centered framework designed to prevent identity backtracking while preserving the pedagogical benefits of AI-powered eye tracking in interactive learning environments. We explore how real-time data anonymization, ethical design principles, and regulatory compliance (such as GDPR) can be integrated to build trust and transparency. We first demonstrate the potential for backtracking student IDs and diagnoses in various scenarios using serious game-based eye-tracking data. We then provide a two-stage privacy-preserving framework that prevents participants from being tracked while still enabling diagnostic classification. The first phase covers four scenarios: I) Predicting disorder diagnoses based on different game levels. II) Predicting student IDs based on different game levels. III) Predicting student IDs based on randomized data. IV) Utilizing K-Means for out-of-sample data. In the second phase, we present a two-stage framework that preserves privacy. We also employ Federated Learning (FL) across multiple clients, incorporating a secure identity management system with dummy IDs and administrator-only access controls. In the first phase, the proposed framework achieved 99.3% accuracy for scenario 1, 63% accuracy for scenario 2, and 99.7% accuracy for scenario 3, successfully identifying and assigning a new student ID in scenario 4. In phase 2, we effectively prevented backtracking and established a secure identity management system with dummy IDs and administrator-only access controls, achieving an overall accuracy of 99.40%.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 13:08:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Rehman",
"Abdul",
""
],
[
"Dæhlen",
"Are",
""
],
[
"Heldal",
"Ilona",
""
],
[
"Lin",
"Jerry Chun-wei",
""
]
] |
2509.05377
|
Thien Phan
|
Duc-Thien Phan, Minh-Duong Nguyen, Quoc-Viet Pham and Huilong Pi
|
Enhancing Gradient Variance and Differential Privacy in Quantum Federated Learning
| null | null | null | null |
quant-ph cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Upon integrating Quantum Neural Network (QNN) as the local model, Quantum Federated Learning (QFL) has recently confronted notable challenges. Firstly, exploration is hindered over sharp minima, decreasing learning performance. Secondly, the steady gradient descent results in more stable and predictable model transmissions over wireless channels, making the model more susceptible to attacks from adversarial entities. Additionally, the local QFL model is vulnerable to noise produced by the quantum device's intermediate noise states, since it requires the use of quantum gates and circuits for training. This local noise becomes intertwined with learning parameters during training, impairing model precision and convergence rate. To address these issues, we propose a new QFL technique that incorporates differential privacy and introduces a dedicated noise estimation strategy to quantify and mitigate the impact of intermediate quantum noise. Furthermore, we design an adaptive noise generation scheme to alleviate privacy threats associated with the vanishing gradient variance phenomenon of QNN and enhance robustness against device noise. Experimental results demonstrate that our algorithm effectively balances convergence, reduces communication costs, and mitigates the adverse effects of intermediate quantum noise while maintaining strong privacy protection. Using real-world datasets, we achieved test accuracy of up to 98.47\% for the MNIST dataset and 83.85\% for the CIFAR-10 dataset while maintaining fast execution times.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 15:29:52 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Phan",
"Duc-Thien",
""
],
[
"Nguyen",
"Minh-Duong",
""
],
[
"Pham",
"Quoc-Viet",
""
],
[
"Pi",
"Huilong",
""
]
] |
2509.05378
|
Andreas Geert Motzfeldt
|
Andreas Motzfeldt, Joakim Edin, Casper L. Christensen, Christian Hardmeier, Lars Maal{\o}e, Anna Rogers
|
Code Like Humans: A Multi-Agent Solution for Medical Coding
|
EMNLP Findings 2025
| null | null | null |
cs.AI cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
In medical coding, experts map unstructured clinical notes to alphanumeric codes for diagnoses and procedures. We introduce Code Like Humans: a new agentic framework for medical coding with large language models. It implements official coding guidelines for human experts, and it is the first solution that can support the full ICD-10 coding system (+70K labels). It achieves the best performance to date on rare diagnosis codes (fine-tuned discriminative classifiers retain an advantage for high-frequency codes, to which they are limited). Towards future work, we also contribute an analysis of system performance and identify its `blind spots' (codes that are systematically undercoded).
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 16:31:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Motzfeldt",
"Andreas",
""
],
[
"Edin",
"Joakim",
""
],
[
"Christensen",
"Casper L.",
""
],
[
"Hardmeier",
"Christian",
""
],
[
"Maaløe",
"Lars",
""
],
[
"Rogers",
"Anna",
""
]
] |
2509.05379
|
Sharif Noor Zisad
|
Sharif Noor Zisad and Ragib Hasan
|
ThreatGPT: An Agentic AI Framework for Enhancing Public Safety through Threat Modeling
| null | null | null | null |
cs.CR cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
As our cities and communities become smarter, the systems that keep us safe, such as traffic control centers, emergency response networks, and public transportation, also become more complex. With this complexity comes a greater risk of security threats that can affect not just machines but real people's lives. To address this challenge, we present ThreatGPT, an agentic Artificial Intelligence (AI) assistant built to help people whether they are engineers, safety officers, or policy makers to understand and analyze threats in public safety systems. Instead of requiring deep cybersecurity expertise, it allows users to simply describe the components of a system they are concerned about, such as login systems, data storage, or communication networks. Then, with the click of a button, users can choose how they want the system to be analyzed by using popular frameworks such as STRIDE, MITRE ATT&CK, CVE reports, NIST, or CISA. ThreatGPT is unique because it does not just provide threat information, but rather it acts like a knowledgeable partner. Using few-shot learning, the AI learns from examples and generates relevant smart threat models. It can highlight what might go wrong, how attackers could take advantage, and what can be done to prevent harm. Whether securing a city's infrastructure or a local health service, this tool adapts to users' needs. In simple terms, ThreatGPT brings together AI and human judgment to make our public systems safer. It is designed not just to analyze threats, but to empower people to understand and act on them, faster, smarter, and with more confidence.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 20:26:54 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zisad",
"Sharif Noor",
""
],
[
"Hasan",
"Ragib",
""
]
] |
2509.05380
|
Yoana Pita Lorenzo Yoana
|
Yoana Pita Lorenzo
|
Cumplimiento del Reglamento (UE) 2024/1689 en rob\'otica y sistemas aut\'onomos: una revisi\'on sistem\'atica de la literatura
|
in Spanish language
| null | null | null |
cs.CY cs.AI cs.CR cs.RO
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
This systematic literature review analyzes the current state of compliance with Regulation (EU) 2024/1689 in autonomous robotic systems, focusing on cybersecurity frameworks and methodologies. Using the PRISMA protocol, 22 studies were selected from 243 initial records across IEEE Xplore, ACM DL, Scopus, and Web of Science. Findings reveal partial regulatory alignment: while progress has been made in risk management and encrypted communications, significant gaps persist in explainability modules, real-time human oversight, and knowledge base traceability. Only 40% of reviewed solutions explicitly address transparency requirements, and 30% implement failure intervention mechanisms. The study concludes that modular approaches integrating risk, supervision, and continuous auditing are essential to meet the AI Act mandates in autonomous robotics.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 20:56:33 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lorenzo",
"Yoana Pita",
""
]
] |
2509.05382
|
Jennifer King
|
Jennifer King, Kevin Klyman, Emily Capstick, Tiffany Saade, Victoria Hsieh
|
User Privacy and Large Language Models: An Analysis of Frontier Developers' Privacy Policies
|
See additional files for appendices
| null | null | null |
cs.CY cs.AI cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Hundreds of millions of people now regularly interact with large language models via chatbots. Model developers are eager to acquire new sources of high-quality training data as they race to improve model capabilities and win market share. This paper analyzes the privacy policies of six U.S. frontier AI developers to understand how they use their users' chats to train models. Drawing primarily on the California Consumer Privacy Act, we develop a novel qualitative coding schema that we apply to each developer's relevant privacy policies to compare data collection and use practices across the six companies. We find that all six developers appear to employ their users' chat data to train and improve their models by default, and that some retain this data indefinitely. Developers may collect and train on personal information disclosed in chats, including sensitive information such as biometric and health data, as well as files uploaded by users. Four of the six companies we examined appear to include children's chat data for model training, as well as customer data from other products. On the whole, developers' privacy policies often lack essential information about their practices, highlighting the need for greater transparency and accountability. We address the implications of users' lack of consent for the use of their chat data for model training, data security issues arising from indefinite chat data retention, and training on children's chat data. We conclude by providing recommendations to policymakers and developers to address the data privacy challenges posed by LLM-powered chatbots.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 01:01:21 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"King",
"Jennifer",
""
],
[
"Klyman",
"Kevin",
""
],
[
"Capstick",
"Emily",
""
],
[
"Saade",
"Tiffany",
""
],
[
"Hsieh",
"Victoria",
""
]
] |
2509.05383
|
Michael Dallaston
|
Michael C Dallaston, Steven A Kedda, Scott W McCue
|
Self-similar rupture of thin films of power-law fluid
|
22 pages, 6 figures
| null | null | null |
cond-mat.soft physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
Models that describe Newtonian liquid films evolving due to the competing effects of surface tension and attractive intermolecular or van der Waals forces are known to rupture in finite time in a self-similar manner. We extend the computation of similarity solutions to non-Newtonian power-law liquid films. The bifurcation diagram, indexed by power law exponent $n$, has a highly nontrivial structure with branches merging via a snaking bifurcation around $n=1$. A countably infinite number of solutions are also found in the extreme shear-thinning ($n\to 0$) limit, in which similarity solutions possess an exponentially small inner region. Numerical simulations are shown to be attracted to the single primary branch of similarity solutions.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 01:41:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Dallaston",
"Michael C",
""
],
[
"Kedda",
"Steven A",
""
],
[
"McCue",
"Scott W",
""
]
] |
2509.05384
|
Shoichi Sato
|
Shoichi Sato, Masaaki Tanaka, and Ryosho Nakane
|
Spin-transport characteristics in a Si-based spin metal-oxide-semiconductor field-effect transistor (spin MOSFET): Bias dependence of the spin polarization in Si and magnetoresistance in spin-valve signals
| null | null | null | null |
cond-mat.mes-hall
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We have studied the spin transport characteristics of a spin metal-oxide-semiconductor field-effect transistor (spin MOSFET), particularly the bias voltage dependence of the electron spin polarization P_S in Si and the magnetoresistance ratio MR in spin-valve signals, to optimize the device performance. The spin MOSFET device has an 8-nm-thick p-Si channel with a back gate (G) and ferromagnetic source / drain (S/D) junctions consisting of Fe/Mg/MgO/SiOx/n+-Si. In addition to transistor characteristics with an on-off ratio of 104, clear spin-valve signals and Hanle spin precession signals were observed at 4 K in a wide range of the source-to-gate V_GS and source-to-drain V_DS bias voltages. We achieved a high P_S of 50% and a high MR of 0.35% as the maximum values in their single-peaked curves plotted as a function of the junction voltage V_J, mainly because the ferromagnetic S/D junction can generate high P_S and the spin diffusion length is very long in the Si inversion channel. These P_S and MR values are the highest ever reported in spin-MOSFETs. Our spin transport model for our spin MOSFET structure was improved in this study by taking into account the electron distribution and band profile of the n+-Si regions in the ferromagnetic S/D junctions, which enables the accurate estimation of P_S. Detailed analyses with various V_GS and V_J clarified that P_S is determined only by V_J. Our analyses also revealed that the main parameters for determining MR, such as P_S and the resistance-area products of the S/D ferromagnetic junctions, have different V_J dependences, leading to the finding that the present device does not exploit the full potential of the ferromagnetic S/D junctions to maximize MR. Based on the results, we discuss the device physics and engineering for further enhancement of MR, with a focus on the electrical and spin-related properties of the ferromagnetic S/D junctions.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 02:25:28 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sato",
"Shoichi",
""
],
[
"Tanaka",
"Masaaki",
""
],
[
"Nakane",
"Ryosho",
""
]
] |
2509.05385
|
Jiacheng Wei
|
Jiacheng Wei, Faguo Wu, Xiao Zhang
|
A Lightweight Framework for Trigger-Guided LoRA-Based Self-Adaptation in LLMs
|
11 pages, 7 figures, conference
| null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language models are unable to continuously adapt and learn from new data during reasoning at inference time. To address this limitation, we propose that complex reasoning tasks be decomposed into atomic subtasks and introduce SAGE, a trigger-guided dynamic fine-tuning framework that enables adaptive updates during reasoning at inference time. SAGE consists of three key components: (1) a Trigger module that detects reasoning failures through multiple evaluation metrics in real time; (2) a Trigger Buffer module that clusters anomaly samples using a streaming clustering process with HDBSCAN, followed by stability checks and similarity-based merging; and (3) a Lora Store module that dynamically optimizes parameter updates with an adapter pool for knowledge retention. Evaluation results show that SAGE demonstrates excellent accuracy, robustness, and stability on the atomic reasoning subtask through dynamic knowledge updating during test time.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 03:23:25 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wei",
