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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.08131 | Mohammad Morid | Mohammad Amin Morid, Robert E Tillman, Eran Halperin | Balanced Area Deprivation Index (bADI): Enhancing social determinants of health indices to strengthen their association with healthcare clinical outcomes, utilization and costs | null | Journal of Public Health Management & Practice, 2025 | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | Background: As value-based care expands across the U.S. healthcare system, reducing health disparities has become a priority. Social determinants of health (SDoH) indices, like the widely used Area Deprivation Index (ADI), guide efforts to manage patient health and costs. However, the ADI's reliance on housing-related variables (e.g., median home value) may reduce its effectiveness, especially in high-cost regions, by masking inequalities and poor health outcomes.
Methods: To overcome these limitations, we developed the balanced ADI (bADI), a new SDoH index that reduces dependence on housing metrics through standardized construction. We evaluated the bADI using data from millions of Medicare Fee-for-Service and Medicare Advantage beneficiaries. Correlation analyses measured its association with clinical outcomes, life expectancy, healthcare use, and cost, and compared results to the ADI.
Results: The bADI showed stronger correlations with clinical outcomes and life expectancy than the ADI. It was less influenced by housing costs in expensive regions and more accurately predicted healthcare use and costs. While ADI-based research suggested both the most and least disadvantaged groups had higher healthcare costs, the bADI revealed a more nuanced pattern, showing more accurate cost differences across groups.
Conclusions: The bADI offers stronger predictive power for healthcare outcomes and spending, making it a valuable tool for accountable care organizations. By reallocating resources from less to more disadvantaged areas, ACOs could use the bADI to promote equity and cost-effective care within population health initiatives. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:34:39 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Morid",
"Mohammad Amin",
""
],
[
"Tillman",
"Robert E",
""
],
[
"Halperin",
"Eran",
""
]
] |
2506.08132 | Erfan Nosrati | Erfan Nosrati, Majid Ghaderi | Congestion-Aware Path Selection for Load Balancing in AI Clusters | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fast training of large machine learning models requires distributed training on AI clusters consisting of thousands of GPUs. The efficiency of distributed training crucially depends on the efficiency of the network interconnecting GPUs in the cluster. These networks are commonly built using RDMA following a Clos-like datacenter topology. To efficiently utilize the network bandwidth, load balancing is employed to distribute traffic across multiple redundant paths. While there exists numerous techniques for load-balancing in traditional datacenters, these are often either optimized for TCP traffic or require specialized network hardware, thus limiting their utility in AI clusters.
This paper presents the design and evaluation of Hopper, a new load-balancing technique optimized for RDMA traffic in AI clusters. Operating entirely at the host level, Hopper requires no specialized hardware or modifications to network switches. It continuously monitors the current path for congestion and dynamically switches traffic to a less congested path when congestion is detected. Furthermore, it incorporates a lightweight mechanism to identify alternative paths and carefully controls the timing of path switching to prevent excessive out-of-order packets.
We evaluated Hopper using ns-3 simulations and a testbed implementation. Our evaluations show that Hopper reduces the average and 99-percentile tail flow completion time by up to 20% and 14%, respectively, compared to state-of-the-art host-based load balancing techniques. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:35:18 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nosrati",
"Erfan",
""
],
[
"Ghaderi",
"Majid",
""
]
] |
2506.08133 | Daniel Erman | Daniel Erman | About how large are algebraic Betti numbers? | 12 pages | null | null | null | math.AC math.AG | http://creativecommons.org/licenses/by/4.0/ | We use Boij-S\"oderberg theory to provide some order of magnitude bounds on algebraic Betti numbers. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:36:40 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Erman",
"Daniel",
""
]
] |
2506.08134 | Samuel Holt | Qiyao Wei, Samuel Holt, Jing Yang, Markus Wulfmeier, Mihaela van der Schaar | The AI Imperative: Scaling High-Quality Peer Review in Machine Learning | 18 pages, 3 figures. Position paper | null | null | null | cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | Peer review, the bedrock of scientific advancement in machine learning (ML), is strained by a crisis of scale. Exponential growth in manuscript submissions to premier ML venues such as NeurIPS, ICML, and ICLR is outpacing the finite capacity of qualified reviewers, leading to concerns about review quality, consistency, and reviewer fatigue. This position paper argues that AI-assisted peer review must become an urgent research and infrastructure priority. We advocate for a comprehensive AI-augmented ecosystem, leveraging Large Language Models (LLMs) not as replacements for human judgment, but as sophisticated collaborators for authors, reviewers, and Area Chairs (ACs). We propose specific roles for AI in enhancing factual verification, guiding reviewer performance, assisting authors in quality improvement, and supporting ACs in decision-making. Crucially, we contend that the development of such systems hinges on access to more granular, structured, and ethically-sourced peer review process data. We outline a research agenda, including illustrative experiments, to develop and validate these AI assistants, and discuss significant technical and ethical challenges. We call upon the ML community to proactively build this AI-assisted future, ensuring the continued integrity and scalability of scientific validation, while maintaining high standards of peer review. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:37:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wei",
"Qiyao",
""
],
[
"Holt",
"Samuel",
""
],
[
"Yang",
"Jing",
""
],
[
"Wulfmeier",
"Markus",
""
],
[
"van der Schaar",
"Mihaela",
""
]
] |
2506.08135 | Hermano Velten | Hermano Velten, Felipe S. Esc\'orcio and Nadson J. S. Trindade | Anisotropic quantum polytropes | 12 pages, 5 figures | null | null | null | gr-qc astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The structure of astrophysical objects is usually modeled under the assumption of hydrostatic equilibrium. However, actual configurations may deviate from perfect spherical or isotropic properties. Consequently, cosmic objects are expected to exhibit some degree of anisotropy. This consideration also extends to hypothetical dark structures, such as dark stars and dark matter halos. Although the nature of dark matter remains unknown, axion-like particles (ALPs) are strong candidates, suggesting that dark matter halos may have originated from bosonic configurations undergoing gravitational collapse, sustained by boson-boson interactions in the condensate state. This system is described by the Gross-Pitaevskii-Poisson equation. Furthermore, within the framework of the Bohm-de Broglie approach, quantum effects,encapsulated in the so-called quantum potential, may play a significant role in equilibrium astrophysical configurations. In this study, we examine a class of static anisotropic boson stars which are non-minimally coupled to gravity. By including all these factors, we derive a generalized Lane-Emden-like equation and conduct a detailed analysis of the maximum degree of anisotropy that such systems can sustain, thereby identifying physically viable equilibrium configurations. Apart from focusing on the impact of anisotropic contributions, we find that for the so-called Quantum Polytropes (when the quantum potential is the main responsible for the equilibrium condition), the anisotropic factor and the gravitational field have opposite roles compared to the classical case. This leads to a new class of hydrostatic equilibrium objects. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:37:44 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Velten",
"Hermano",
""
],
[
"Escórcio",
"Felipe S.",
""
],
[
"Trindade",
"Nadson J. S.",
""
]
] |
2506.08136 | Zefang Liu | Zefang Liu, Yinzhu Quan | EconWebArena: Benchmarking Autonomous Agents on Economic Tasks in Realistic Web Environments | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce EconWebArena, a benchmark for evaluating autonomous agents on complex, multimodal economic tasks in realistic web environments. The benchmark comprises 360 curated tasks from 82 authoritative websites spanning domains such as macroeconomics, labor, finance, trade, and public policy. Each task challenges agents to navigate live websites, interpret structured and visual content, interact with real interfaces, and extract precise, time-sensitive data through multi-step workflows. We construct the benchmark by prompting multiple large language models (LLMs) to generate candidate tasks, followed by rigorous human curation to ensure clarity, feasibility, and source reliability. Unlike prior work, EconWebArena emphasizes fidelity to authoritative data sources and the need for grounded web-based economic reasoning. We evaluate a diverse set of state-of-the-art multimodal LLMs as web agents, analyze failure cases, and conduct ablation studies to assess the impact of visual grounding, plan-based reasoning, and interaction design. Our results reveal substantial performance gaps and highlight persistent challenges in grounding, navigation, and multimodal understanding, positioning EconWebArena as a rigorous testbed for economic web intelligence. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:39:48 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Liu",
"Zefang",
""
],
[
"Quan",
"Yinzhu",
""
]
] |
2506.08138 | William Gebhardt | William Gebhardt, Alexander G. Ororbia, Nathan McDonald, Clare Thiem, Jack Lombardi | A Practical Guide to Tuning Spiking Neuronal Dynamics | null | null | null | null | cs.NE q-bio.NC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we examine fundamental elements of spiking neural networks (SNNs) as well as how to tune them. Concretely, we focus on two different foundational neuronal units utilized in SNNs -- the leaky integrate-and-fire (LIF) and the resonate-and-fire (RAF) neuron. We explore key equations and how hyperparameter values affect behavior. Beyond hyperparameters, we discuss other important design elements of SNNs -- the choice of input encoding and the setup for excitatory-inhibitory populations -- and how these impact LIF and RAF dynamics. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:41:11 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Gebhardt",
"William",
""
],
[
"Ororbia",
"Alexander G.",
