id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.14451 | Aditya Shourya | Aditya Shourya, Michel Dumontier, Chang Sun | Adapting Lightweight Vision Language Models for Radiological Visual Question Answering | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent advancements in vision-language systems have improved the accuracy of Radiological Visual Question Answering (VQA) Models. However, some challenges remain across each stage of model development: limited expert-labeled images hinders data procurement at scale; the intricate and nuanced patterns of radiological images make modeling inherently difficult; and the lack of evaluation evaluation efforts makes it difficult to identify cases where the model might be ill-conditioned. In this study, we fine-tune a lightweight 3B parameter vision-language model for Radiological VQA, demonstrating that small models, when appropriately tuned with curated data, can achieve robust performance across both open- and closed-ended questions. We propose a cost-effective training pipeline from synthetic question-answer pair generation to multi-stage fine-tuning on specialised radiological domain-targeted datasets (e.g., ROCO v2.0, MedPix v2.0). Our results show that despite operating at a fraction of the scale of state-of-the-art models such as LLaVA-Med, our model achieves promising performance given its small parameter size and the limited scale of training data. We introduce a lightweight saliency-based diagnostic tool that enables domain experts to inspect VQA model performance and identify ill-conditioned failure modes through saliency analysis. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:15:08 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shourya",
"Aditya",
""
],
[
"Dumontier",
"Michel",
""
],
[
"Sun",
"Chang",
""
]
] |
2506.14452 | Yongtao Liu | Yongtao Liu, Anton V. Ievlev, Eugene A. Eliseev, Nana Sun, Kazuki Okamoto, Hiroshi Funakubo, Anna N. Morozovska, Sergei Kalinin | Polarization switching on the open surfaces of the wurtzite ferroelectric nitrides: ferroelectric subsystems and electrochemical reactivity | 22 pages; 7 figures | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Binary ferroelectric nitrides are promising materials for information technologies and power electronics. However, polarization switching in these materials is highly unusual. From the structural perspective, polarization reversal is associated with the change of the effective polarity at the surfaces and interfaces from N-to-M terminated, suggesting strong coupling between ferroelectric and chemical phenomena. Phenomenologically, macroscopic studies demonstrate the presence of complex time dependent phenomena including wake-up. Here, we explore the polarization switching using the multidimensional high-resolution piezoresponse force microscopy (PFM) and spectroscopy, detecting both the evolution of induced ferroelectric domain, electromechanical response, and surface deformation during first-order reversal curve measurements. We demonstrate the presence of two weakly coupled ferroelectric subsystems and the bias-induced electrochemical reactivity. The observed behaviors are very similar to the recent studies of other wurtzite system but additionally include electrochemical reactivity, suggesting the universality of these behaviors for the wurtzite binary ferroelectrics. These studies suggest potential of high-resolution multimodal PFM spectroscopies to resolve complex coupled polarization dynamics in materials. Furthermore, these PFM based studies are fully consistent with the recent electron microscopy observations of the shark-teeth like ferroelectric domains in nitrides. Hence, we believe that these studies establish the universal phenomenological picture of polarization switching in binary wurtzite. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:15:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Yongtao",
""
],
[
"Ievlev",
"Anton V.",
""
],
[
"Eliseev",
"Eugene A.",
""
],
[
"Sun",
"Nana",
""
],
[
"Okamoto",
"Kazuki",
""
],
[
"Funakubo",
"Hiroshi",
""
],
[
"Morozovska",
"Anna N.",
""
],
[
"Kalinin",
"Sergei",
""
]
] |
2506.14453 | Matteo Torzoni | Matteo Torzoni, Domenico Maisto, Andrea Manzoni, Francesco Donnarumma, Giovanni Pezzulo, Alberto Corigliano | Active Digital Twins via Active Inference | null | null | null | null | cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Digital twins are transforming engineering and applied sciences by enabling real-time monitoring, simulation, and predictive analysis of physical systems and processes. However, conventional digital twins rely primarily on passive data assimilation, which limits their adaptability in uncertain and dynamic environments. This paper introduces the active digital twin paradigm, based on active inference. Active inference is a neuroscience-inspired, Bayesian framework for probabilistic reasoning and predictive modeling that unifies inference, decision-making, and learning under a unique, free energy minimization objective. By formulating the evolution of the active digital twin as a partially observable Markov decision process, the active inference agent continuously refines its generative model through Bayesian updates and forecasts future states and observations. Decision-making emerges from an optimization process that balances pragmatic exploitation (maximizing goal-directed utility) and epistemic exploration or information gain (actively resolving uncertainty). Actions are dynamically planned to minimize expected free energy, which quantifies both the divergence between predicted and preferred future observations, and the epistemic value of expected information gain about hidden states. This approach enables a new level of autonomy and resilience in digital twins, offering superior spontaneous exploration capabilities. The proposed framework is assessed on the health monitoring and predictive maintenance of a railway bridge. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:16:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Torzoni",
"Matteo",
""
],
[
"Maisto",
"Domenico",
""
],
[
"Manzoni",
"Andrea",
""
],
[
"Donnarumma",
"Francesco",
""
],
[
"Pezzulo",
"Giovanni",
""
],
[
"Corigliano",
"Alberto",
""
]
] |
2506.14454 | Miguel A. S. Pinto | Nuno Barros e S\'a, Miguel A. S. Pinto | Clarifying the relation between covariantly conserved currents and Noether's theorem | 22 pages; comments welcome | null | null | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | On one hand, proper conservation laws, that is, the existence of local functions of the fields whose divergence vanishes on-shell, are associated with global symmetries and are a consequence of Noether's first theorem. On the other hand, covariant conservation laws, i.e., the same type of laws but featuring a covariant divergence, are associated with local symmetries and are a consequence (under certain conditions) of Noether's second theorem. In this paper, we clarify under which conditions Noether's second theorem, which is mostly known for providing relations between the equations of motion of a theory that possesses some local symmetries, can be used to generate covariantly conservation laws. We also investigate the relation between the conserved current of a theory with global symmetry and the covariantly conserved current of the corresponding local symmetry. We treat these issues using only the tools of Calculus, in the fashion of Noether's original article, avoiding the somewhat complicated mathematical tools that plague much of the literature on this topic, thus making the article readable for the general physicist. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:16:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sá",
"Nuno Barros e",
""
],
[
"Pinto",
"Miguel A. S.",
""
]
] |
2506.14455 | Ricardo Ruiz Baier | Neela Nataraj, Ricardo Ruiz-Baier, Aamir Yousuf | Unified numerical analysis for thermoelastic diffusion and thermo-poroelasticity of thin plates | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | We investigate a coupled hyperbolic-parabolic system modeling thermoelastic diffusion (resp. thermo-poroelasticity) in plates, consisting of a fourth-order hyperbolic partial differential equation for plate deflection and two second-order parabolic partial differential equations for the first moments of temperature and chemical potential (resp. pore pressure). The unique solvability of the system is established via Galerkin approach, and the additional regularity of the solution is obtained under appropriately strengthened data. For numerical approximation, we employ the Newmark method for time discretization of the hyperbolic term and a continuous interior penalty scheme for the spatial discretization of displacement. For the parabolic equations that represent the first moments of temperature and chemical potential (resp. pore pressure), we use the Crank--Nicolson method for time discretization and conforming finite elements for spatial discretization. The convergence of the fully discrete scheme with quasi-optimal rates in space and time is established. The numerical experiments demonstrate the effectiveness of the 2D Kirchhoff--Love plate model in capturing thermoelastic diffusion and thermo-poroelastic behavior in specific materials. We illustrate that as plate thickness decreases, the two-dimensional simulations closely approximate the results of three-dimensional problem. Finally, the numerical experiments also validate the theoretical rates of convergence. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:16:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nataraj",
"Neela",
""
],
[
"Ruiz-Baier",
"Ricardo",
""
],
[
"Yousuf",
"Aamir",
""
]
] |
2506.14456 | Elija Perrier | Elija Perrier | Hamiltonian Formalism for Comparing Quantum and Classical Intelligence | This is the version accepted at AGI 25 (camera ready length limit of 10 pages plus references and appendices). Further work detailing bounds and limitations is in preparation. Comments and criticisms welcome | null | null | null | quant-ph cs.AI | http://creativecommons.org/licenses/by/4.0/ | The prospect of AGI instantiated on quantum substrates motivates the development of mathematical frameworks that enable direct comparison of their operation in classical and quantum environments. To this end, we introduce a Hamiltonian formalism for describing classical and quantum AGI tasks as a means of contrasting their interaction with the environment. We propose a decomposition of AGI dynamics into Hamiltonian generators for core functions such as induction, reasoning, recursion, learning, measurement, and memory. This formalism aims to contribute to the development of a precise mathematical language for how quantum and classical agents differ via environmental interaction. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:17:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Perrier",
"Elija",
""
]
] |
2506.14457 | Freya Behrens | Freya Behrens, Lenka Zdeborov\'a | Dataset distillation for memorized data: Soft labels can leak held-out teacher knowledge | 9 pages, 21 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Dataset distillation aims to compress training data into fewer examples via a teacher, from which a student can learn effectively. While its success is often attributed to structure in the data, modern neural networks also memorize specific facts, but if and how such memorized information is can transferred in distillation settings remains less understood. In this work, we show that students trained on soft labels from teachers can achieve non-trivial accuracy on held-out memorized data they never directly observed. This effect persists on structured data when the teacher has not generalized.To analyze it in isolation, we consider finite random i.i.d. datasets where generalization is a priori impossible and a successful teacher fit implies pure memorization. Still, students can learn non-trivial information about the held-out data, in some cases up to perfect accuracy. In those settings, enough soft labels are available to recover the teacher functionally - the student matches the teacher's predictions on all possible inputs, including the held-out memorized data. We show that these phenomena strongly depend on the temperature with which the logits are smoothed, but persist across varying network capacities, architectures and dataset compositions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:18:37 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Behrens",
"Freya",
""
],
[
"Zdeborová",
"Lenka",
""
]
] |
2506.14458 | Xiaofu Peng | Xiao-Fu Peng | Macroscopicity of quantum superposition | 18 pages, 3 figures, 2 tables | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | In this paper, we define a universal measure of macroscopicity $\beta$ in the form of, ({\em experimentally observed coherence time})/({\em characteristic time required to repeatably distinguish the components of a quantum superposition state}), to divide the quantum/classical (Q-C) boundary. The analysis shows that, for a matter-wave interferometer, covering systems from electron to macromolecular, the measure of its macroscopicity is expressed by $\beta=p\theta d/8\hbar$, and it consistently yields $\beta<1$ by investigating the coherence data reported in the literature. The result implies that limited advancement has been made toward exploring the Q-C boundary in such experiments. Additionally, for a Ramsey-like interferometer, the measure of its macroscopicity is expressed as $\beta=8\pi^3\alpha\nu_0r_s^2\nu_s^2T/ (9c^2)$. After analyzing the reported experimental data, we find that a maximum value of $\beta \sim 300$ is achieved in the vertical atomic fountain experiments. Therefore, searching the Q-C boundary in the atomic fountain may be a more proper direction. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:18:51 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Peng",
"Xiao-Fu",
""
]
] |
2506.14459 | Mahin Montasir Afif | Md. Mortuza Ahmmed, Abdullah Al Noman, Mahin Montasir Afif, K. M. Tahsin Kabir, Md. Mostafizur Rahman, Mufti Mahmud | A Model-Mediated Stacked Ensemble Approach for Depression Prediction Among Professionals | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Depression is a significant mental health concern, particularly in professional environments where work-related stress, financial pressure, and lifestyle imbalances contribute to deteriorating well-being. Despite increasing awareness, researchers and practitioners face critical challenges in developing accurate and generalizable predictive models for mental health disorders. Traditional classification approaches often struggle with the complexity of depression, as it is influenced by multifaceted, interdependent factors, including occupational stress, sleep patterns, and job satisfaction. This study addresses these challenges by proposing a stacking-based ensemble learning approach to improve the predictive accuracy of depression classification among professionals. The Depression Professional Dataset has been collected from Kaggle. The dataset comprises demographic, occupational, and lifestyle attributes that influence mental well-being. Our stacking model integrates multiple base learners with a logistic regression-mediated model, effectively capturing diverse learning patterns. The experimental results demonstrate that the proposed model achieves high predictive performance, with an accuracy of 99.64% on training data and 98.75% on testing data, with precision, recall, and F1-score all exceeding 98%. These findings highlight the effectiveness of ensemble learning in mental health analytics and underscore its potential for early detection and intervention strategies. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:19:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ahmmed",
"Md. Mortuza",
""
],
[
"Noman",
"Abdullah Al",
""
],
[
"Afif",
"Mahin Montasir",
""
],
[
"Kabir",
"K. M. Tahsin",
""
],
[
"Rahman",
"Md. Mostafizur",
""
],
[
"Mahmud",
"Mufti",
""
]
] |
2506.14460 | Junbin Qiu | Junbin Qiu, Zhengpeng Xie, Xiangda Yan, Yongjie Yang, Yao Shu | Zeroth-Order Optimization is Secretly Single-Step Policy Optimization | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Zeroth-Order Optimization (ZOO) provides powerful tools for optimizing functions where explicit gradients are unavailable or expensive to compute. However, the underlying mechanisms of popular ZOO methods, particularly those employing randomized finite differences, and their connection to other optimization paradigms like Reinforcement Learning (RL) are not fully elucidated. This paper establishes a fundamental and previously unrecognized connection: ZOO with finite differences is equivalent to a specific instance of single-step Policy Optimization (PO). We formally unveil that the implicitly smoothed objective function optimized by common ZOO algorithms is identical to a single-step PO objective. Furthermore, we show that widely used ZOO gradient estimators, are mathematically equivalent to the REINFORCE gradient estimator with a specific baseline function, revealing the variance-reducing mechanism in ZOO from a PO perspective.Built on this unified framework, we propose ZoAR (Zeroth-Order Optimization with Averaged Baseline and Query Reuse), a novel ZOO algorithm incorporating PO-inspired variance reduction techniques: an averaged baseline from recent evaluations and query reuse analogous to experience replay. Our theoretical analysis further substantiates these techniques reduce variance and enhance convergence. Extensive empirical studies validate our theory and demonstrate that ZoAR significantly outperforms other methods in terms of convergence speed and final performance. Overall, our work provides a new theoretical lens for understanding ZOO and offers practical algorithmic improvements derived from its connection to PO. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:20:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Qiu",
"Junbin",
""
],
[
"Xie",
"Zhengpeng",
""
],
[
"Yan",
"Xiangda",
""
],
[
"Yang",
"Yongjie",
""
],
[
"Shu",
"Yao",
""
]
] |
2506.14461 | Fangting Zhou | Fangting Zhou, Bal\'azs Kulcs\'ar and Jiaming Wu | Collaborative Charging Scheduling via Balanced Bounding Box Methods | null | null | null | null | math.OC cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Electric mobility faces several challenges, most notably the high cost of infrastructure development and the underutilization of charging stations. The concept of shared charging offers a promising solution. The paper explores sustainable urban logistics through horizontal collaboration between two fleet operators and addresses a scheduling problem for the shared use of charging stations. To tackle this, the study formulates a collaborative scheduling problem as a bi-objective nonlinear integer programming model, in which each company aims to minimize its own costs, creating inherent conflicts that require trade-offs. The Balanced Bounding Box Methods (B3Ms) are introduced in order to efficiently derive the efficient frontier, identifying a reduced set of representative solutions. These methods enhance computational efficiency by selectively disregarding closely positioned and competing solutions, preserving the diversity and representativeness of the solutions over the efficient frontier. To determine the final solution and ensure balanced collaboration, cooperative bargaining methods are applied. Numerical case studies demonstrate the viability and scalability of the developed methods, showing that the B3Ms can significantly reduce computational time while maintaining the integrity of the frontier. These methods, along with cooperative bargaining, provide an effective framework for solving various bi-objective optimization problems, extending beyond the collaborative scheduling problem presented here. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:20:26 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhou",
"Fangting",
""
],
[
"Kulcsár",
"Balázs",
""
],
[
"Wu",
"Jiaming",
""
]
] |
2506.14462 | Luca Pignatelli | Riccardo Cristoferi, Luca Pignatelli | Phase separation in multiply periodic materials with fine microstructures | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a Cahn-Hilliard model for phase separation in composite materials with multiple periodic microstructures. These are modeled by considering a highly oscillating potential. The focus of this paper is in the case where the scales of the microstructures are smaller than that of phase separation. We provide a compactness result and prove that the {\Gamma}-limit of the energy is a multiple of the perimeter. In particular, using the recently introduced unfolding operator for multiple scales, we show that the taking the limit of all of the scales together is equivalent to taking one limit at the time, starting from the smaller scale and keeping the larger fixed. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:20:45 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Cristoferi",
"Riccardo",
""
],