"Jiacheng",
""
],
[
"Wu",
"Faguo",
""
],
[
"Zhang",
"Xiao",
""
]
] |
2509.05386
|
Vladimir Pokrovskii
|
Vladimir Pokrovskii
|
A note on the mechanism of substitution of labour with capital in the production processes
|
8 pages, no figures
| null | null | null |
physics.soc-ph econ.GN q-fin.EC
|
http://creativecommons.org/licenses/by/4.0/
|
Considering the production processes, it was noted that the use of various equipment leads to an increase in output -- the phenomenon that is usually described as the substitution of labor with capital. The proposed theory of substitution is based on the assumption that not the quantity of capital (production equipment) does substitute labor, but rather its ability to operate similar to the workers. This is the true content of the substitution of labor by capital. To formulate a correct mechanism of substitution requires considering three factors of production: the amount of production equipment (capital $K$), human activity (labor $L$), and the substitutive capacity of equipment (substitutive work $P$). The technological properties of production equipment are characterized by the technological coefficients $\overline \lambda$ and $\overline \varepsilon$, indicating the amount of labor and energy required to engaged with a unit of equipment. The production function can assume various forms, none of which coincide with the popular Cobb-Douglas expression, which seems to be erroneous in its core.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 06:02:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Pokrovskii",
"Vladimir",
""
]
] |
2509.05387
|
Evgeny Asmolov S
|
Evgeny S. Asmolov, Elena F. Silkina, Olga I. Vinogradova
|
Diffusioosmosis of electrolyte solutions in uniformly charged channels
|
15 pages, 12 figures
| null | null | null |
cond-mat.soft
|
http://creativecommons.org/licenses/by/4.0/
|
When the concentration of electrolyte solution varies along the channel the forces arise that drag the fluid toward the higher or lower concentration region inducing a flow termed diffusio-osmotic. This article investigates a flow that emerges in channels with constant density of surface charge {\sigma} and thin compared to their thickness electrostatic diffuse layers. An equation for the fluid flow rate Q is derived and used to describe analytically the flux of ions, and local potentials and concentrations. This equation, which allows to treat the diffusio-osmotic problems without tedious and time consuming computations, clarifies that the global flow rate is controlled only by the surface charge and concentration drop between the channel ends, and indicates that there always exist two different values of {\sigma} that correspond to a particular Q. Our theory provides a simple explanation of the directions of the fluid flow rate and ionic flux depending on the surface charge and diffusivity of ions, predicts a non-linear concentration distribution along the channel caused by convection, and relates it to the local potential changes by a compact formula. We also present and interpret the variations of the diffusio-osmotic velocity profiles and the apparent slip velocity along the channel and show that the latter is highly non-uniform and could even becomes alternating. The relevance of our results for diffusio-osmotic experiments and for some electrochemistry and membrane science issues is discussed briefly.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 06:38:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Asmolov",
"Evgeny S.",
""
],
[
"Silkina",
"Elena F.",
""
],
[
"Vinogradova",
"Olga I.",
""
]
] |
2509.05388
|
Alberto Badias
|
Juan Olalla-Pombo, Alberto Bad\'ias, Miguel \'Angel Sanz-G\'omez, Jos\'e Mar\'ia Ben\'itez, Francisco Javier Mont\'ans
|
Augmented Structure Preserving Neural Networks for cell biomechanics
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Cell biomechanics involve a great number of complex phenomena that are fundamental to the evolution of life itself and other associated processes, ranging from the very early stages of embryo-genesis to the maintenance of damaged structures or the growth of tumors. Given the importance of such phenomena, increasing research has been dedicated to their understanding, but the many interactions between them and their influence on the decisions of cells as a collective network or cluster remain unclear. We present a new approach that combines Structure Preserving Neural Networks, which study cell movements as a purely mechanical system, with other Machine Learning tools (Artificial Neural Networks), which allow taking into consideration environmental factors that can be directly deduced from an experiment with Computer Vision techniques. This new model, tested on simulated and real cell migration cases, predicts complete cell trajectories following a roll-out policy with a high level of accuracy. This work also includes a mitosis event prediction model based on Neural Networks architectures which makes use of the same observed features.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 07:32:50 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Olalla-Pombo",
"Juan",
""
],
[
"Badías",
"Alberto",
""
],
[
"Sanz-Gómez",
"Miguel Ángel",
""
],
[
"Benítez",
"José María",
""
],
[
"Montáns",
"Francisco Javier",
""
]
] |
2509.05389
|
Dina Razafindralandy
|
Oscar Cosserat and Dina Razafindralandy and Can Sel\c{c}uk
|
Vorticity-dependent and symmetry-preserving LES models
| null | null | null | null |
math-ph math.MP physics.flu-dyn
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
We propose Lie symmetry-preserving turbulence models for the incompressible Navier-Stokes equations, within the Large Eddy Simulation framework. These models depend on both the filtered strain-rate tensor and the filtered vorticity tensor. Particular emphasis is placed on models that additionally ensure stability.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 08:14:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Cosserat",
"Oscar",
""
],
[
"Razafindralandy",
"Dina",
""
],
[
"Selçuk",
"Can",
""
]
] |
2509.05390
|
Sebastian Porsdam Mann
|
Clint Hurshman, Sebastian Porsdam Mann, Julian Savulescu and Brian D. Earp
|
Authorship Without Writing: Large Language Models and the Senior Author Analogy
|
28 pages, 0 figures
| null | null | null |
cs.CY cs.AI cs.CL
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The use of large language models (LLMs) in bioethical, scientific, and medical writing remains controversial. While there is broad agreement in some circles that LLMs cannot count as authors, there is no consensus about whether and how humans using LLMs can count as authors. In many fields, authorship is distributed among large teams of researchers, some of whom, including paradigmatic senior authors who guide and determine the scope of a project and ultimately vouch for its integrity, may not write a single word. In this paper, we argue that LLM use (under specific conditions) is analogous to a form of senior authorship. On this view, the use of LLMs, even to generate complete drafts of research papers, can be considered a legitimate form of authorship according to the accepted criteria in many fields. We conclude that either such use should be recognized as legitimate, or current criteria for authorship require fundamental revision. AI use declaration: GPT-5 was used to help format Box 1. AI was not used for any other part of the preparation or writing of this manuscript.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 10:11:56 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hurshman",
"Clint",
""
],
[
"Mann",
"Sebastian Porsdam",
""
],
[
"Savulescu",
"Julian",
""
],
[
"Earp",
"Brian D.",
""
]
] |
2509.05391
|
Christian Masuhr
|
Christian Masuhr, Julian Koch, Thorsten Sch\"uppstuhl
|
Evaluating Magic Leap 2 Tool Tracking for AR Sensor Guidance in Industrial Inspections
| null | null | null | null |
cs.RO cs.HC cs.MM
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Rigorous evaluation of commercial Augmented Reality (AR) hardware is crucial, yet public benchmarks for tool tracking on modern Head-Mounted Displays (HMDs) are limited. This paper addresses this gap by systematically assessing the Magic Leap 2 (ML2) controllers tracking performance. Using a robotic arm for repeatable motion (EN ISO 9283) and an optical tracking system as ground truth, our protocol evaluates static and dynamic performance under various conditions, including realistic paths from a hydrogen leak inspection use case. The results provide a quantitative baseline of the ML2 controller's accuracy and repeatability and present a robust, transferable evaluation methodology. The findings provide a basis to assess the controllers suitability for the inspection use case and similar industrial sensor-based AR guidance tasks.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 10:13:28 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Masuhr",
"Christian",
""
],
[
"Koch",
"Julian",
""
],
[
"Schüppstuhl",
"Thorsten",
""
]
] |
2509.05392
|
Mohamed Chatti
|
Qurat Ul Ain and Mohamed Amine Chatti and Jean Qussa and Amr Shakhshir and Rawaa Alatrash and Shoeb Joarder
|
An Optimized Pipeline for Automatic Educational Knowledge Graph Construction
|
Accepted at IJCKG 2025
| null | null | null |
cs.CY cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
The automatic construction of Educational Knowledge Graphs (EduKGs) is essential for domain knowledge modeling by extracting meaningful representations from learning materials. Despite growing interest, identifying a scalable and reliable approach for automatic EduKG generation remains a challenge. In an attempt to develop a unified and robust pipeline for automatic EduKG construction, in this study we propose a pipeline for automatic EduKG construction from PDF learning materials. The process begins with generating slide-level EduKGs from individual pages/slides, which are then merged to form a comprehensive EduKG representing the entire learning material. We evaluate the accuracy of the EduKG generated from the proposed pipeline in our MOOC platform, CourseMapper. The observed accuracy, while indicative of partial success, is relatively low particularly in the educational context, where the reliability of knowledge representations is critical for supporting meaningful learning. To address this, we introduce targeted optimizations across multiple pipeline components. The optimized pipeline achieves a 17.5% improvement in accuracy and a tenfold increase in processing efficiency. Our approach offers a holistic, scalable and end-to-end pipeline for automatic EduKG construction, adaptable to diverse educational contexts, and supports improved semantic representation of learning content.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 10:29:46 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ain",
"Qurat Ul",
""
],
[
"Chatti",
"Mohamed Amine",
""
],
[
"Qussa",
"Jean",
""
],
[
"Shakhshir",
"Amr",
""
],
[
"Alatrash",
"Rawaa",
""
],
[
"Joarder",
"Shoeb",
""
]
] |
2509.05393
|
Mohamed Chatti
|
Rawaa Alatrash and Mohamed Amine Chatti and Nasha Wibowo and Qurat Ul Ain
|
Inferring Prerequisite Knowledge Concepts in Educational Knowledge Graphs: A Multi-criteria Approach
|
Accepted at IJCKG 2025
| null | null | null |
cs.CY cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Educational Knowledge Graphs (EduKGs) organize various learning entities and their relationships to support structured and adaptive learning. Prerequisite relationships (PRs) are critical in EduKGs for defining the logical order in which concepts should be learned. However, the current EduKG in the MOOC platform CourseMapper lacks explicit PR links, and manually annotating them is time-consuming and inconsistent. To address this, we propose an unsupervised method for automatically inferring concept PRs without relying on labeled data. We define ten criteria based on document-based, Wikipedia hyperlink-based, graph-based, and text-based features, and combine them using a voting algorithm to robustly capture PRs in educational content. Experiments on benchmark datasets show that our approach achieves higher precision than existing methods while maintaining scalability and adaptability, thus providing reliable support for sequence-aware learning in CourseMapper.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 10:37:58 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Alatrash",
"Rawaa",
""
],
[
"Chatti",
"Mohamed Amine",
""
],
[
"Wibowo",
"Nasha",
""
],
[
"Ain",
"Qurat Ul",
""
]
] |
2509.05394
|
Zoltan Toth-Czifra
|
Zoltan Toth-Czifra
|
Reverse Browser: Vector-Image-to-Code Generator
|
Submitted to AIWare 2025 ArXiv Track
| null | null | null |
cs.SE cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Automating the conversion of user interface design into code (image-to-code or image-to-UI) is an active area of software engineering research. However, the state-of-the-art solutions do not achieve high fidelity to the original design, as evidenced by benchmarks. In this work, I approach the problem differently: I use vector images instead of bitmaps as model input. I create several large datasets for training machine learning models. I evaluate the available array of Image Quality Assessment (IQA) algorithms and introduce a new, multi-scale metric. I then train a large open-weights model and discuss its limitations.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 11:13:40 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Toth-Czifra",
"Zoltan",
""
]
] |
2509.05395
|
M AbuGhanem
|
M. AbuGhanem
|
Practical Fidelity Limits of Toffoli Gates in Superconducting Quantum Processors
| null | null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
High-fidelity multi-qubit gates are a critical resource for near-term quantum computing, as they underpin the execution of both quantum algorithms and fault-tolerant protocols. The Toffoli gate (CCNOT), in particular, plays a central role in quantum error correction and quantum arithmetic, yet its efficient implementation on current quantum hardware remains limited by noise and connectivity constraints. In this work, we present a hardware-aware characterization of the Toffoli gate using optimized, connectivity-compliant decompositions executed on IBM's 127-qubit superconducting quantum processors. Our study integrates state preparation, gate synthesis, and quantum state/process tomography (QST/QPT) to evaluate fidelity across three distinct classes of input states: Greenberger-Horne-Zeilinger (GHZ), W, and the uniform superposition of all three-qubit computational basis states -- under noise-free simulation, noise-aware emulation, and real hardware execution. For GHZ states, we report state fidelities of 98.442% (noise-free simulation), 81.470% (noise-aware quantum emulation), and 56.368% (real quantum hardware). For W states, state fidelities are 98.739%, 79.900%, and 63.689%, respectively, and for the uniform superposition state, we observe state fidelities of 99.490%, 85.469%, and 61.161%. Comparative QPT experiments yield process fidelities of 98.976% (noise-free) and 80.160% (noise-aware emulation). Our results empirically characterize state-dependent error patterns in multi-qubit circuits and quantify trade-offs between gate decomposition strategies and native hardware performance, offering practical insights for scalable, hardware-efficient quantum circuit design.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 12:58:13 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"AbuGhanem",