""
],
[
"McDonald",
"Nathan",
""
],
[
"Thiem",
"Clare",
""
],
[
"Lombardi",
"Jack",
""
]
] |
2506.08139 | Aviad Susman | Aviad Susman, Mayte Su\'arez-Fari\~nas, Joseph T Colonel | Nearness of Neighbors Attention for Regression in Supervised Finetuning | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | It is common in supervised machine learning to combine the feature extraction capabilities of neural networks with the predictive power of traditional algorithms, such as k-nearest neighbors (k-NN) or support vector machines. This procedure involves performing supervised fine-tuning (SFT) on a domain-appropriate feature extractor, followed by training a traditional predictor on the resulting SFT embeddings. When used in this manner, traditional predictors often deliver increased performance over the SFT model itself, despite the fine-tuned feature extractor yielding embeddings specifically optimized for prediction by the neural network's final dense layer. This suggests that directly incorporating traditional algorithms into SFT as prediction layers may further improve performance. However, many traditional algorithms have not been implemented as neural network layers due to their non-differentiable nature and their unique optimization requirements. As a step towards solving this problem, we introduce the Nearness of Neighbors Attention (NONA) regression layer. NONA uses the mechanics of neural network attention and a novel learned attention-masking scheme to yield a differentiable proxy of the k-NN regression algorithm. Results on multiple unstructured datasets show improved performance over both dense layer prediction and k-NN on SFT embeddings for regression. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:41:48 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Susman",
"Aviad",
""
],
[
"Suárez-Fariñas",
"Mayte",
""
],
[
"Colonel",
"Joseph T",
""
]
] |
2506.08140 | Yifei Li | Yifei Li, Hanane Nour Moussa, Ziru Chen, Shijie Chen, Botao Yu, Mingyi Xue, Benjamin Burns, Tzu-Yao Chiu, Vishal Dey, Zitong Lu, Chen Wei, Qianheng Zhang, Tianyu Zhang, Song Gao, Xuhui Huang, Xia Ning, Nesreen K. Ahmed, Ali Payani, Huan Sun | AutoSDT: Scaling Data-Driven Discovery Tasks Toward Open Co-Scientists | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | Despite long-standing efforts in accelerating scientific discovery with AI, building AI co-scientists remains challenging due to limited high-quality data for training and evaluation. To tackle this data scarcity issue, we present AutoSDT, an automatic pipeline that collects high-quality coding tasks in real-world data-driven discovery workflows. AutoSDT leverages the coding capabilities and parametric knowledge of LLMs to search for diverse sources, select ecologically valid tasks, and synthesize accurate task instructions and code solutions. Using our pipeline, we construct AutoSDT-5K, a dataset of 5,404 coding tasks for data-driven discovery that covers four scientific disciplines and 756 unique Python packages. To the best of our knowledge, AutoSDT-5K is the only automatically collected and the largest open dataset for data-driven scientific discovery. Expert feedback on a subset of 256 tasks shows the effectiveness of AutoSDT: 93% of the collected tasks are ecologically valid, and 92.2% of the synthesized programs are functionally correct. Trained on AutoSDT-5K, the Qwen2.5-Coder-Instruct LLM series, dubbed AutoSDT-Coder, show substantial improvement on two challenging data-driven discovery benchmarks, ScienceAgentBench and DiscoveryBench. Most notably, AutoSDT-Coder-32B reaches the same level of performance as GPT-4o on ScienceAgentBench with a success rate of 7.8%, doubling the performance of its base model. On DiscoveryBench, it lifts the hypothesis matching score to 8.1, bringing a 17.4% relative improvement and closing the gap between open-weight models and GPT-4o. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:41:50 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Li",
"Yifei",
""
],
[
"Moussa",
"Hanane Nour",
""
],
[
"Chen",
"Ziru",
""
],
[
"Chen",
"Shijie",
""
],
[
"Yu",
"Botao",
""
],
[
"Xue",
"Mingyi",
""
],
[
"Burns",
"Benjamin",
""
],
[
"Chiu",
"Tzu-Yao",
""
],
[
"Dey",
"Vishal",
""
],
[
"Lu",
"Zitong",
""
],
[
"Wei",
"Chen",
""
],
[
"Zhang",
"Qianheng",
""
],
[
"Zhang",
"Tianyu",
""
],
[
"Gao",
"Song",
""
],
[
"Huang",
"Xuhui",
""
],
[
"Ning",
"Xia",
""
],
[
"Ahmed",
"Nesreen K.",
""
],
[
"Payani",
"Ali",
""
],
[
"Sun",
"Huan",
""
]
] |
2506.08141 | B Lekshmi | B Lekshmi, Sushanta Tripathy, Kiran Jain and Alexei Pevtsov | Dynamics of Subsurface Flows in Solar Active Regions During the May 2024 Storm | 12 pages, 6 figures, accepted for publication in The Astrophysical Journal | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | In May 2024, the Sun exhibited intense magnetic activity, marked by numerous high-intensity flares resulting from the interaction and merging of NOAA ARs 13664 and 13668 in the southern hemisphere and AR 13663 in the northern hemisphere. Notably, AR 13664 displayed an extended lifetime, remaining visible after a full solar rotation and continuing to produce significant flaring activity. In this study, we investigate the evolution of sub-photospheric plasma flows associated with these ARs during their disk passage using ring-diagram analysis of SDO/HMI Dopplergrams. We analyze flow divergence, vorticity, and kinetic helicity across depths from the surface to 25 Mm, revealing pronounced temporal and depth-dependent variations. Our observations indicate that the majority of flares occur on the days when the Normalized Helicity Gradient Variance, a measure of kinetic helicity spread, peaks or on the following day. Furthermore, we examine the relationship between subsurface flow dynamics and surface magnetic properties of these complex active regions to understand the interaction between them. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:42:49 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Lekshmi",
"B",
""
],
[
"Tripathy",
"Sushanta",
""
],
[
"Jain",
"Kiran",
""
],
[
"Pevtsov",
"Alexei",
""
]
] |
2506.08142 | Michael McGuffin | Michael J. McGuffin, Jean-Marc Robert, Kazuki Ikeda | How to Write a Simulator for Quantum Circuits from Scratch: A Tutorial | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This tutorial guides a competent programmer through the crafting of a quantum circuit simulator from scratch, even for readers with almost no prior experience in quantum computing. Open source simulators for quantum circuits already exist, but a deeper understanding is gained by writing ones own. With roughly 1000-2000 lines of code, one can simulate Hadamard, Pauli X, Y, Z, SWAP, and other quantum logic gates, with arbitrary combinations of control and anticontrol qubits, on circuits of up to 20+ qubits, with no special libraries, on a personal computer. We explain key algorithms for a simulator, in particular: qubit-wise multiplication for updating the state vector, and partial trace for finding a reduced density matrix. We also discuss optimizations, and how to compute qubit phase, purity, and other statistics. A complete example implementation in JavaScript is available at https://github.com/MJMcGuffin/muqcs.js , which also demonstrates how to compute von Neumann entropy, concurrence (to quantify entanglement), and magic, while remaining much smaller and easier to study than other popular software packages. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:43:46 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"McGuffin",
"Michael J.",
""
],
[
"Robert",
"Jean-Marc",
""
],
[
"Ikeda",
"Kazuki",
""
]
] |
2506.08143 | Francesco Tonin | Francesco Tonin, Alex Lambert, Johan A. K. Suykens, Volkan Cevher | Accelerating Spectral Clustering under Fairness Constraints | ICML 2025 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Fairness of decision-making algorithms is an increasingly important issue. In this paper, we focus on spectral clustering with group fairness constraints, where every demographic group is represented in each cluster proportionally as in the general population. We present a new efficient method for fair spectral clustering (Fair SC) by casting the Fair SC problem within the difference of convex functions (DC) framework. To this end, we introduce a novel variable augmentation strategy and employ an alternating direction method of multipliers type of algorithm adapted to DC problems. We show that each associated subproblem can be solved efficiently, resulting in higher computational efficiency compared to prior work, which required a computationally expensive eigendecomposition. Numerical experiments demonstrate the effectiveness of our approach on both synthetic and real-world benchmarks, showing significant speedups in computation time over prior art, especially as the problem size grows. This work thus represents a considerable step forward towards the adoption of fair clustering in real-world applications. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:46:27 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tonin",
"Francesco",
""
],
[
"Lambert",
"Alex",
""
],
[
"Suykens",
"Johan A. K.",
""
],
[
"Cevher",
"Volkan",
""
]
] |
2506.08144 | Armen Sedrakian | Mark Alford, Arus Harutyunyan, Armen Sedrakian, Stefanos Tsiopelas | Bulk Viscosity of Two-Flavor Color Superconducting Quark Matter in Neutron Star Mergers | 27 pages, 9 figures | null | null | null | nucl-th astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | This work investigates the bulk viscosity of warm, dense,
neutrino-transparent, color-superconducting quark matter, where
damping of density oscillations in the kHz frequency range arises
from weak-interaction-driven direct Urca processes involving
quarks. We study the two-flavor red-green paired
color-superconducting (2SC) phase, while allowing for the presence
of unpaired strange quarks and blue color light quarks of all
flavors. Our calculations are based on the SU(3) Nambu-Jona-Lasinio
(NJL) model, extended to include both vector interactions and the
't Hooft determinant term. The primary focus is on how variations in
the NJL Lagrangian parameters - specifically, the diquark and vector
coupling strengths - affect both the static properties of quark
matter, such as its equation of state and composition, and its
dynamical behavior, including bulk viscosity and associated damping
timescales. We find that the bulk viscosity and corresponding
damping timescale can change by more than an order of magnitude upon
varying the vector coupling by a factor of two at high densities and
by a lesser degree at lower densities. This sensitivity primarily
arises from the susceptibility of 2SC matter, with a smaller
contribution from modifications to the weak interaction rates. In
comparison, changes in the diquark coupling have a more limited
impact. The damping of density oscillations in 2SC matter is similar
quantitatively to nucleonic matter and can be a leading mechanism of
dissipation in merging hybrid stars containing color superconducting
cores. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:48:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Alford",
"Mark",
""
],
[