[
"Pignatelli",
"Luca",
""
]
] |
2506.14463 | Mourad Choulli | Mourad Choulli | Upper bound on the multiplicity of eigenvalues of the Sch\"odinger-Dirichlet operator in dimension two | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish an upper bound on the multiplicity of eigenvalues of the Sch\"odinger-Dirichlet operator in dimension two. We give a proof based on a generalized Morse Lemma due to Cheng \cite{Ch}. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:25:02 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Choulli",
"Mourad",
""
]
] |
2506.14464 | Maximilian Baronig | Maximilian Baronig, Yeganeh Bahariasl, Ozan \"Ozdenizci, Robert Legenstein | A Scalable Hybrid Training Approach for Recurrent Spiking Neural Networks | null | null | null | null | cs.NE cs.AI cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Recurrent spiking neural networks (RSNNs) can be implemented very efficiently in neuromorphic systems. Nevertheless, training of these models with powerful gradient-based learning algorithms is mostly performed on standard digital hardware using Backpropagation through time (BPTT). However, BPTT has substantial limitations. It does not permit online training and its memory consumption scales linearly with the number of computation steps. In contrast, learning methods using forward propagation of gradients operate in an online manner with a memory consumption independent of the number of time steps. These methods enable SNNs to learn from continuous, infinite-length input sequences. Yet, slow execution speed on conventional hardware as well as inferior performance has hindered their widespread application. In this work, we introduce HYbrid PRopagation (HYPR) that combines the efficiency of parallelization with approximate online forward learning. Our algorithm yields high-throughput online learning through parallelization, paired with constant, i.e., sequence length independent, memory demands. HYPR enables parallelization of parameter update computation over the sub sequences for RSNNs consisting of almost arbitrary non-linear spiking neuron models. We apply HYPR to networks of spiking neurons with oscillatory subthreshold dynamics. We find that this type of neuron model is particularly well trainable by HYPR, resulting in an unprecedentedly low task performance gap between approximate forward gradient learning and BPTT. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:27:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Baronig",
"Maximilian",
""
],
[
"Bahariasl",
"Yeganeh",
""
],
[
"Özdenizci",
"Ozan",
""
],
[
"Legenstein",
"Robert",
""
]
] |
2506.14465 | Francesco Giuli Dr. | Francesco Giuli | Precision and rare ElectroWeak processes | 11 pages, proceeding of the Thirteenth Annual Large Hadron Collider Physics (LHCP2025) | null | null | null | hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This proceeding summarizes recent measurements of electroweak boson and diboson production. The results, based on data from LHC Run 2 and Run 3 at $\sqrt{s}=13$ and $13.6\,\mathrm{TeV}$, probe the Standard Model with unprecedented precision. These include inclusive and differential cross-section measurements of $W$, $Z$, $WZ$, $WW$, and $ZZ$ final states, constraints on Parton Distribution Functions, and interpretations within Effective Field Theories. A consistent picture with theory predictions emerges, laying the groundwork for future precision studies. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:28:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Giuli",
"Francesco",
""
]
] |
2506.14466 | Gao Xingan | Xingan Gao and Xiaobing Sun and Sicong Cao and Kaifeng Huang and Di Wu and Xiaolei Liu and Xingwei Lin and Yang Xiang | MalGuard: Towards Real-Time, Accurate, and Actionable Detection of Malicious Packages in PyPI Ecosystem | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Malicious package detection has become a critical task in ensuring the security and stability of the PyPI. Existing detection approaches have focused on advancing model selection, evolving from traditional machine learning (ML) models to large language models (LLMs). However, as the complexity of the model increases, the time consumption also increases, which raises the question of whether a lightweight model achieves effective detection. Through empirical research, we demonstrate that collecting a sufficiently comprehensive feature set enables even traditional ML models to achieve outstanding performance. However, with the continuous emergence of new malicious packages, considerable human and material resources are required for feature analysis. Also, traditional ML model-based approaches lack of explainability to malicious packages.Therefore, we propose a novel approach MalGuard based on graph centrality analysis and the LIME (Local Interpretable Model-agnostic Explanations) algorithm to detect malicious packages.To overcome the above two challenges, we leverage graph centrality analysis to extract sensitive APIs automatically to replace manual analysis. To understand the sensitive APIs, we further refine the feature set using LLM and integrate the LIME algorithm with ML models to provide explanations for malicious packages. We evaluated MalGuard against six SOTA baselines with the same settings. Experimental results show that our proposed MalGuard, improves precision by 0.5%-33.2% and recall by 1.8%-22.1%. With MalGuard, we successfully identified 113 previously unknown malicious packages from a pool of 64,348 newly-uploaded packages over a five-week period, and 109 out of them have been removed by the PyPI official. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:30:56 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Gao",
"Xingan",
""
],
[
"Sun",
"Xiaobing",
""
],
[
"Cao",
"Sicong",
""
],
[
"Huang",
"Kaifeng",
""
],
[
"Wu",
"Di",
""
],
[
"Liu",
"Xiaolei",
""
],
[
"Lin",
"Xingwei",
""
],
[
"Xiang",
"Yang",
""
]
] |
2506.14467 | Cecilia Morales | Nico Zevallos, Cecilia G. Morales, Andrew Orekhov, Tejas Rane, Hernando Gomez, Francis X. Guyette, Michael R. Pinsky, John Galeotti, Artur Dubrawski, Howie Choset | Automatic Cannulation of Femoral Vessels in a Porcine Shock Model | 2 pages, 2 figures, conference | Hamlyn Symposium on Medical Robotics 2025 | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Rapid and reliable vascular access is critical in trauma and critical care. Central vascular catheterization enables high-volume resuscitation, hemodynamic monitoring, and advanced interventions like ECMO and REBOA. While peripheral access is common, central access is often necessary but requires specialized ultrasound-guided skills, posing challenges in prehospital settings. The complexity arises from deep target vessels and the precision needed for needle placement. Traditional techniques, like the Seldinger method, demand expertise to avoid complications. Despite its importance, ultrasound-guided central access is underutilized due to limited field expertise. While autonomous needle insertion has been explored for peripheral vessels, only semi-autonomous methods exist for femoral access. This work advances toward full automation, integrating robotic ultrasound for minimally invasive emergency procedures. Our key contribution is the successful femoral vein and artery cannulation in a porcine hemorrhagic shock model. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:31:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zevallos",
"Nico",
""
],
[
"Morales",
"Cecilia G.",
""
],
[
"Orekhov",
"Andrew",
""
],
[
"Rane",
"Tejas",
""
],
[
"Gomez",
"Hernando",
""
],
[
"Guyette",
"Francis X.",
""
],
[
"Pinsky",
"Michael R.",
""
],
[
"Galeotti",
"John",
""
],
[
"Dubrawski",
"Artur",
""
],
[
"Choset",
"Howie",
""
]
] |
2506.14468 | Xinglong Mao | Xinglong Mao, Shifeng Liu, Sirui Zhao, Tong Xu, Enhong Chen | MERba: Multi-Receptive Field MambaVision for Micro-Expression Recognition | null | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Micro-expressions (MEs) are brief, involuntary facial movements that reveal genuine emotions, holding significant potential in psychological diagnosis and criminal investigations. Despite notable advances in automatic ME recognition (MER), existing methods still struggle to jointly capture localized muscle activations and global facial dependencies, both critical for recognizing subtle emotional cues. To tackle this challenge, we propose MERba, a novel multi-receptive field architecture tailored for MER. MERba introduces a series of Local-Global Feature Integration stages, where fine-grained motion features are first extracted by local extractors containing MambaVision Mixers within non-overlapping windows, and then global dependencies across these regions are modeled via lightweight self-attention layers. This hierarchical design enables a progressive transition from localized perception to holistic facial understanding. Furthermore, we introduce an asymmetric multi-scanning strategy to eliminate redundant scanning directions and enhance local spatial perception. To address the high inter-class similarity among negative MEs, we introduce a Dual-Granularity Classification Module that decouples the recognition process into a coarse-to-fine paradigm. Experiments on two benchmark MER datasets demonstrate that MERba outperforms existing methods, with ablation studies confirming the effectiveness of each proposed component. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:31:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mao",
"Xinglong",
""
],
[
"Liu",
"Shifeng",
""
],
[
"Zhao",
"Sirui",
""
],
[
"Xu",
"Tong",
""
],
[
"Chen",
"Enhong",
""
]
] |
2506.14469 | Jared Miller | Jared Miller, Maitraya Avadhut Desai, Xiuqiang He, Roy S. Smith, Gabriela Hug | Network-Independent Incremental Passivity Conditions for Grid-Forming Inverter Control | 7 pages, 4 figures | null | null | null | eess.SY cs.SY math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Grid-forming inverters control the power transfer between the AC and DC sides of an electrical grid while maintaining the frequency and voltage of the AC side. This paper focuses on ensuring large-signal stability of an electrical grid with inverter-interfaced renewable sources. We prove that the Hybrid-Angle Control (HAC) scheme for grid-forming inverters can exhibit incremental passivity properties between current and voltage at both the AC and DC ports. This incremental passivity can be certified through decentralized conditions. Inverters operating under HAC can, therefore, be connected to other passive elements (e.g. transmission lines) with an immediate guarantee of global transient stability regardless of the network topology or parameters. Passivity of Hybrid Angle Control is also preserved under small-signal (linearized) analyses, in contrast to conventional proportional droop laws that are passivity-short at low frequencies. Passivity and interconnected-stability properties are demonstrated through an example case study. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:31:52 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Miller",
"Jared",
""
],
[
"Desai",
"Maitraya Avadhut",
""
],
[
"He",
"Xiuqiang",
""
],
[
"Smith",
"Roy S.",
""
],
[
"Hug",
"Gabriela",
""
]
] |
2506.14470 | Zixian Zhang | Zixian Zhang, Takfarinas Saber | AST-Enhanced or AST-Overloaded? The Surprising Impact of Hybrid Graph Representations on Code Clone Detection | null | null | null | null | cs.AI cs.SE | http://creativecommons.org/licenses/by-sa/4.0/ | As one of the most detrimental code smells, code clones significantly increase software maintenance costs and heighten vulnerability risks, making their detection a critical challenge in software engineering. Abstract Syntax Trees (ASTs) dominate deep learning-based code clone detection due to their precise syntactic structure representation, but they inherently lack semantic depth. Recent studies address this by enriching AST-based representations with semantic graphs, such as Control Flow Graphs (CFGs) and Data Flow Graphs (DFGs). However, the effectiveness of various enriched AST-based representations and their compatibility with different graph-based machine learning techniques remains an open question, warranting further investigation to unlock their full potential in addressing the complexities of code clone detection. In this paper, we present a comprehensive empirical study to rigorously evaluate the effectiveness of AST-based hybrid graph representations in Graph Neural Network (GNN)-based code clone detection. We systematically compare various hybrid representations ((CFG, DFG, Flow-Augmented ASTs (FA-AST)) across multiple GNN architectures. Our experiments reveal that hybrid representations impact GNNs differently: while AST+CFG+DFG consistently enhances accuracy for convolution- and attention-based models (Graph Convolutional Networks (GCN), Graph Attention Networks (GAT)), FA-AST frequently introduces structural complexity that harms performance. Notably, GMN outperforms others even with standard AST representations, highlighting its superior cross-code similarity detection and reducing the need for enriched structures. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:35:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Zixian",
""
],
[
"Saber",
"Takfarinas",
""
]
] |
2506.14471 | Yikang Zhou | Yikang Zhou, Tao Zhang, Dizhe Zhang, Shunping Ji, Xiangtai Li, Lu Qi | Dense360: Dense Understanding from Omnidirectional Panoramas | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal Large Language Models (MLLMs) require comprehensive visual inputs to achieve dense understanding of the physical world. While existing MLLMs demonstrate impressive world understanding capabilities through limited field-of-view (FOV) visual inputs (e.g., 70 degree), we take the first step toward dense understanding from omnidirectional panoramas. We first introduce an omnidirectional panoramas dataset featuring a comprehensive suite of reliability-scored annotations. Specifically, our dataset contains 160K panoramas with 5M dense entity-level captions, 1M unique referring expressions, and 100K entity-grounded panoramic scene descriptions. Compared to multi-view alternatives, panoramas can provide more complete, compact, and continuous scene representations through equirectangular projections (ERP). However, the use of ERP introduces two key challenges for MLLMs: i) spatial continuity along the circle of latitude, and ii) latitude-dependent variation in information density. We address these challenges through ERP-RoPE, a position encoding scheme specifically designed for panoramic ERP. In addition, we introduce Dense360-Bench, the first benchmark for evaluating MLLMs on omnidirectional captioning and grounding, establishing a comprehensive framework for advancing dense visual-language understanding in panoramic settings. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:35:23 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhou",
"Yikang",
""
],
[
"Zhang",
"Tao",
""
],
[
"Zhang",
"Dizhe",
""
],
[
"Ji",
"Shunping",
""
],
[
"Li",
"Xiangtai",
""
],
[
"Qi",
"Lu",
""
]
] |
2506.14472 | Fabien Bernier | Fabien Bernier, Maxime Cordy, Yves Le Traon | Leveraging External Factors in Household-Level Electrical Consumption Forecasting using Hypernetworks | ECML PKDD 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Accurate electrical consumption forecasting is crucial for efficient energy management and resource allocation. While traditional time series forecasting relies on historical patterns and temporal dependencies, incorporating external factors -- such as weather indicators -- has shown significant potential for improving prediction accuracy in complex real-world applications. However, the inclusion of these additional features often degrades the performance of global predictive models trained on entire populations, despite improving individual household-level models. To address this challenge, we found that a hypernetwork architecture can effectively leverage external factors to enhance the accuracy of global electrical consumption forecasting models, by specifically adjusting the model weights to each consumer.
We collected a comprehensive dataset spanning two years, comprising consumption data from over 6000 luxembourgish households and corresponding external factors such as weather indicators, holidays, and major local events. By comparing various forecasting models, we demonstrate that a hypernetwork approach outperforms existing methods when associated to external factors, reducing forecasting errors and achieving the best accuracy while maintaining the benefits of a global model. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:35:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bernier",
"Fabien",
""
],
[
"Cordy",
"Maxime",
""
],
[
"Traon",
"Yves Le",
""
]
] |
2506.14473 | Zhijing Wan | Zhijing Wan, Zhixiang Wang, Zheng Wang, Xin Xu, Shin'ichi Satoh | Foundation Model Insights and a Multi-Model Approach for Superior Fine-Grained One-shot Subset Selection | 18 pages, 10 figures, accepted by ICML 2025 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One-shot subset selection serves as an effective tool to reduce deep learning training costs by identifying an informative data subset based on the information extracted by an information extractor (IE). Traditional IEs, typically pre-trained on the target dataset, are inherently dataset-dependent. Foundation models (FMs) offer a promising alternative, potentially mitigating this limitation. This work investigates two key questions: (1) Can FM-based subset selection outperform traditional IE-based methods across diverse datasets? (2) Do all FMs perform equally well as IEs for subset selection? Extensive experiments uncovered surprising insights: FMs consistently outperform traditional IEs on fine-grained datasets, whereas their advantage diminishes on coarse-grained datasets with noisy labels. Motivated by these finding, we propose RAM-APL (RAnking Mean-Accuracy of Pseudo-class Labels), a method tailored for fine-grained image datasets. RAM-APL leverages multiple FMs to enhance subset selection by exploiting their complementary strengths. Our approach achieves state-of-the-art performance on fine-grained datasets, including Oxford-IIIT Pet, Food-101, and Caltech-UCSD Birds-200-2011. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:37:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wan",
"Zhijing",
""
],
[
"Wang",
"Zhixiang",
""
],
[
"Wang",
"Zheng",
""
],
[
"Xu",
"Xin",
""
],
[
"Satoh",
"Shin'ichi",
""
]
] |
2506.14474 | Eyal German | Eyal German, Sagiv Antebi, Edan Habler, Asaf Shabtai, Yuval Elovici | LexiMark: Robust Watermarking via Lexical Substitutions to Enhance Membership Verification of an LLM's Textual Training Data | null | null | null | null | cs.CL cs.CR | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) can be trained or fine-tuned on data obtained without the owner's consent. Verifying whether a specific LLM was trained on particular data instances or an entire dataset is extremely challenging. Dataset watermarking addresses this by embedding identifiable modifications in training data to detect unauthorized use. However, existing methods often lack stealth, making them relatively easy to detect and remove. In light of these limitations, we propose LexiMark, a novel watermarking technique designed for text and documents, which embeds synonym substitutions for carefully selected high-entropy words. Our method aims to enhance an LLM's memorization capabilities on the watermarked text without altering the semantic integrity of the text. As a result, the watermark is difficult to detect, blending seamlessly into the text with no visible markers, and is resistant to removal due to its subtle, contextually appropriate substitutions that evade automated and manual detection. We evaluated our method using baseline datasets from recent studies and seven open-source models: LLaMA-1 7B, LLaMA-3 8B, Mistral 7B, Pythia 6.9B, as well as three smaller variants from the Pythia family (160M, 410M, and 1B). Our evaluation spans multiple training settings, including continued pretraining and fine-tuning scenarios. The results demonstrate significant improvements in AUROC scores compared to existing methods, underscoring our method's effectiveness in reliably verifying whether unauthorized watermarked data was used in LLM training. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:41:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"German",
"Eyal",
""
],
[
"Antebi",
"Sagiv",
""
],
[
"Habler",
"Edan",
""
],
[
"Shabtai",
"Asaf",
""
],
[
"Elovici",
"Yuval",
""
]
] |
2506.14475 | Alexandros Kehagias | Alex Kehagias and Antonio Riotto | The AdS Perspective on the Nonlinear Tails in Black Hole Ringdown | 7 pages | null | null | null | gr-qc astro-ph.CO hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | Black holes gradually settle into their static configuration by emitting gravitational waves, whose amplitude diminish over time according to a power-law decay at fixed spatial locations. We show that the nonlinear tails in the presence of a quadratic source, which have been recently found to potentially dominate over the linear ones, can be simply derived from the AdS$_2$$\times$S$^2$ spacetime perspective with their amplitudes being related to the Aretakis constants. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:43:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kehagias",