"M.",
""
]
] |
2509.05396
|
Andrea Wynn
|
Andrea Wynn and Harsh Satija and Gillian Hadfield
|
Talk Isn't Always Cheap: Understanding Failure Modes in Multi-Agent Debate
|
ICML MAS Workshop 2025
| null | null | null |
cs.CL cs.AI cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
While multi-agent debate has been proposed as a promising strategy for improving AI reasoning ability, we find that debate can sometimes be harmful rather than helpful. The prior work has exclusively focused on debates within homogeneous groups of agents, whereas we explore how diversity in model capabilities influences the dynamics and outcomes of multi-agent interactions. Through a series of experiments, we demonstrate that debate can lead to a decrease in accuracy over time -- even in settings where stronger (i.e., more capable) models outnumber their weaker counterparts. Our analysis reveals that models frequently shift from correct to incorrect answers in response to peer reasoning, favoring agreement over challenging flawed reasoning. These results highlight important failure modes in the exchange of reasons during multi-agent debate, suggesting that naive applications of debate may cause performance degradation when agents are neither incentivized nor adequately equipped to resist persuasive but incorrect reasoning.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 13:47:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wynn",
"Andrea",
""
],
[
"Satija",
"Harsh",
""
],
[
"Hadfield",
"Gillian",
""
]
] |
2509.05397
|
Matthew Lai
|
Matthew Lai, Keegan Go, Zhibin Li, Torsten Kroger, Stefan Schaal, Kelsey Allen, Jonathan Scholz
|
RoboBallet: Planning for Multi-Robot Reaching with Graph Neural Networks and Reinforcement Learning
|
Published in Science Robotics
|
RoboBallet: Planning for multirobot reaching with graph neural networks and reinforcement learning. Sci. Robot. 10, eads1204(2025)
|
10.1126/scirobotics.ads1204
| null |
cs.RO cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Modern robotic manufacturing requires collision-free coordination of multiple robots to complete numerous tasks in shared, obstacle-rich workspaces. Although individual tasks may be simple in isolation, automated joint task allocation, scheduling, and motion planning under spatio-temporal constraints remain computationally intractable for classical methods at real-world scales. Existing multi-arm systems deployed in the industry rely on human intuition and experience to design feasible trajectories manually in a labor-intensive process. To address this challenge, we propose a reinforcement learning (RL) framework to achieve automated task and motion planning, tested in an obstacle-rich environment with eight robots performing 40 reaching tasks in a shared workspace, where any robot can perform any task in any order. Our approach builds on a graph neural network (GNN) policy trained via RL on procedurally-generated environments with diverse obstacle layouts, robot configurations, and task distributions. It employs a graph representation of scenes and a graph policy neural network trained through reinforcement learning to generate trajectories of multiple robots, jointly solving the sub-problems of task allocation, scheduling, and motion planning. Trained on large randomly generated task sets in simulation, our policy generalizes zero-shot to unseen settings with varying robot placements, obstacle geometries, and task poses. We further demonstrate that the high-speed capability of our solution enables its use in workcell layout optimization, improving solution times. The speed and scalability of our planner also open the door to new capabilities such as fault-tolerant planning and online perception-based re-planning, where rapid adaptation to dynamic task sets is required.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 14:28:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lai",
"Matthew",
""
],
[
"Go",
"Keegan",
""
],
[
"Li",
"Zhibin",
""
],
[
"Kroger",
"Torsten",
""
],
[
"Schaal",
"Stefan",
""
],
[
"Allen",
"Kelsey",
""
],
[
"Scholz",
"Jonathan",
""
]
] |
2509.05398
|
Sheila Wafula Ms
|
Sheila Wafula, Blessed Madukoma
|
Unmasking COVID-19 Vulnerability in Nigeria: Mapping Risks Beyond Urban Hotspots
|
8 pages, 6 figures. Submission to NeurIPS 2025 in preparation
| null | null | null |
cs.CY cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The COVID-19 pandemic has presented significant challenges in Nigeria's public health systems since the first case reported on February 27, 2020. This study investigates key factors that contribute to state vulnerability, quantifying them through a composite risk score integrating population density (weight 0.2), poverty (0.4), access to healthcare (0.3), and age risk (0.1), adjusted by normalized case rates per 100,000. States were categorized into low-, medium-, and high-density areas to analyze trends and identify hotspots using geographic information system (GIS) mapping. The findings reveal that high-density urban areas, such as Lagos, accounting for 35.4% of national cases, had the highest risk scores (Lagos: 673.47 vs. national average: 28.16). These results align with global and local studies on the spatial variability of COVID-19 in Nigeria, including international frameworks such as the CDC Social Vulnerability Index. Google Trends data highlight variations in public health awareness, serving as a supplementary analysis to contextualize vulnerability. The risk score provides a prioritization tool for policymakers to allocate testing, vaccines, and healthcare resources to high-risk areas, though data gaps and rural underreporting call for further research. This framework can extend to other infectious diseases, offering lessons for future pandemics in resource-limited settings.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 14:45:01 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wafula",
"Sheila",
""
],
[
"Madukoma",
"Blessed",
""
]
] |
2509.05399
|
Henry Graf\'e
|
Henry Graf\'e, Hugo Van hamme
|
Graph Connectionist Temporal Classification for Phoneme Recognition
|
Accepted to the IEEE Automatic Speech Recognition and Understanding Workshop (ASRU 2025)
| null | null | null |
eess.AS cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Automatic Phoneme Recognition (APR) systems are often trained using pseudo phoneme-level annotations generated from text through Grapheme-to-Phoneme (G2P) systems. These G2P systems frequently output multiple possible pronunciations per word, but the standard Connectionist Temporal Classification (CTC) loss cannot account for such ambiguity during training. In this work, we adapt Graph Temporal Classification (GTC) to the APR setting. GTC enables training from a graph of alternative phoneme sequences, allowing the model to consider multiple pronunciations per word as valid supervision. Our experiments on English and Dutch data sets show that incorporating multiple pronunciations per word into the training loss consistently improves phoneme error rates compared to a baseline trained with CTC. These results suggest that integrating pronunciation variation into the loss function is a promising strategy for training APR systems from noisy G2P-based supervision.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 15:20:59 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Grafé",
"Henry",
""
],
[
"Van hamme",
"Hugo",
""
]
] |
2509.05400
|
M. R. S. Hawkins
|
M.R.S. Hawkins and J. Garc\'ia-Bellido
|
A critical analysis of the recent OGLE limits on stellar mass primordial black holes in the halo of the Milky Way
|
8 pages, 4 figures
| null | null |
IFT-UAM/CSIC-25-91
|
astro-ph.GA astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
This paper is a response to recent claims that a population of primordial black holes in the Galactic halo has been ruled out by the OGLE collaboration. This claim was based on the latest results from the OGLE microlensing survey towards the Large Magellanic Cloud which failed to detect even the number of events expected from known stellar populations. In particular, their results are completely inconsistent with the results of the MACHO survey which detected a population of compact bodies in the Galactic halo which could not be accounted for by any known stellar population. The discrepancy between the results of these two groups has a long history, and includes problems such as different choice of photometric passbands, quality of light curves, microlensing event selection, detection efficiency, self lensing and halo models. In this paper it is demonstrated that these issues not only account for the discrepancy between the OGLE and MACHO results, but imply that the OGLE observations can put no meaningful constraints on a population of primordial black holes in the Galactic halo.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 15:37:25 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hawkins",
"M. R. S.",
""
],
[
"García-Bellido",
"J.",
""
]
] |
2509.05401
|
Mark Linton
|
Mark Linton, Mausumi Dikpati, Rachel Howe
|
Solar Interior
| null |
Space Physics and Aeronomy Collection Volume 1: Solar Physics and Solar Wind, Geophysical Monograph 258 (Wiley and Sons Inc, 2021)
|
10.1002/9781119815600.ch7
| null |
astro-ph.SR
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
This chapter introduces the reader to the solar interior, in particular the convection zone. The first section explores the solar cycle and the dynamo models that have been studied to explain this cycle. The second section explores helioseismology observations and analysis of the solar interior, and reviews the fundamental knowledge that has been gained from these studies. The third and final section reviews observations and theory of magnetic fields emerging from the convection zone into the solar corona.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 16:48:19 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Linton",
"Mark",
""
],
[
"Dikpati",
"Mausumi",
""
],
[
"Howe",
"Rachel",
""
]
] |
2509.05402
|
Bernd Kellner
|
Bernd C. Kellner
|
Wilson's theorem modulo higher prime powers II: Bernoulli numbers and polynomials
|
16 pages, 1 table, 1 figure
| null | null | null |
math.NT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
By recent work of the author, Wilson's theorem as well as the Wilson quotient can be described by supercongruences of power sums of Fermat quotients modulo every higher prime power. We translate these congruences into congruences of power sums and Bernoulli numbers. This together provides relatively short proofs of the congruences compared to former approaches. As an application, we compute, e.g., the Wilson quotient up to $\pmod{p^4}$ and equivalently the factorial $(p-1)!$ up to $\pmod{p^5}$, which can be extended to any higher prime power with some effort. As a by-product, we determine some power sums of the Fermat quotients up to $\pmod{p^4}$.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 17:52:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kellner",
"Bernd C.",
""
]
] |
2509.05403
|
Ivan Kramarenko
|
I. G. Kramarenko, J. Rosdahl, J. Blaizot, J. Matthee, H. Katz, C. Di Cesare
|
Hydrogen-Alpha as a Tracer of Star Formation in the SPHINX Cosmological Simulations
|
Submitted to A&A
| null | null | null |
astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
The Hydrogen-alpha (Ha) emission line in galaxies is a powerful tracer of their recent star formation activity. With the advent of JWST, we are now able to routinely observe Ha in galaxies at high redshifts (z > 3) and thus measure their star-formation rates (SFRs). However, using "classical" SFR(Ha) calibrations to derive the SFRs leads to biased results because high-redshift galaxies are commonly characterized by low metallicities and bursty star-formation histories, affecting the conversion factor between the Ha luminosity and the SFR. In this work, we develop a set of new SFR(Ha) calibrations that allow us to predict the SFRs of Ha-emitters at z > 3 with minimal error. We use the SPHINX cosmological simulations to select a sample of star-forming galaxies representative of the Ha-emitter population observed with JWST. We then derive linear corrections to the classical SFR(Ha) calibrations, taking into account variations in the physical properties (e.g., stellar metallicities) among individual galaxies. We obtain two new SFR(Ha) calibrations that, compared to the classical calibrations, reduce the root mean squared error (RMSE) in the predicted SFRs by $\Delta$RMSE $\approx$ 0.04 dex and $\Delta$RMSE $\approx$ 0.06 dex, respectively. Using the recent JWST NIRCam/grism observations of Ha-emitters at z ~ 6, we show that the new calibrations affect the high-redshift galaxy population statistics: (i) the estimated cosmic star-formation density decreases by $\Delta\rho$(SFR) $\approx$ 12%, and (ii) the observed slope of the star-formation main sequence increases by $\Delta$ $\partial$log SFR / $\partial$log M* = 0.08 $\pm$ 0.02.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kramarenko",
"I. G.",
""
],
[
"Rosdahl",
"J.",
""
],
[
"Blaizot",
"J.",
""
],
[
"Matthee",
"J.",
""
],
[
"Katz",
"H.",
""
],
[
"Di Cesare",
"C.",
""
]
] |
2509.05404
|
Thierry Nicolas Kaldenbach
|
Thierry N. Kaldenbach, Isaac D. Smith, Hendrik Poulsen Nautrup, Matthias Heller, Hans J. Briegel
|
Efficient Preparation of Resource States for Hamiltonian Simulation and Universal Quantum Computation
|
36 pages, 22 figures
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
The direct compilation of algorithm-specific graph states in measurement-based quantum computation (MBQC) can lead to resource reductions in terms of circuit depth, entangling gates, and even the number of physical qubits. In this work, we extend previous studies on algorithm-tailored graph states to periodic sequences of generalized Pauli rotations, which commonly appear in, e.g., Trotterized Hamiltonian simulation. We first implement an enhanced simulated-annealing-based algorithm to find optimal periodic graph states within local-Clifford (LC-)MBQC. In addition, we derive a novel scheme for the preparation of resource states based on a graph state and a ladder of CNOT gates, which we term anticommutation-based (AC-)MBQC, since it uncovers a direct relationship between the graph state and the anticommutation matrix for the set of Hamilonians generating the computation. We also deploy our two approaches to derive universal resource states from minimal universal sets of generating Hamiltonians, thus providing a straightforward algorithm for finding the former. Finally, we demonstrate and compare both of our methods based on various examples from condensed matter physics and universal quantum computation.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kaldenbach",
"Thierry N.",
""
],
[
"Smith",
"Isaac D.",
""
],
[
"Nautrup",
"Hendrik Poulsen",
""
],
[
"Heller",
"Matthias",
""
],
[
"Briegel",
"Hans J.",
""
]
] |
2509.05405
|
Noah Franz
|
Noah Franz, Kate D Alexander, Sebastian Gomez, Collin T Christy, Tanmoy Laskar, Sjoert van Velzen, Nicholas Earl, Suvi Gezari, Mitchell Karmen, Raffaella Margutti, Jeniveve Pearson, V. Ashley Villar, and Ann I Zabludoff
|
The Open mulTiwavelength Transient Event Repository (OTTER): Infrastructure Release and Tidal Disruption Event Catalog
|
Submitted to ApJ. The OTTER web interface is available at https://otter.idies.jhu.edu and the API documentation (including example python notebooks demonstrating usage) is available at https://astro-otter.readthedocs.io. Comments are welcome! Please submit any comments and feedback on GitHub at https://github.com/astro-otter/otter/issues/new/choose