"Harutyunyan",
"Arus",
""
],
[
"Sedrakian",
"Armen",
""
],
[
"Tsiopelas",
"Stefanos",
""
]
] |
2506.08145 | Constantine Yannouleas | Constantine Yannouleas and Uzi Landman | Wave-function microscopy: Derivation and anatomy of exact algebraic spinful wave functions and full Wigner-molecular spectra of a few highly correlated rapidly rotating ultracold fermionic atoms | 16 pages with 5 color figures. Accepted for publication in Phys. Rev. A. For related papers, see https://sites.gatech.edu/cyannouleas/ | null | null | null | cond-mat.quant-gas nucl-th quant-ph | http://creativecommons.org/licenses/by/4.0/ | Exploring strongly correlated spinful states of few fermionic ultracold atoms in a rapidly rotating trap, an example of which was recently realized for two fermionic $^6$Li atoms in an optical tweezer, we derive analytical (algebraic) total-spin-eigenstate wavefunctions through the development and employment of a theoretical platform that integrates exact numerical diagonalization (full configuration interaction, FCI) with symbolic language processing. For such rapid rotations, where the atoms occupy the lowest Landau level (LLL), the obtained algebraic expressions can address the full LLL spectrum in all its complexity, demonstrating that their spatial, spectral, and spin characteristics manifest formation of collectively rotating and vibrating Wigner molecules. The explicitly exhibited analytic wavefunctions (for two and three spinful $^6$Li atoms) reproduce precisely the corresponding numerical FCI results, and they are shown to reach beyond the limited range of applicability of previous Jastrow-type treatments. These results, and their extension to bosonic systems, provide the impetus and analysis tools for future experimental and theoretical simulations of larger mesoscopic systems | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:49:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yannouleas",
"Constantine",
""
],
[
"Landman",
"Uzi",
""
]
] |
2506.08146 | Francisco Sahli Costabal | Vahidullah Ta\c{c}, Amirhossein Amiri-Hezaveh, Manuel K. Rausch, Grace N. Bechtel, Francisco Sahli Costabal, Adrian Buganza Tepole | Fully data-driven inverse hyperelasticity with hyper-network neural ODE fields | null | null | null | null | cs.LG physics.comp-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We propose a new framework for identifying mechanical properties of heterogeneous materials without a closed-form constitutive equation. Given a full-field measurement of the displacement field, for instance as obtained from digital image correlation (DIC), a continuous approximation of the strain field is obtained by training a neural network that incorporates Fourier features to effectively capture sharp gradients in the data. A physics-based data-driven method built upon ordinary neural differential equations (NODEs) is employed to discover constitutive equations. The NODE framework can represent arbitrary materials while satisfying constraints in the theory of constitutive equations by default. To account for heterogeneity, a hyper-network is defined, where the input is the material coordinate system, and the output is the NODE-based constitutive equation. The parameters of the hyper-network are optimized by minimizing a multi-objective loss function that includes penalty terms for violations of the strong form of the equilibrium equations of elasticity and the associated Neumann boundary conditions. We showcase the framework with several numerical examples, including heterogeneity arising from variations in material parameters, spatial transitions from isotropy to anisotropy, material identification in the presence of noise, and, ultimately, application to experimental data. As the numerical results suggest, the proposed approach is robust and general in identifying the mechanical properties of heterogeneous materials with very few assumptions, making it a suitable alternative to classical inverse methods. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:50:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Taç",
"Vahidullah",
""
],
[
"Amiri-Hezaveh",
"Amirhossein",
""
],
[
"Rausch",
"Manuel K.",
""
],
[
"Bechtel",
"Grace N.",
""
],
[
"Costabal",
"Francisco Sahli",
""
],
[
"Tepole",
"Adrian Buganza",
""
]
] |
2506.08147 | Muhammad Usman | Muhammad Usman, Muhammad Ahmad, M. Shahiki Tash, Irina Gelbukh, Rolando Quintero Tellez and Grigori Sidorov | Multilingual Hate Speech Detection in Social Media Using Translation-Based Approaches with Large Language Models | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Social media platforms are critical spaces for public discourse, shaping opinions and community dynamics, yet their widespread use has amplified harmful content, particularly hate speech, threatening online safety and inclusivity. While hate speech detection has been extensively studied in languages like English and Spanish, Urdu remains underexplored, especially using translation-based approaches. To address this gap, we introduce a trilingual dataset of 10,193 tweets in English (3,834 samples), Urdu (3,197 samples), and Spanish (3,162 samples), collected via keyword filtering, with a balanced distribution of 4,849 Hateful and 5,344 Not-Hateful labels. Our methodology leverages attention layers as a precursor to transformer-based models and large language models (LLMs), enhancing feature extraction for multilingual hate speech detection. For non-transformer models, we use TF-IDF for feature extraction. The dataset is benchmarked using state-of-the-art models, including GPT-3.5 Turbo and Qwen 2.5 72B, alongside traditional machine learning models like SVM and other transformers (e.g., BERT, RoBERTa). Three annotators, following rigorous guidelines, ensured high dataset quality, achieving a Fleiss' Kappa of 0.821. Our approach, integrating attention layers with GPT-3.5 Turbo and Qwen 2.5 72B, achieves strong performance, with macro F1 scores of 0.87 for English (GPT-3.5 Turbo), 0.85 for Spanish (GPT-3.5 Turbo), 0.81 for Urdu (Qwen 2.5 72B), and 0.88 for the joint multilingual model (Qwen 2.5 72B). These results reflect improvements of 8.75% in English (over SVM baseline 0.80), 8.97% in Spanish (over SVM baseline 0.78), 5.19% in Urdu (over SVM baseline 0.77), and 7.32% in the joint multilingual model (over SVM baseline 0.82). Our framework offers a robust solution for multilingual hate speech detection, fostering safer digital communities worldwide. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:53:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Usman",
"Muhammad",
""
],
[
"Ahmad",
"Muhammad",
""
],
[
"Tash",
"M. Shahiki",
""
],
[
"Gelbukh",
"Irina",
""
],
[
"Tellez",
"Rolando Quintero",
""
],
[
"Sidorov",
"Grigori",
""
]
] |
2506.08148 | Yurii Ignat'ev | Yurii Ignat'ev | From Geometry to Physics -- AZ: touches to the portrait | 13 pages, 3 figures, 11 references | null | null | null | physics.hist-ph gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The process launched by Lobachevsky. The movement of the Kazan school of geometry towards physics. Personal memories of Alexei Zinovievich Petrov, the great Kazan geometer and theoretical physicist, who became the Author's Guiding Star. KeyWords: Alexey Zinovievich Petrov, geometry, general theory of relativity, Kazan University, Department of Relativity and Gravitation Theory, methods of teaching exact sciences. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:55:09 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ignat'ev",
"Yurii",
""
]
] |
2506.08149 | Hang Wang | Hang Wang, Dechen Gao, Junshan Zhang | Ego-centric Learning of Communicative World Models for Autonomous Driving | null | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | We study multi-agent reinforcement learning (MARL) for tasks in complex high-dimensional environments, such as autonomous driving. MARL is known to suffer from the \textit{partial observability} and \textit{non-stationarity} issues. To tackle these challenges, information sharing is often employed, which however faces major hurdles in practice, including overwhelming communication overhead and scalability concerns. By making use of generative AI embodied in world model together with its latent representation, we develop {\it CALL}, \underline{C}ommunic\underline{a}tive Wor\underline{l}d Mode\underline{l}, for MARL, where 1) each agent first learns its world model that encodes its state and intention into low-dimensional latent representation with smaller memory footprint, which can be shared with other agents of interest via lightweight communication; and 2) each agent carries out ego-centric learning while exploiting lightweight information sharing to enrich her world model, and then exploits its generalization capacity to improve prediction for better planning. We characterize the gain on the prediction accuracy from the information sharing and its impact on performance gap. Extensive experiments are carried out on the challenging local trajectory planning tasks in the CARLA platform to demonstrate the performance gains of using \textit{CALL}. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:56:40 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Hang",
""
],
[
"Gao",
"Dechen",
""
],
[
"Zhang",
"Junshan",
""
]
] |
2506.08150 | Susana Hahn Martin Lunas | Arvid Becker, Pedro Cabalar, Martin Di\'eguez, Javier Romero, Susana Hahn, Torsten Schaub | Compiling Metric Temporal Answer Set Programming | null | null | null | null | cs.AI cs.LO | http://creativecommons.org/licenses/by/4.0/ | We develop a computational approach to Metric Answer Set Programming (ASP) to allow for expressing quantitative temporal constrains, like durations and deadlines. A central challenge is to maintain scalability when dealing with fine-grained timing constraints, which can significantly exacerbate ASP's grounding bottleneck. To address this issue, we leverage extensions of ASP with difference constraints, a simplified form of linear constraints, to handle time-related aspects externally. Our approach effectively decouples metric ASP from the granularity of time, resulting in a solution that is unaffected by time precision. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:56:57 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Becker",
"Arvid",
""
],
[
"Cabalar",
"Pedro",
""
],
[
"Diéguez",
"Martin",
""
],
[
"Romero",
"Javier",
""
],
[
"Hahn",
"Susana",
""
],
[
"Schaub",
"Torsten",
""
]
] |
2506.08151 | Rafa{\l} Pyzik | Rafa{\l} Pyzik | Treewidth of Outer $k$-Planar Graphs | null | null | null | null | cs.DM math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Treewidth is an important structural graph parameter that quantifies how closely a graph resembles a tree-like structure. It has applications in many algorithmic and combinatorial problems. In this paper, we study treewidth of outer $k$-planar graphs - graphs admitting a convex drawing where all vertices lie on a circle and each edge crosses at most $k$ other edges. We also consider a more general class of outer min-$k$-planar graphs, which are graphs admitting a convex drawing where for every crossing of two edges at least one of these edges is crossed at most $k$ times.