"Alex",
""
],
[
"Riotto",
"Antonio",
""
]
] |
2506.14476 | Ziyue Lin | Ziyue Lin and Yi Shan and Lin Gao and Xinghua Jia and Siming Chen | SimSpark: Interactive Simulation of Social Media Behaviors | 32 pages, 7 figures | Proc. ACM Hum.-Comput. Interact. 9, 2, Article CSCW168 (April 2025), 32 pages | 10.1145/3711066 | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding user behaviors on social media has garnered significant scholarly attention, enhancing our comprehension of how virtual platforms impact society and empowering decision-makers. Simulating social media behaviors provides a robust tool for capturing the patterns of social media behaviors, testing hypotheses, and predicting the effects of various interventions, ultimately contributing to a deeper understanding of social media environments. Moreover, it can overcome difficulties associated with utilizing real data for analysis, such as data accessibility issues, ethical concerns, and the complexity of processing large and heterogeneous datasets. However, researchers and stakeholders need more flexible platforms to investigate different user behaviors by simulating different scenarios and characters, which is not possible yet. Therefore, this paper introduces SimSpark, an interactive system including simulation algorithms and interactive visual interfaces which is capable of creating small simulated social media platforms with customizable characters and social environments. We address three key challenges: generating believable behaviors, validating simulation results, and supporting interactive control for generation and results analysis. A simulation workflow is introduced to generate believable behaviors of agents by utilizing large language models. A visual interface enables real-time parameter adjustment and process monitoring for customizing generation settings. A set of visualizations and interactions are also designed to display the models' outputs for further analysis. Effectiveness is evaluated through case studies, quantitative simulation model assessments, and expert interviews. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:50:07 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lin",
"Ziyue",
""
],
[
"Shan",
"Yi",
""
],
[
"Gao",
"Lin",
""
],
[
"Jia",
"Xinghua",
""
],
[
"Chen",
"Siming",
""
]
] |
2506.14477 | Jingqi Yang | Jingqi Yang, Zhilong Song, Jiawei Chen, Mingli Song, Sheng Zhou, linjun sun, Xiaogang Ouyang, Chun Chen, Can Wang | GUI-Robust: A Comprehensive Dataset for Testing GUI Agent Robustness in Real-World Anomalies | 10 pages, 4 figures, submitted to NIPS 2025 | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | The development of high-quality datasets is crucial for benchmarking and advancing research in Graphical User Interface (GUI) agents. Despite their importance, existing datasets are often constructed under idealized conditions, overlooking the diverse anomalies frequently encountered in real-world deployments. To address this limitation, we introduce GUI-Robust, a novel dataset designed for comprehensive GUI agent evaluation, explicitly incorporating seven common types of anomalies observed in everyday GUI interactions. Furthermore, we propose a semi-automated dataset construction paradigm that collects user action sequences from natural interactions via RPA tools and then generate corresponding step and task descriptions for these actions with the assistance of MLLMs. This paradigm significantly reduces annotation time cost by a factor of over 19 times. Finally, we assess state-of-the-art GUI agents using the GUI-Robust dataset, revealing their substantial performance degradation in abnormal scenarios. We anticipate that our work will highlight the importance of robustness in GUI agents and inspires more future research in this direction. The dataset and code are available at https://github.com/chessbean1/GUI-Robust.. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:50:35 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yang",
"Jingqi",
""
],
[
"Song",
"Zhilong",
""
],
[
"Chen",
"Jiawei",
""
],
[
"Song",
"Mingli",
""
],
[
"Zhou",
"Sheng",
""
],
[
"sun",
"linjun",
""
],
[
"Ouyang",
"Xiaogang",
""
],
[
"Chen",
"Chun",
""
],
[
"Wang",
"Can",
""
]
] |
2506.14478 | Anthony Sanchez | Ko W. Ohm, Anthony Sanchez | Quantitative finiteness of hyperplanes in hybrid manifolds | 42 pages | null | null | null | math.DS math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove a quantitative finiteness theorem for the number of totally geodesic hyperplanes of non-arithmetic hyperbolic $n$-manifolds that arise from a gluing construction of Gromov and Piatetski-Shapiro for $n\ge3$. This extends work of Lindenstrauss-Mohammadi in dimension 3. This follows from effective density theorem for periodic orbits of $\mathrm{SO}(n-1,1)$ acting on quotients of $\mathrm{SO}(n,1)$ by a lattice for $n\ge3$. The effective density result uses a number of a ideas including Margulis functions, a restricted projection theorem, and an effective equidistribution result for measures on the horospherical subgroup that are nearly full dimensional. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:57:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ohm",
"Ko W.",
""
],
[
"Sanchez",
"Anthony",
""
]
] |
2506.14479 | Wonyoung Kim | Wonyoung Kim | Adaptive Data Augmentation for Thompson Sampling | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In linear contextual bandits, the objective is to select actions that maximize cumulative rewards, modeled as a linear function with unknown parameters. Although Thompson Sampling performs well empirically, it does not achieve optimal regret bounds. This paper proposes a nearly minimax optimal Thompson Sampling for linear contextual bandits by developing a novel estimator with the adaptive augmentation and coupling of the hypothetical samples that are designed for efficient parameter learning. The proposed estimator accurately predicts rewards for all arms without relying on assumptions for the context distribution. Empirical results show robust performance and significant improvement over existing methods. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:57:33 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kim",
"Wonyoung",
""
]
] |
2506.14480 | Alexander M\"uller-Hermes | Francesca La Piana, Alexander M\"uller-Hermes | Annihilating and breaking Lorentz cone entanglement | 27 pages | null | null | null | quant-ph math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Linear maps between finite-dimensional ordered vector spaces with orders induced by proper cones $C_A$ and $C_B$ are called entanglement breaking if their partial application sends the maximal tensor product $K\otimes_{\max} C_A$ into the minimal tensor product $K\otimes_{\min} C_B$ for any proper cone $K$. We study the larger class of Lorentz-entanglement breaking maps where $K$ is restricted to be a Lorentz cone of any dimension, i.e., any cone over a Euclidean ball. This class of maps appeared recently in the study of asymptotic entanglement annihilation and it is dual to the linear maps factoring through Lorentz cones. Our main results establish connections between these classes of maps and operator ideals studied in the theory of Banach spaces. For operators $u:X\rightarrow Y$ between finite-dimensional normed spaces $X$ and $Y$ we consider so-called central maps which are positive with respect to the cones $C_A=C_X$ and $C_B=C_Y$. We show how to characterize when such a map factors through a Lorentz cone and when it is Lorentz-entanglement breaking by using the Hilbert-space factorization norm $\gamma_2$ and its dual $\gamma^*_2$. We also study the class of Lorentz-entanglement annihilating maps whose local application sends the Lorentzian tensor product $C_A\otimes_{L} C_A$ into the minimal tensor product $C_B\otimes_{\min} C_B$. When $C_A$ is a cone over a finite-dimensional normed space and $C_B$ is a Lorentz cone itself, the central maps of this kind can be characterized by the $2$-summing norm $\pi_2$. Finally, we prove interesting connections between these classes of maps for general cones, and we identify examples with particular properties, e.g., cones with an analogue of the $2$-summing property. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:58:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"La Piana",
"Francesca",
""
],
[
"Müller-Hermes",
"Alexander",
""
]
] |
2506.14481 | Sofie K\"olling | Sofie K\"olling, \.Inan\c{c} Adagideli, Alexander Brinkman | Signature of current-induced nuclear spin polarization in (Bi$_{1-x}$Sb$_{x}$)$_2$Te$_3$ | 7 + 8 pages, 3 + 4 figures | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In systems with spin-momentum locking, such as the surface states of three-dimensional topological insulators, a charge current is spin-polarized and spin-flip interactions between electron and nuclear spins can transfer this polarization to the nuclear spin system. When a nonzero bias voltage is applied, the nuclear polarization reaches a steady-state value. This polarization emerges as an effective in-plane magnetic field acting on electrons, called the Overhauser field, which causes an offset in-plane magnetoresistance perpendicular to the current, visible in experiments. The in-plane offset is measured in the three-dimensional topological insulator \bsttight, and the magnitude of the magnetic field offset is compared to the Overhauser field. We attribute the observed magnetic field offset to current-induced nuclear polarization in \bsttight, which forms an important step towards experimentally realizing an entropic inductor. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:58:18 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kölling",
"Sofie",
""
],
[
"Adagideli",
"İnanç",
""
],
[
"Brinkman",
"Alexander",
""
]
] |
2506.14482 | Eli Visbal | Eli Visbal, Greg L. Bryan, Zoltan Haiman | Chemical Signatures of Population III Stars in Damped Lyman-$\alpha$ Absorption Systems at $z \approx 6$ | 18 pages, 6 figures, submitted to JCAP | null | null | null | astro-ph.GA astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | Recently, Sodini et al. (2024) presented a sample of OI damped Lyman-$\alpha$ absorption system (DLA) analogs at $z\sim6$ that contain possible chemical signatures of Population III (Pop III) stars. In this paper, we use an N-body simulation-based semi-analytic model of the first stars and galaxies to predict the impact of Pop III stars on high-redshift DLAs. These Pop III DLA predictions are the first to include a number of important physical effects such as Lyman-Werner (LW) feedback, reionization, and external metal enrichment (all of which account for three-dimensional spatial fluctuations caused by halo clustering). We predict the abundance of DLAs as a function of their carbon-to-oxygen ratios ([C/O]). We find that our fiducial model is strongly ruled out by the data as it contains too few high-[C/O] DLAs, which have metals primarily from Pop III stars. However, increasing the delay time between Pop III and metal-enriched star formation due to supernovae feedback leads to better agreement with the data. Our results suggest that DLA analogs at $z\sim6$ are a promising probe of Pop III star formation for two key reasons. First, for reasonable parameter choices there are significant numbers of DLAs with metals primarily originating from Pop III stars. Second, we find that the number of DLAs with substantial Pop III contributions depends strongly on the Pop III star formation efficiency and the delay time between Pop III and metal-enriched star formation. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 12:59:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Visbal",
"Eli",
""
],
[
"Bryan",
"Greg L.",
""
],
[
"Haiman",
"Zoltan",
""
]
] |
2506.14483 | C\'ecile Bouette | C. Bouette, L. Luzzi, M. Bloch | Covert Capacity of AWGN Channels under Average Error Probability | To be presented at ISIT 2025 | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We derive upper and lower bounds for the covert capacity of Additive White Gaussian Noise channels when measuring reliability in terms of the average error probability and covertness in terms of Kullback-Leibler divergence. This characterization confirms the absence of strong converse for this setting in both the reliability and covertness parameters. The crux of our approach is to analyze a codebook of BPSK-modulated codewords carefully augmented with "all-zero" codewords. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:00:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bouette",
"C.",
""
],
[
"Luzzi",
"L.",
""
],
[
"Bloch",
"M.",
""
]
] |
2506.14484 | Ritwick Sarkar | Ritwick Sarkar, Sreya Chatterjee, and Urna Basu | From Attraction to Repulsion: Emergent Interactions in Harmonically Coupled Active Binary System | 23 pages, 10 figures | null | null | null | cond-mat.stat-mech cond-mat.soft | http://creativecommons.org/licenses/by/4.0/ | We investigate the emergent interactions between two active Brownian particles coupled by an attractive harmonic potential and in contact with a thermal reservoir. By analyzing the stationary distribution of their separation, we demonstrate that the effective interaction can be either attractive or repulsive, depending on the interplay between activity, coupling strength, and temperature. Notably, we find that an effective short-range repulsion emerges in the strong and moderate-coupling regimes, when the temperature is below some threshold value, which we characterize analytically. In the strong-coupling regime, the repulsion emerges due to the difference in the self-propulsion speeds of the particles. We also compute the short-time position distribution of the centroid of the coupled particles, which shows strongly non-Gaussian fluctuations at low temperatures. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:05:34 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sarkar",
"Ritwick",
""
],
[
"Chatterjee",
"Sreya",
""
],
[
"Basu",
"Urna",
""
]
] |
2506.14485 | Sascha Rechenberger | Sascha Rechenberger and Thom Fr\"uhwirth | Optimized Execution of FreeCHR | This is a preprint of a paper submitted to the 39th Workshop on (Constraint and Functional) Logic Programming (WLP 2025) | null | null | null | cs.PL | http://creativecommons.org/licenses/by/4.0/ | Constraint Handling Rules (CHR) is a rule-based programming language that rewrites collections of constraints. It is typically embedded into a general-purpose language. There exists a plethora of implementations for numerous host languages. However, the existing implementations often re-invent the method of embedding, which impedes maintenance and weakens assertions of correctness. To formalize and thereby unify the embedding of a ground subset of CHR into arbitrary host languages, we introduced the framework FreeCHR and proved it to be a valid representation of classical CHR. For the sake of simplicity, abstract implementations of our framework did not yet include a concrete matching algorithm nor optimizations. In this paper, we introduce an improved execution algorithm for FreeCHR. We also provide an evaluation of the algorithm via benchmarks which suggest the effectiveness of our implemented optimizations. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:07:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Rechenberger",
"Sascha",
""
],
[
"Frühwirth",
"Thom",
""
]
] |
2506.14486 | Tom\'a\v{s} Je\v{z}o | S. Devoto, T. Jezo, S. Kallweit, C. Schwan | MATRIX HAWAII: PineAPPL interpolation grids with MATRIX | 29 pages, 13 figures | null | null | MS-TP-25-08, ZU-TH 41/25 | hep-ph | http://creativecommons.org/licenses/by/4.0/ | We present an interface between PineAPPL and Matrix, which allows fully differential cross sections to be calculated in the form of interpolation grids, accurate at next-to-next-to-leading order (NNLO) in QCD and next-to-leading order in electroweak (EW) theory. This interface is the first publicly available tool to calculate interpolation grids at NNLO QCD accuracy for a wide set of processes. Interpolation grids provide the functionality to compute predictions for arbitrary parton distribution functions (PDFs) as well as PDF uncertainties without the need to repeat the actual calculation. Another important application of the these grids is to perform global analyses of PDFs using exact NNLO calculations instead of $K$-factors, which have several drawbacks. This exact treatment of NNLO corrections is also an important prerequisite for fitting PDFs at next-to-next-to-next-to-leading order level with reliable uncertainties. The new version of the Matrix code interfaced to PineAPPL, as well as the grids produced for this publication, are available on the Matrix website and on PloughShare, respectively. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:08:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Devoto",
"S.",
""
],
[
"Jezo",
"T.",
""
],
[
"Kallweit",
"S.",
""
],
[
"Schwan",
"C.",
""
]
] |
2506.14487 | Paolo Franceschi | Paolo Franceschi, Marco Faroni, Stefano Baraldo, Anna Valente | ros2 fanuc interface: Design and Evaluation of a Fanuc CRX Hardware Interface in ROS2 | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces the ROS2 control and the Hardware Interface (HW) integration for the Fanuc CRX- robot family. It explains basic implementation details and communication protocols, and its integration with the Moveit2 motion planning library. We conducted a series of experiments to evaluate relevant performances in the robotics field. We tested the developed ros2_fanuc_interface for four relevant robotics cases: step response, trajectory tracking, collision avoidance integrated with Moveit2, and dynamic velocity scaling, respectively. Results show that, despite a non-negligible delay between command and feedback, the robot can track the defined path with negligible errors (if it complies with joint velocity limits), ensuring collision avoidance. Full code is open source and available at https://github.com/paolofrance/ros2_fanuc_interface. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:08:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Franceschi",
"Paolo",
""
],
[
"Faroni",
"Marco",
""
],
[
"Baraldo",
"Stefano",
""
],
[
"Valente",
"Anna",
""
]
] |
2506.14488 | Dong Xu | Dong Xu, Zhangfan Yang, Ka-chun Wong, Zexuan Zhu, Jiangqiang Li and Junkai Ji | Reimagining Target-Aware Molecular Generation through Retrieval-Enhanced Aligned Diffusion | 13 pages, 5 figures | null | null | null | q-bio.BM cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Breakthroughs in high-accuracy protein structure prediction, such as AlphaFold, have established receptor-based molecule design as a critical driver for rapid early-phase drug discovery. However, most approaches still struggle to balance pocket-specific geometric fit with strict valence and synthetic constraints. To resolve this trade-off, a Retrieval-Enhanced Aligned Diffusion termed READ is introduced, which is the first to merge molecular Retrieval-Augmented Generation with an SE(3)-equivariant diffusion model. Specifically, a contrastively pre-trained encoder aligns atom-level representations during training, then retrieves graph embeddings of pocket-matched scaffolds to guide each reverse-diffusion step at inference. This single mechanism can inject real-world chemical priors exactly where needed, producing valid, diverse, and shape-complementary ligands. Experimental results demonstrate that READ can achieve very competitive performance in CBGBench, surpassing state-of-the-art generative models and even native ligands. That suggests retrieval and diffusion can be co-optimized for faster, more reliable structure-based drug design. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:09:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Dong",
""
],
[
"Yang",
"Zhangfan",
""
],
[
"Wong",
"Ka-chun",
""
],
[
"Zhu",
"Zexuan",
""
],
[
"Li",
"Jiangqiang",
""
],
[
"Ji",
"Junkai",
""
]
] |
2506.14489 | Jonas Sander | Felix Maurer, Jonas Sander, Thomas Eisenbarth | ReDASH: Fast and efficient Scaling in Arithmetic Garbled Circuits for Secure Outsourced Inference | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | ReDash extends Dash's arithmetic garbled circuits to provide a more flexible and efficient framework for secure outsourced inference. By introducing a novel garbled scaling gadget based on a generalized base extension for the residue number system, ReDash removes Dash's limitation of scaling exclusively by powers of two. This enables arbitrary scaling factors drawn from the residue number system's modular base, allowing for tailored quantization schemes and more efficient model evaluation.