| null | null | null |
astro-ph.HE astro-ph.IM
|
http://creativecommons.org/licenses/by/4.0/
|
Multiwavelength analyses of astrophysical transients are essential for understanding the physics of these events. To make such analyses more efficient and effective, we present the Open mulTiwavelength Transient Event Repository (OTTER), a publicly available catalog of published transient event metadata and photometry. Unlike previous efforts, our data schema is optimized for the storage of multiwavelength photometric datasets spanning the entire electromagnetic spectrum. Open source software, including an application programming interface (API) and web application, are available for viewing, accessing, and analyzing the dataset. For the initial release of OTTER, we present the largest ever photometric archive of tidal disruption events (TDEs), including $\gtrsim 80,000$ observations of 232 TDEs spanning from radio to X-ray wavelengths. We demonstrate the power of this infrastructure through four example analyses of the TDE population. We plan to maintain this dataset as more TDEs are discovered in the future and encourage other users to contribute by uploading newly published data via our web application. The infrastructure was built with the goal of archiving additional transient data (supernovae, gamma-ray bursts, fast blue optical transients, fast radio bursts, etc.) in the future. The web application is available at https://otter.idies.jhu.edu and the API documentation is available at https://astro-otter.readthedocs.io.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Franz",
"Noah",
""
],
[
"Alexander",
"Kate D",
""
],
[
"Gomez",
"Sebastian",
""
],
[
"Christy",
"Collin T",
""
],
[
"Laskar",
"Tanmoy",
""
],
[
"van Velzen",
"Sjoert",
""
],
[
"Earl",
"Nicholas",
""
],
[
"Gezari",
"Suvi",
""
],
[
"Karmen",
"Mitchell",
""
],
[
"Margutti",
"Raffaella",
""
],
[
"Pearson",
"Jeniveve",
""
],
[
"Villar",
"V. Ashley",
""
],
[
"Zabludoff",
"Ann I",
""
]
] |
2509.05406
|
Zexuan Wu
|
Zexuan Wu, Subo Dong, Andrew P. Gould, Przemek Mr\'oz, A. M\'erand
|
"Industrial-Scale" Black Hole Selection without a Satellite
| null | null | null | null |
astro-ph.IM astro-ph.HE astro-ph.SR
|
http://creativecommons.org/licenses/by/4.0/
|
The forthcoming GRAVITY+ instrument promises to usher in an era of "industrial-scale" mass measurements of isolated black holes (BHs), with the potential to assemble a sample of many tens of BHs via interferometric microlensing over several years. A key challenge will be selecting interferometric follow-up targets from an order-of-magnitude larger pool of ongoing microlensing events when using traditional selection based on event timescale alone. Gould (2023) proposed a criterion optimized for BH selection using space-based microlens parallax measurements enabled by a satellite. We adapt it to work with microlens parallax constraints obtainable from ground-based data only. Using Galactic simulations, we show that our selection criterion is highly efficient, expecting to detect about a dozen BHs per year with GRAVITY+ from following up ~35 selected events.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:01 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wu",
"Zexuan",
""
],
[
"Dong",
"Subo",
""
],
[
"Gould",
"Andrew P.",
""
],
[
"Mróz",
"Przemek",
""
],
[
"Mérand",
"A.",
""
]
] |
2509.05407
|
Ana Glidden
|
Ana Glidden, Sukrit Ranjan, Sara Seager, N\'estor Espinoza, Ryan J. MacDonald, Natalie H. Allen, Caleb I. Ca\~nas, David Grant, Am\'elie Gressier, Kevin B. Stevenson, Natasha E. Batalha, Nikole K. Lewis, Douglas Long, Hannah R. Wakeford, Lili Alderson, Ryan C. Challener, Knicole Col\'on, Jingcheng Huang, Zifan Lin, Dana R. Louie, Elijah Mullens, Kristin S. Sotzen, Jeff A. Valenti, Daniel Valentine, Mark Clampin, C. Matt Mountain, Marshall Perrin, and Roeland P. van der Marel
|
JWST-TST DREAMS: Secondary Atmosphere Constraints for the Habitable Zone Planet TRAPPIST-1 e
|
29 pages, 8 figures, accepted for publication in ApJL
| null |
10.3847/2041-8213/adf62e
| null |
astro-ph.EP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The TRAPPIST-1 system offers one of the best opportunities to characterize temperate terrestrial planets beyond our own solar system. Within the TRAPPIST-1 system, planet e stands out as highly likely to sustain surface liquid water if it possesses an atmosphere. Recently, we reported the first JWST/NIRSpec PRISM transmission spectra of TRAPPIST-1 e, revealing significant stellar contamination, which varied between the four visits. Here, we assess the range of planetary atmospheres consistent with our transmission spectrum. We explore a wide range of atmospheric scenarios via a hierarchy of forward modeling and retrievals. We do not obtain strong evidence for or against an atmosphere. Our results weakly disfavor CO$_2$-rich atmospheres for pressures corresponding to the surface of Venus and Mars and the cloud tops of Venus at 2$\sigma$. We exclude H$_2$-rich atmospheres containing CO$_2$ and CH$_4$ in agreement with past work, but find that higher mean molecular weight, N$_2$-rich atmospheres with trace CO$_2$ and CH$_4$ are permitted by the data. Both a bare rock and N$_2$-rich atmospheric scenario provide adequate fits to the data, but do not fully explain all features, which may be due to either uncorrected stellar contamination or atmospheric signals. Ongoing JWST observations of TRAPPIST-1 e, exploiting consecutive transits with TRAPPIST-1 b, will offer stronger constraints via a more effective stellar contamination correction. The present work is part of the JWST Telescope Scientist Team (JWST-TST) Guaranteed Time Observations, which is performing a Deep Reconnaissance of Exoplanet Atmospheres through Multi-instrument Spectroscopy (DREAMS).
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:01 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Glidden",
"Ana",
""
],
[
"Ranjan",
"Sukrit",
""
],
[
"Seager",
"Sara",
""
],
[
"Espinoza",
"Néstor",
""
],
[
"MacDonald",
"Ryan J.",
""
],
[
"Allen",
"Natalie H.",
""
],
[
"Cañas",
"Caleb I.",
""
],
[
"Grant",
"David",
""
],
[
"Gressier",
"Amélie",
""
],
[
"Stevenson",
"Kevin B.",
""
],
[
"Batalha",
"Natasha E.",
""
],
[
"Lewis",
"Nikole K.",
""
],
[
"Long",
"Douglas",
""
],
[
"Wakeford",
"Hannah R.",
""
],
[
"Alderson",
"Lili",
""
],
[
"Challener",
"Ryan C.",
""
],
[
"Colón",
"Knicole",
""
],
[
"Huang",
"Jingcheng",
""
],
[
"Lin",
"Zifan",
""
],
[
"Louie",
"Dana R.",
""
],
[
"Mullens",
"Elijah",
""
],
[
"Sotzen",
"Kristin S.",
""
],
[
"Valenti",
"Jeff A.",
""
],
[
"Valentine",
"Daniel",
""
],
[
"Clampin",
"Mark",
""
],
[
"Mountain",
"C. Matt",
""
],
[
"Perrin",
"Marshall",
""
],
[
"van der Marel",
"Roeland P.",
""
]
] |
2509.05408
|
Subham Roy
|
Daniel Robbins and Subham Roy
|
SymTFT actions, Condensable algebras and Categorical anomaly resolutions
| null | null | null | null |
hep-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We investigate symmetry topological field theories (SymTFTs) of non-abelian and non-invertible symmetries and the different Lagrangian algebras associated with a given Drinfeld center. For several examples we analyze the condensable algebras of the Drinfeld center to identify the intrinsically gapless symmetry protected topological (igSPT) phases. In previous work, the relation between igSPT phases and resolving anomalies by embedding an anomalous symmetry inside a larger fusion category was demonstrated. Here we present more examples of this mechanism that involve both group-like and categorical symmetries.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:02 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Robbins",
"Daniel",
""
],
[
"Roy",
"Subham",
""
]
] |
2509.05409
|
Nikita Schmal
|
Henning Bahl, Tilman Plehn, Nikita Schmal
|
Unbinning global LHC analyses
|
30 pages, 16 figures
| null | null | null |
hep-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Simulation-based inference has been shown to outperform traditional, histogram-based inference in numerous phenomenological and experimental studies. So far, these analyses have focused on individual high-profile processes. We study the combination of four different di-boson processes in terms of the Standard Model Effective Field Theory. Our results demonstrate how simulation-based inference also wins over traditional methods for global LHC analyses.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:02 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Bahl",
"Henning",
""
],
[
"Plehn",
"Tilman",
""
],
[
"Schmal",
"Nikita",
""
]
] |
2509.05410
|
Jiabin Yu
|
Jiabin Yu, Biao Lian, Shinsei Ryu
|
Wilson-Loop-Ideal Bands and General Idealization
|
10+20 pages, 2+0 figures
| null | null | null |
cond-mat.mes-hall cond-mat.str-el
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Quantum geometry is universally bounded from below by Wilson-loop windings. In this work, we define an isolated set of bands to be Wilson-loop-ideal, if their quantum metric saturates the Wilson-loop lower bound. The definition naturally incorporates the known Chern-ideal and Euler-ideal bands, and allows us to define other types of ideal bands, such as Kane-Mele $Z_2$-ideal bands. In particular, we find that an isolated set of two $Z_2$-ideal bands with non-singular nonabelian Berry curvature always admits a Chern-ideal gauge (i.e., effectively behaving as two decoupled Chern-ideal bands), even in the absence of any global good quantum number (such as spin). This enables the direct construction of fractional topological insulator wavefunctions. We further propose a general framework of constructing monotonic flows that achieve Wilson-loop-ideal states starting from non-ideal bands through band mixing, where Wilson-loop-ideal states are not energy eigenstates but have smooth projectors similar to isolated bands. We apply the constructed flows to the realistic model of $3.89^\circ$ twisted bilayer MoTe$_2$ and a moir\'e Rashba model, and numerically find Chern-ideal and $Z_2$-ideal states, respectively, with relative error in the integrated quantum metric below $5\times 10^{-3}$. Our general definition of Wilson-loop-ideal bands and general procedure of constructing Wilson-loop-ideal states provide a solid basis for future study of novel correlated physics.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:03 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yu",
"Jiabin",
""
],
[
"Lian",
"Biao",
""
],
[
"Ryu",
"Shinsei",
""
]
] |
2509.05411
|
Zichang Hao
|
Zichang Hao, Wei Jie Chan, Ching Hua Lee
|
Interacting many-body non-Hermitian systems as Markov chains
|
32 pages, 11 figures
| null | null | null |
cond-mat.other cond-mat.stat-mech quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Rich phenomenology emerges at the intersection of non-Hermiticity and many-body dynamics, yet physically realizable implementations remain challenging. In this work, we propose a general formalism that maps non-Hermitian many-body Hamiltonians to the Laplacians of Markov chains, such that wavefunction amplitudes are re-interpreted as stochastic many-body configuration probabilities. Despite explicitly preserving all state transition processes and inheriting analogous non-Hermitian localization and state-space fragmentation, our Markov chain processes exhibit distinct steady-state behavior independently of energetic considerations that govern quantum evolution. We demonstrate our framework with two contrasting representative scenarios, one involving asymmetric (biased) propagation with exclusion interactions, and the other involving flipping pairs of adjacent spins (agents). These results reveal robust and distinctive signatures of non-Hermitian phenomena in classical stochastic settings such as ecological and social networks, and provide a versatile framework for studying non-reciprocal many-body dynamics across and beyond physics.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:03 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hao",
"Zichang",
""
],
[
"Chan",
"Wei Jie",
""
],
[
"Lee",
"Ching Hua",
""
]
] |
2509.05412
|
Henry Maxfield
|
Jesse Held, Henry Maxfield
|
Gravitational Hilbert spaces: invariant and co-invariant states, inner products, gauge-fixing, and BRST
|
75 pages + appendices, with detailed 12 page summary of ideas in section 1
| null | null | null |
hep-th gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
Hilbert spaces in theories of gravity are notoriously subtle due to the Hamiltonian constraints, particularly regarding the inner product. To demystify this subject, we review and extend a collection of ideas in canonical gravity, and connect to the sum-over-histories approach by clarifying the Hilbert space interpretation of various gravitational path integrals. We use one-dimensional (or mini-superspace) models as the simplest context to exemplify the conceptual ideas. We emphasise that a physical Hilbert space can be defined either by requiring states to be annihilated by constraint operators (e.g., the Wheeler-DeWitt equation) or by equivalence relations between wavefunctions, and explain that these two approaches are related by an inner product. We advocate that the group averaging procedure constructs the correct physical inner product. The Klein-Gordon inner product is not positive-definite, which we explain as arising from a bad gauge choice; nonetheless, it agrees with group averaging when such a problem is absent. These concepts are all embedded in the BRST/BFV formalism, which provides a systematic way to construct these and other physically equivalent inner products (e.g., from maximal-volume gauge and Gaussian averaged gauges). Finally we discuss the application of these ideas in the semi-classical approximation, including non-perturbative gravitational effects.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:04 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Held",
"Jesse",
""
],
[
"Maxfield",
"Henry",
""
]
] |
2509.05413
|
Netta Engelhardt
|
Netta Engelhardt and Hong Liu
|
The Making of von Neumann Algebras from Bulk Focusing
|
27 pages, 3 figures
| null | null | null |
hep-th gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The single-trace, infinite-N algebra of an arbitrary region may or may not be a von Neumann algebra depending on the GNS sector. In this paper we identify the holographic dual of this mechanism as a consequence of the focusing of null geodesics; more precisely, this GNS sector-dependence corresponds to the well-known difference between null congruences fired from the bulk and those fired from the boundary. As part of establishing this property, we give a rigorous formulation and proof of causal wedge reconstruction for those general boundary subregions whose single trace algebras support von Neumann algebras at large-N. We discuss a possible finite-N extension and interpretation of our results as an explanation for the Hawking area theorem.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:05 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Engelhardt",
"Netta",
""
],
[
"Liu",
"Hong",
""
]
] |
2509.05414
|
N\'estor Espinoza
|
N\'estor Espinoza, Natalie H. Allen, Ana Glidden, Nikole K. Lewis, Sara Seager, Caleb I. Ca\~nas, David Grant, Am\'elie Gressier, Shelby Courreges, Kevin B. Stevenson, Sukrit Ranjan, Knicole Col\'on, Brett M. Morris, Ryan J. MacDonald, Douglas Long, Hannah R. Wakeford, Jeff A. Valenti, Lili Alderson, Natasha E. Batalha, Ryan C. Challener, Jingcheng Huang, Zifan Lin, Dana R. Louie, Elijah Mullens, Daniel Valentine, C. Matt Mountain, Laurent Pueyo, Marshall D. Perrin, Andrea Bellini, Jens Kammerer, Mattia Libralato, Isabel Rebollido, Emily Rickman, Sangmo Tony Sohn, Roeland P. van der Marel