Firman, Gutowski, Kryven, Okada and Wolff [GD 2024] proved that every outer $k$-planar graph has treewidth at most $1.5k+2$ and provided a lower bound of $k+2$ for even $k$. We establish a lower bound of $1.5k+0.5$ for every odd $k$. Additionally, they showed that every outer min-$k$-planar graph has treewidth at most $3k+1$. We improve this upper bound to $3 \cdot \lfloor 0.5k \rfloor+4$.
Our approach also allows us to upper bound the separation number, a parameter closely related to treewidth, of outer min-$k$-planar graphs by $2 \cdot \lfloor 0.5k \rfloor+4$. This improves the previous bound of $2k+1$ and achieves a bound with an optimal multiplicative constant. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:58:23 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Pyzik",
"Rafał",
""
]
] |
2506.08152 | Luis Fernando Lozano | Luis Fernando Lozano, Grigori Chapiro and Dan Marchesin | The Riemann problem for three-phase foam flow in porous media | null | null | null | null | math.AP math-ph math.MP physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Gas injection in the context of the three-phase flow in porous media appears in applications such as Enhanced Oil Recovery, aquifer remediation, and carbon capture, utilization, and storage (CCUS). In general, this technique suffers from a difficulty related to excessive gas mobility, which can be circumvented by using foam. This study addresses the non-linear system of differential equations describing the three-phase foam flow based on Corey relative permeability functions. A major obstacle is an umbilic point, where the characteristic wave velocities for different families coincide, complicating the identification of stable wave structures.
We developed a methodology to solve the Riemann problem describing the three-phase foam displacement in the case when the gas viscosity exceeds that of oil and water. To allow the analysis, we assume foam in local equilibrium (or maximum foam texture), resulting in a constant mobility reduction factor (MRF). These simplifications allowed the classification of possible solutions for the injection of foamed gas and water mixtures under a wide range of initial conditions within the framework of non-classical Conservation Law Theory. As a relevant industrial application of the proposed solution, we investigate the conditions resulting in oil bank formation. Besides improving the general physical understanding of foam flow in a porous medium, this analysis can be applied to calibrate numerical simulators and perform uncertainty quantification. Our analytical estimates were validated through numerical simulations. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 18:59:02 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Lozano",
"Luis Fernando",
""
],
[
"Chapiro",
"Grigori",
""
],
[
"Marchesin",
"Dan",
""
]
] |
2506.08153 | Renato Cordeiro Ferreira | Renato Cordeiro Ferreira (1,2,3,4) ((1) University of S\~ao Paulo, (2) Jheronimus Academy of Data Science, (3) Technical University of Eindhoven, (4) Tilburg University) | A Metrics-Oriented Architectural Model to Characterize Complexity on Machine Learning-Enabled Systems | 4 pages, 3 figures (2 diagrams, 1 table), to be published in CAIN 2025 | null | null | null | cs.SE cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | How can the complexity of ML-enabled systems be managed effectively? The goal of this research is to investigate how complexity affects ML-Enabled Systems (MLES). To address this question, this research aims to introduce a metrics-based architectural model to characterize the complexity of MLES. The goal is to support architectural decisions, providing a guideline for the inception and growth of these systems. This paper showcases the first step for creating the metrics-based architectural model: an extension of a reference architecture that can describe MLES to collect their metrics. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:02:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ferreira",
"Renato Cordeiro",
"",
"University of São Paulo",
"Jheronimus Academy of Data Science",
"Technical University of Eindhoven",
"Tilburg University"
]
] |
2506.08154 | Abigail Watkins | Sebastian Heinrich, Julia Plavnik, Ingo Runkel, Abigail Watkins | Generalised Orbifolds and G-equivariantisation | 34 pages | null | null | null | math.QA hep-th math.CT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a construction motivated by topological field theory, a so-called orbifold datum $\mathbb{A}$ in a ribbon category $C$ allows one to define a new ribbon category $C_{\mathbb{A}}$. If $C$ is the neutral component of a $G$-crossed ribbon category $B$, and $\mathbb{A}$ is an orbifold datum in $C$ defined in terms of $B$, one finds that $C_{\mathbb{A}}$ is equivalent to the equivariantisation $B^G$ of $B$ as a ribbon category. We give a constructive proof of this equivalence. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:05:15 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Heinrich",
"Sebastian",
""
],
[
"Plavnik",
"Julia",
""
],
[
"Runkel",
"Ingo",
""
],
[
"Watkins",
"Abigail",
""
]
] |
2506.08155 | Sammy No\"el Paris\'e | Sammy Noel Parise, Raphael Dube-Demers, David Turgeon, Alireza Geravand, Sophie LaRochelle, and Wei Shi | Fabry-Perot-Insensitive Edge Coupling for Robust PIC Characterization | 8 pages and 6 figures | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | In this work, we investigate the impact of Fabry-Perot (FP) cavities formed between a fiber array and a silicon chip during edge-coupled testing of photonic integrated circuits (PICs). Our results show that subwavelength variations in the distance between the fiber array and the device under test induce coupling efficiency modulations of several tenths of a decibel - challenging current alignment precision and undermining measurement repeatability in wafer-level probing. To address these FP-induced fluctuations, we introduce a complementary dual-measurement technique that shifts the chip-fiber separation by a quarter wavelength, thereby generating phase-opposite modulation artifacts that cancel upon averaging. This approach allows a higher tolerance in probe-to-PIC positioning by substantially eliminating the FP modulation. We also demonstrate how it can be leveraged to enhance repeatability and accurate positioning during alignment, reducing the insertion loss variability between multiple measurements from 0.34 to 0.04 dB. Such precision is essential for emerging applications like quantum photonics, silicon photonic sensors, and high-bandwidth optical interconnects where measurement accuracy directly impacts device performance characterization. Ultimately, our method offers a robust and accurate solution for high-throughput PIC characterization without requiring index-matching gel, which is impractical for wafer-level testing. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:05:44 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Parise",
"Sammy Noel",
""
],
[
"Dube-Demers",
"Raphael",
""
],
[
"Turgeon",
"David",
""
],
[
"Geravand",
"Alireza",
""
],
[
"LaRochelle",
"Sophie",
""
],
[
"Shi",
"Wei",
""
]
] |
2506.08156 | Travis Leadbetter | Travis Leadbetter, Prashant K. Purohit, Celia Reina | On a structure preserving closure of Langevin dynamics | null | null | null | null | cond-mat.stat-mech | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Given a particle system obeying overdamped Langevin dynamics, we demonstrate that it is always possible to construct a thermodynamically consistent macroscopic model which obeys a gradient flow with respect to its non-equilibrium free energy. To do so, we significantly extend the recent Stochastic Thermodynamics with Internal Variables (STIV) framework, a method for producing macroscopic thermodynamic models far-from-equilibrium from the underlying mesoscopic dynamics and an approximate probability density of states parameterized with so-called internal variables. Though originally explored for Gaussian probability distributions, we here allow for an arbitrary choice of the approximate probability density while retaining a gradient flow dynamics. This greatly extends its range of applicability and automatically ensures consistency with the second law of thermodynamics, without the need for secondary verification. We demonstrate numerical convergence, in the limit of increasing internal variables, to the true probability density of states for both a multi-modal relaxation problem, a protein diffusing on a strand of DNA, and for an externally driven particle in a periodic landscape. Finally, we provide a reformulation of STIV with the quasi-equilibrium approximations in terms of the averages of observables of the mesostate, and show that these, too, obey a gradient flow. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:06:10 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Leadbetter",
"Travis",
""
],
[
"Purohit",
"Prashant K.",
""
],
[
"Reina",
"Celia",
""
]
] |
2506.08157 | Parham Oveissi | Ryan DeBoskey, Parham Oveissi, Venkat Narayanaswamy, and Ankit Goel | An In-situ Solid Fuel Ramjet Thrust Monitoring and Regulation Framework Using Neural Networks and Adaptive Control | null | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | Controlling the complex combustion dynamics within solid fuel ramjets (SFRJs) remains a critical challenge limiting deployment at scale. This paper proposes the use of a neural network model to process in-situ measurements for monitoring and regulating SFRJ thrust with a learning-based adaptive controller. A neural network is trained to estimate thrust from synthetic data generated by a feed-forward quasi-one-dimensional SFRJ model with variable inlet control. An online learning controller based on retrospective cost optimization is integrated with the quasi-one-dimensional SFRJ model to regulate the thrust. Sensitivity studies are conducted on both the neural network and adaptive controller to identify optimal hyperparameters. Numerical simulation results indicate that the combined neural network and learning control framework can effectively regulate the thrust produced by the SFRJ model using limited in-situ data. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:06:51 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"DeBoskey",
"Ryan",
""
],
[
"Oveissi",
"Parham",
""
],
[
"Narayanaswamy",
"Venkat",
""
],
[
"Goel",
"Ankit",
""
]
] |