Through the new $\text{ScaleQuant}^+$ quantization mechanism, ReDash supports optimized modular bases that can significantly reduce the overhead of arithmetic operations during convolutional neural network inference. ReDash achieves up to a 33-fold speedup in overall inference time compared to Dash Despite these enhancements, ReDash preserves the robust security guarantees of arithmetic garbling. By delivering both performance gains and quantization flexibility, ReDash expands the practicality of garbled convolutional neural network inference. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:09:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Maurer",
"Felix",
""
],
[
"Sander",
"Jonas",
""
],
[
"Eisenbarth",
"Thomas",
""
]
] |
2506.14490 | Solomiya Mizyuk | Solomiya Mizyuk | Virtual invariants of Quot schemes of points on threefolds | 49 pages, comments welcome | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct an almost perfect obstruction theory of virtual dimension zero on the Quot scheme parametrizing zero-dimensional quotients of a locally free sheaf on a smooth projective $3$-fold. This gives a virtual class in degree zero and therefore allows one to define virtual invariants of the Quot scheme. We compute these invariants proving a conjecture by Ricolfi. The computation is done by reducing to the toric case via cobordism theory and a degeneration argument. The toric case is solved by reducing to the computation on the Quot scheme of points on $\mathbb{A}^3$ via torus localization, the torus-equivariant Siebert's formula for almost perfect obstruction theories and the torus-equivariant Jouanolou trick. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:10:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mizyuk",
"Solomiya",
""
]
] |
2506.14491 | Wojciech Julian Pasek | Wojciech Julian Pasek, Soufiane Hajji, Oussama Tata, Laurent Cerutti, Fernando Gonzalez-posada Flores, Simon Hurand, Abdelouahed El Fatimy and Thierry Taliercio | Multisubband Plasmons in InAs/GaSb Broken Gap Quantum Well | null | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-subband plasmon (MSP) modes in heavily doped InAs/GaSb broken-gap quantum wells grown via molecular beam epitaxy (MBE) are investigated. An $8$-band $\vec{k} \cdot \vec{p}$ semiclassical model accurately predicts ellipsometric spectra, reflecting strong subband hybridization and non-parabolicity. In contrast, single-band plasmon models show qualitative discrepancies with experiment, even with adjusted effective masses. These findings highlight the potential of broken-gap wells for quantum technologies leveraging interband coupling and wavefunction hybridization. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:11:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pasek",
"Wojciech Julian",
""
],
[
"Hajji",
"Soufiane",
""
],
[
"Tata",
"Oussama",
""
],
[
"Cerutti",
"Laurent",
""
],
[
"Flores",
"Fernando Gonzalez-posada",
""
],
[
"Hurand",
"Simon",
""
],
[
"Fatimy",
"Abdelouahed El",
""
],
[
"Taliercio",
"Thierry",
""
]
] |
2506.14492 | Jan Hermann | Sebastian Ehlert, Jan Hermann, Thijs Vogels, Victor Garcia Satorras, Stephanie Lanius, Marwin Segler, Derk P. Kooi, Kenji Takeda, Chin-Wei Huang, Giulia Luise, Rianne van den Berg, Paola Gori-Giorgi, Amir Karton | Accurate Chemistry Collection: Coupled cluster atomization energies for broad chemical space | 9 pages plus references, 6 figures, dataset on Zenodo | null | null | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate thermochemical data with sub-chemical accuracy (i.e., within $\pm$1 kcal mol$^{-1}$ from sufficiently accurate experimental or theoretical reference data) is essential for the development and improvement of computational chemistry methods. Challenging thermochemical properties such as heats of formation and total atomization energies (TAEs) are of particular interest because they rigorously test the ability of computational chemistry methods to accurately describe complex chemical transformations involving multiple bond rearrangements. Yet, existing thermochemical datasets that confidently reach this level of accuracy are limited in either size or scope. Datasets with highly accurate reference values include a small number of data points, and larger datasets provide less accurate data or only cover a narrow portion of the chemical space. The existing datasets are therefore insufficient for developing data-driven methods with predictive accuracy over a large chemical space. The Microsoft Research Accurate Chemistry Collection (MSR-ACC) will address this challenge. Here, it offers the MSR-ACC/TAE25 dataset of 76,879 total atomization energies obtained at the CCSD(T)/CBS level via the W1-F12 thermochemical protocol. The dataset is constructed to exhaustively cover chemical space for all elements up to argon by enumerating and sampling chemical graphs, thus avoiding bias towards any particular subspace of the chemical space (such as drug-like, organic, or experimentally observed molecules). With this first dataset in MSR-ACC, we enable data-driven approaches for developing predictive computational chemistry methods with unprecedented accuracy and scope. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:12:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ehlert",
"Sebastian",
""
],
[
"Hermann",
"Jan",
""
],
[
"Vogels",
"Thijs",
""
],
[
"Satorras",
"Victor Garcia",
""
],
[
"Lanius",
"Stephanie",
""
],
[
"Segler",
"Marwin",
""
],
[
"Kooi",
"Derk P.",
""
],
[
"Takeda",
"Kenji",
""
],
[
"Huang",
"Chin-Wei",
""
],
[
"Luise",
"Giulia",
""
],
[
"Berg",
"Rianne van den",
""
],
[
"Gori-Giorgi",
"Paola",
""
],
[
"Karton",
"Amir",
""
]
] |
2506.14493 | Jiyuan Fu | Jiyuan Fu, Kaixun Jiang, Lingyi Hong, Jinglun Li, Haijing Guo, Dingkang Yang, Zhaoyu Chen, Wenqiang Zhang | LingoLoop Attack: Trapping MLLMs via Linguistic Context and State Entrapment into Endless Loops | null | null | null | null | cs.CL cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal Large Language Models (MLLMs) have shown great promise but require substantial computational resources during inference. Attackers can exploit this by inducing excessive output, leading to resource exhaustion and service degradation. Prior energy-latency attacks aim to increase generation time by broadly shifting the output token distribution away from the EOS token, but they neglect the influence of token-level Part-of-Speech (POS) characteristics on EOS and sentence-level structural patterns on output counts, limiting their efficacy. To address this, we propose LingoLoop, an attack designed to induce MLLMs to generate excessively verbose and repetitive sequences. First, we find that the POS tag of a token strongly affects the likelihood of generating an EOS token. Based on this insight, we propose a POS-Aware Delay Mechanism to postpone EOS token generation by adjusting attention weights guided by POS information. Second, we identify that constraining output diversity to induce repetitive loops is effective for sustained generation. We introduce a Generative Path Pruning Mechanism that limits the magnitude of hidden states, encouraging the model to produce persistent loops. Extensive experiments demonstrate LingoLoop can increase generated tokens by up to 30 times and energy consumption by a comparable factor on models like Qwen2.5-VL-3B, consistently driving MLLMs towards their maximum generation limits. These findings expose significant MLLMs' vulnerabilities, posing challenges for their reliable deployment. The code will be released publicly following the paper's acceptance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:14:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fu",
"Jiyuan",
""
],
[
"Jiang",
"Kaixun",
""
],
[
"Hong",
"Lingyi",
""
],
[
"Li",
"Jinglun",
""
],
[
"Guo",
"Haijing",
""
],
[
"Yang",
"Dingkang",
""
],
[
"Chen",
"Zhaoyu",
""
],
[
"Zhang",
"Wenqiang",
""
]
] |
2506.14494 | Zahra Mobini Mobini | Zahra Mobini, Ahmet Hasim Gokceoglu, Li Wang, Gunnar Peters and Hien Quoc Ngo | Fronthaul-Aware User-Centric Generalized Cell-Free Massive MIMO Systems | null | null | null | null | cs.IT eess.SP math.IT | http://creativecommons.org/licenses/by/4.0/ | We consider fronthaul-limited generalized zeroforcing-based cell-free massive multiple-input multiple-output (CF-mMIMO) systems with multiple-antenna users and multipleantenna access points (APs) relying on both cooperative beamforming (CB) and user-centric (UC) clustering. The proposed framework is very general and can be degenerated into different special cases, such as pure CB/pure UC clustering, or fully centralized CB/fully distributed beamforming. We comprehensively analyze the spectral efficiency (SE) performance of the system wherein the users use the minimum mean-squared errorbased successive interference cancellation (MMSE-SIC) scheme to detect the desired signals. Specifically, we formulate an optimization problem for the user association and power control for maximizing the sum SE. The formulated problem is under per-AP transmit power and fronthaul constraints, and is based on only long-term channel state information (CSI). The challenging formulated problem is transformed into tractable form and a novel algorithm is proposed to solve it using minorization maximization (MM) technique. We analyze the trade-offs provided by the CF-mMIMO system with different number of CB clusters, hence highlighting the importance of the appropriate choice of CB design for different system setups. Numerical results show that for the centralized CB, the proposed power optimization provides nearly 59% improvement in the average sum SE over the heuristic approach, and 312% improvement, when the distributed beamforming is employed. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:15:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mobini",
"Zahra",
""
],
[
"Gokceoglu",
"Ahmet Hasim",
""
],
[
"Wang",
"Li",
""
],
[
"Peters",
"Gunnar",
""
],
[
"Ngo",
"Hien Quoc",
""
]
] |
2506.14495 | Yu Qi | Yu Qi, Lipeng Gu, Honghua Chen, Liangliang Nan, and Mingqiang Wei | I Speak and You Find: Robust 3D Visual Grounding with Noisy and Ambiguous Speech Inputs | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing 3D visual grounding methods rely on precise text prompts to locate objects within 3D scenes. Speech, as a natural and intuitive modality, offers a promising alternative. Real-world speech inputs, however, often suffer from transcription errors due to accents, background noise, and varying speech rates, limiting the applicability of existing 3DVG methods. To address these challenges, we propose \textbf{SpeechRefer}, a novel 3DVG framework designed to enhance performance in the presence of noisy and ambiguous speech-to-text transcriptions. SpeechRefer integrates seamlessly with xisting 3DVG models and introduces two key innovations. First, the Speech Complementary Module captures acoustic similarities between phonetically related words and highlights subtle distinctions, generating complementary proposal scores from the speech signal. This reduces dependence on potentially erroneous transcriptions. Second, the Contrastive Complementary Module employs contrastive learning to align erroneous text features with corresponding speech features, ensuring robust performance even when transcription errors dominate. Extensive experiments on the SpeechRefer and peechNr3D datasets demonstrate that SpeechRefer improves the performance of existing 3DVG methods by a large margin, which highlights SpeechRefer's potential to bridge the gap between noisy speech inputs and reliable 3DVG, enabling more intuitive and practical multimodal systems. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:17:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Qi",
"Yu",
""
],
[
"Gu",
"Lipeng",
""
],
[
"Chen",
"Honghua",
""
],
[
"Nan",
"Liangliang",
""
],
[
"Wei",
"Mingqiang",
""
]
] |
2506.14496 | Melanie Schranz | Muhammad Atta Ur Rahman, Melanie Schranz | LLM-Powered Swarms: A New Frontier or a Conceptual Stretch? | This is the author's version of a paper submitted to IEEE Intelligent Systems. 6 Tables, 3 Figures | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Swarm intelligence traditionally refers to systems of simple, decentralized agents whose local interactions lead to emergent, collective behavior. Recently, the term 'swarm' has been extended to describe AI systems like OpenAI's Swarm, where large language models (LLMs) act as collaborative agents. This paper contrasts traditional swarm algorithms with LLM-driven swarms exploring how decentralization, scalability, and emergence are redefined in modern artificial intelligence (AI). We implement and compare both paradigms using Boids and Ant Colony Optimization (ACO), evaluating latency, resource usage, and behavioral accuracy. The suitability of both cloud-based and local LLMs is assessed for the agent-based use in swarms. Although LLMs offer powerful reasoning and abstraction capabilities, they introduce new constraints in computation and coordination that challenge traditional notions of swarm design. This study highlights the opportunities and limitations of integrating LLMs into swarm systems and discusses the evolving definition of 'swarm' in modern AI research. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:18:34 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Rahman",
"Muhammad Atta Ur",
""
],
[
"Schranz",
"Melanie",
""
]
] |
2506.14497 | Franco Matzkin | Franco Matzkin, Agostina Larrazabal, Diego H Milone, Jose Dolz, Enzo Ferrante | Towards Reliable WMH Segmentation under Domain Shift: An Application Study using Maximum Entropy Regularization to Improve Uncertainty Estimation | 32 pages, 7 figures | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Accurate segmentation of white matter hyperintensities (WMH) is crucial for clinical decision-making, particularly in the context of multiple sclerosis. However, domain shifts, such as variations in MRI machine types or acquisition parameters, pose significant challenges to model calibration and uncertainty estimation. This study investigates the impact of domain shift on WMH segmentation by proposing maximum-entropy regularization techniques to enhance model calibration and uncertainty estimation, with the purpose of identifying errors post-deployment using predictive uncertainty as a proxy measure that does not require ground-truth labels. To do this, we conducted experiments using a U-Net architecture to evaluate these regularization schemes on two publicly available datasets, assessing performance with the Dice coefficient, expected calibration error, and entropy-based uncertainty estimates. Our results show that entropy-based uncertainty estimates can anticipate segmentation errors, and that maximum-entropy regularization further strengthens the correlation between uncertainty and segmentation performance while also improving model calibration under domain shift. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:21:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Matzkin",
"Franco",
""
],
[
"Larrazabal",
"Agostina",
""
],
[
"Milone",
"Diego H",
""
],
[
"Dolz",
"Jose",
""
],
[
"Ferrante",
"Enzo",
""
]
] |
2506.14498 | Jacob Pedersen | Henry Thompson, Matteo Bonanomi, Jacob Pedersen, Oksana Plekan, Nitish Pal, Cesare Grazioli, Kevin C. Prince, Bruno N. C. Tenorio, Michele Devetta, Davide Faccial\`a, Caterina Vozzi, Paolo Piseri, Miltcho B. Danailov, Alexander Demidovich, Alexander D. Brynes, Alberto Simoncig, Marco Zangrando, Marcello Coreno, Raimund Feifel, Richard J. Squibb, David M. P. Holland, Felix Allum, Daniel Rolles, Piero Decleva, Michael S. Schuurman, Ruharid J. G. Forbes, Sonia Coriani, Carlo Callegari, Russell S. Minns, Michele Di Fraia | Shake-down spectroscopy as state- and site-specific probe of ultrafast chemical dynamics | null | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Tracking the multifarious ultrafast electronic and structural changes occurring in a molecule during a photochemical transformation is a challenging endeavor that benefits from recent experimental and computational progress in time-resolved techniques. Measurements of valence electronic states, which provide a global picture of the bonding structure of the molecule, and core electronic states, which provide insight into the local environment, traditionally require different approaches and are often studied separately. Here, we demonstrate that X-ray pulses from a seeded free-electron laser (FEL) enable the measurement of high-resolution, time-resolved X-ray photoelectron spectra (XPS) that capture weak satellite states resulting from shake-down processes in a valence-excited molecule. This approach effectively combines the advantages of both valence- and core-state investigations. We applied this method to investigate photoexcited CS$_2$ molecules, where the role of internal conversion (IC) and intersystem crossing (ISC) in determining the pre-dissociation dynamics is controversial. We present XPS spectra from photoexcited CS$_2$, obtained at the FERMI FEL. High-resolution measurements, compared to the corresponding spectra obtained from accurate multireference quantum chemical calculations, reveal that shake-down satellite channels are highly sensitive to both valence electronic and geometric changes. Previous studies of the pre-dissociation dynamics have led to uncertain assignments around the branching between singlet and triplet excited states. We derived a propensity rule that demonstrates the spin-selectivity of the shake-downs. This selectivity allows us to unequivocally assign contributions from the bright and dark singlet excited states with populations tracked along the pre-dissociation dynamic pathway. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:23:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Thompson",
"Henry",
""
],
[
"Bonanomi",
"Matteo",
""
],
[
"Pedersen",
"Jacob",
""
],
[
"Plekan",
"Oksana",
""
],
[
"Pal",
"Nitish",
""
],
[
"Grazioli",
"Cesare",
""
],
[
"Prince",
"Kevin C.",
""
],
[
"Tenorio",
"Bruno N. C.",
""
],
[
"Devetta",
"Michele",
""
],
[
"Faccialà",
"Davide",
""
],
[
"Vozzi",
"Caterina",
""
],
[
"Piseri",
"Paolo",
""
],
[
"Danailov",
"Miltcho B.",
""
],
[
"Demidovich",
"Alexander",
""
],
[
"Brynes",
"Alexander D.",
""
],
[
"Simoncig",
"Alberto",
""
],
[
"Zangrando",
"Marco",
""
],
[
"Coreno",
"Marcello",
""
],
[
"Feifel",
"Raimund",
""
],
[
"Squibb",
"Richard J.",
""
],
[
"Holland",
"David M. P.",
""
],
[
"Allum",
"Felix",
""
],
[
"Rolles",
"Daniel",
""
],
[
"Decleva",
"Piero",
""
],
[
"Schuurman",
"Michael S.",
""
],
[
"Forbes",
"Ruharid J. G.",
""
],
[
"Coriani",
"Sonia",
""
],
[
"Callegari",
"Carlo",
""
],
[
"Minns",
"Russell S.",
""
],
[
"Di Fraia",
"Michele",
""
]
] |
2506.14499 | Pablo Oliveira | P. R. A de Oliveira, I. Coelho, G. Felix, P. Venezuela, F. Stavale | In situ growth of a type-II ZnO/ZnS heterostructure:From stability to band-offset | 5 Figures. Supporting information available under request | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/publicdomain/zero/1.0/ | We have successfully obtained a ZnO/ZnS heterostructure by heating a ZnS(001) single crystal in a controlled impurities-free oxygen atmosphere. Combining X-ray photoelectron spectroscopy (XPS), and atomic force microscopy (AFM), we explore the stability, electronic structure, and morphology of that interface. Our XPS measurements reveal a binding energy shift of the core-level peaks, indicating a band-bending effect due to the formation of a hybrid ZnO/ZnS interface. In addition, AFM measurements show that exposure of ZnS single-crystal to an oxygen atmosphere leads to the formation of ZnO/ZnS-like islands. Interestingly, our band-offset estimation suggest a type-II heterostructure arrangement with suitable electronic edges positions that turn ZnO/ZnS heterostructure a promising platform for catalytic applications, particularly hydrogen and oxygen evolution reactions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:24:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"de Oliveira",
"P. R. A",
""
],
[
"Coelho",
"I.",
""
],
[
"Felix",
"G.",
""
],
[
"Venezuela",
"P.",
""
],
[
"Stavale",
"F.",
""
]
] |
2506.14500 | Zhen Chen | Zi-Yang Zhan and Zhen Chen | Versatile SPH Open Boundary Conditions for Multiphase Flows in Extreme Condition | 42 pages, 30 figures | null | null | null | physics.flu-dyn physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although the Smoothed Particle Hydrodynamics (SPH) method has been demonstrated as a promising numerical solver for multiphase flow problems due to its Lagrangian nature, its application to complex channel flow may encounter additional issues such as the open boundary condition and numerical instability in the extreme flow state. The present work aims to establish a general SPH algorithm for accurate and stable simulation of complex multiphase flows in configurations with open boundaries. The general scheme of weakly compressible SPH is adopted with special treatments implemented to ease the numerical oscillation in the density discontinuity scenario, and the turbulent model is implemented for interpretations of extreme flow conditions in high Reynolds numbers. Then, the conventional open boundary condition is fine-tuned by two new algorithms to guarantee the numerical stability in the inflow and the outflow regions. Firstly, a density relaxation is proposed to alleviate the pressure instability in the inflow region, which improves the smoothness of the particle pre-processing procedure. Secondly, the particle shifting technique with adaptive damper is implemented to adjust the magnitude of correction in the outflow region, which helps to suppress the velocity oscillation near the outlet. Validations of the proposed algorithms are carried out through four classic numerical examples, presenting appealing agreements with the analytical solutions and thus demonstrating the versatility in various flow conditions. Then, the robustness of the method is established through the turbulent multiphase channel flow cases and the horizontal slug channel flow with large density ratios. These results shed light on the value of the proposed algorithm as a general solver for complex multiphase flow problems with open boundaries. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:26:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhan",
"Zi-Yang",
""
],
[
"Chen",
"Zhen",
""
]
] |
2506.14501 | Wei Chen | Luis F. C\'ardenas-Castillo, Shuai Zhang, Fernando L. Freire Jr., Wei Chen | Absorbance marker: Detection of quantum geometry and spread of Wannier function in disordered 2D semiconductors | 8 pages, 2 figures | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | The optical absorbance of 2D semiconductors is generalized to individual lattice sites through the topological marker formalism, yielding an absorbance marker. This marker allows to investigate the atomic scale variation of absorbance caused by impurities, thereby quantifies the influence of disorder on the quantum geometry and the spread of Wannier functions of valence band states. Applying this marker to transition metal dichalcogenides reveals a very localized suppression of absorbance caused by potential impurities, rendering a reduction of absorbance in the macroscopic scale proportional to the impurity density, in good agreement with the experimental results of plasma-treated WS$_{2}$. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:27:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Cárdenas-Castillo",
"Luis F.",
""
],
[
"Zhang",
"Shuai",
""
],
[
"Freire",
"Fernando L.",
"Jr."