|
JWST-TST DREAMS: NIRSpec/PRISM Transmission Spectroscopy of the Habitable Zone Planet TRAPPIST-1 e
|
Published in ApJ Letters, 16 pages, 4 figures (not including appendix). Check the companion paper (Glidden+2025) for secondary atmospheric interpretation. Data and scripts to reproduce all figures: https://zenodo.org/records/16125662
| null |
10.3847/2041-8213/adf42e
| null |
astro-ph.EP astro-ph.IM astro-ph.SR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
TRAPPIST-1 e is one of the very few rocky exoplanets that is both amenable to atmospheric characterization and that resides in the habitable zone of its star -- located at a distance from its star such that it might, with the right atmosphere, sustain liquid water on its surface. Here, we present a set of 4 JWST/NIRSpec PRISM transmission spectra of TRAPPIST-1 e obtained from mid to late 2023. Our transmission spectra exhibit similar levels of stellar contamination as observed in prior works for other planets in the TRAPPIST-1 system (Lim et al, 2023; Radica et al., 2024), but over a wider wavelength range, showcasing the challenge of characterizing the TRAPPIST-1 planets even at relatively long wavelengths (3-5 um). While we show that current stellar modeling frameworks are unable to explain the stellar contamination features in our spectra, we demonstrate that we can marginalize over those features instead using Gaussian Processes, which enables us to perform novel exoplanet atmospheric inferences with our transmission spectra. In particular, we are able to rule out cloudy, primary H$_2$-dominated ($\gtrsim$ 80$\%$ by volume) atmospheres at better than a 3$\sigma$ level. Constraints on possible secondary atmospheres on TRAPPIST-1 e are presented in a companion paper (Glidden et al., 2025). Our work showcases how JWST is breaking ground into the precisions needed to constrain the atmospheric composition of habitable-zone rocky exoplanets.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Espinoza",
"Néstor",
""
],
[
"Allen",
"Natalie H.",
""
],
[
"Glidden",
"Ana",
""
],
[
"Lewis",
"Nikole K.",
""
],
[
"Seager",
"Sara",
""
],
[
"Cañas",
"Caleb I.",
""
],
[
"Grant",
"David",
""
],
[
"Gressier",
"Amélie",
""
],
[
"Courreges",
"Shelby",
""
],
[
"Stevenson",
"Kevin B.",
""
],
[
"Ranjan",
"Sukrit",
""
],
[
"Colón",
"Knicole",
""
],
[
"Morris",
"Brett M.",
""
],
[
"MacDonald",
"Ryan J.",
""
],
[
"Long",
"Douglas",
""
],
[
"Wakeford",
"Hannah R.",
""
],
[
"Valenti",
"Jeff A.",
""
],
[
"Alderson",
"Lili",
""
],
[
"Batalha",
"Natasha E.",
""
],
[
"Challener",
"Ryan C.",
""
],
[
"Huang",
"Jingcheng",
""
],
[
"Lin",
"Zifan",
""
],
[
"Louie",
"Dana R.",
""
],
[
"Mullens",
"Elijah",
""
],
[
"Valentine",
"Daniel",
""
],
[
"Mountain",
"C. Matt",
""
],
[
"Pueyo",
"Laurent",
""
],
[
"Perrin",
"Marshall D.",
""
],
[
"Bellini",
"Andrea",
""
],
[
"Kammerer",
"Jens",
""
],
[
"Libralato",
"Mattia",
""
],
[
"Rebollido",
"Isabel",
""
],
[
"Rickman",
"Emily",
""
],
[
"Sohn",
"Sangmo Tony",
""
],
[
"van der Marel",
"Roeland P.",
""
]
] |
2509.05415
|
Fulya K{\i}ro\u{g}lu
|
Fulya K{\i}ro\u{g}lu, Kyle Kremer, Frederic A. Rasio
|
Beyond Hierarchical Mergers: Accretion-Driven Origins of Massive, Highly Spinning Black Holes in Dense Star Clusters
|
Submitted for publication in The Astrophysical Journal Letters. 10 pages, 4 figures, 2 tables
| null | null | null |
astro-ph.HE astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
GW231123, the most massive binary black hole (BBH) merger detected by LIGO/Virgo/KAGRA, highlights the need to understand the origins of massive, high-spin stellar black holes (BHs). Dense star clusters provide natural environments for forming such systems, beyond the limits of standard massive star evolution to core collapse. While repeated BBH mergers can grow BHs through dynamical interactions (the so-called "hierarchical merger" channel), most star clusters with masses $\lesssim 10^6\,M_\odot$ have escape speeds too low to retain higher-generation BHs, limiting growth into or beyond the mass gap. In contrast, BH--star collisions with subsequent accretion of the collision debris can grow and retain BHs irrespective of the cluster escape speed. Using $N$-body (Cluster Monte Carlo) simulations, we study BH growth and spin evolution through this process and we find that accretion can drive BH masses up to at least $\sim200\,M_\odot$, with spins set by the details of the growth history. BHs up to about $150\,M_\odot$ can reach dimensionless spins $\chi \gtrsim 0.7$ via single coherent episodes, while more massive BHs form through multiple stochastic accretion events and eventually spin down to $\chi \lesssim 0.4$. These BHs later form binaries through dynamical encounters, producing BBH mergers that contribute up to $\sim10\%$ of all detectable events, comparable to predictions for the hierarchical channel. However, the two pathways predict distinct signatures: hierarchical mergers yield more unequal mass ratios, whereas accretion-grown BHs preferentially form near-equal-mass binaries. The accretion-driven channel allows dense clusters with low escape speeds, such as globular clusters, to produce highly spinning BBHs with both components in or above the mass gap, providing a natural formation pathway to GW231123-like systems.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:09 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kıroğlu",
"Fulya",
""
],
[
"Kremer",
"Kyle",
""
],
[
"Rasio",
"Frederic A.",
""
]
] |
2509.05416
|
Hadrien Paugnat
|
Hadrien Paugnat (UCLA), Tommaso Treu (UCLA) and Daniel Gilman (UChicago)
|
Macromodel-free flux-ratio prediction in quadruply imaged quasars with local constraints from lensed arcs
|
23 pages, 13 figures, submitted to Physical Review D
| null | null | null |
astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Strong gravitational lensing is a powerful cosmological probe, providing a direct tool to unveil the properties of dark matter (DM) on sub-galactic scales. In particular, flux-ratio anomalies in quadruply imaged quasars ("quads") can reveal the presence of dark substructure, such that population-level statistics can be used to constrain the particle nature of DM. Current methods, however, rely on globally parametrized models ("macromodels") of the lens mass distribution, which impose rigid physical assumptions on the deflection field. Given the high stakes, it is important to develop complementary methods that do not require the assumption of a macromodel. One promising avenue consists of modeling the resolved emission from the quasar host galaxy (lensed arcs) using a local lensing formalism like the Curved Arc Basis (CAB) description. In this paper, we test the ability of CAB models to predict flux ratios from mock imaging data. We find that CAB model-predicted flux ratios accurately reproduce the expected values, with a typical precision of $\sim 3-5\%$. While a macromodel-based approach yields smaller uncertainties, as expected, the CAB method permits a more flexible, local description of the deflection field, thus being more robust to angular structure in the main deflector mass profile, in particular avoiding false-positive detections of flux-ratio anomalies that can arise with overly simplistic parametrizations. On the other hand, by injecting individual DM halos near quasar images, we demonstrate that CAB models do not absorb the local lensing perturbations from DM substructure, and can therefore distinguish flux-ratio anomalies caused by DM substructure from other sources of small-scale perturbation. We conclude that CAB models can be used to infer DM properties from flux-ratio anomaly statistics with minimal assumptions, complementing the traditional macromodel based approach.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:09 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Paugnat",
"Hadrien",
"",
"UCLA"
],
[
"Treu",
"Tommaso",
"",
"UCLA"
],
[
"Gilman",
"Daniel",
"",
"UChicago"
]
] |
2509.05417
|
Anna-Christina Eilers
|
Anna-Christina Eilers, Minghao Yue, Jorryt Matthee, Joseph F. Hennawi, Frederick B. Davies, Robert A. Simcoe, Richard Teague, Rongmon Bordoloi, Gabriel Brammer, Yi Kang, Daichi Kashino, Ruari Mackenzie, Rohan P. Naidu, Benjamin Navarrete
|
The Light Echo of a High-Redshift Quasar mapped with Lyman-$\alpha$ Tomography
|
Submitted to AAS Journals
| null | null | null |
astro-ph.GA astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
Ultra-violet (UV) radiation from accreting black holes ionizes the intergalactic gas around early quasars, carving out highly ionized bubbles in their surroundings. Any changes in a quasar's luminosity are therefore predicted to produce outward-propagating ionization gradients, affecting the Lyman-$\alpha$ (Ly$\alpha$) absorption opacity near the quasar's systemic redshift. This "proximity effect" is well-documented in rest-UV quasar spectra but only provides a one-dimensional probe along our line-of-sight. Here we present deep spectroscopic observations with the James Webb Space Telescope (JWST) of galaxies in the background of a super-luminous quasar at $z_{\rm QSO}\approx6.3$, which reveal the quasar's "light echo" with Ly$\alpha$ tomography in the transverse direction. This transverse proximity effect is detected for the first time towards multiple galaxy sightlines, allowing us to map the extent and geometry of the quasar's ionization cone. We obtain constraints on the orientation and inclination of the cone, as well as an upper limit on the obscured solid angle fraction of $f_{\rm obsc}<91\%$. Additionally, we find a timescale of the quasar's UV radiation of $t_{\rm QSO}=10^{5.6^{+0.1}_{-0.3}}$ years, which is significantly shorter than would be required to build up the central supermassive black hole (SMBH) with conventional growth models, but is consistent with independent measurements of the quasars' duty cycle. Our inferred obscured fraction disfavors a scenario where short quasar lifetimes can be explained exclusively by geometric obscuration, and instead supports the idea that radiatively inefficient accretion or growth in initially heavily enshrouded cocoons plays a pivotal role in early SMBH growth. Our results pave the way for novel studies of quasars' ionizing geometries and radiative histories at early cosmic times.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:00:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Eilers",
"Anna-Christina",
""
],
[
"Yue",
"Minghao",
""
],
[
"Matthee",
"Jorryt",
""
],
[
"Hennawi",
"Joseph F.",
""
],
[
"Davies",
"Frederick B.",
""
],
[
"Simcoe",
"Robert A.",
""
],
[
"Teague",
"Richard",
""
],
[
"Bordoloi",
"Rongmon",
""
],
[
"Brammer",
"Gabriel",
""
],
[
"Kang",
"Yi",
""
],
[
"Kashino",
"Daichi",
""
],
[
"Mackenzie",
"Ruari",
""
],
[
"Naidu",
"Rohan P.",
""
],
[
"Navarrete",
"Benjamin",
""
]
] |
2509.05418
|
Robert Plato
|
Robert Plato
|
A class of regularization schemes for linear ill-posed problems in Banach spaces under low order source conditions
| null | null | null | null |
math.NA cs.NA
|
http://creativecommons.org/licenses/by/4.0/
|
In this work we consider, in a Banach space framework, the regularization of linear ill-posed problems. Our focus is on the recovery of solutions that have a logarithmic source representation. Such cases typically occur in exponentially ill-posed problems like the backwards heat equation. The mathematical framework on logarithms of operators is provided, and convergence rates for a class of parametric regularization schemes are deduced. This class includes the iterated version of Lavrentiev's method and the method of the abstract Cauchy problem. The presentation includes both a priori and a posteriori parameter choice strategies. Finally, we present an example for a logarithmic source representable function with respect to the integration operator.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:01:04 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Plato",
"Robert",
""
]
] |
2509.05419
|
William Ortol\'a Leonard
|
William Ortol\'a Leonard and Zachary Slepian
|
Analytical Template for the 4-Point Correlation Function Covariance Beyond the Gaussian Random Field ${\rm I}$: 1-Loop Corrections involving Second-Order Densities
|
37 pages, 7 figures
| null | null | null |
astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Analytical templates for the covariance matrix of the 4-Point Correlation Function (4PCF) have been developed in the past assuming a Gaussian Random Field (GRF). In this work, we present the first non-Gaussian calculation of the 4PCF covariance, incorporating 1-loop corrections using the second-order density contrast. Furthermore, we introduce a non-trivial galaxy bias scheme at second order. To simplify the calculation, we decompose the covariance into five distinct structures, and then exploit the isotropic basis functions of Cahn & Slepian (2023). This approach reduces the complexity of the high-dimensional integrals naively involved, enabling the angular parts to be performed and leaving us with low-dimensional radial integrals. This analytical template will provide a more accurate characterization of the statistical errors on the 4PCF, improving both the parity-odd and parity-even analyses. This is the first paper in a two-part series.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:02:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Leonard",
"William Ortolá",
""
],
[
"Slepian",
"Zachary",
""
]
] |
2509.05420
|
Benjamin Savinson
|
Benjamin Savinson, David J. Norris, Siddhartha Mishra and Samuel Lanthaler
|
Universality of physical neural networks with multivariate nonlinearity
| null | null | null | null |
physics.optics cs.AI physics.class-ph physics.comp-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The enormous energy demand of artificial intelligence is driving the development of alternative hardware for deep learning. Physical neural networks try to exploit physical systems to perform machine learning more efficiently. In particular, optical systems can calculate with light using negligible energy. While their computational capabilities were long limited by the linearity of optical materials, nonlinear computations have recently been demonstrated through modified input encoding. Despite this breakthrough, our inability to determine if physical neural networks can learn arbitrary relationships between data -- a key requirement for deep learning known as universality -- hinders further progress. Here we present a fundamental theorem that establishes a universality condition for physical neural networks. It provides a powerful mathematical criterion that imposes device constraints, detailing how inputs should be encoded in the tunable parameters of the physical system. Based on this result, we propose a scalable architecture using free-space optics that is provably universal and achieves high accuracy on image classification tasks. Further, by combining the theorem with temporal multiplexing, we present a route to potentially huge effective system sizes in highly practical but poorly scalable on-chip photonic devices. Our theorem and scaling methods apply beyond optical systems and inform the design of a wide class of universal, energy-efficient physical neural networks, justifying further efforts in their development.