2506.08158 | Lijing Zhu | Lijing Zhu, Qizhen Lan, Qing Tian, Wenbo Sun, Li Yang, Lu Xia, Yixin Xie, Xi Xiao, Tiehang Duan, Cui Tao, Shuteng Niu | ETT-CKGE: Efficient Task-driven Tokens for Continual Knowledge Graph Embedding | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Continual Knowledge Graph Embedding (CKGE) seeks to integrate new knowledge while preserving past information. However, existing methods struggle with efficiency and scalability due to two key limitations: (1) suboptimal knowledge preservation between snapshots caused by manually designed node/relation importance scores that ignore graph dependencies relevant to the downstream task, and (2) computationally expensive graph traversal for node/relation importance calculation, leading to slow training and high memory overhead. To address these limitations, we introduce ETT-CKGE (Efficient, Task-driven, Tokens for Continual Knowledge Graph Embedding), a novel task-guided CKGE method that leverages efficient task-driven tokens for efficient and effective knowledge transfer between snapshots. Our method introduces a set of learnable tokens that directly capture task-relevant signals, eliminating the need for explicit node scoring or traversal. These tokens serve as consistent and reusable guidance across snapshots, enabling efficient token-masked embedding alignment between snapshots. Importantly, knowledge transfer is achieved through simple matrix operations, significantly reducing training time and memory usage. Extensive experiments across six benchmark datasets demonstrate that ETT-CKGE consistently achieves superior or competitive predictive performance, while substantially improving training efficiency and scalability compared to state-of-the-art CKGE methods. The code is available at: https://github.com/lijingzhu1/ETT-CKGE/tree/main | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:07:59 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zhu",
"Lijing",
""
],
[
"Lan",
"Qizhen",
""
],
[
"Tian",
"Qing",
""
],
[
"Sun",
"Wenbo",
""
],
[
"Yang",
"Li",
""
],
[
"Xia",
"Lu",
""
],
[
"Xie",
"Yixin",
""
],
[
"Xiao",
"Xi",
""
],
[
"Duan",
"Tiehang",
""
],
[
"Tao",
"Cui",
""
],
[
"Niu",
"Shuteng",
""
]
] |
2506.08159 | Gabriele Grillo | Gabriele Grillo, Matteo Muratori, Troy Petitt, Nikita Simonov | A porous medium equation with rough weights: sharp Widder theory | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We establish an optimal \emph{Widder theory} for a weighted porous medium equation with rough and inhomogeneous density that may be singular at a point and tends to zero at spatial infinity. Specifically, for this equation, we identify a class $X$ of initial measure data that give rise to very weak solutions, we show that non-negative very weak solutions necessarily admit an initial trace in $X$ at time $t=0$, and we prove that any two non-negative solutions having the same initial trace are equal. The corresponding theory for the classical (unweighted) equation was established by exploiting various properties that are not available in our weighted setting, such as the continuity of solutions, the explicit scale invariance of the equation, Aleksandrov's reflection principle, and the Aronson--B\'enilan inequality. Therefore, to complete the Widder theory, we must devise several proofs by means of entirely new methods. We also establish an optimal quantitative \emph{a priori} smoothing estimate for unsigned local solutions without resorting to scale invariance, which seems to be new in this form even for the classical porous medium equation. Finally, we show that non-negative very weak solutions are always locally bounded, and in particular that they have locally finite energy. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:10:41 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Grillo",
"Gabriele",
""
],
[
"Muratori",
"Matteo",
""
],
[
"Petitt",
"Troy",
""
],
[
"Simonov",
"Nikita",
""
]
] |
2506.08160 | Wolfgang Staubach | Eric Schippers and Wolfgang Staubach | Scattering theory on Riemann surfaces I: Schiffer operators, cohomology, and index theorems | The paper "A scattering theory of harmonic one-forms on Riemann surfaces" [arXiv:2112.00835] was divided into four parts. This is part 3 in that division which has been revised and made self-contained | null | null | null | math.DG math.AG math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider a compact Riemann surface $\mathscr{R}$ with a complex of non-intersecting Jordan curves, whose complement is a pair of Riemann surfaces with boundary, each of which may be possibly disconnected. We investigate conformally invariant integral operators of Schiffer, which act on $L^{2}$ anti-holomorphic one-forms on one of these surfaces with boundary and produce holomorphic one-forms on the disjoint union. These operators arise in potential theory, boundary value problems, approximation theory, and conformal field theory, and are closely related to a kind of Cauchy operator. We develop an extensive calculus for the Schiffer and Cauchy operators, including a number of adjoint identities for the Schiffer operators. In the case that the Jordan curves are quasicircles, we derive a Plemelj-Sokhotski jump formula for Dirichlet-bounded functions. We generalize a theorem of Napalkov and Yulmukhametov, which shows that a certain Schiffer operator is an isomorphism for quasicircles. Finally, we characterize the kernels and images, and derive index theorems for the Schiffer operators, which will in turn connect conformal invariants to topological invariants. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:12:27 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Schippers",
"Eric",
""
],
[
"Staubach",
"Wolfgang",
""
]
] |
2506.08161 | Jakub Boksansky | Jakub Bok\v{s}ansk\'y, Daniel Meister, Carsten Benthin | GATE: Geometry-Aware Trained Encoding | null | null | null | null | cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The encoding of input parameters is one of the fundamental building blocks of neural network algorithms. Its goal is to map the input data to a higher-dimensional space, typically supported by trained feature vectors. The mapping is crucial for the efficiency and approximation quality of neural networks. We propose a novel geometry-aware encoding called GATE that stores feature vectors on the surface of triangular meshes. Our encoding is suitable for neural rendering-related algorithms, for example, neural radiance caching. It also avoids limitations of previous hash-based encoding schemes, such as hash collisions, selection of resolution versus scene size, and divergent memory access. Our approach decouples feature vector density from geometry density using mesh colors, while allowing for finer control over neural network training and adaptive level-of-detail. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:13:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bokšanský",
"Jakub",
""
],
[
"Meister",
"Daniel",
""
],
[
"Benthin",
"Carsten",
""
]
] |
2506.08162 | Jenna Moore | Jenna E. Moore, Seth H. Cohen, Philip Mauskopf, Evan Scannapieco | Integrated Galaxy Light from Stacking $10^5$ Random Pointings in the Dark Energy Survey Data | 13 pages, 9 figures, accepted for publication in The Astrophysical Journal | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We present a new technique for measuring the integrated galaxy light (IGL) with stacked image data from the Dark Energy Survey (DES). We extract $1\times1$ arcminute cutouts from nearly 100,000 randomly selected positions in the g, r, i, z, and Y bands from the DES data release 2 (DR2) maps. We generate source catalogs and masks for each cutout and the images are subsequently stacked to generate deep images of the sky both with and without sources. The IGL is then calculated by taking the difference in average brightness between stacks that contain galaxies and stacks in which galaxies have been masked. We find IGL values of $g = 4.27 \pm 0.28, r = 6.97 \pm 0.42, i = 8.66 \pm 0.53, z = 10.16 \pm 0.7,$ and $Y = 13.78 \pm 2.35$ nW/m$^2$/sr. These measurements, which require no foreground estimation or removal, are in agreement with previously reported IGL values derived from galaxy number counts and other methods. This stacking technique reduces the sensitivity to diffuse local backgrounds but is not sensitive to large-scale diffuse extragalactic background light. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:21:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Moore",
"Jenna E.",
""
],
[
"Cohen",
"Seth H.",
""
],
[
"Mauskopf",
"Philip",
""
],
[
"Scannapieco",
"Evan",
""
]
] |
2506.08163 | Harshvardhan Takawale | Harshvardhan Takawale and Nirupam Roy | Spectral Domain Neural Reconstruction for Passband FMCW Radars | arXiv admin note: substantial text overlap with arXiv:2503.23313 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present SpINRv2, a neural framework for high-fidelity volumetric reconstruction using Frequency-Modulated Continuous-Wave (FMCW) radar. Extending our prior work (SpINR), this version introduces enhancements that allow accurate learning under high start frequencies-where phase aliasing and sub-bin ambiguity become prominent. Our core contribution is a fully differentiable frequency-domain forward model that captures the complex radar response using closed-form synthesis, paired with an implicit neural representation (INR) for continuous volumetric scene modeling. Unlike time-domain baselines, SpINRv2 directly supervises the complex frequency spectrum, preserving spectral fidelity while drastically reducing computational overhead. Additionally, we introduce sparsity and smoothness regularization to disambiguate sub-bin ambiguities that arise at fine range resolutions. Experimental results show that SpINRv2 significantly outperforms both classical and learning-based baselines, especially under high-frequency regimes, establishing a new benchmark for neural radar-based 3D imaging. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:21:27 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Takawale",
"Harshvardhan",
""
],
[
"Roy",
"Nirupam",
""
]
] |