],
[
"Chen",
"Wei",
""
]
] |
2506.14502 | Junru Yu | Xiao Wang, Junru Yu, Jun Huang, Qiong Wu, Ljubo Vacic, Changyin Sun | Toward Safety-First Human-Like Decision Making for Autonomous Vehicles in Time-Varying Traffic Flow | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the recent advancements in artificial intelligence technologies have shown great potential in improving transport efficiency and safety, autonomous vehicles(AVs) still face great challenge of driving in time-varying traffic flow, especially in dense and interactive situations. Meanwhile, human have free wills and usually do not make the same decisions even situate in the exactly same scenarios, leading to the data-driven methods suffer from poor migratability and high search cost problems, decreasing the efficiency and effectiveness of the behavior policy. In this research, we propose a safety-first human-like decision-making framework(SF-HLDM) for AVs to drive safely, comfortably, and social compatiblely in effiency. The framework integrates a hierarchical progressive framework, which combines a spatial-temporal attention (S-TA) mechanism for other road users' intention inference, a social compliance estimation module for behavior regulation, and a Deep Evolutionary Reinforcement Learning(DERL) model for expanding the search space efficiently and effectively to make avoidance of falling into the local optimal trap and reduce the risk of overfitting, thus make human-like decisions with interpretability and flexibility. The SF-HLDM framework enables autonomous driving AI agents dynamically adjusts decision parameters to maintain safety margins and adhering to contextually appropriate driving behaviors at the same time. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:28:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Xiao",
""
],
[
"Yu",
"Junru",
""
],
[
"Huang",
"Jun",
""
],
[
"Wu",
"Qiong",
""
],
[
"Vacic",
"Ljubo",
""
],
[
"Sun",
"Changyin",
""
]
] |
2506.14503 | Baris Bozkurt | Baris Bozkurt (College of Interdisciplinary Studies, Zayed University, Dubai, United Arab Emirates) | An Open Research Dataset of the 1932 Cairo Congress of Arab Music | 14 pages, 4 figures, 4 tables | null | null | null | cs.SD cs.DL eess.AS | http://creativecommons.org/licenses/by-sa/4.0/ | This paper introduces ORD-CC32 , an open research dataset derived from the 1932 Cairo Congress of Arab Music recordings, a historically significant collection representing diverse Arab musical traditions. The dataset includes structured metadata, melodic and rhythmic mode tags (maqam and iqa), manually labeled tonic information, and acoustic features extracted using state-of-the-art pitch detection methods. These resources support computational studies of tuning, temperament, and regional variations in Arab music. A case study using pitch histograms demonstrates the potential for data-driven analysis of microtonal differences across regions. By making this dataset openly available, we aim to enable interdisciplinary research in computational ethnomusicology, music information retrieval (MIR), cultural studies, and digital heritage preservation. ORD-CC32 is shared on Zenodo with tools for feature extraction and metadata retrieval. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:28:27 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bozkurt",
"Baris",
"",
"College of Interdisciplinary Studies, Zayed University, Dubai, United Arab Emirates"
]
] |
2506.14504 | Emmanuel Deruty | Emmanuel Deruty | Evolving music theory for emerging musical languages | In Music 2025, Innovation in Music Conference. 20-22 June, 2025, Bath Spa University, Bath, UK | null | null | null | cs.SD | http://creativecommons.org/licenses/by/4.0/ | This chapter reconsiders the concept of pitch in contemporary popular music (CPM), particularly in electronic contexts where traditional assumptions may fail. Drawing on phenomenological and inductive methods, it argues that pitch is not an ontologically objective property but a perceptual construct shaped by listeners and conditions. Analyses of quasi-harmonic tones reveal that a single tone can convey multiple pitches, giving rise to tonal fission. The perception of pitch may also be multistable, varying for the same listener over time. In this framework, the tuning system may emerge from a tone's internal structure. A parallel with the coastline paradox supports a model of pitch grounded in perceptual variability, challenging inherited theoretical norms. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:29:07 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Deruty",
"Emmanuel",
""
]
] |
2506.14505 | George Georgiou | George Georgiou, Georgios Linardopoulos and Dimitrios Zoakos | Holography of a novel codimension-2 defect CFT | 27 pages, 3 figures | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We propose and study a new holographic duality between a non-supersymmetric defect conformal field theory (dCFT) and its gravity dual. On the gravity side, the defect is realised by a novel solution of a D5 probe brane embedded in the $AdS_5\times S^5$ geometry. The D5 brane wraps an $S^2\subset S^5$ and accommodates $k$ units of flux through the $S^2$. The symmetry of the induced on the brane metric is $AdS_3\times S^1\times S^2$. The brane ends on an $\mathbb{R}^{(1,1)}$ subspace of the 4-dimensional $AdS_5$ boundary resulting to a codimension-2 defect. We first prove that our brane configuration is stable by showing that the masses of all the fluctuations of the transverse to the brane coordinates respect the B-F bound. On the field theory side, the 2-dimensional defect is described by a classical solution whose precise form we determine. Subsequently, we calculate the one-point functions of the energy-momentum tensor and of the chiral primary operators (CPOs), first at strong and then at weak coupling. In an appropriate limit, we find compelling agreement between the weak and strong coupling results. Furthermore, we also extract one of the B-type Weyl anomaly coefficients. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:29:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Georgiou",
"George",
""
],
[
"Linardopoulos",
"Georgios",
""
],
[
"Zoakos",
"Dimitrios",
""
]
] |
2506.14506 | Shantanu Desai | Yuva Himanshu Pallam, Shantanu Desai | An independent search for Jovian neutrinos using BOREXINO data | 16 pages, 14 figures | null | null | null | hep-ex astro-ph.EP astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | In a recent study, arXiv:2401.13043 found evidence for a 6% flux contribution from Jupiter to the total flux rate time series data from the BOREXINO solar neutrino experiment, specifically during the time intervals 2019-2021 and 2011-2013. The significance of this detection was estimated to be around $2\sigma$. We reanalyze the BOREXINO data and independently confirm the Jovian signal with the same amplitude and significance as that obtained in arXiv:2401.13043. However, using the same technique, we also find a spurious flux contribution from Venus and Saturn (at $\sim 2\sigma$ significance), whereas prima facie one should not expect any signal from any other planet. We then implement Bayesian model comparison to ascertain whether the BOREXINO data contain an additional contribution from Jupiter, Venus or Saturn. We find Bayes factors of less than five for an additional contribution from Jupiter, and less than or close to one for Venus and Saturn. This implies that the evidence for an additional contribution from Jupiter is very marginal. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:30:58 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pallam",
"Yuva Himanshu",
""
],
[
"Desai",
"Shantanu",
""
]
] |
2506.14507 | Nitesh Subedi | Nitesh Subedi, Adam Haroon, Shreyan Ganguly, Samuel T.K. Tetteh, Prajwal Koirala, Cody Fleming, Soumik Sarkar | Can Pretrained Vision-Language Embeddings Alone Guide Robot Navigation? | 6 figures, 2 tables, Accepted to Robotics: Science and Systems (RSS) 2025 Workshop on Robot Planning in the Era of Foundation Models (FM4RoboPlan) | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Foundation models have revolutionized robotics by providing rich semantic representations without task-specific training. While many approaches integrate pretrained vision-language models (VLMs) with specialized navigation architectures, the fundamental question remains: can these pretrained embeddings alone successfully guide navigation without additional fine-tuning or specialized modules? We present a minimalist framework that decouples this question by training a behavior cloning policy directly on frozen vision-language embeddings from demonstrations collected by a privileged expert. Our approach achieves a 74% success rate in navigation to language-specified targets, compared to 100% for the state-aware expert, though requiring 3.2 times more steps on average. This performance gap reveals that pretrained embeddings effectively support basic language grounding but struggle with long-horizon planning and spatial reasoning. By providing this empirical baseline, we highlight both the capabilities and limitations of using foundation models as drop-in representations for embodied tasks, offering critical insights for robotics researchers facing practical design tradeoffs between system complexity and performance in resource-constrained scenarios. Our code is available at https://github.com/oadamharoon/text2nav | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:31:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Subedi",
"Nitesh",
""
],
[
"Haroon",
"Adam",
""
],
[
"Ganguly",
"Shreyan",
""
],
[
"Tetteh",
"Samuel T. K.",
""
],
[
"Koirala",
"Prajwal",
""
],
[
"Fleming",
"Cody",
""
],
[
"Sarkar",
"Soumik",
""
]
] |
2506.14508 | Styliani Christina Fragkouli | Styliani-Christina Fragkouli, Somya Iqbal, Lisa Crossman, Barbara Gravel, Nagat Masued, Mark Onders, Devesh Haseja, Alex Stikkelman, Alfonso Valencia, Tom Lenaerts, Fotis Psomopoulos, Pilib \'O Broin, N\'uria Queralt-Rosinach, and Davide Cirillo | An ELIXIR scoping review on domain-specific evaluation metrics for synthetic data in life sciences | null | null | null | null | q-bio.OT | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Synthetic data has emerged as a powerful resource in life sciences, offering solutions for data scarcity, privacy protection and accessibility constraints. By creating artificial datasets that mirror the characteristics of real data, allows researchers to develop and validate computational methods in controlled environments. Despite its promise, the adoption of synthetic data in Life Sciences hinges on rigorous evaluation metrics designed to assess their fidelity and reliability. To explore the current landscape of synthetic data evaluation metrics in several Life Sciences domains, the ELIXIR Machine Learning Focus Group performed a systematic review of the scientific literature following the PRISMA guidelines. Six critical domains were examined to identify current practices for assessing synthetic data. Findings reveal that, while generation methods are rapidly evolving, systematic evaluation is often overlooked, limiting researchers ability to compare, validate, and trust synthetic datasets across different domains. This systematic review underscores the urgent need for robust, standardized evaluation approaches that not only bolster confidence in synthetic data but also guide its effective and responsible implementation. By laying the groundwork for establishing domain-specific yet interoperable standards, this scoping review paves the way for future initiatives aimed at enhancing the role of synthetic data in scientific discovery, clinical practice and beyond. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:31:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fragkouli",
"Styliani-Christina",
""
],
[
"Iqbal",
"Somya",
""
],
[
"Crossman",
"Lisa",
""
],
[
"Gravel",
"Barbara",
""
],
[
"Masued",
"Nagat",
""
],
[
"Onders",
"Mark",
""
],
[
"Haseja",
"Devesh",
""
],
[
"Stikkelman",
"Alex",
""
],
[
"Valencia",
"Alfonso",
""
],
[
"Lenaerts",
"Tom",
""
],
[
"Psomopoulos",
"Fotis",
""
],
[
"Broin",
"Pilib Ó",
""
],
[
"Queralt-Rosinach",
"Núria",
""
],
[
"Cirillo",
"Davide",
""
]
] |
2506.14509 | Michiel Smet | Michiel Smet | Hermitian Cubic norm structures and groups of relative rank one | 32 pages | null | null | null | math.GR math.RA | http://creativecommons.org/licenses/by/4.0/ | Hermitian cubic norm structures were recently introduced in order to study the class of skew-dimension one structurable algebras (which are typically only defined over fields of characteristic different from $2$ and $3$) over arbitrary rings and fields. Here, we generalize the quartic norm for these algebras and show that elements for which the quartic norm is invertible are conjugate invertible. This leads to the notion of a division hermitian cubic norm structure, defined as one in which the quartic norm is anisotropic. We classify such structures in terms of the Tits index of an associated (rank one) adjoint simple linear algebraic group and show that any adjoint linear algebraic group with such a Tits index defines a corresponding hermitian cubic norm structure. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:31:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Smet",
"Michiel",
""
]
] |
2506.14510 | Gautam Nagaraj | Gautam Nagaraj, Robin Ciardullo, Caryl Gronwall, Vandana Ramakrishnan, Kyoung-Soo Lee, Eric Gawiser, Nicole M. Firestone, Govind Ramgopal, J. Aguilar, Steven Ahlen, Davide Bianchi, David Brooks, Francisco Javier Castander, Todd Claybaugh, Andrei Cuceu, Axel de la Macorra, Arjun Dey, Biprateep Dey, Peter Doel, Jaime Forero-Romero, Enrique Gaztanaga, Satya Gontcho A Gontcho, Gaston Gutierrez, Hiram K. Herrera-Alcantar, Klaus Honscheid, Mustapha Ishak, Robert Kehoe, David Kirkby, T. Kisner, Anthony Kremin, Martin Landriau, Laurent Le Guillou, Michael Levi, Christophe Magneville, Marc Manera, Paul Martini, Aaron M. Meisner, Ramon Miquel, John Moustakas, Nathalie Palanque-Delabrouille, Francisco Prada, Ignasi Perez-Rafols, Graziano Rossi, Lado Samushia, Eusebio Sanchez, David J. Schlegel, Michael F. Schubnell, Hee-Jong Seo, Joseph H. Silber, David Sprayberry, Gregory Tarle, Francisco Valdes, Benjamin A. Weaver, Martin White, Rongpu Zhou, Hu Zou | ODIN: The LAE Ly{\alpha} Luminosity Function over Cosmic Time and Environmental Density | 22 pages, 16 figures, 6 tables, submitted to ApJ | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The ubiquity and relative ease of discovery make $2\lesssim z\lesssim 5$ Ly$\alpha$ emitting galaxies (LAEs) ideal tracers for cosmology. In addition, because Ly$\alpha$ is a resonance line, but frequently observed at large equivalent width, it is potentially a probe of galaxy evolution. The LAE Ly$\alpha$ luminosity function (LF) is an essential measurement for making progress on both of these aspects. Although several studies have computed the LAE LF, very few have delved into how the function varies with environment. The large area and depth of the One-hundred-deg2 DECam Imaging in Narrowbands (ODIN) survey makes such measurements possible at the cosmic noon redshifts of z~2.4, ~3.1, and ~4.5. In this initial work, we present algorithms to rigorously compute the LAE LF and test our methods on the ~16,000 ODIN LAEs found in the extended COSMOS field. Using these limited samples, we find slight evidence that protocluster environments either suppress the numbers of very faint and very bright LAEs or enhance medium-bright LAEs in comparison to the field. We also find that the LF decreases in number density and evolves towards a steeper faint-end slope over cosmic time from z~4.5 to z~2.4. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:35:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nagaraj",
"Gautam",
""
],
[
"Ciardullo",
"Robin",
""
],
[
"Gronwall",
"Caryl",
""
],
[
"Ramakrishnan",
"Vandana",
""
],
[
"Lee",
"Kyoung-Soo",
""
],
[
"Gawiser",
"Eric",
""
],
[
"Firestone",
"Nicole M.",
""
],
[
"Ramgopal",
"Govind",
""
],
[
"Aguilar",
"J.",
""
],
[
"Ahlen",
"Steven",
""
],
[
"Bianchi",
"Davide",
""
],
[
"Brooks",
"David",
""
],
[
"Castander",
"Francisco Javier",
""
],
[
"Claybaugh",
"Todd",
""
],
[
"Cuceu",
"Andrei",
""
],
[
"de la Macorra",
"Axel",
""
],
[
"Dey",
"Arjun",
""
],
[
"Dey",
"Biprateep",
""
],
[
"Doel",
"Peter",
""
],
[
"Forero-Romero",
"Jaime",
""
],
[
"Gaztanaga",
"Enrique",
""
],
[
"Gontcho",
"Satya Gontcho A",
""
],
[
"Gutierrez",
"Gaston",
""
],
[
"Herrera-Alcantar",
"Hiram K.",
""
],
[
"Honscheid",
"Klaus",
""
],
[
"Ishak",
"Mustapha",
""
],
[
"Kehoe",
"Robert",
""
],
[
"Kirkby",
"David",
""
],
[
"Kisner",
"T.",
""
],
[
"Kremin",
"Anthony",
""
],
[
"Landriau",
"Martin",
""
],
[
"Guillou",
"Laurent Le",
""
],
[
"Levi",
"Michael",
""
],
[
"Magneville",
"Christophe",
""
],
[
"Manera",
"Marc",
""
],
[
"Martini",
"Paul",
""
],
[
"Meisner",
"Aaron M.",
""
],
[
"Miquel",
"Ramon",
""
],
[
"Moustakas",
"John",
""
],
[
"Palanque-Delabrouille",
"Nathalie",
""
],
[
"Prada",
"Francisco",
""
],
[
"Perez-Rafols",
"Ignasi",
""
],
[
"Rossi",
"Graziano",
""
],
[
"Samushia",
"Lado",
""
],
[
"Sanchez",
"Eusebio",
""
],
[
"Schlegel",
"David J.",
""
],
[
"Schubnell",
"Michael F.",
""
],
[
"Seo",
"Hee-Jong",
""
],
[
"Silber",
"Joseph H.",
""
],
[
"Sprayberry",
"David",
""
],
[
"Tarle",
"Gregory",
""
],
[
"Valdes",
"Francisco",
""
],
[
"Weaver",
"Benjamin A.",
""
],
[
"White",
"Martin",
""
],
[
"Zhou",
"Rongpu",
""
],
[
"Zou",
"Hu",
""
]
] |
2506.14511 | Zhiwen Shao | Zhiwen Shao, Yifan Cheng, Feiran Li, Yong Zhou, Xuequan Lu, Yuan Xie, and Lizhuang Ma | MOL: Joint Estimation of Micro-Expression, Optical Flow, and Landmark via Transformer-Graph-Style Convolution | This paper has been accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence | null | 10.1109/TPAMI.2025.3581162 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Facial micro-expression recognition (MER) is a challenging problem, due to transient and subtle micro-expression (ME) actions. Most existing methods depend on hand-crafted features, key frames like onset, apex, and offset frames, or deep networks limited by small-scale and low-diversity datasets. In this paper, we propose an end-to-end micro-action-aware deep learning framework with advantages from transformer, graph convolution, and vanilla convolution. In particular, we propose a novel F5C block composed of fully-connected convolution and channel correspondence convolution to directly extract local-global features from a sequence of raw frames, without the prior knowledge of key frames. The transformer-style fully-connected convolution is proposed to extract local features while maintaining global receptive fields, and the graph-style channel correspondence convolution is introduced to model the correlations among feature patterns. Moreover, MER, optical flow estimation, and facial landmark detection are jointly trained by sharing the local-global features. The two latter tasks contribute to capturing facial subtle action information for MER, which can alleviate the impact of insufficient training data. Extensive experiments demonstrate that our framework (i) outperforms the state-of-the-art MER methods on CASME II, SAMM, and SMIC benchmarks, (ii) works well for optical flow estimation and facial landmark detection, and (iii) can capture facial subtle muscle actions in local regions associated with MEs. The code is available at https://github.com/CYF-cuber/MOL. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:35:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shao",
"Zhiwen",
""
],
[
"Cheng",
"Yifan",
""
],
[
"Li",
"Feiran",
""
],
[
"Zhou",
"Yong",
""
],
[
"Lu",
"Xuequan",
""
],
[
"Xie",
"Yuan",
""
],
[
"Ma",
"Lizhuang",
""
]
] |
2506.14512 | Zijian Song | Zijian Song, Xiaoxin Lin, Qiuming Huang, Guangrun Wang, Liang Lin | SIRI-Bench: Challenging VLMs' Spatial Intelligence through Complex Reasoning Tasks | 16 pages, 9 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) are experiencing rapid advancements in complex reasoning, exhibiting remarkable generalization in mathematics and programming. In contrast, while spatial intelligence is fundamental for Vision-Language Models (VLMs) in real-world interaction, the systematic evaluation of their complex reasoning ability within spatial contexts remains underexplored. To bridge this gap, we introduce SIRI-Bench, a benchmark designed to evaluate VLMs' spatial intelligence through video-based reasoning tasks. SIRI-Bench comprises nearly 1K video-question-answer triplets, where each problem is embedded in a realistic 3D scene and captured by video. By carefully designing questions and corresponding 3D scenes, our benchmark ensures that solving the questions requires both spatial comprehension for extracting information and high-level reasoning for deriving solutions, making it a challenging benchmark for evaluating VLMs. To facilitate large-scale data synthesis, we develop an Automatic Scene Creation Engine. This engine, leveraging multiple specialized LLM agents, can generate realistic 3D scenes from abstract math problems, ensuring faithfulness to the original descriptions. Experimental results reveal that state-of-the-art VLMs struggle significantly on SIRI-Bench, underscoring the challenge of spatial reasoning. We hope that our study will bring researchers' attention to spatially grounded reasoning and advance VLMs in visual problem-solving. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:40:00 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Song",