|
[
{
"version": "v1",
"created": "Sat, 06 Sep 2025 14:08:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Savinson",
"Benjamin",
""
],
[
"Norris",
"David J.",
""
],
[
"Mishra",
"Siddhartha",
""
],
[
"Lanthaler",
"Samuel",
""
]
] |
2509.05421
|
Hantian Zhang
|
Pia Bredt, Tatsuya Banno, Marius H\"ofer, Syuhei Iguro, Wolfgang Kilian, Yang Ma, J\"urgen Reuter, Hantian Zhang
|
Chasing the two-Higgs-doublet model via electroweak corrections at $e^+e^-$ colliders
| null | null | null |
CERN-TH-2025-181, DESY-25-121, SI-HEP-2025-19, P3H-25-063, IRMP-CP3-25-30, KA-TP-28-2025
|
hep-ph hep-ex
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present a comprehensive study of Higgs boson production associated with a neutrino pair at $e^+e^-$ colliders ($e^+ e^- \to h \, \nu \bar{\nu}$) at the next-to-leading-order accuracy in both the Standard Model and the two-Higgs-doublet model. We show that new physics effects from the extended Higgs sector can be probed through electroweak corrections, which lead to several percent deviations from the Standard Model predictions in total cross sections and differential distributions, even in the alignment limit. This highlights the potential of precision studies at future $e^+e^-$ colliders for searching new physics.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:05:33 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Bredt",
"Pia",
""
],
[
"Banno",
"Tatsuya",
""
],
[
"Höfer",
"Marius",
""
],
[
"Iguro",
"Syuhei",
""
],
[
"Kilian",
"Wolfgang",
""
],
[
"Ma",
"Yang",
""
],
[
"Reuter",
"Jürgen",
""
],
[
"Zhang",
"Hantian",
""
]
] |
2509.05422
|
William Ortol\'a Leonard
|
William Ortol\'a Leonard and Zachary Slepian
|
Analytical Template for the 4-Point Correlation Function Covariance Beyond the Gaussian Random Field ${\rm II}$: 1-Loop Corrections with Third-Order Densities
|
34 pages, 8 figures
| null | null | null |
astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Analytical templates for the 4-Point Correlation Function (4PCF) covariance matrix have been developed in the past assuming a Gaussian Random Field (GRF). In this work, we present the second part of the beyond GRF calculation of the 4PCF covariance, incorporating 1-loop corrections stemming from the third-order density contrast. Furthermore, we introduce a non-trivial galaxy biasing scheme at third order. To simplify the calculation, we decompose the covariance into three distinct structures and leverage the isotropic basis of Cahn & Slepian (2023). This approach reduces the complexity of the high-dimensional integrals that would be naively involved, enabling the angular parts to be performed and leaving us with low-dimensional radial integrals. This analytical template will provide a more accurate characterization of the statistical errors on the 4PCF, improving our ability to probe both its parity-even and parity-odd modes. This is the second and final paper in a two-part series.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:05:53 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Leonard",
"William Ortolá",
""
],
[
"Slepian",
"Zachary",
""
]
] |
2509.05423
|
Wako Ishibashi
|
W. Ishibashi, A. C. Fabian, R. Maiolino, Y. Gursahani, C. S. Reynolds
|
Another view into JWST-discovered X-ray weak AGNs via radiative dusty feedback
|
Accepted for publication in MNRAS
| null | null | null |
astro-ph.GA astro-ph.HE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
JWST has revealed a previously unknown population of low-luminosity active galactic nuclei (AGN) in the early Universe. These JWST-AGN at high redshifts are characterised by a set of peculiar properties, including unusually weak X-ray emission. Here we investigate the apparent lack of X-ray emission in the framework of the ``AGN radiative dusty feedback'' scenario based on the effective Eddington limit for dust. We analyse how the boundary in the $N_\mathrm{H} - \lambda$ plane, defined by the column density versus the Eddington ratio, is modified as a function of the dusty gas parameters (metallicity, dust grain size and composition). Low metallicity gas with little dust content tends to survive against radiation pressure, and likely accumulates in the nuclear region. We suggest that such dust-poor gas can provide long-lived absorption and may lead to heavy X-ray obscuration, as observed in early JWST-AGN. The blowout vs. stalling condition of the obscuring clouds indicates that higher metallicities are required to eject heavier column densities, while large columns of gas can stall in low metallicity environments. Therefore the metallicity may play a key role in the AGN radiative dusty feedback scenario. We discuss how other peculiar properties of JWST-AGN -- such as Balmer absorption features and weak radio emission -- may be naturally interpreted within the same physical framework.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:07:08 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ishibashi",
"W.",
""
],
[
"Fabian",
"A. C.",
""
],
[
"Maiolino",
"R.",
""
],
[
"Gursahani",
"Y.",
""
],
[
"Reynolds",
"C. S.",
""
]
] |
2509.05424
|
Alessandro Soncini
|
Alessandro Soncini, Kieran Hymas, Jonas Braun, Yannik F. Schneider, Simone Calvello, Amer Baniodeh, Yanhua Lan, Wolfgang Wernsdorfer, Marco Affronte, Christopher E. Anson, Annie K. Powell
|
Giant Molecular Toroidal Moment Amenable to Direct Observation in a Fe$_{10}$Dy$_{10}$ Ring
|
Main manuscript: 20 pages, 5 figures. Supplementary notes: 17 pages, 4 figures
| null | null | null |
cond-mat.mes-hall physics.atm-clus
|
http://creativecommons.org/licenses/by/4.0/
|
In single molecule toroics (SMTs) atomic spins and orbital currents generate magnetic vortices known as toroidal moments $\boldsymbol{\tau}$, endowed with both magnetic and electric dipole symmetries, which can enable spin control via magnetoelectric effects as well as the development of robust qubits. In the archetypal Dy$_3$ SMT, $\boldsymbol{\tau}$ is challenging to detect and control. Larger molecular rings can offer an enhanced toroidal response amenable to direct observation and manipulation. Here we report SMT properties for the $3d$-$4f$ icosanuclear molecular ring Fe$_{10}$Dy$_{10}$, displaying toroidal excitations of unprecedented magnitude and energy dispersion spanning a $\sim$62 billion dimensional toroidal space. We show these properties can be modeled using an ab initio-parameterised transfer matrix approach yielding excellent agreement with experiments. To assess the bulk toroidal polarization attainable in this system, we introduce the molar toroidal susceptibility $\xi$, a thermodynamic linear response function measuring the SMT finite-temperature toroidal polarization induced by a magnetic field with a small non-vanishing curl. Direct calculation of $\xi$ for Fe$_{10}$Dy$_{10}$ reveals a significant finite-temperature ground state toroidal polarization which should be amenable to experimental detection via spatially-focused magnetic field curls, as attainable e.g. using focused femtosecond laser pulses. Our findings could thus pave the way for direct observation and manipulation of molecular toroidal moments.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:08:05 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Soncini",
"Alessandro",
""
],
[
"Hymas",
"Kieran",
""
],
[
"Braun",
"Jonas",
""
],
[
"Schneider",
"Yannik F.",
""
],
[
"Calvello",
"Simone",
""
],
[
"Baniodeh",
"Amer",
""
],
[
"Lan",
"Yanhua",
""
],
[
"Wernsdorfer",
"Wolfgang",
""
],
[
"Affronte",
"Marco",
""
],
[
"Anson",
"Christopher E.",
""
],
[
"Powell",
"Annie K.",
""
]
] |
2509.05425
|
Jessica Lundin PhD
|
Jessica M. Lundin, Ada Zhang, David Adelani, Cody Carroll
|
No Translation Needed: Forecasting Quality from Fertility and Metadata
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We show that translation quality can be predicted with surprising accuracy \textit{without ever running the translation system itself}. Using only a handful of features, token fertility ratios, token counts, and basic linguistic metadata (language family, script, and region), we can forecast ChrF scores for GPT-4o translations across 203 languages in the FLORES-200 benchmark. Gradient boosting models achieve favorable performance ($R^{2}=0.66$ for XX$\rightarrow$English and $R^{2}=0.72$ for English$\rightarrow$XX). Feature importance analyses reveal that typological factors dominate predictions into English, while fertility plays a larger role for translations into diverse target languages. These findings suggest that translation quality is shaped by both token-level fertility and broader linguistic typology, offering new insights for multilingual evaluation and quality estimation.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:11:49 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lundin",
"Jessica M.",
""
],
[
"Zhang",
"Ada",
""
],
[
"Adelani",
"David",
""
],
[
"Carroll",
"Cody",
""
]
] |
2509.05426
|
Pengfei Cai
|
Pengfei Cai, Anas Abdallah, Pratheepa Jeganathan
|
Sparse Seemingly Unrelated Regression (SSUR) Copula Mixed Models for Multivariate Loss Reserving
| null | null | null | null |
stat.ME stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Insurance companies often operate across multiple interrelated lines of business (LOBs), and accounting for dependencies between them is essential for accurate reserve estimation and risk capital determination. In our previous work on the Extended Deep Triangle (EDT), we demonstrated that a more flexible model that uses multiple companies' data reduces reserve prediction error and increases diversification benefits. However, the EDT's limitation lies in its limited interpretability of the dependence structure, which is an important feature needed by insurers to guide strategic decisions. Motivated by the need for interpretability and flexibility, this paper proposes a Seemingly Unrelated Regression (SUR) copula mixed model to handle heterogeneous data across multiple companies. The model incorporates random effects to capture company-specific heterogeneity, uses flexible marginal distributions across LOBs, and treats development and accident year effects as fixed effects with shrinkage via LASSO to enhance robustness. We estimate the model using an iterative two-stage procedure and generate predictive reserve distributions via a modified bootstrap that accounts for systematic effects, dependence structure, and sparse fixed-effect coefficients. Through simulation studies and real data from the National Association of Insurance Commissioners, we show that the proposed model outperforms the SUR copula regression model in terms of reserve accuracy and generates larger risk capital gain. Overall, the SUR copula mixed model achieves better predictive performance, greater risk diversification, and retains interpretability.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:12:08 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Cai",
"Pengfei",
""
],
[
"Abdallah",
"Anas",
""
],
[
"Jeganathan",
"Pratheepa",
""
]
] |
2509.05427
|
Alexander A. Mushtukov
|
Alexander A. Mushtukov, Emir Tataroglu, Alex J. Cooper, Sergey S. Tsygankov
|
Gamma-ray lines, electron-positron annihilation, and possible radio emission in X-ray pulsars
|
submitted to MNRAS, 11 pages, 8 figures, comments are welcome
| null | null | null |
astro-ph.HE astro-ph.SR
|
http://creativecommons.org/licenses/by/4.0/
|
Accretion onto neutron stars (NSs) in X-ray pulsars (XRPs) results in intense X-ray emission, and under specific conditions, high-energy nuclear interactions that produce gamma-ray photons at discrete energies. These interactions are enabled by the high free-fall velocities of accreting nuclei near the NS surface and give rise to characteristic gamma-ray lines, notably at 2.2 MeV, 5.5 MeV, and 67.5 MeV. We investigate the production mechanisms of these lines and estimate the resulting gamma-ray luminosities, accounting for the suppression effects of radiative deceleration in bright XRPs and the creation of electron-positron pairs in strong magnetic fields. The resulting annihilation of these pairs leads to a secondary emission line at $\sim 511$ keV. We also discuss the possibility that non-stationary pair creation in the polar cap region could drive coherent radio emission, though its detectability in accreting systems remains uncertain. Using a numerical framework incorporating general relativistic light bending and magnetic absorption, we compute the escape fraction of photons and distinguish between actual and apparent gamma-ray luminosities. Our results identify the parameter space - defined by magnetic field strength, accretion luminosity, and NS compactness - where these gamma-ray signatures may be observable by upcoming MeV gamma-ray missions. In particular, we highlight the diagnostic potential of detecting gravitationally redshifted gamma-ray lines and annihilation features for probing the mass-radius relation and magnetospheric structure of NSs.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:13:39 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Mushtukov",
"Alexander A.",
""
],
[
"Tataroglu",
"Emir",
""
],
[
"Cooper",
"Alex J.",
""
],
[
"Tsygankov",
"Sergey S.",
""
]
] |
2509.05428
|
Elena Bratkovskaya
|
Mahbobeh Jafarpour, Vadym Voronyuk, Klaus Werner, Elena Bratkovskaya, Damien Vintache
|
Disentangling Initial-State and Evolution Effects in Heavy-Ion Collisions Using EPOS and PHSD
|
25 pages, 30 figures
| null | null | null |
hep-ph hep-ex nucl-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this study we examine the impact of the initial stage and dynamical evolution on final-state observables in heavy-ion collisions. For this goal we develop a novel approach, EPOSir+PHSDe, which employs EPOS initial conditions as the starting point for parton and hadron evolution within the PHSD microscopic transport approach. By examining the space-time evolution of matter in this model and comparing to EPOS (which starts with an S-matrix approach for parallel scatterings for the initial conditions and uses a hydrodynamic evolution for the quark-gluon plasma stage with the UrQMD as afterburner) and PHSD (which starts with primary high energy $NN$ scattering realized via the LUND string model and continues with fully microscopic transport dynamics for strongly interacting partonic and hadronic matter), we identify the key differences in the final particle distributions among the three approaches. Our analysis focuses on rapidity, transverse momentum spectra, and flow harmonics $v_2$ for Au+Au collisions at the invariant energy of $\sqrt{s_{NN}}=200$ GeV. We find a dominant influence of dynamical evolution over the initial conditions on the final observables.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:16:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Jafarpour",
"Mahbobeh",
""
],
[
"Voronyuk",
"Vadym",