2506.08164 | Hadi Reisizadeh | Hadi Reisizadeh, Jinghan Jia, Zhiqi Bu, Bhanukiran Vinzamuri, Anil Ramakrishna, Kai-Wei Chang, Volkan Cevher, Sijia Liu, Mingyi Hong | BLUR: A Bi-Level Optimization Approach for LLM Unlearning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Enabling large language models (LLMs) to unlearn knowledge and capabilities acquired during training has proven vital for ensuring compliance with data regulations and promoting ethical practices in generative AI. Although there are growing interests in developing various unlearning algorithms, it remains unclear how to best formulate the unlearning problem. The most popular formulation uses a weighted sum of forget and retain loss, but it often leads to performance degradation due to the inherent trade-off between forget and retain losses. In this work, we argue that it is important to model the hierarchical structure of the unlearning problem, where the forget problem (which \textit{unlearns} certain knowledge and/or capabilities) takes priority over the retain problem (which preserves model utility). This hierarchical structure naturally leads to a bi-level optimization formulation where the lower-level objective focuses on minimizing the forget loss, while the upper-level objective aims to maintain the model's utility. Based on this new formulation, we propose a novel algorithm, termed Bi-Level UnleaRning (\texttt{BLUR}), which not only possesses strong theoretical guarantees but more importantly, delivers superior performance. In particular, our extensive experiments demonstrate that \texttt{BLUR} consistently outperforms all the state-of-the-art algorithms across various unlearning tasks, models, and metrics. Codes are available at https://github.com/OptimAI-Lab/BLURLLMUnlearning. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:23:05 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Reisizadeh",
"Hadi",
""
],
[
"Jia",
"Jinghan",
""
],
[
"Bu",
"Zhiqi",
""
],
[
"Vinzamuri",
"Bhanukiran",
""
],
[
"Ramakrishna",
"Anil",
""
],
[
"Chang",
"Kai-Wei",
""
],
[
"Cevher",
"Volkan",
""
],
[
"Liu",
"Sijia",
""
],
[
"Hong",
"Mingyi",
""
]
] |
2506.08165 | James Hateley Iv | James Hateley | Compact Temporal Geometry and the $T^2$ Framework for Quantum Gravity | null | null | null | null | hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We introduce a two-dimensional temporal framework in which time is represented by a compact manifold $T^2 = (t_1, t_2)$, with $t_1$ encoding classical causal structure and $t_2$ representing quantum coherence. This construction unifies unitary evolution, decoherence, measurement collapse, and gravitational dynamics within a consistent geometric and algebraic formalism. Compactification of the coherence time $t_2$ yields a minimal temporal resolution $\Delta t_2 \sim \sqrt{\alpha'}$, leading to a discretized spectrum of temporal modes and regularized ultraviolet behavior in quantum field theory and string-theoretic gravity. We formulate an extended Schr\"odinger equation and generalized Lindblad dynamics on $T^2$, and demonstrate the compatibility of this structure with local gauge symmetry through a complexified BRST quantization procedure. Using para-Hermitian geometry and generalized complex structures, we derive a covariant formulation of temporal T-duality that accommodates both Lorentzian and Euclidean signatures. The $T^2$ framework provides new insights into modular thermodynamics, black hole entropy, and the emergence of classical time from quantum coherence, offering a compact and quantized model of temporal geometry rooted in string theory and quantum gravity. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:24:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Hateley",
"James",
""
]
] |
2506.08166 | Wolfgang Staubach | Eric Schippers and Wolfgang Staubach | Scattering theory on Riemann surfaces II: The scattering matrix and generalized period mappings | The paper "A scattering theory of harmonic one-forms on Riemann surfaces" [arXiv:2112.00835] was divided into four parts. This is part 4 in that division which has been revised and made self-contained | null | null | null | math.DG math.AG math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct a scattering theory for harmonic one-forms on Riemann surfaces, obtained from boundary value problems involving systems of curves and the jump problem. We obtain an explicit expression for the scattering matrix in terms of integral operators which we call Schiffer operators, and show that the matrix is unitary. We also obtain a general association of positive polarizing Lagrangian spaces to bordered Riemann surfaces, which unifies the classical polarizations for compact surfaces of algebraic geometry with the infinite-dimensional period map of the universal Teichm\"uller space. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:25:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Schippers",
"Eric",
""
],
[
"Staubach",
"Wolfgang",
""
]
] |
2506.08167 | Sunny Gupta | Sunny Gupta, Nikita Jangid, Amit Sethi | UniVarFL: Uniformity and Variance Regularized Federated Learning for Heterogeneous Data | null | null | null | null | cs.LG cs.AI cs.CV cs.DC | http://creativecommons.org/licenses/by/4.0/ | Federated Learning (FL) often suffers from severe performance degradation when faced with non-IID data, largely due to local classifier bias. Traditional remedies such as global model regularization or layer freezing either incur high computational costs or struggle to adapt to feature shifts. In this work, we propose UniVarFL, a novel FL framework that emulates IID-like training dynamics directly at the client level, eliminating the need for global model dependency. UniVarFL leverages two complementary regularization strategies during local training: Classifier Variance Regularization, which aligns class-wise probability distributions with those expected under IID conditions, effectively mitigating local classifier bias; and Hyperspherical Uniformity Regularization, which encourages a uniform distribution of feature representations across the hypersphere, thereby enhancing the model's ability to generalize under diverse data distributions. Extensive experiments on multiple benchmark datasets demonstrate that UniVarFL outperforms existing methods in accuracy, highlighting its potential as a highly scalable and efficient solution for real-world FL deployments, especially in resource-constrained settings. Code: https://github.com/sunnyinAI/UniVarFL | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:25:35 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Gupta",
"Sunny",
""
],
[
"Jangid",
"Nikita",
""
],
[
"Sethi",
"Amit",
""
]
] |
2506.08168 | Axel van de Walle | Axel van de Walle | Pilot-waves and copilot-particles: A novel approach to objective collapse | 6 pages, 3 figures | null | null | null | cond-mat.stat-mech quant-ph | http://creativecommons.org/licenses/by/4.0/ | We propose an extension of Schr\"odinger's equation that incorporates the macroscopic measurement-induced wavefunction collapse phenomenon. Our approach relies on a hybrid between Bohm-de Broglie pilot-wave and objective collapse theories. The Bohmian particle is guided by the wavefunction and, conversely, the wavefunction gradually localizes towards the particle's position. As long as the particle can visit any state, as in a typical microscopic system, the localization effect does not favor any particular quantum state and, on average, the usual Schr\"odinger-like time evolution results. However, when the wavefunction develops spatially well-separated lobes, as would happen during a macroscopic measurement, the Bohmian particle can remain trapped in one lobe, and the wavefunction eventually localizes there. The end result, in macroscopic systems, is a wavefunction collapse that is consistent with Born's rule. We illustrate the theory with a simple double-slit experiment simulation. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:27:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"van de Walle",
"Axel",
""
]
] |
2506.08169 | Jingqiao Tang | Jingqiao Tang (1), Ryan Bausback (1), Feng Bao (1), Richard Archibald (2) ((1) Department of Mathematics at Florida State University, Tallahassee, Florida, USA, (2) Division of Computer Science and Mathematics, Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA) | Federated Learning on Stochastic Neural Networks | 25 pages, 19 figures, Submitted to Journal of Machine Learning for Modeling and Computing | null | null | null | cs.LG cs.DC | http://creativecommons.org/licenses/by/4.0/ | Federated learning is a machine learning paradigm that leverages edge computing on client devices to optimize models while maintaining user privacy by ensuring that local data remains on the device. However, since all data is collected by clients, federated learning is susceptible to latent noise in local datasets. Factors such as limited measurement capabilities or human errors may introduce inaccuracies in client data. To address this challenge, we propose the use of a stochastic neural network as the local model within the federated learning framework. Stochastic neural networks not only facilitate the estimation of the true underlying states of the data but also enable the quantification of latent noise. We refer to our federated learning approach, which incorporates stochastic neural networks as local models, as Federated stochastic neural networks. We will present numerical experiments demonstrating the performance and effectiveness of our method, particularly in handling non-independent and identically distributed data. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:28:39 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tang",
"Jingqiao",
"",
"Department of Mathematics at Florida State University, Tallahassee, Florida, USA"
],
[
"Bausback",
"Ryan",
"",
"Department of Mathematics at Florida State University, Tallahassee, Florida, USA"
],
[
"Bao",
"Feng",
"",
"Department of Mathematics at Florida State University, Tallahassee, Florida, USA"
],
[
"Archibald",
"Richard",
"",
"Division of Computer Science and Mathematics, Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA"
]
] |
2506.08170 | Charles Cheung | Amani Kiruga, Charles Cheung, Dmytro Filin, Parinaz Barakhshan, Akshay Bhosale, Vipul Badhan, Bindiya Arora, Rudolf Eigenmann, Marianna S. Safronova | Portal for High-Precision Atomic Data and Computation | 11 pages, 6 figures | null | null | null | physics.atom-ph physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | We've developed a scalable and sustainable online atomic data portal with an automated interface for easy update and addition of new data. The current portal provides energies, transition matrix elements, transition rates, radiative lifetimes, branching ratios, polarizabilities, hyperfine constants, and other data, for 28 atoms and ions. It also features an interactive polarizability plotting interface for neutral atoms and singly-charged ions. The data production is supported by recent developments of open-access atomic software based on our research codes, including new workflow algorithms, which allow large volumes of such data to be generated with automated accuracy assessments. This entails a new method of comparing our calculated values with data from the NIST Atomic Spectra Database. All calculated values include estimated uncertainties. Data for more systems will be added in the future. Experimental values are included with references, where high-precision data are available. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:30:40 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Kiruga",