"Zijian",
""
],
[
"Lin",
"Xiaoxin",
""
],
[
"Huang",
"Qiuming",
""
],
[
"Wang",
"Guangrun",
""
],
[
"Lin",
"Liang",
""
]
] |
2506.14513 | Hania Ghouse | Farha Abdul Wasay, Mohammed Abdul Rahman, Hania Ghouse | GAMORA: A Gesture Articulated Meta Operative Robotic Arm for Hazardous Material Handling in Containment-Level Environments | null | null | null | null | cs.RO cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | The convergence of robotics and virtual reality (VR) has enabled safer and more efficient workflows in high-risk laboratory settings, particularly virology labs. As biohazard complexity increases, minimizing direct human exposure while maintaining precision becomes essential. We propose GAMORA (Gesture Articulated Meta Operative Robotic Arm), a novel VR-guided robotic system that enables remote execution of hazardous tasks using natural hand gestures. Unlike existing scripted automation or traditional teleoperation, GAMORA integrates the Oculus Quest 2, NVIDIA Jetson Nano, and Robot Operating System (ROS) to provide real-time immersive control, digital twin simulation, and inverse kinematics-based articulation. The system supports VR-based training and simulation while executing precision tasks in physical environments via a 3D-printed robotic arm. Inverse kinematics ensure accurate manipulation for delicate operations such as specimen handling and pipetting. The pipeline includes Unity-based 3D environment construction, real-time motion planning, and hardware-in-the-loop testing. GAMORA achieved a mean positional discrepancy of 2.2 mm (improved from 4 mm), pipetting accuracy within 0.2 mL, and repeatability of 1.2 mm across 50 trials. Integrated object detection via YOLOv8 enhances spatial awareness, while energy-efficient operation (50% reduced power output) ensures sustainable deployment. The system's digital-physical feedback loop enables safe, precise, and repeatable automation of high-risk lab tasks. GAMORA offers a scalable, immersive solution for robotic control and biosafety in biomedical research environments. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:40:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wasay",
"Farha Abdul",
""
],
[
"Rahman",
"Mohammed Abdul",
""
],
[
"Ghouse",
"Hania",
""
]
] |
2506.14514 | Ga\"el Buldgen | Gael Buldgen, Gloria. Canocchi, Arthur. Le Saux, Vladimir A. Baturin, Regner Trampedach, Anna V. Oreshina, Sergey V. Ayukov, Anil Pradhan, Jean-Christophe Pain, Masanobu Kunitomo, Thierry Appourchaux, Rafael A. Garcia, Morgan Deal, Nicolas Grevesse, Arlette Noels, Joergen Christensen-Dalsgaard, Tristan Guillot, Devesh Nandal, J\'er\^ome B\'etrisey, Christophe Blancard, James Colgan, Philippe Coss\'e Christopher, J. Fontes, Ludovic Petitdemange, Charly Pincon | The Future of Solar modelling: requirements for a new generation of solar models | Accepted for publication in Solar Physics | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Helioseismology and solar modelling have enjoyed a golden era thanks to decades-long surveys from ground-based networks such as for example GONG, BiSON, IRIS and the SOHO and SDO space missions which have provided high-quality helioseismic observations that supplemented photometric, gravitational, size and shape, limb-darkening and spectroscopic constraints as well as measurements of neutrino fluxes. However, the success of solar models is also deeply rooted in progress in fundamental physics (equation of state of the solar plasma, high-quality atomic physics computations and opacities, description of convection and the role of macroscopic transport processes of angular momentum and chemicals, such as for example meridional circulation, internal gravity waves, shear-induced turbulence or even convection. In this paper, we briefly outline some key areas of research that deserve particular attention in solar modelling. We discuss the current uncertainties that need to be addressed, how these limit our predictions from solar models and their impact on stellar evolution in general. We outline potential strategies to mitigate them and how multidisciplinary approaches will be needed in the future to tackle them. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:40:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Buldgen",
"Gael",
""
],
[
"Canocchi",
"Gloria.",
""
],
[
"Saux",
"Arthur. Le",
""
],
[
"Baturin",
"Vladimir A.",
""
],
[
"Trampedach",
"Regner",
""
],
[
"Oreshina",
"Anna V.",
""
],
[
"Ayukov",
"Sergey V.",
""
],
[
"Pradhan",
"Anil",
""
],
[
"Pain",
"Jean-Christophe",
""
],
[
"Kunitomo",
"Masanobu",
""
],
[
"Appourchaux",
"Thierry",
""
],
[
"Garcia",
"Rafael A.",
""
],
[
"Deal",
"Morgan",
""
],
[
"Grevesse",
"Nicolas",
""
],
[
"Noels",
"Arlette",
""
],
[
"Christensen-Dalsgaard",
"Joergen",
""
],
[
"Guillot",
"Tristan",
""
],
[
"Nandal",
"Devesh",
""
],
[
"Bétrisey",
"Jérôme",
""
],
[
"Blancard",
"Christophe",
""
],
[
"Colgan",
"James",
""
],
[
"Christopher",
"Philippe Cossé",
""
],
[
"Fontes",
"J.",
""
],
[
"Petitdemange",
"Ludovic",
""
],
[
"Pincon",
"Charly",
""
]
] |
2506.14515 | Prabhav Sanga | Prabhav Sanga, Jaskaran Singh, Arun K. Dubey | Train Once, Forget Precisely: Anchored Optimization for Efficient Post-Hoc Unlearning | Accepted at ICML MUGen'25 | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As machine learning systems increasingly rely on data subject to privacy regulation, selectively unlearning specific information from trained models has become essential. In image classification, this involves removing the influence of particular training samples, semantic classes, or visual styles without full retraining. We introduce \textbf{Forget-Aligned Model Reconstruction (FAMR)}, a theoretically grounded and computationally efficient framework for post-hoc unlearning in deep image classifiers. FAMR frames forgetting as a constrained optimization problem that minimizes a uniform-prediction loss on the forget set while anchoring model parameters to their original values via an $\ell_2$ penalty. A theoretical analysis links FAMR's solution to influence-function-based retraining approximations, with bounds on parameter and output deviation. Empirical results on class forgetting tasks using CIFAR-10 and ImageNet-100 demonstrate FAMR's effectiveness, with strong performance retention and minimal computational overhead. The framework generalizes naturally to concept and style erasure, offering a scalable and certifiable route to efficient post-hoc forgetting in vision models. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:40:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sanga",
"Prabhav",
""
],
[
"Singh",
"Jaskaran",
""
],
[
"Dubey",
"Arun K.",
""
]
] |
2506.14516 | Oleg Zendel | Kun Ran, Shuoqi Sun, Khoi Nguyen Dinh Anh, Damiano Spina and Oleg Zendel | RMIT-ADM+S at the SIGIR 2025 LiveRAG Challenge | Accepted for oral presentation at SIGIR 2025 LiveRAG | null | null | null | cs.IR | http://creativecommons.org/licenses/by-sa/4.0/ | This paper presents the RMIT--ADM+S participation in the SIGIR 2025 LiveRAG Challenge. Our Generation-Retrieval-Augmented Generation (GRAG) approach relies on generating a hypothetical answer that is used in the retrieval phase, alongside the original question. GRAG also incorporates a pointwise large language model (LLM)-based re-ranking step prior to final answer generation. We describe the system architecture and the rationale behind our design choices. In particular, a systematic evaluation using the Grid of Points (GoP) framework and N-way ANOVA enabled comparison across multiple configurations, including query variant generation, question decomposition, rank fusion strategies, and prompting techniques for answer generation. Our system achieved a Relevance score of 1.199 and a Faithfulness score of 0.477 on the private leaderboard, placing among the top four finalists in the LiveRAG 2025 Challenge. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:41:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ran",
"Kun",
""
],
[
"Sun",
"Shuoqi",
""
],
[
"Anh",
"Khoi Nguyen Dinh",
""
],
[
"Spina",
"Damiano",
""
],
[
"Zendel",
"Oleg",
""
]
] |
2506.14517 | Sinan Genc | Sinan Genc, Oguzhan Yucel, Furkan Aglarci, Carlos Rodriguez-Fernandez, Alpay Yilmaz, Humeyra Caglayan, Serkan Ates and Alpan Bek | Disorder-Engineered Hybrid Plasmonic Cavities for Emission Control of Defects in hBN | null | null | null | null | physics.optics physics.app-ph quant-ph | http://creativecommons.org/licenses/by/4.0/ | Defect-based quantum emitters in hexagonal boron nitride (hBN) are promising building blocks for scalable quantum photonics due to their stable single-photon emission at room temperature. However, enhancing their emission intensity and controlling the decay dynamics remain significant challenges. This study demonstrates a low-cost, scalable fabrication approach to integrate plasmonic nanocavities with defect-based quantum emitters in hBN nanoflakes. Using the thermal dewetting process, we realize two distinct configurations: stochastic Ag nanoparticles (AgNPs) on hBN flakes and hybrid plasmonic nanocavities formed by AgNPs on top of hBN flakes supported on gold/silicon dioxide (Au/SiO2) substrates. While AgNPs on bare hBN yield up to a two-fold photoluminescence (PL) enhancement with reduced emitter lifetimes, the hybrid nanocavity architecture provides a dramatic, up to 100-fold PL enhancement and improved uniformity across multiple. emitters, all without requiring deterministic positioning. Finite-difference time-domain (FDTD) simulations and time-resolved PL measurements confirm size-dependent control over decay dynamics and cavity-emitter interactions. Our versatile solution overcomes key quantum photonic device development challenges, including material integration, emission intensity optimization, and spectral multiplexity. Future work will explore potential applications in integrated photonic circuits hosting on-chip quantum systems and hBN-based label-free single-molecule detection through such quantum nanoantennas. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:46:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Genc",
"Sinan",
""
],
[
"Yucel",
"Oguzhan",
""
],
[
"Aglarci",
"Furkan",
""
],
[
"Rodriguez-Fernandez",
"Carlos",
""
],
[
"Yilmaz",
"Alpay",
""
],
[
"Caglayan",
"Humeyra",
""
],
[
"Ates",
"Serkan",
""
],
[
"Bek",
"Alpan",
""
]
] |
2506.14518 | Elif Y{\i}lmaz | Elif Y{\i}lmaz and Christos Dimitrakakis | Two-Player Zero-Sum Games with Bandit Feedback | null | null | null | null | cs.LG cs.GT | http://creativecommons.org/licenses/by/4.0/ | We study a two-player zero-sum game (TPZSG) in which the row player aims to maximize their payoff against an adversarial column player, under an unknown payoff matrix estimated through bandit feedback. We propose and analyze two algorithms: ETC-TPZSG, which directly applies ETC to the TPZSG setting and ETC-TPZSG-AE, which improves upon it by incorporating an action pair elimination (AE) strategy that leverages the $\varepsilon$-Nash Equilibrium property to efficiently select the optimal action pair. Our objective is to demonstrate the applicability of ETC in a TPZSG setting by focusing on learning pure strategy Nash Equilibrium. A key contribution of our work is a derivation of instance-dependent upper bounds on the expected regret for both algorithms, has received limited attention in the literature on zero-sum games. Particularly, after $T$ rounds, we achieve an instance-dependent regret upper bounds of $O(\Delta + \sqrt{T})$ for ETC-TPZSG and $O(\frac{\log (T \Delta^2)}{\Delta})$ for ETC-TPZSG-AE, where $\Delta$ denotes the suboptimality gap. Therefore, our results indicate that ETC-based algorithms perform effectively in adversarial game settings, achieving regret bounds comparable to existing methods while providing insights through instance-dependent analysis. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:46:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yılmaz",
"Elif",
""
],
[
"Dimitrakakis",
"Christos",
""
]
] |
2506.14519 | Wei-Cong Jing | W. C. Jing, J. L. Han, C. Wang, P. F. Wang, T. Wang, N. N. Cai, J. Xu, Z. L. Yang, D. J. Zhou, Yi Yan, W. Q. Su, X. Y. Gao, L. Xie | FAST Pulsar Database: II. Scattering profiles of 122 Pulsars | Subband profiles of each pulsar are shown in the paper. Accepted by RAA | null | null | null | astro-ph.HE astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The turbulent ionized interstellar medium diffracts radio waves and makes them propagate in multiple paths. The pulse-broadening observed at low frequencies results from the scattering effect of interstellar clouds of ionized gas. During the Galactic Plane Pulsar Snapshot (GPPS) survey and other projects by using the Five-hundred-meter Aperture Spherical radio Telescope (FAST), we detect the pulse-broadening for 122 pulsars in the radio frequency band between 1.0 and 1.5 GHz, including 60 newly discovered pulsars in the GPPS survey and 62 previously known pulsars. We find that a more accurate dispersion measure can be obtained from aligning the front edge of the scattered subband pulses at the 1/4 or 1/2 peak level for most pulsars with one dominant component in the intrinsic profile, and the best DM values from aligning the intrinsic profile components from the model-fitting. From the pulse profiles at a few subbands we derive the pulse-broadening timescale and the scattering spectral index. These scattering parameters are measured for the first time for 93 pulsars. For 29 pulsars with previously detected scattering features, our measurements of the pulse-broadening timescale are consistent with results in the literature. We find that pulsars behind spiral arms show a stronger scattering effect due to greater density fluctuations in the arm regions. With a properly derived dispersion measure and careful calibration, we also present polarization profiles for 41 pulsars in three subbands of FAST observations. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:46:57 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jing",
"W. C.",
""
],
[
"Han",
"J. L.",
""
],
[
"Wang",
"C.",
""
],
[
"Wang",
"P. F.",
""
],
[
"Wang",
"T.",
""
],
[
"Cai",
"N. N.",
""
],
[
"Xu",
"J.",
""
],
[
"Yang",
"Z. L.",
""
],
[
"Zhou",
"D. J.",
""
],
[
"Yan",
"Yi",
""
],
[
"Su",
"W. Q.",
""
],
[
"Gao",
"X. Y.",
""
],
[
"Xie",
"L.",
""
]
] |
2506.14520 | Fabian Rupp | Matthias R\"oger and Fabian Rupp | Elastic bilayer membranes under confinement -- Existence, regularity, and rigidity | 35 pages, comments welcome! | null | null | null | math.AP math.DG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Motivated by applications to cell biology, we study the constrained minimization of the Helfrich energy among closed surfaces confined to a container. We show existence of minimizers in the class of bubble trees of spherical weak branched immersions and derive the Euler--Lagrange equations which involve a measure-valued Lagrange multiplier that is concentrated on the coincidence set with the container boundary. We provide a careful analysis of this elliptic system and prove optimal regularity for solutions throughout the branch points. For surfaces confined in the unit ball we show that the minimization problem behaves rather rigid and identify a parameter range for which minimizers are always round spheres. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:48:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Röger",
"Matthias",
""
],
[
"Rupp",
"Fabian",
""
]
] |
2506.14521 | Korbinian Pfab | Korbinian Pfab and Marcel Rothering | Towards Improved Research Methodologies for Industrial AI: A case study of false call reduction | Submitted and accepted to IEEE COMPSAC 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Are current artificial intelligence (AI) research methodologies ready to create successful, productive, and profitable AI applications? This work presents a case study on an industrial AI use case called false call reduction for automated optical inspection to demonstrate the shortcomings of current best practices. We identify seven weaknesses prevalent in related peer-reviewed work and experimentally show their consequences. We show that the best-practice methodology would fail for this use case. We argue amongst others for the necessity of requirement-aware metrics to ensure achieving business objectives, clear definitions of success criteria, and a thorough analysis of temporal dynamics in experimental datasets. Our work encourages researchers to critically assess their methodologies for more successful applied AI research. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:48:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pfab",
"Korbinian",
""
],
[
"Rothering",
"Marcel",
""
]
] |
2506.14522 | Chaipattana Juntarapomdach | Chaipattana Juntarapomdach and Teeradej Kittipassorn | Transparent Rectangle Visibility Graphs | 17 pages, 21 figures, submitted | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | A transparent rectangle visibility graph (TRVG) is a graph whose vertices can be represented by a collection of non-overlapping rectangles in the plane whose sides are parallel to the axes such that two vertices are adjacent if and only if there is a horizontal or vertical line intersecting the interiors of their rectangles.
We show that every threshold graph, tree, cycle, rectangular grid graph, triangular grid graph and hexagonal grid graph is a TRVG. We also obtain a maximum number of edges of a bipartite TRVG and characterize complete bipartite TRVGs. More precisely, a bipartite TRVG with $n$ vertices has at most $2n-2$ edges. The complete bipartite graph $K_{p,q}$ is a TRVG if and only if $\min\{p,q\} \le 2$ or $(p,q) \in \{(3,3), (3,4)\}$. We prove similar results for the torus. Moreover, we study whether powers of cycles and their complements are TRVGs. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:49:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Juntarapomdach",
"Chaipattana",
""
],
[
"Kittipassorn",
"Teeradej",
""
]
] |
2506.14523 | G\'eza T\'oth | G\'eza T\'oth, J\'ozsef Pitrik | Several types of quantum Wasserstein distance based on an optimization over separable states | 10 pages, revtex 4.2 | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We consider several definitions of the quantum Wasserstein distance based on an optimization over general bipartite quantum states with given marginals. Then, we examine the quantities obtained after the optimization is carried out over bipartite separable states instead. We prove that several of these quantities are equal to each other. Thus, we connect several approaches in the literature. As a byproduct, we show that the Uhlmann-Jozsa quantum fidelity can also be written as an optimization over separable states with given marginals. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:50:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tóth",
"Géza",
""
],
[
"Pitrik",
"József",
""
]
] |
2506.14524 | Nadezhda Alsahanova | Nadezhda Alsahanova, Pavel Bartenev, Maksim Sharaev, Milos Ljubisavljevic, Taleb Al. Mansoori, Yauhen Statsenko | Integrating Radiomics with Deep Learning Enhances Multiple Sclerosis Lesion Delineation | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Background: Accurate lesion segmentation is critical for multiple sclerosis (MS) diagnosis, yet current deep learning approaches face robustness challenges.
Aim: This study improves MS lesion segmentation by combining data fusion and deep learning techniques.
Materials and Methods: We suggested novel radiomic features (concentration rate and R\'enyi entropy) to characterize different MS lesion types and fused these with raw imaging data. The study integrated radiomic features with imaging data through a ResNeXt-UNet architecture and attention-augmented U-Net architecture. Our approach was evaluated on scans from 46 patients (1102 slices), comparing performance before and after data fusion.