""
],
[
"Werner",
"Klaus",
""
],
[
"Bratkovskaya",
"Elena",
""
],
[
"Vintache",
"Damien",
""
]
] |
2509.05430
|
Laurence Datrier
|
Laurence Datrier, Geoffrey Lovelace, Joshua R. Smith, Andrew Saenz, Amber Romero
|
Site Evaluation and Cost Estimation for Cosmic Explorer
|
4 pages, 3 figures, conference proceedings for GR24/Amaldi16 meeting (submitted)
| null | null | null |
physics.ins-det astro-ph.IM
|
http://creativecommons.org/licenses/by/4.0/
|
Cosmic Explorer (CE) is a proposed next generation gravitational-wave observatory that would be sited in the United States. As of 2025, CE is in its design phase, with plans to begin operations in the 2030s together with the Einstein Telescope in Europe. CE's reference design consists of two widely separated L-shaped detectors, one with 20km arms and one with 40km arms, each based on technology proven by the National Science Foundation's highly successful Laser Interferometer Gravitational Wave Observatory (LIGO). There are unique challenges associated with identifying locations suitable for hosting Cosmic Explorer in the conterminous United States, not least of which is the order of magnitude upscaling of the observatory with respect to the 4km LIGO observatories. Cosmic Explorer's approach to site evaluation integrates physical, social and cultural criteria. Here we present improvements to the Cosmic Explorer Location Search (CELS) code used to identify and assess locations where CE would have low construction costs incurred by the geology, geography and topography of the land. We also report on efforts to integrate astrophysical requirements established by the Cosmic Explorer Science Traceability Matrix into the site evaluation process. National-level results are presented and combined with results from a related National Suitability Analysis to provide a list of locations that are preliminarily promising for a 40km CE.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:23:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Datrier",
"Laurence",
""
],
[
"Lovelace",
"Geoffrey",
""
],
[
"Smith",
"Joshua R.",
""
],
[
"Saenz",
"Andrew",
""
],
[
"Romero",
"Amber",
""
]
] |
2509.05431
|
Tania-Amanda Nkoyo Fredrick Eneye
|
GodsGift Uzor, Tania-Amanda Nkoyo Fredrick Eneye, Chukwuebuka Ijezue
|
Advanced Brain Tumor Segmentation Using EMCAD: Efficient Multi-scale Convolutional Attention Decoding
| null | null | null | null |
cs.CV cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Brain tumor segmentation is a critical pre-processing step in the medical image analysis pipeline that involves precise delineation of tumor regions from healthy brain tissue in medical imaging data, particularly MRI scans. An efficient and effective decoding mechanism is crucial in brain tumor segmentation especially in scenarios with limited computational resources. However these decoding mechanisms usually come with high computational costs. To address this concern EMCAD a new efficient multi-scale convolutional attention decoder designed was utilized to optimize both performance and computational efficiency for brain tumor segmentation on the BraTs2020 dataset consisting of MRI scans from 369 brain tumor patients. The preliminary result obtained by the model achieved a best Dice score of 0.31 and maintained a stable mean Dice score of 0.285 plus/minus 0.015 throughout the training process which is moderate. The initial model maintained consistent performance across the validation set without showing signs of over-fitting.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:23:47 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Uzor",
"GodsGift",
""
],
[
"Eneye",
"Tania-Amanda Nkoyo Fredrick",
""
],
[
"Ijezue",
"Chukwuebuka",
""
]
] |
2509.05432
|
Ziad Ghanem
|
Ziad Ghanem
|
Explicit Generators for the Unit Group of the Burnside ring
| null | null | null | null |
math.RA math.RT
|
http://creativecommons.org/licenses/by/4.0/
|
To the best of our knowledge, there is no explicit, constructive description of the generating set for the unit group $A(G)^\times$ of the Burnside ring associated with a finite group $G$. We resolve this long-standing open question, proving that $A(G)^\times$ is generated by the set of \emph{basic degrees} -- canonical Burnside ring elements arising from the $G$-equivariant degree of the identity map on irreducible $G$-representations. In particular, we demonstrate that every unit in $A(G)$ is realized as the equivariant degree of a linear $G$-isomorphism on a suitable orthogonal $G$-representation which, in turn, can be described as the Burnside ring product of a finite number of basic degrees, establishing a concrete link between the multiplicative structure of the Burnside ring and the field of equivariant topology.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:23:50 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ghanem",
"Ziad",
""
]
] |
2509.05433
|
Rui Chen
|
Rui Chen, Domenico Chiaradia, Antonio Frisoli, and Daniele Leonardis
|
HapMorph: A Pneumatic Framework for Multi-Dimensional Haptic Property Rendering
|
20 pages, 5 figures
| null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Haptic interfaces that can simultaneously modulate multiple physical properties remain a fundamental challenge in human-robot interaction. Existing systems typically allow the rendering of either geometric features or mechanical properties, but rarely both, within wearable form factors. Here, we introduce HapMorph, a pneumatic framework that enables continuous, simultaneous modulation of object size and stiffness through antagonistic fabric-based pneumatic actuators (AFPAs). We implemented a HapMorph protoytpe designed for hands interaction achieving size variation from 50 to 104 mm, stiffness modulation up to 4.7 N/mm and mass of the wearable parts of just 21 g. Through systematic characterization, we demonstrate decoupled control of size and stiffness properties via dual-chamber pressure regulation. Human perception studies with 10 participants reveal that users can distinguish nine discrete states across three size categories and three stiffness levels with 89.4% accuracy and 6.7 s average response time. We further demonstrate extended architectures that combine AFPAs with complementary pneumatic structures to enable shape or geometry morphing with concurrent stiffness control. Our results establish antagonistic pneumatic principle as a pathway toward next-generation haptic interfaces, capable of multi-dimensiona rendering properties within practical wearable constraints.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:26:58 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chen",
"Rui",
""
],
[
"Chiaradia",
"Domenico",
""
],
[
"Frisoli",
"Antonio",
""
],
[
"Leonardis",
"Daniele",
""
]
] |
2509.05434
|
Jenny Greene
|
Jenny E. Greene, David J. Setton, Lukas J. Furtak, Rohan P. Naidu, Marta Volonteri, Pratika Dayal, Ivo Labbe, Pieter van Dokkum, Rachel Bezanson, Gabriel Brammer, Sam E. Cutler, Karl Glazebrook, Anna de Graaff, Michaela Hirschmann, Raphael E. Hviding, Vasily Kokorev, Joel Leja, Hanpu Liu, Yilun Ma, Jorryt Matthee, Themiya Nanayakkara, Pascal A. Oesch, Richard Pan, Sedona H. Price, Justin S. Spilker, Bingjie Wang, John R. Weaver, Katherine E. Whitaker, Christina C. Williams, and Adi Zitrin
|
What you see is what you get: empirically measured bolometric luminosities of Little Red Dots
|
11 pages, 3 figures, 1 table, posted following referee comments
| null | null | null |
astro-ph.GA astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
New populations of red active galactic nuclei (known as ``Little Red Dots'') discovered by JWST exhibit remarkable spectral energy distributions. Leveraging X-ray through far-infrared observations of two of the most luminous known Little Red Dots, we directly their bolometric luminosities. We find evidence that more than half of the bolometric luminosity likely emerges in the rest-frame optical, with $L_{\rm bol}/L_{5100} = 5$, roughly half the value for ``standard'' Active Galactic Nuclei. Meanwhile, the X-ray emitting corona, UV-emitting black-body, and reprocessed mid to far-infrared emission are all considerably sub-dominant, assuming that the far-infrared luminosity is well below current measured limits. We present new bolometric corrections that dramatically lower inferred bolometric luminosities by a factor of ten compared to published values in the literature. These bolometric corrections are in accord with expectations from models in which gas absorption and reprocessing are responsible for the red rest-frame optical colors of Little Red Dots. We discuss how this lowered luminosity scale suggests a lower mass scale for the population by at least an order of magnitude {\bf (e.g., $\sim 10^5-10^7~{\rm M_{\odot}}$ black holes, and $\sim 10^8~{\rm M_{\odot}}$ galaxies)}, alleviating tensions with clustering, overmassive black holes, and the integrated black hole mass density in the Universe.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:30:31 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Greene",
"Jenny E.",
""
],
[
"Setton",
"David J.",
""
],
[
"Furtak",
"Lukas J.",
""
],
[
"Naidu",
"Rohan P.",
""
],
[
"Volonteri",
"Marta",
""
],
[
"Dayal",
"Pratika",
""
],
[
"Labbe",
"Ivo",
""
],
[
"van Dokkum",
"Pieter",
""
],
[
"Bezanson",
"Rachel",
""
],
[
"Brammer",
"Gabriel",
""
],
[
"Cutler",
"Sam E.",
""
],
[
"Glazebrook",
"Karl",
""
],
[
"de Graaff",
"Anna",
""
],
[
"Hirschmann",
"Michaela",
""
],
[
"Hviding",
"Raphael E.",
""
],
[
"Kokorev",
"Vasily",
""
],
[
"Leja",
"Joel",
""
],
[
"Liu",
"Hanpu",
""
],
[
"Ma",
"Yilun",
""
],
[
"Matthee",
"Jorryt",
""
],
[
"Nanayakkara",
"Themiya",
""
],
[
"Oesch",
"Pascal A.",
""
],
[
"Pan",
"Richard",
""
],
[
"Price",
"Sedona H.",
""
],
[
"Spilker",
"Justin S.",
""
],
[
"Wang",
"Bingjie",
""
],
[
"Weaver",
"John R.",
""
],
[
"Whitaker",
"Katherine E.",
""
],
[
"Williams",
"Christina C.",
""
],
[
"Zitrin",
"Adi",
""
]
] |
2509.05435
|
David Vegh
|
David Vegh
|
A folded string dual for the Sachdev-Ye-Kitaev model
|
8 pages, 8 figures
| null | null |
QMUL-PH-25-25
|
hep-th cond-mat.str-el
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose a folded string moving in rigid AdS$_2$ with imaginary radius squared as a dual of the Sachdev-Ye-Kitaev (SYK) model at its conformal fixed point. In standard AdS$_2$, the string is represented by two massless particles connected by straight string segments. The particles move at the speed of light, abruptly reversing direction at turning points. We describe the system using the lightcone coordinates of these points, with the Poisson structure obtained from the Peierls bracket. In AdS$_2$ with imaginary radius squared, quantization of the string's mass-squared in momentum-fraction space yields a P\"oschl-Teller equation, reproducing the SYK operator spectrum.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:32:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Vegh",
"David",
""
]
] |
2509.05436
|
Kip S. Thorne
|
James B. Hartle (1) and Kip S. Thorne (2) ((1) University of California Santa Barbara, (2) California Institute of Technology)
|
Slowly Rotating Relativistic Stars: VII. Gravitational Radiation from the Quasi-Radial Modes
|
The research reported in this paper was completed in 1972. The paper itself was completed and put on the arXiv in 2025. For explanation, see the box on page 2 following the Abstract
| null | null | null |
gr-qc astro-ph.HE astro-ph.SR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
When a relativistic star rotates slowly and rigidly, centrifugal forces flatten it slightly, thereby catalyzing a small admixture of quadrupolar vibration into its radial modes and a damping of the resulting quasi-radial modes. The damping rate $1/\tau$ of each quasi-radial mode divided by its frequency $\sigma^{(0)}$ is given by ${(1/\tau) / \sigma^{(0)} }= \beta (\sigma^{(0)})^3 \Omega^4 R^8/M \;, $where $\Omega$, $R$ and $M$ are the angular velocity, radius and mass of the star, and $\beta$ is a dimensionless number that depends on the mode, on relativistic corrections, and on the structure of the star, and is typically of order unity for the fundamental quasi-radial mode in the nonrelativistic limit. In this paper we develop equations and an algorithm for computing the emitted waves and the resulting damping factor $\beta$. The rotation is treated to second-order in the angular velocity and the pulsation amplitudes are assumed small and linearized, but no other approximations are made.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:38:50 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hartle",
"James B.",
"",
"University of California Santa Barbara"
],
[
"Thorne",
"Kip S.",
"",
"California Institute of Technology"
]
] |
2509.05437
|
He Wang
|
Yang Gao, Feiyu Li, Yang Liu, Zhen Yang, Jiayu Ding, Wuerkaixi Nuerbolati, Ruixia Wang, Tang Su, Yanjun Ma, Yirong Jin, Haifeng Yu, He Wang, and Fei Yan
|
Mitigating Measurement Crosstalk via Pulse Shaping
|
5 pages, 3 figures
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Quantum error correction protocols require rapid and repeated qubit measurements. While multiplexed readout in superconducting quantum systems improves efficiency, fast probe pulses introduce spectral broadening, leading to signal leakage into neighboring readout resonators. This crosstalk results in qubit dephasing and degraded readout fidelity. Here, we introduce a pulse shaping technique inspired by the derivative removal by adiabatic gate (DRAG) protocol to suppress measurement crosstalk during fast readout. By engineering a spectral notch at neighboring resonator frequencies, the method effectively mitigates spurious signal interference. Our approach integrates seamlessly with existing readout architectures, enabling fast, low-crosstalk multiplexed measurements without additional hardware overhead -- a critical advancement for scalable quantum computing.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:42:35 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Gao",
"Yang",
""
],
[
"Li",
"Feiyu",
""
],
[
"Liu",
"Yang",
""
],
[
"Yang",
"Zhen",
""
],
[
"Ding",
"Jiayu",
""
],
[
"Nuerbolati",
"Wuerkaixi",
""
],
[
"Wang",
"Ruixia",
""
],
[
"Su",
"Tang",
""
],
[
"Ma",
"Yanjun",
""
],
[
"Jin",
"Yirong",
""
],
[
"Yu",
"Haifeng",
""
],
[
"Wang",
"He",
""
],
[
"Yan",
"Fei",
""
]
] |
2509.05438
|
Sarah Gavin
|
S. Carin Gavin, William Koll, Moumita Kar, Yiying Liu, Anushka Dasgupta, Ethan Garvey, Thomas W. Song, Chunxi Zhou, Brendan P. Kerwin, Jash Jain, Tobin J. Marks, Mark C. Hersam, George C. Schatz, Jay A. Gupta, Nathaniel P. Stern
|
Modifying the Optical Emission of Vanadyl Phthalocyanine via Molecular Self-Assembly on van der Waals Materials
|
8 pages, 5 figures