"Amani",
""
],
[
"Cheung",
"Charles",
""
],
[
"Filin",
"Dmytro",
""
],
[
"Barakhshan",
"Parinaz",
""
],
[
"Bhosale",
"Akshay",
""
],
[
"Badhan",
"Vipul",
""
],
[
"Arora",
"Bindiya",
""
],
[
"Eigenmann",
"Rudolf",
""
],
[
"Safronova",
"Marianna S.",
""
]
] |
2506.08171 | Youcheng Sun | Daniel Koh, Yannic Noller, Corina S. Pasareanu, Adrians Skapars, Youcheng Sun | Worst-Case Symbolic Constraints Analysis and Generalisation with Large Language Models | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have been successfully applied to a variety of coding tasks, including code generation, completion, and repair. However, more complex symbolic reasoning tasks remain largely unexplored by LLMs. This paper investigates the capacity of LLMs to reason about worst-case executions in programs through symbolic constraints analysis, aiming to connect LLMs and symbolic reasoning approaches. Specifically, we define and address the problem of worst-case symbolic constraints analysis as a measure to assess the comprehension of LLMs. We evaluate the performance of existing LLMs on this novel task and further improve their capabilities through symbolic reasoning-guided fine-tuning, grounded in SMT (Satisfiability Modulo Theories) constraint solving and supported by a specially designed dataset of symbolic constraints. Experimental results show that our solver-aligned model, WARP-1.0-3B, consistently surpasses size-matched and even much larger baselines, demonstrating that a 3B LLM can recover the very constraints that pin down an algorithm's worst-case behaviour through reinforcement learning methods. These findings suggest that LLMs are capable of engaging in deeper symbolic reasoning, supporting a closer integration between neural network-based learning and formal methods for rigorous program analysis. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:33:30 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Koh",
"Daniel",
""
],
[
"Noller",
"Yannic",
""
],
[
"Pasareanu",
"Corina S.",
""
],
[
"Skapars",
"Adrians",
""
],
[
"Sun",
"Youcheng",
""
]
] |
2506.08172 | Gerardo Aleman Manzanarez | Gerardo Aleman Manzanarez, Nora de la Cruz Arana, Jorge Garcia Flores, Yobany Garcia Medina, Raul Monroy, Nathalie Pernelle | Can Artificial Intelligence Write Like Borges? An Evaluation Protocol for Spanish Microfiction | 28 pages, 16 figures. Submitted to Applied Sciences | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Automated story writing has been a subject of study for over 60 years. Large language models can generate narratively consistent and linguistically coherent short fiction texts. Despite these advancements, rigorous assessment of such outputs for literary merit - especially concerning aesthetic qualities - has received scant attention. In this paper, we address the challenge of evaluating AI-generated microfictions and argue that this task requires consideration of literary criteria across various aspects of the text, such as thematic coherence, textual clarity, interpretive depth, and aesthetic quality. To facilitate this, we present GrAImes: an evaluation protocol grounded in literary theory, specifically drawing from a literary perspective, to offer an objective framework for assessing AI-generated microfiction. Furthermore, we report the results of our validation of the evaluation protocol, as answered by both literature experts and literary enthusiasts. This protocol will serve as a foundation for evaluating automatically generated microfictions and assessing their literary value. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:34:13 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Manzanarez",
"Gerardo Aleman",
""
],
[
"Arana",
"Nora de la Cruz",
""
],
[
"Flores",
"Jorge Garcia",
""
],
[
"Medina",
"Yobany Garcia",
""
],
[
"Monroy",
"Raul",
""
],
[
"Pernelle",
"Nathalie",
""
]
] |
2506.08173 | Truong-Son Hy | Nguyen Phu Vinh, Anh Chung Hoang, Chris Ngo, Truong-Son Hy | Repeton: Structured Bug Repair with ReAct-Guided Patch-and-Test Cycles | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have shown strong capabilities in code generation and comprehension, yet their application to complex software engineering tasks often suffers from low precision and limited interpretability. We present Repeton, a fully open-source framework that leverages LLMs for precise and automated code manipulation in real-world Git repositories. Rather than generating holistic fixes, Repeton operates through a structured patch-and-test pipeline: it iteratively diagnoses issues, proposes code changes, and validates each patch through automated testing. This stepwise process is guided by lightweight heuristics and development tools, avoiding reliance on embedding-based retrieval systems. Evaluated on the SWE-bench Lite benchmark, our method shows good performance compared to RAG-based methods in both patch validity and interpretability. By decomposing software engineering tasks into modular, verifiable stages, Repeton provides a practical path toward scalable and transparent autonomous debugging. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:36:40 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Vinh",
"Nguyen Phu",
""
],
[
"Hoang",
"Anh Chung",
""
],
[
"Ngo",
"Chris",
""
],
[
"Hy",
"Truong-Son",
""
]
] |
2506.08175 | Meesoon Ha | Jaeseok Hur, Meesoon Ha, and Hawoong Jeong | Anomaly, Class Division, and Decoupling in Wealth Dynamics | 18 pages,3 figures (main) and 11 figures (SM) | null | null | null | cond-mat.stat-mech physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Economic inequality is shaped by the agent-network structure, the interaction between agents, and the individual agent's ability. We provide a comprehensive picture of anomalous diffusion, economic class division, and bimodal wealth distribution in an agent-based model, where the allocation of heterogeneous agent abilities/growth rates is tuned in sparse regular networks.In particular, we focus on the statistical characteristics of logarithmic scaled normalized wealth distributions with two ability parameters, assortativity $\mathcal{A}$ and concentration $\mathcal{R}$. For the set of $(\mathcal{A},\mathcal{R})$, temporal behaviors of log-wealth distributions reveal that the decoupling between different ability groups depends primarily on $\mathcal{R}$ and long-term inequality depends mainly on $\mathcal{A}$. In other words, class division and decoupling are driven by $\mathcal{R}$, while the super-diffusive nature in the leading class is driven by $\mathcal{A}$. Our findings highlight that hierarchical segregation of abilities, rather than ability differences alone, is a key driver of economic class stratification. Finally, our model provides a minimal, yet powerful framework for understanding the bimodal global income distribution observed over the past half century and highlights the critical role of network-level segregation in shaping economic inequality. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:39:21 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Hur",
"Jaeseok",
""
],
[
"Ha",
"Meesoon",
""
],
[
"Jeong",
"Hawoong",
""
]
] |
2506.08176 | Anh Nguyen | Anh V Nguyen, Diego Klabjan | FedGA-Tree: Federated Decision Tree using Genetic Algorithm | null | null | null | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, with rising concerns for data privacy, Federated Learning has gained prominence, as it enables collaborative training without the aggregation of raw data from participating clients. However, much of the current focus has been on parametric gradient-based models, while nonparametric counterparts such as decision tree are relatively understudied. Existing methods for adapting decision trees to Federated Learning generally combine a greedy tree-building algorithm with differential privacy to produce a global model for all clients. These methods are limited to classification trees and categorical data due to the constraints of differential privacy. In this paper, we explore an alternative approach that utilizes Genetic Algorithm to facilitate the construction of personalized decision trees and accommodate categorical and numerical data, thus allowing for both classification and regression trees. Comprehensive experiments demonstrate that our method surpasses decision trees trained solely on local data and a benchmark algorithm. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:39:22 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nguyen",
"Anh V",
""
],
[
"Klabjan",
"Diego",
""
]
] |
2506.08177 | Abby Boehm | V. Abby Boehm, Darryl Z. Seligman, and Nikole K. Lewis | Constraining Ongoing Volcanic Outgassing Rates and Interior Compositions of Extrasolar Planets with Mass Measurements of Plasma Tori | 28 pages, 9 figures, 2 tables. Accepted for publication in The Astrophysical Journal Letters | null | null | null | astro-ph.EP physics.space-ph | http://creativecommons.org/licenses/by/4.0/ | We present a novel method of constraining volcanic activity on extrasolar terrestrial worlds via characterization of circumstellar plasma tori. Our work generalizes the physics of the Io plasma torus to propose a hypothetical circumstellar plasma torus generated by exoplanetary volcanism. The quasi-steady torus mass is determined by a balance between material injection and ejection rates from volcanic activity and corotating magnetospheric convection, respectively. By estimating the Alfv\'en surfaces of planet-hosting stars, we calculate the torus mass-removal timescale for a number of exoplanets with properties amenable to plasma torus construction. Assuming a uniform toroidal geometry comparable to Io's "warm" torus, we calculate quasi-steady torus masses inferable from the optical depth of atomic spectral features in torus-contaminated stellar spectra. The calculated quasi-steady masses can be used to constrain the volcanic outgassing rates of each species detected in the torus, providing quantitative estimates of bulk volcanic activity and interior composition with minimal assumptions. Such insight into the interior state of an exoplanet is otherwise accessible only after destruction via tidal forces. We demonstrate the feasibility of our method by showcasing known exoplanets which are susceptible to tidal heating and could generate readily-detectable tori with realistic outgassing rates of order 1 ton s$^{-1}$, comparable to the Io plasma torus mass injection rate. This methodology may be applied to stellar spectra measured with ultraviolet instruments with sufficient resolution to detect atomic lines and sensitivity to recover the ultraviolet continuum of GKM dwarf stars. This further motivates the need for ultraviolet instrumentation above Earth's atmosphere. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:41:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Boehm",