Results: The radiomics-enhanced ResNeXt-UNet demonstrated high segmentation accuracy, achieving significant improvements in precision and sensitivity over the MRI-only baseline and a Dice score of 0.774$\pm$0.05; p<0.001 according to Bonferroni-adjusted Wilcoxon signed-rank tests. The radiomics-enhanced attention-augmented U-Net model showed a greater model stability evidenced by reduced performance variability (SDD = 0.18 $\pm$ 0.09 vs. 0.21 $\pm$ 0.06; p=0.03) and smoother validation curves with radiomics integration.
Conclusion: These results validate our hypothesis that fusing radiomics with raw imaging data boosts segmentation performance and stability in state-of-the-art models. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:50:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Alsahanova",
"Nadezhda",
""
],
[
"Bartenev",
"Pavel",
""
],
[
"Sharaev",
"Maksim",
""
],
[
"Ljubisavljevic",
"Milos",
""
],
[
"Mansoori",
"Taleb Al.",
""
],
[
"Statsenko",
"Yauhen",
""
]
] |
2506.14525 | Boyong He | Zhuoyue Tan, Boyong He, Yuxiang Ji and Liaoni Wu | VisLanding: Monocular 3D Perception for UAV Safe Landing via Depth-Normal Synergy | Accepted by IROS2025 | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents VisLanding, a monocular 3D perception-based framework for safe UAV (Unmanned Aerial Vehicle) landing. Addressing the core challenge of autonomous UAV landing in complex and unknown environments, this study innovatively leverages the depth-normal synergy prediction capabilities of the Metric3D V2 model to construct an end-to-end safe landing zones (SLZ) estimation framework. By introducing a safe zone segmentation branch, we transform the landing zone estimation task into a binary semantic segmentation problem. The model is fine-tuned and annotated using the WildUAV dataset from a UAV perspective, while a cross-domain evaluation dataset is constructed to validate the model's robustness. Experimental results demonstrate that VisLanding significantly enhances the accuracy of safe zone identification through a depth-normal joint optimization mechanism, while retaining the zero-shot generalization advantages of Metric3D V2. The proposed method exhibits superior generalization and robustness in cross-domain testing compared to other approaches. Furthermore, it enables the estimation of landing zone area by integrating predicted depth and normal information, providing critical decision-making support for practical applications. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:51:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tan",
"Zhuoyue",
""
],
[
"He",
"Boyong",
""
],
[
"Ji",
"Yuxiang",
""
],
[
"Wu",
"Liaoni",
""
]
] |
2506.14526 | Markus P. Mueller | Caroline L. Jones, Albert Aloy, Gerard Higgins, Markus P. Mueller | Certified randomness from quantum speed limits | 10+4 pages, 7 figures | null | null | null | quant-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum speed limits are usually regarded as fundamental restrictions, constraining the amount of computation that can be achieved within some given time and energy. Complementary to this intuition, here we show that these limitations are also of operational value: they enable the secure generation of certified randomness. We consider a prepare-and-measure scenario with some (experimentally determined or promised) upper bound on the energy uncertainty of the average prepared quantum state, but without any further assumptions on the devices, Hilbert space or Hamiltonian. Given that we can freely choose the time at which to apply the untrusted preparation procedure, we show that this scenario admits the generation of randomness that is secure against adversaries with additional classical information. We show how to determine the amount of certified randomness given the observed correlations, discuss how interactions with the environment are taken into account, and sketch a conceivable experimental implementation. Remarkably, even single-mode coherent states admit this kind of certification of non-zero randomness in some parameter regimes, reinforcing ongoing approaches to demonstrate versions of nonclassicality in the simple harmonic oscillator. Our results extend existing efforts to devise semi-device-independent protocols grounded in reasonable physical assumptions, and they contribute to the understanding of time-energy uncertainty relations via their operational consequences. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:52:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jones",
"Caroline L.",
""
],
[
"Aloy",
"Albert",
""
],
[
"Higgins",
"Gerard",
""
],
[
"Mueller",
"Markus P.",
""
]
] |
2506.14527 | Yong-Jiang Xu | Yong-Jiang Xu, Luo-Geng Tian, Yuan Li, Jin-Yun Wu, Yong-Lu Liu, and Ming-Qiu Huang | Scalar Triple-Heavy Tetraquark States With Quark Content $cc\bar{c}\bar{s}$ | 20 pages,9 figures | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study the scalar triple-heavy tetraquark states with quark content $cc\bar{c}\bar{s}$, $\eta_{c}D_{s}$ molecular state and $[cc]_{A(T)}[\bar{c}\bar{s}]_{A(T)}$ compact tetraquark states, by the QCD sum rule method. First, we construct the needed interpolating currents, $J^{M}(x)$, $J^{T_{1}}(x)$, and $J^{T_{2}}(x)$. Then, we derive the sum rules for the masses and the current coupling constants. Finally, we numerically analyze these sum rules, and find $m_{M}=4.9392^{+0.0851}_{-0.0817}~\mbox{GeV}$, and $\lambda_{M}=2.8857^{+0.5729}_{-0.4928}\times10^{-2}~\mbox{GeV}^{5}$ for the mass and the current coupling constant of the $\eta_{c}D_{s}$ molecular state, $m_{T_{1}}=5.0774^{+0.0708}_{-0.0641}~\mbox{GeV}$, and $\lambda_{T_{1}}=1.0436^{+0.1862}_{-0.1573}\times10^{-1}~\mbox{GeV}^{5}$ for the mass and the current coupling constant of the $[cc]_{A}[\bar{c}\bar{s}]_{A}$ compact tetraquark state, $m_{T_{2}}=5.0679^{+0.0839}_{-0.0721}~\mbox{GeV}$, and $\lambda_{T_{2}}=2.0316^{+0.4119}_{-0.3119}\times10^{-1}~\mbox{GeV}^{5}$ for the mass and the current coupling constant of the $[cc]_{T}[\bar{c}\bar{s}]_{T}$ compact tetraquark state. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:52:45 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Yong-Jiang",
""
],
[
"Tian",
"Luo-Geng",
""
],
[
"Li",
"Yuan",
""
],
[
"Wu",
"Jin-Yun",
""
],
[
"Liu",
"Yong-Lu",
""
],
[
"Huang",
"Ming-Qiu",
""
]
] |
2506.14528 | Claudia Schoemann | Claudia Schoemann | The kernel of the Gysin homomorphism for smooth projective curves | 7 pages | null | null | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | Let $S$ be a smooth projective connected surface over an algebraically closed field $k$ and $\Sigma$ the linear system of a very ample divisor $D$ on $S$. Let $d:=\dim(\Sigma)$ be the dimension of $\Sigma$ and $\phi_{\Sigma}: S \hookrightarrow \mathbb{P}^{d}$ the closed embedding of $S$ into $\mathbb{P}^{d}$, induced by $\Sigma$. For any closed point $t\in\Sigma \cong\mathbb{P}^{d^*}$, let $C_t$ be the corresponding hyperplane section on $S$, and let $ r_t:C_t\hookrightarrow S$ be the closed embedding of the curve $C_t$ into $S$. Let $\Delta:= \{t \in \Sigma: C_t \text{ is singular}\}$ be the discriminant locus of $\Sigma$ and let $U :=\Sigma\setminus \Delta$.
For $t \in U$, the kernel of the Gysin homomorphism of the Chow groups of $0$-cycles of degree zero, from $CH_0(C_t)_{deg=0}$ to $CH_0(S)_{deg=0}$ is the countable union of shifts of a certain abelian subvariety $A_t$ inside $J(C_t)$, the Jacobian of the curve $C_t$ (\cite{PS24} for $k \cong \mathbb{C}$, \cite{SW25} for $k \cong \overline{\mathbb{F}_q((t))}$). We prove that for every closed point $t \in U$ either $A_t$ coincides with the abelian variety $B_t$ inside $J(C_t)$ corresponding to the vanishing cohomology $H^1(C_t, k')_{van}$, where $k'$ is the minimal field of definition of $k$, and then the Gysin kernel is a countable union of shifts of $B_t$, or $A_t = 0$, in which case the Gysin kernel is countable. Using the language of algebraic stacks as a generalisation of algebraic varieties this is done by constructing an increasing filtration of Zariski countable open substacks $U_i, i \in I,$ of $U$, where $I$ is a countable set and by applying a convergence argument. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:53:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Schoemann",
"Claudia",
""
]
] |
2506.14529 | Xiaohan Zheng | Xiaohan Zheng, Lanning Wei, Yong Li and Quanming Yao | Automated Decision-Making on Networks with LLMs through Knowledge-Guided Evolution | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Effective decision-making on networks often relies on learning from graph-structured data, where Graph Neural Networks (GNNs) play a central role, but they take efforts to configure and tune. In this demo, we propose LLMNet, showing how to design GNN automated through Large Language Models. Our system develops a set of agents that construct graph-related knowlege bases and then leverages Retrieval-Augmented Generation (RAG) to support automated configuration and refinement of GNN models through a knowledge-guided evolution process. These agents, equipped with specialized knowledge bases, extract insights into tasks and graph structures by interacting with the knowledge bases. Empirical results show LLMNet excels in twelve datasets across three graph learning tasks, validating its effectiveness of GNN model designing. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:53:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zheng",
"Xiaohan",
""
],
[
"Wei",
"Lanning",
""
],
[
"Li",
"Yong",
""
],
[
"Yao",
"Quanming",
""
]
] |
2506.14530 | Anastasis Kratsios | Anastasis Kratsios, Tin Sum Cheng, Aurelien Lucchi, Haitz S\'aez de Oc\'ariz Borde | Sharp Generalization Bounds for Foundation Models with Asymmetric Randomized Low-Rank Adapters | null | null | null | null | stat.ML cs.AI cs.LG cs.NE math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Low-Rank Adaptation (LoRA) has emerged as a widely adopted parameter-efficient fine-tuning (PEFT) technique for foundation models. Recent work has highlighted an inherent asymmetry in the initialization of LoRA's low-rank factors, which has been present since its inception and was presumably derived experimentally. This paper focuses on providing a comprehensive theoretical characterization of asymmetric LoRA with frozen random factors. First, while existing research provides upper-bound generalization guarantees based on averages over multiple experiments, the behaviour of a single fine-tuning run with specific random factors remains an open question. We address this by investigating the concentration of the typical LoRA generalization gap around its mean. Our main upper bound reveals a sample complexity of $\tilde{\mathcal{O}}\left(\frac{\sqrt{r}}{\sqrt{N}}\right)$ with high probability for rank $r$ LoRAs trained on $N$ samples. Additionally, we also determine the fundamental limits in terms of sample efficiency, establishing a matching lower bound of $\mathcal{O}\left(\frac{1}{\sqrt{N}}\right)$. By more closely reflecting the practical scenario of a single fine-tuning run, our findings offer crucial insights into the reliability and practicality of asymmetric LoRA. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:55:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kratsios",
"Anastasis",
""
],
[
"Cheng",
"Tin Sum",
""
],
[
"Lucchi",
"Aurelien",
""
],
[
"Borde",
"Haitz Sáez de Ocáriz",
""
]
] |
2506.14531 | Yawen Ma | Yawen Ma, Anastasia Ushakova, Kate Cain, Gabriel Wallin | A statistical framework for dynamic cognitive diagnosis in digital learning environments | null | null | null | null | stat.AP stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reading is foundational for educational, employment, and economic outcomes, but a persistent proportion of students globally struggle to develop adequate reading skills. Some countries promote digital tools to support reading development, alongside regular classroom instruction. Such tools generate rich log data capturing students' behaviour and performance. This study proposes a dynamic cognitive diagnostic modeling (CDM) framework based on restricted latent class models to trace students' time-varying skills mastery using log files from digital tools. Unlike traditional CDMs that require expert-defined skill-item mappings (Q-matrix), our approach jointly estimates the Q-matrix and latent skill profiles, integrates log-derived covariates (e.g., reattempts, response times, counts of mastered items) and individual characteristics, and models transitions in mastery using a Bayesian estimation approach. Applied to real-world data, the model demonstrates practical value in educational settings by effectively uncovering individual skill profiles and the skill-item mappings. Simulation studies confirm robust recovery of Q-matrix structures and latent profiles with high accuracy under varied sample sizes, item counts and different sparsity of Q-matrices. The framework offers a data-driven, time-dependent restricted latent class modeling approach to understanding early reading development. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:57:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ma",
"Yawen",
""
],
[
"Ushakova",
"Anastasia",
""
],
[
"Cain",
"Kate",
""
],
[
"Wallin",
"Gabriel",
""
]
] |
2506.14532 | Can Zheng | Can Zheng and Jiguang He and Chung G. Kang and Guofa Cai and Zitong Yu and Merouane Debbah | M2BeamLLM: Multimodal Sensing-empowered mmWave Beam Prediction with Large Language Models | 13 pages, 20 figures | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | This paper introduces a novel neural network framework called M2BeamLLM for beam prediction in millimeter-wave (mmWave) massive multi-input multi-output (mMIMO) communication systems. M2BeamLLM integrates multi-modal sensor data, including images, radar, LiDAR, and GPS, leveraging the powerful reasoning capabilities of large language models (LLMs) such as GPT-2 for beam prediction. By combining sensing data encoding, multimodal alignment and fusion, and supervised fine-tuning (SFT), M2BeamLLM achieves significantly higher beam prediction accuracy and robustness, demonstrably outperforming traditional deep learning (DL) models in both standard and few-shot scenarios. Furthermore, its prediction performance consistently improves with increased diversity in sensing modalities. Our study provides an efficient and intelligent beam prediction solution for vehicle-to-infrastructure (V2I) mmWave communication systems. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:58:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zheng",
"Can",
""
],
[
"He",
"Jiguang",
""
],
[
"Kang",
"Chung G.",
""
],
[
"Cai",
"Guofa",
""
],
[
"Yu",
"Zitong",
""
],
[
"Debbah",
"Merouane",
""
]
] |
2506.14533 | Jincheng Yang | Matei P. Coiculescu and Jincheng Yang | Conditional Liouville theorems for the Navier-Stokes equations | 24 pages, 1 figure | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel approach to the Liouville problem for the stationary Navier-Stokes equations. As an application of our method, we prove conditional Liouville theorems with assumptions on the antiderivative of the velocity that represent substantial improvements on what was heretofore known. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 13:59:23 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Coiculescu",
"Matei P.",
""
],
[
"Yang",
"Jincheng",
""
]
] |
2506.14534 | Cl\'ement Yvernes | Cl\'ement Yvernes, Emilie Devijver, Eric Gaussier | Complete Characterization for Adjustment in Summary Causal Graphs of Time Series | Accepted at the 41st Conference on Uncertainty in Artificial Intelligence (UAI) | null | null | null | math.ST cs.AI stat.TH | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The identifiability problem for interventions aims at assessing whether the total causal effect can be written with a do-free formula, and thus be estimated from observational data only. We study this problem, considering multiple interventions, in the context of time series when only an abstraction of the true causal graph, in the form of a summary causal graph, is available. We propose in particular both necessary and sufficient conditions for the adjustment criterion, which we show is complete in this setting, and provide a pseudo-linear algorithm to decide whether the query is identifiable or not. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:00:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yvernes",
"Clément",
""
],
[
"Devijver",
"Emilie",
""
],
[
"Gaussier",
"Eric",
""
]
] |
2506.14535 | Joaquin Bogado PhD | Jos\'e Manuel Su\'arez and Luis Mariano Bibb\'o and Joaquin Bogado and Alejandro Fernandez | Automatic Qiskit Code Refactoring Using Large Language Models | Submitted for review to "Taller Latinoamericano de Ingenier\'ia de Software Cu\'antico" (https://www.ripaisc.net/call-for-papers-tlisc-2025/) | null | null | null | cs.SE cs.AI cs.ET | http://creativecommons.org/licenses/by-sa/4.0/ | As quantum software frameworks evolve, developers face increasing challenges in maintaining compatibility with rapidly changing APIs. In this work, we present a novel methodology for refactoring Qiskit code using large language models (LLMs). We begin by extracting a taxonomy of migration scenarios from the different sources of official Qiskit documentation (such as release notes), capturing common patterns such as migration of functionality to different modules and deprecated usage. This taxonomy, along with the original Python source code, is provided as input to an LLM, which is then tasked with identifying instances of migration scenarios in the code and suggesting appropriate refactoring solutions. Our approach is designed to address the context length limitations of current LLMs by structuring the input and reasoning process in a targeted, efficient manner. The results demonstrate that LLMs, when guided by domain-specific migration knowledge, can effectively assist in automating Qiskit code migration. This work contributes both a set of proven prompts and taxonomy for Qiskit code migration from earlier versions to version 0.46 and a methodology to asses the capabilities of LLMs to assist in the migration of quantum code. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:00:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Suárez",
"José Manuel",
""
],
[
"Bibbó",
"Luis Mariano",
""
],
[
"Bogado",
"Joaquin",
""
],
[
"Fernandez",
"Alejandro",
""
]
] |
2506.14536 | Conor Rosato | Conor Rosato, Joshua Murphy, Simon Maskell, John Harris | Assessing the Impact of Vaccination on Rotavirus Transmission Dynamics Using Bayesian Inference | Accepted to EUSIPCO 2025 conference | null | null | null | q-bio.PE stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The introduction of the rotavirus vaccine in the United Kingdom (UK) in 2013 led to a noticeable decline in laboratory reports in subsequent years. To assess the impact of vaccination on rotavirus transmissibility we calibrated a stochastic compartmental epidemiological model using Sequential Monte Carlo (SMC) methods. Our analysis focuses on estimating the time-varying transmissibility parameter and documenting its evolution before and after vaccine rollout. We observe distinct periods of increasing and decreasing transmissibility, reflecting the dynamic response of rotavirus spread to immunization efforts. These findings improve our understanding of vaccination-driven shifts in disease transmission and provide a quantitative framework for evaluating long-term epidemiological trends. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:01:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Rosato",
"Conor",
""
],
[
"Murphy",
"Joshua",
""
],
[
"Maskell",
"Simon",
""
],
[
"Harris",
"John",
""
]
] |
2506.14537 | TzuMiao Chou | Tzu-Miao Chou | Topological Contextuality and Quantum Representations | null | null | null | null | quant-ph hep-th | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Quantum contextuality, a fundamental feature distinguishing quantum theory from classical models, is investigated via algebraic and topological structures inherent in modular tensor categories. This work rigorously demonstrates that braid group representations induced by modular categories, particularly those associated with SU(2) at level k and Fibonacci anyon models, exhibit state-dependent contextuality characterized by violations of noncontextuality inequalities. By explicitly constructing these unitary representations on fusion spaces, the study establishes a direct correspondence between braiding operations and logical contextuality scenarios. The results offer a comprehensive topological framework to classify and quantify contextuality in low-dimensional quantum systems, elucidating its role as a resource in topological quantum computation and advancing the interface between quantum algebra, topology, and quantum foundations. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:01:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chou",
"Tzu-Miao",
""
]
] |