| null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vanadyl phthalocyanine (VOPc) is a promising organic molecule for applications in quantum information because of its thermal stability, efficient processing, and potential as a spin qubit. The deposition of VOPc in different molecular orientations allows the properties to be customized for integration into various devices. However, such customization has yet to be fully leveraged to alter its intrinsic properties, particularly optical emission. Normally, VOPc films on dielectric substrates emit a broad photoluminescence peak in the near-infrared range, attributed to transitions in the Pc ring from its pi orbital structure. In this work, we demonstrate that the dominant optical transition of VOPc can be shifted by over 250 meV through the controlled deposition of thin films on van der Waals material substrates. The weak interactions with van der Waals materials allow the molecules to uniquely self-assemble, resulting in modified optical behavior modulated by molecular phase and thickness. This work connects the self-assembling properties of molecules with their altered electronic structures and the resulting optical emission.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:47:01 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Gavin",
"S. Carin",
""
],
[
"Koll",
"William",
""
],
[
"Kar",
"Moumita",
""
],
[
"Liu",
"Yiying",
""
],
[
"Dasgupta",
"Anushka",
""
],
[
"Garvey",
"Ethan",
""
],
[
"Song",
"Thomas W.",
""
],
[
"Zhou",
"Chunxi",
""
],
[
"Kerwin",
"Brendan P.",
""
],
[
"Jain",
"Jash",
""
],
[
"Marks",
"Tobin J.",
""
],
[
"Hersam",
"Mark C.",
""
],
[
"Schatz",
"George C.",
""
],
[
"Gupta",
"Jay A.",
""
],
[
"Stern",
"Nathaniel P.",
""
]
] |
2509.05439
|
Matthew Frazier
|
Matthew Frazier and Guillaume Bal
|
Quantum anomalous Hall phases in gated rhombohedral graphene
|
21 pages, 7 figures
| null | null | null |
cond-mat.mes-hall math-ph math.MP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We consider a coupled system of Dirac operators that finds applications as a macroscopic model of spin and valley polarized gated rhombohedral graphene (RHG) with an arbitrary number of layers as well as in replica models of Floquet topological insulators. We classify all quantum anomalous Hall phases that are compatible with the model and show that a bulk-edge correspondence between bulk phases and chiral edge states carrying a quantized anomalous Hall charge applies. When the displacement field is sufficiently small compared to the inter-layer coupling in the RHG application, we retrieve the known phases where the charge is given by the number of graphene layers. When the displacement field increases, we identify all possible topological phase transitions and corresponding quantized chiral edge charges. Numerical simulations confirm the theoretical findings.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:47:18 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Frazier",
"Matthew",
""
],
[
"Bal",
"Guillaume",
""
]
] |
2509.05440
|
Logan Lawrence
|
Logan Lawrence, Ashton Williamson, Alexander Shelton
|
Direct-Scoring NLG Evaluators Can Use Pairwise Comparisons Too
|
12 pages, 18 tables, 1 figure
| null | null | null |
cs.CL cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
As large-language models have been increasingly used as automatic raters for evaluating free-form content, including document summarization, dialog, and story generation, work has been dedicated to evaluating such models by measuring their correlations with human judgment. For \textit{sample-level} performance, methods which operate by using pairwise comparisons between machine-generated text perform well but often lack the ability to assign absolute scores to individual summaries, an ability crucial for use cases that require thresholding. In this work, we propose a direct-scoring method which uses synthetic summaries to act as pairwise machine rankings at test time. We show that our method performs comparably to state-of-the-art pairwise evaluators in terms of axis-averaged sample-level correlations on the SummEval (\textbf{+0.03}), TopicalChat (\textbf{-0.03}), and HANNA (\textbf{+0.05}) meta-evaluation benchmarks, and release the synthetic in-context summaries as data to facilitate future work.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:48:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lawrence",
"Logan",
""
],
[
"Williamson",
"Ashton",
""
],
[
"Shelton",
"Alexander",
""
]
] |
2509.05442
|
Konstantinos Kolokythas Dr
|
Konstantinos Kolokythas, Tiziana Venturi, Kenda Knowles, Marcus Br\"uggen, Francesco de Gasperin, S. Precious Sikhosana, Tracy E. Clarke, Oleg Smirnov, S. Ilani Loubser, and Kavilian Moodley
|
The MeerKAT Galaxy Cluster Legacy Survey -- II. Catalogue of the diffuse radio emission in MeerKAT-GCLS clusters
|
Accepted for publication by MNRAS, 19 manuscript pages with 2 tables and 9 figures, along with 51 pages of appendices that include a literature review on 56 systems and 56 figures
| null | null | null |
astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present a follow-up study focused on cluster-scale diffuse radio emissions in 115 galaxy clusters in the Southern sky, using full-resolution (7.8") and tapered low-resolution (15") images from the MeerKAT Galaxy Cluster Legacy Survey (MGCLS). In this MGCLS follow-up paper, we update and quantitatively characterise the presence of diffuse radio emission and provide detailed catalogue information on its radio properties at 1.28 GHz. As the MGCLS sample is heterogeneous and was initially used as a test bed to reveal the scientific potential of MeerKAT, the reported numbers are subject to this special case. More than half ($\sim$54%; 62/115) of the observed MGCLS clusters present diffuse cluster radio emission, with the total number of diffuse radio sources or candidates detected being 103. Including candidates, we find that radio relics are the most frequently detected diffuse sources in MGCLS at 53% (55/103), followed by halos at 32% (33/103) and mini-halos at 10% (11/103), with only 3% (3/103) being Phoenixes. The sizes of the diffuse radio structures and candidates range from 55 kpc to over 2 Mpc, with $P_{1.28GHz}$ radio power ranging from $10^{22}$ W Hz$^{-1}$ to greater than $10^{25}$ W Hz$^{-1}$. In-band radio spectral index estimates reveal revived radio plasma emissions that exhibit steep radio spectral indices down to $\alpha^{1656}_{908} \sim -3.5$. Mini-halos and their respective candidates are detected mainly in low-mass ($M_{500}$ $\leq$ 5$\times10^{14}$ M$_\odot$) and low-power ($P_{1.4GHz}$ $\leq$ $10^{23}$ W Hz$^{-1}$) systems. We suggest the presence of a statistically significant correlation between the 1.4 GHz radio power and the cluster mass for mini-halo (and candidate) systems.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:49:20 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kolokythas",
"Konstantinos",
""
],
[
"Venturi",
"Tiziana",
""
],
[
"Knowles",
"Kenda",
""
],
[
"Brüggen",
"Marcus",
""
],
[
"de Gasperin",
"Francesco",
""
],
[
"Sikhosana",
"S. Precious",
""
],
[
"Clarke",
"Tracy E.",
""
],
[
"Smirnov",
"Oleg",
""
],
[
"Loubser",
"S. Ilani",
""
],
[
"Moodley",
"Kavilian",
""
]
] |
2509.05443
|
R. Noah Padgett
|
R. Noah Padgett
|
Multidimensional constructs and moderated linear and nonlinear factor analysis
|
17 pages, 2 figures
| null | null | null |
stat.ME stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Multidimensional factor models with moderations on all model parameters have so far been limited to single-factor and two-factor models. This does not align well with existing psychological measures, which are commonly intended to assess 3-5 dimensions of a latent construct. In this paper, we introduce a penalized maximum likelihood approach for multidimensional MNLFA that permits moderation of item intercepts, loadings, residual variances, factor means, variances, and correlations across three or more latent factors. Our approach incorporates ridge, lasso, and alignment penalties to stabilize estimation and detect partial measurement non-invariance while preserving model interpretability. We derive closed-form analytic gradients of the likelihood, eliminating the need for costly numerical or MCMC-based approximations, and demonstrate how this dramatically improves computational efficiency. Through simulation and application, we illustrate that penalized MNLFA recovers complex moderation patterns in multidimensional constructs and provides a scalable alternative to Bayesian implementations. We conclude by discussing the theoretical implications of penalization for measurement invariance, computational considerations, and future directions for extending the framework to categorical indicators, longitudinal data, and applied research contexts.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:49:54 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Padgett",
"R. Noah",
""
]
] |
2509.05444
|
Yili Hong
|
Jared M. Clark, Jie Min, Yueyao Wang, Yili Hong, George Ostrouchov
|
Modeling Spatially Correlated Failure-time Data Under Two Distance Functions with an Application to Titan GPU Data
|
27 pages, 9 figures
| null | null | null |
stat.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
One common approach to statistical analysis of spatially correlated data relies on defining a correlation structure based solely on unknown parameters and the physical distance between the locations of observed values. However, some data have a complex spatial structure that cannot be adequately described with the physical distance alone. In this work, the spatial failure-time data of focus contains information on GPUs that are connected through a network fabric topology that differs from their physical layout and that is expected to introduce additional correlations. The proposed lifetime regression model includes random effects capturing the dependency due to physical location as well as random effects explaining the dependency due to logical connections between GPUs. The analysis of this GPU dataset serves as an example of models with multiple spatial random effects and the ideas presented can be extended to other applications with complex spatial structures. A Bayesian modeling scheme is recommended for this class of analyses. The examples in this work use the software package, Stan, to produce Markov chain Monte Carlo draws for parameter estimation. This modeling effort is validated through simulation which demonstrates accuracy in statistical inference. We also apply the developed framework to the large-scale Titan GPU failure time data.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:51:36 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Clark",
"Jared M.",
""
],
[
"Min",
"Jie",
""
],
[
"Wang",
"Yueyao",
""
],
[
"Hong",
"Yili",
""
],
[
"Ostrouchov",
"George",
""
]
] |
2509.05445
|
Grzegorz Sroka
|
Grzegorz Sroka and S{\l}awomir T. Wierzcho\'n
|
Robustness and Invariance of Hybrid Metaheuristics under Objective Function Transformations
| null | null | null | null |
cs.NE
|
http://creativecommons.org/licenses/by/4.0/
|
This paper evaluates the robustness and structural invariance of hybrid population-based metaheuristics under various objective space transformations. A lightweight plug-and-play hybridization operator is applied to nineteen state-of-the-art algorithms-including differential evolution (DE), particle swarm optimization (PSO), and recent bio-inspired methods-without modifying their internal logic. Benchmarking on the CEC-2017 suite across four dimensions (10, 30, 50, 100) is performed under five transformation types: baseline, translation, scaling, rotation, and constant shift. Statistical comparisons based on Wilcoxon and Friedman tests, Bayesian dominance analysis, and convergence trajectory profiling consistently show that differential-based hybrids (e.g., hIMODE, hSHADE, hDMSSA) maintain high accuracy, stability, and invariance under all tested deformations. In contrast, classical algorithms-especially PSO- and HHO-based variants-exhibit significant performance degradation under non-separable or distorted landscapes. The findings confirm the superiority of adaptive, structurally resilient hybrids for real-world optimization tasks subject to domain-specific transformations.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:55:33 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sroka",
"Grzegorz",
""
],
[
"Wierzchoń",
"Sławomir T.",
""
]
] |
2509.05446
|
Ahmed Faizul Haque Dhrubo Mr.
|
Iftekhar Haider Chowdhury, Zaed Ikbal Syed, Ahmed Faizul Haque Dhrubo, Mohammad Abdul Qayum
|
Dynamic Sensitivity Filter Pruning using Multi-Agent Reinforcement Learning For DCNN's
|
This paper includes figures and two tables, and our work outperforms the existing research that has been published in a journal
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Deep Convolutional Neural Networks have achieved state of the art performance across various computer vision tasks, however their practical deployment is limited by computational and memory overhead. This paper introduces Differential Sensitivity Fusion Pruning, a novel single shot filter pruning framework that focuses on evaluating the stability and redundancy of filter importance scores across multiple criteria. Differential Sensitivity Fusion Pruning computes a differential sensitivity score for each filter by fusing the discrepancies among gradient based sensitivity, first order Taylor expansion, and KL divergence of activation distributions. An exponential scaling mechanism is applied to emphasize filters with inconsistent importance across metrics, identifying candidates that are structurally unstable or less critical to the model performance. Unlike iterative or reinforcement learning based pruning strategies, Differential Sensitivity Fusion Pruning is efficient and deterministic, requiring only a single forward-backward pass for scoring and pruning. Extensive experiments across varying pruning rates between 50 to 70 percent demonstrate that Differential Sensitivity Fusion Pruning significantly reduces model complexity, achieving over 80 percent Floating point Operations Per Seconds reduction while maintaining high accuracy. For instance, at 70 percent pruning, our approach retains up to 98.23 percent of baseline accuracy, surpassing traditional heuristics in both compression and generalization. The proposed method presents an effective solution for scalable and adaptive Deep Convolutional Neural Networks compression, paving the way for efficient deployment on edge and mobile platforms.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 18:55:45 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chowdhury",
"Iftekhar Haider",
""
],
[
"Syed",
"Zaed Ikbal",
""
],
[
"Dhrubo",
"Ahmed Faizul Haque",
""
],
[
"Qayum",
"Mohammad Abdul",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.