2506.14538 | Mohamed Hamza Bandukara | Mohamed H Bandukara, Nikos Tzevelekos | A Logic For Fresh Labelled Transition Systems | 39 pages | null | null | null | cs.LO cs.FL | http://creativecommons.org/licenses/by/4.0/ | We introduce a Hennessy-Milner logic with recursion for Fresh Labelled Transition Systems (FLTSs). These are nominal labelled transition systems which keep track of the history, i.e. of data values seen so far, and can capture fresh data generation. In particular, FLTSs generalise the computations of Fresh-Register Automata, which in turn are one of the simplest classes of history-dependent automata operating on infinite input alphabets. Each automaton comes equipped with a finite set of registers where it can store data values and compare them with others from the input. In addition, the automaton can accept an input just if it be fresh: not seen in the computation before. The logic we introduce can express a variety of properties, such as the existence of an infinite path of distinct data values or the existence of a finite path where some taint property is violated. We study the model checking problem and its complexity via reduction to parity games and, using nominal sets techniques, provide an exponential upper bound for it. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:01:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bandukara",
"Mohamed H",
""
],
[
"Tzevelekos",
"Nikos",
""
]
] |
2506.14539 | Yeonghwan Shin` | Daewon Kang, YeongHwan Shin, Doyeon Kim, Kyu-Hwan Jung, Meong Hi Son | Doppelg\"anger Method: Breaking Role Consistency in LLM Agent via Prompt-based Transferable Adversarial Attack | null | null | null | null | cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | Since the advent of large language models, prompt engineering now enables the rapid, low-effort creation of diverse autonomous agents that are already in widespread use. Yet this convenience raises urgent concerns about the safety, robustness, and behavioral consistency of the underlying prompts, along with the pressing challenge of preventing those prompts from being exposed to user's attempts. In this paper, we propose the ''Doppelg\"anger method'' to demonstrate the risk of an agent being hijacked, thereby exposing system instructions and internal information. Next, we define the ''Prompt Alignment Collapse under Adversarial Transfer (PACAT)'' level to evaluate the vulnerability to this adversarial transfer attack. We also propose a ''Caution for Adversarial Transfer (CAT)'' prompt to counter the Doppelg\"anger method. The experimental results demonstrate that the Doppelg\"anger method can compromise the agent's consistency and expose its internal information. In contrast, CAT prompts enable effective defense against this adversarial attack. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:01:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kang",
"Daewon",
""
],
[
"Shin",
"YeongHwan",
""
],
[
"Kim",
"Doyeon",
""
],
[
"Jung",
"Kyu-Hwan",
""
],
[
"Son",
"Meong Hi",
""
]
] |
2506.14541 | Rongchang Lu | Rongchang Lu, Tianduo Luo, Yunzhi Zhang, Conghan Yue, Pei Yang, Guibao Liu and Changyang Gu | Exploring Diffusion with Test-Time Training on Efficient Image Restoration | Submitted to The 8th Chinese Conference on Pattern Recognition and Computer Vision (2025). Contact to nomodeset@qq.com. Source code will open in 4 months | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image restoration faces challenges including ineffective feature fusion, computational bottlenecks and inefficient diffusion processes. To address these, we propose DiffRWKVIR, a novel framework unifying Test-Time Training (TTT) with efficient diffusion. Our approach introduces three key innovations: (1) Omni-Scale 2D State Evolution extends RWKV's location-dependent parameterization to hierarchical multi-directional 2D scanning, enabling global contextual awareness with linear complexity O(L); (2) Chunk-Optimized Flash Processing accelerates intra-chunk parallelism by 3.2x via contiguous chunk processing (O(LCd) complexity), reducing sequential dependencies and computational overhead; (3) Prior-Guided Efficient Diffusion extracts a compact Image Prior Representation (IPR) in only 5-20 steps, proving 45% faster training/inference than DiffIR while solving computational inefficiency in denoising. Evaluated across super-resolution and inpainting benchmarks (Set5, Set14, BSD100, Urban100, Places365), DiffRWKVIR outperforms SwinIR, HAT, and MambaIR/v2 in PSNR, SSIM, LPIPS, and efficiency metrics. Our method establishes a new paradigm for adaptive, high-efficiency image restoration with optimized hardware utilization. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:01:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lu",
"Rongchang",
""
],
[
"Luo",
"Tianduo",
""
],
[
"Zhang",
"Yunzhi",
""
],
[
"Yue",
"Conghan",
""
],
[
"Yang",
"Pei",
""
],
[
"Liu",
"Guibao",
""
],
[
"Gu",
"Changyang",
""
]
] |
2506.14542 | Shuyang Xie | Xie Shuyang, Zhou Jie, Xu Bo, Wang Jun, Xu Renjing | MobileHolo: A Lightweight Complex-Valued Deformable CNN for High-Quality Computer-Generated Hologram | 8 pages, 9 figures | null | null | null | physics.optics cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Holographic displays have significant potential in virtual reality and augmented reality owing to their ability to provide all the depth cues. Deep learning-based methods play an important role in computer-generated holograms (CGH). During the diffraction process, each pixel exerts an influence on the reconstructed image. However, previous works face challenges in capturing sufficient information to accurately model this process, primarily due to the inadequacy of their effective receptive field (ERF). Here, we designed complex-valued deformable convolution for integration into network, enabling dynamic adjustment of the convolution kernel's shape to increase flexibility of ERF for better feature extraction. This approach allows us to utilize a single model while achieving state-of-the-art performance in both simulated and optical experiment reconstructions, surpassing existing open-source models. Specifically, our method has a peak signal-to-noise ratio that is 2.04 dB, 5.31 dB, and 9.71 dB higher than that of CCNN-CGH, HoloNet, and Holo-encoder, respectively, when the resolution is 1920$\times$1072. The number of parameters of our model is only about one-eighth of that of CCNN-CGH. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:02:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shuyang",
"Xie",
""
],
[
"Jie",
"Zhou",
""
],
[
"Bo",
"Xu",
""
],
[
"Jun",
"Wang",
""
],
[
"Renjing",
"Xu",
""
]
] |
2506.14543 | Julia Novo | Bosco Garc\'ia-Archilla, Alicia Garc\'ia-Mascaraque and Julia Novo | Using BDF schemes in the temporal integration of POD-ROM methods | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | In this paper we consider the numerical approximation of a semilinear reaction-diffusion model problem (PDEs) by means of reduced order methods (ROMs) based on proper orthogonal decomposition (POD). We focus on the time integration of the fully discrete reduced order model. Most of the analysis in the literature has been carried out for the implicit Euler method as time integrator. We integrate in time the reduced order model with the BDF-q time stepping ($1\le q\le 5$) and prove optimal rate of convergence of order $q$ in time. Our set of snapshots is obtained from finite element approximations to the original model problem computed at different times. These finite element approximations can be obtained with any time integrator. The POD method is based on first order difference quotients of the snapshots. The reason for doing this is twofold. On the one hand, the use of difference quotients allow us to provide pointwise-in-time error bounds. On the other, the use of difference quotients is essential to get the expected rate $q$ in time since we apply that the BDF-q time stepping, $1\le q\le 5$, can be written as a linear combination of first order difference quotients. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:03:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"García-Archilla",
"Bosco",
""
],
[
"García-Mascaraque",
"Alicia",
""
],
[
"Novo",
"Julia",
""
]
] |
2506.14544 | Micha{\l} Skrzypczak | Antonio Casares, Pierre Ohlmann, Micha{\l} Skrzypczak, Igor Walukiewicz | Infinite lexicographic products of positional objectives | null | null | null | null | cs.GT | http://creativecommons.org/licenses/by/4.0/ | This paper contributes to the study of positional determinacy of infinite duration games played on potentially infinite graphs. Recently, [Ohlmann, TheoretiCS 2023] established that positionality of prefix-independent objectives is preserved by finite lexicographic products. We propose two different notions of infinite lexicographic products indexed by arbitrary ordinals, and extend Ohlmann's result by proving that they also preserve positionality. In the context of one-player positionality, this extends positional determinacy results of [Gr\"adel and Walukiewicz, Logical Methods in Computer Science 2006] to edge-labelled games and arbitrarily many priorities for both Max-Parity and Min-Parity. Moreover, we show that the Max-Parity objectives over countable ordinals are complete for the infinite levels of the difference hierarchy over $\Sigma^0_2$ and that Min-Parity is complete for the class $\Sigma^0_3$. We obtain therefore positional languages that are complete for all those levels, as well as new insights about closure under unions and neutral letters. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:04:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Casares",
"Antonio",
""
],
[
"Ohlmann",
"Pierre",
""
],
[
"Skrzypczak",
"Michał",
""
],
[
"Walukiewicz",
"Igor",
""
]
] |
2506.14545 | Tommy Lundemo | Tommy Lundemo | On the Residue Sequence in Logarithmic Topological Cyclic Homology | 29 pages, comments welcome! | null | null | null | math.AT math.AG math.KT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As a localizing invariant, THH participates in localization sequences of cyclotomic spectra. We resolve a conjecture of Rognes by relating these to residue sequences in logarithmic THH. Consequently, logarithmic THH, TR, and TC serve as strict generalizations of the constructions of Hesselholt--Madsen and Blumberg--Mandell, which moreover enjoy localization sequences without the regularity hypotheses usually required for d\'evissage. Combined with work of Ramzi--Sosnilo--Winges, our results imply that there exists a stable infinity-category C such that THH(C), TR(C), and TC(C) realize the relevant logarithmic term for specific log structures, such as the natural ones on discrete valuation rings, connective complex K-theory, and truncated Brown--Peterson spectra. Finally, we conjecture that the category C can be chosen to reflect the additional structure present on the logarithmic terms, and we give evidence for this in the case of discrete valuation rings. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:04:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lundemo",
"Tommy",
""
]
] |
2506.14546 | Katya Fouka | Katya Fouka, Athreya Shankar, Ting Rei Tan and Arghavan Safavi-Naini | Multilevel Electromagnetically Induced Transparency Cooling | null | null | null | null | physics.atom-ph quant-ph | http://creativecommons.org/licenses/by/4.0/ | Electromagnetically Induced Transparency (EIT) cooling is a well-established method for preparing trapped ion systems in their motional ground state. However, isolating a three-level system, as required for EIT cooling, is often challenging or impractical. In this work, we extend the EIT cooling framework to multilevel systems where the number of ground states exceeds the number of excited states, ensuring the presence of at least one dark state. We develop a formalism to accurately determine the cooling rate in the weak sideband coupling regime and provide an approximate estimate for cooling rates beyond this regime, without the need for explicit simulation of the motional degree of freedom. We clarify the connection between the cooling rate and the absorption spectrum, offering a pathway for efficient near-ground-state cooling of ions with complex electronic structures. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:04:35 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fouka",
"Katya",
""
],
[
"Shankar",
"Athreya",
""
],
[
"Tan",
"Ting Rei",
""
],
[
"Safavi-Naini",
"Arghavan",
""
]
] |
2506.14547 | Jos\'e Batista | Jos\'e Pedro Alvarinhas Batista, Matthieu J. Verstraete and Alo\"is Castellano | Thermal Conductivity Of Monolayer Hexagonal Boron Nitride: Four-Phonon Scattering And Quantum Sampling Effects | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Monolayer hexagonal boron nitride is a prototypical planar 2-dimensional system material and has been the subject of many investigations of its exceptional vibrational, spectroscopic and transport properties. The lattice thermal conductivity remains quite uncertain, with theoretical and experimental reports varying between 218 and 1060 Wm-1K-1. It has a strong temperature evolution and is sensitive to strain effects and isotope concentrations. While the impact of isotope scattering has been widely studied and is well understood, nuclear quantum effects and 4-phonon scattering have so far been neglected. Monolayer hexagonal boron nitride is composed of light elements, and further has its 3-phonon scattering phase space restricted by mirror plane symmetry, so these effects may be of similar order as isotope scattering, and would lead to a completely different understanding of the fundamental processes limiting the lattice thermal conductivity for this system. In this work, we use both classical and path-integral molecular dynamics, in conjunction with the Temperature Dependent Effective Potential method, to compute temperature-dependent renormalized phonons including isotope scattering, 3-phonon scattering, 4-phonon scattering and nuclear quantum effects. We show the impact of the latter two on the lattice thermal conductivity for a large temperature range, as well as their impact on the phonon lifetimes. Overall, our work provides a robust framework for calculations of the lattice thermal conductivity in solids, providing quantitative improvements and physical understanding that help explain the variety of results found in the literature. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:04:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Batista",
"José Pedro Alvarinhas",
""
],
[
"Verstraete",
"Matthieu J.",
""
],
[
"Castellano",
"Aloïs",
""
]
] |
2506.14548 | Pradeep Kumar Yadav | Pradeep Kumar Yadav, Shradha Mishra, and Sanjay Puri | Coarsening Kinetics in Active Model B+: Macroscale and Microscale Phase Separation | null | null | null | null | cond-mat.soft | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We perform a comprehensive numerical investigation of the coarsening kinetics of active Brownian particles modeled by the {\it Active Model B+} (AMB+). This model was introduced by Tjhung et al. [Phys. Rev. X {\bf 8}, 031080 (2018)] and is a generalization of Model B for a conserved order parameter, with two additional activity terms. These terms correspond to rotation-free current (of strength $\lambda$) and rotational current (of strength $\xi$). We find that the presence of rotational current $(\xi \neq 0)$ significantly affects growth kinetics. Depending on the parameter values, AMB+ exhibits either {\it macroscale phase separation} (MPS) or {\it microscale phase separation} ($\mu$PS). We present detailed results for the kinetics of MPS and $\mu$PS in AMB+ with critical composition. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:04:56 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yadav",
"Pradeep Kumar",
""
],
[
"Mishra",
"Shradha",
""
],
[
"Puri",
"Sanjay",
""
]
] |
2506.14549 | Yong Liu | Yong Liu, Wenpeng Xiao, Qianqian Wang, Junlin Chen, Shiyin Wang, Yitong Wang, Xinglong Wu, Yansong Tang | DreamLight: Towards Harmonious and Consistent Image Relighting | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce a model named DreamLight for universal image relighting in this work, which can seamlessly composite subjects into a new background while maintaining aesthetic uniformity in terms of lighting and color tone. The background can be specified by natural images (image-based relighting) or generated from unlimited text prompts (text-based relighting). Existing studies primarily focus on image-based relighting, while with scant exploration into text-based scenarios. Some works employ intricate disentanglement pipeline designs relying on environment maps to provide relevant information, which grapples with the expensive data cost required for intrinsic decomposition and light source. Other methods take this task as an image translation problem and perform pixel-level transformation with autoencoder architecture. While these methods have achieved decent harmonization effects, they struggle to generate realistic and natural light interaction effects between the foreground and background. To alleviate these challenges, we reorganize the input data into a unified format and leverage the semantic prior provided by the pretrained diffusion model to facilitate the generation of natural results. Moreover, we propose a Position-Guided Light Adapter (PGLA) that condenses light information from different directions in the background into designed light query embeddings, and modulates the foreground with direction-biased masked attention. In addition, we present a post-processing module named Spectral Foreground Fixer (SFF) to adaptively reorganize different frequency components of subject and relighted background, which helps enhance the consistency of foreground appearance. Extensive comparisons and user study demonstrate that our DreamLight achieves remarkable relighting performance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:05:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Yong",
""
],
[
"Xiao",
"Wenpeng",
""
],
[
"Wang",
"Qianqian",
""
],
[
"Chen",
"Junlin",
""
],
[
"Wang",
"Shiyin",
""
],
[
"Wang",
"Yitong",
""
],
[
"Wu",
"Xinglong",
""
],
[
"Tang",
"Yansong",
""
]
] |
2506.14550 | Katsumasa Yoshioka | Katsumasa Yoshioka, Taro Wakamura, Takuya Okamoto, Norio Kumada | Super-diffusive sub-picosecond extraction of hot carriers in black phosphorous | 22 pages, 4 figures, Supplementary information | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Harvesting hot carriers before they lose energy to the lattice is a critical route toward surpassing the conventional thermodynamic limit in optical-to-electrical (O-E) conversion. However, photocurrent from such hot carriers has remained challenging to directly detect because they equilibrate on picosecond timescales, outpacing conventional electronic measurement. Here, by employing terahertz electronics with sub-picosecond temporal resolution, we directly monitor hot-carrier-driven O-E conversion in black phosphorus (BP). Photoexcitation near the metal contact under zero source-drain bias generates an ultrafast photocurrent with a decay time of ~400 fs, orders of magnitude faster than the typical sub-nanosecond energy relaxation in BP, demonstrating a measured 3 dB bandwidth of 260 GHz with an intrinsic limit of ~600 GHz. Notably, this photocurrent flows via energetic holes toward the contact electrode, regardless of the equilibrium carrier type, revealing a super-diffusive hot-carrier extraction mechanism. Furthermore, we show that the ultrafast hot-carrier contribution can coexist with the much slower cold-carrier contribution based on the photovoltaic effect, demonstrating that hot carriers can be harvested without discarding lower-energy carriers. These findings highlight the potential of sub-picosecond hot-carrier extraction to expand the O-E conversion bandwidth without sacrificing efficiency, bridging fundamental hot-carrier physics with ultrahigh-speed technological applications. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:05:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yoshioka",
"Katsumasa",
""
],
[
"Wakamura",
"Taro",
""
],
[
"Okamoto",
"Takuya",
""
],
[
"Kumada",
"Norio",
""
]
] |
2506.14551 | Alex Newkirk | Alex C. Newkirk, Jared Fernandez, Jonathan Koomey, Imran Latif, Emma Strubell, Arman Shehabi, Constantine Samaras | Empirically-Calibrated H100 Node Power Models for Reducing Uncertainty in AI Training Energy Estimation | 4 figures, 22 pages | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | As AI's energy demand continues to grow, it is critical to enhance the understanding of characteristics of this demand, to improve grid infrastructure planning and environmental assessment. By combining empirical measurements from Brookhaven National Laboratory during AI training on 8-GPU H100 systems with open-source benchmarking data, we develop statistical models relating computational intensity to node-level power consumption. We measure the gap between manufacturer-rated thermal design power (TDP) and actual power demand during AI training. Our analysis reveals that even computationally intensive workloads operate at only 76% of the 10.2 kW TDP rating. Our architecture-specific model, calibrated to floating-point operations, predicts energy consumption with 11.4% mean absolute percentage error, significantly outperforming TDP-based approaches (27-37% error). We identified distinct power signatures between transformer and CNN architectures, with transformers showing characteristic fluctuations that may impact grid stability. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 14:07:27 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Newkirk",
"Alex C.",
""
],
[
"Fernandez",
"Jared",
""
],
[
"Koomey",
"Jonathan",
""
],
[
"Latif",
"Imran",
""
],
[
"Strubell",
"Emma",
""
],
[
"Shehabi",
"Arman",
""
],
[
"Samaras",
"Constantine",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.