"V. Abby",
""
],
[
"Seligman",
"Darryl Z.",
""
],
[
"Lewis",
"Nikole K.",
""
]
] |
2506.08178 | Benjamin Haake | Nils Carqueville, Benjamin Haake | 2-Group Symmetries of 3-dimensional Defect TQFTs and Their Gauging | 63 pages | null | null | null | math.QA hep-th math-ph math.MP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | A large class of symmetries of topological quantum field theories is naturally described by functors into higher categories of topological defects. Here we study 2-group symmetries of 3-dimensional TQFTs. We explain that these symmetries can be gauged to produce new TQFTs iff certain defects satisfy the axioms of orbifold data. In the special case of Reshetikhin-Turaev theories coming from $G$-crossed braided fusion categories $\mathcal C^\times_G$, we show that there are 0- and 1-form symmetries which have no obstructions to gauging. We prove that gauging the 0-form $G$-symmetry on the neutral component $\mathcal C_e$ of $\mathcal C^\times_G$ produces its equivariantisation $(\mathcal C^\times_G)^G$, which in turn features a generalised symmetry whose gauging recovers $\mathcal C_e$. If $G$ is commutative, the latter symmetry reduces to a 1-form symmetry involving the Pontryagin dual group. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:42:49 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Carqueville",
"Nils",
""
],
[
"Haake",
"Benjamin",
""
]
] |
2506.08179 | Vahid Garousi | Sasidhar Matta, Vahid Garousi | MBTModelGenerator: A software tool for reverse engineering of Model-based Testing (MBT) models from clickstream data of web applications | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Automated testing has become a standard practice in software engineering, yet the creation of test models and suites remains labor-intensive. To reduce this effort, we developed an open-source tool that automatically generates Model-Based Testing (MBT) models from clickstream data collected during user interaction with web applications. The tool captures UI events, transforms them into state-transition models, and exports the result in a format compatible with the GraphWalker MBT tool. This enables immediate test execution without the need for manual model creation. The approach lowers the barrier to MBT adoption by leveraging actual usage behavior and reducing the reliance on upfront modeling. This technical report documents the system requirements, design decisions, implementation details, testing process, and empirical evaluation of the tool, which is publicly available as open-source. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:44:10 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Matta",
"Sasidhar",
""
],
[
"Garousi",
"Vahid",
""
]
] |
2506.08180 | Felipe Navarete | Felipe Navarete, Sean D. Points, Augusto Damineli | IRAS 16475-4609: A Young Compact HII Region Sculpting Its Molecular Environment | Accepted for publication in the Astronomical Journal. 18 pages, 11 figures | null | null | null | astro-ph.GA astro-ph.IM astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | We present a near-infrared spectroscopic and imaging analysis of the star-forming region IRAS 16475-4609, based on TripleSpec/SOAR spectroscopy and NEWFIRM/CTIO imaging, complemented by archival radio and sub-millimeter data. Our spectroscopic analysis indicates that the central source is an early B-type star (B0-B0.7V) powering a compact HII region characterized by strong HI and HeI recombination lines, and molecular H$_2$ emission. We derive a distance of 3.51$\pm$0.74 kpc, consistent with the position of the Scutum-Crux near arm at Galactic longitudes of $\sim$340$^\circ$. At this distance, the ionized gas traced by Brackett-$\gamma$ emission has a radius of 0.27$\pm$0.06 pc, placing the source in a transition phase between ultra-compact and compact HII regions. From radio data, we estimate an ionizing photon flux of N$_{ly}$=(2.3$\pm$0.3)$\times10^{47}$ photons s$^{-1}$, and an electron temperature of T$_e$=(5.4$\pm$0.2)$\times$10$^3$ K for the ionized gas. The analysis also reveals an obscured high-density molecular clump southwest of the HII region, coincident with an ATLASGAL sub-millimeter peak, indicating a potential site of ongoing and triggered star formation as the ionization front advances into the surrounding molecular material. These results suggest that IRAS 16475-4609 is a young high-mass star-forming region with stellar feedback actively shaping its environment, offering valuable insight into the early evolution of compact HII regions. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:44:55 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Navarete",
"Felipe",
""
],
[
"Points",
"Sean D.",
""
],
[
"Damineli",
"Augusto",
""
]
] |
2506.08181 | Douglas Goncalves | Douglas S. Gon\c{c}alves and Max L. N. Gon\c{c}alves and Jefferson G. Melo | A Cubic Regularization Method for Multiobjective Optimization | null | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | This work introduces a new cubic regularization method for nonconvex unconstrained multiobjective optimization problems. At each iteration of the method, a model associated with the cubic regularization of each component of the objective function is minimized. This model allows approximations for the first- and second-order derivatives, which must satisfy suitable error conditions. One interesting feature of the proposed algorithm is that the regularization parameter of the model and the accuracy of the derivative approximations are jointly adjusted using a nonmonotone line search criterion. Implementations of the method, where derivative information is computed using finite difference strategies, are discussed. It is shown that, under the assumption that the Hessians of the objectives are globally Lipschitz continuous, the method requires at most $\mathcal{O}(C\epsilon^{-3/2})$ iterations to generate an $\epsilon$-approximate Pareto critical. In particular, if the first- and second-order derivative information is computed using finite differences based solely on function values, the method requires at most $\mathcal{O}(n^{1-\beta}\epsilon^{-3/2})$ iterations, corresponding to $\mathcal{O}\left(mn^{3-\beta}\varepsilon^{-\frac{3}{2}}\right)$ function evaluations, where \(n\) is the dimension of the domain of the objective function, $m$ is the number of objectives, and $\beta \in [0,1]$ is a constant associated with the stepsize used in the finite-difference approximation. We further discuss the global convergence and local convergence rate of the method. Specifically, under the local convexity assumption, we show that the method achieves superlinear convergence when the first derivative is computed exactly, and quadratic convergence when both first- and second-order derivatives are exact. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:45:42 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Gonçalves",
"Douglas S.",
""
],
[
"Gonçalves",
"Max L. N.",
""
],
[
"Melo",
"Jefferson G.",
""
]
] |
2506.08182 | Tyler LeBlond | Tyler LeBlond and Ryan S. Bennink | Quantum Resource Comparison for Two Leading Surface Code Lattice Surgery Approaches | 12 pages, 4 figures, 5 tables | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hamiltonian simulation is one of the most promising candidates for the demonstration of quantum advantage within the next ten years, and several studies have proposed end-to-end resource estimates for executing such algorithms on fault-tolerant quantum processors. Usually, these resource estimates are based upon the assumption that quantum error correction is implemented using the surface code, and that the best surface code compilation scheme involves serializing input circuits by eliminating all Clifford gates. This transformation is thought to make best use of the native multi-body measurement (lattice surgery) instruction set available to surface codes. Some work, however, has suggested that direct compilation from Clifford+T to lattice surgery operations may be beneficial for circuits that have high degrees of logical parallelism. In this study, we analyze the resource costs for implementing Hamiltonian simulation using example approaches from each of these leading surface code compilation families. The Hamiltonians whose dynamics we consider are those of the transverse-field Ising model in several geometries, the Kitaev honeycomb model, and the $\mathrm{\alpha-RuCl_3}$ complex under a time-varying magnetic field. We show, among other things, that the optimal scheme depends on whether Hamiltonian simulation is implemented using the quantum signal processing or Trotter-Suzuki algorithms, with Trotterization benefiting by orders of magnitude from direct Clifford+T compilation for these applications. Our results suggest that surface code quantum computers should not have a one-size-fits-all compilation scheme, but that smart compilers should predict the optimal scheme based upon high-level quantities from logical circuits such as average circuit density, numbers of logical qubits, and T fraction. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:46:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"LeBlond",
"Tyler",
""
],
[
"Bennink",
"Ryan S.",
""
]
] |
2506.08183 | Isha Puri | Isha Puri and David Cox | A System for Accurate Tracking and Video Recordings of Rodent Eye Movements using Convolutional Neural Networks for Biomedical Image Segmentation | null | 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Honolulu, HI, USA, 2018, pp. 3590-3593 | 10.1109/EMBC.2018.8513072 | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Research in neuroscience and vision science relies heavily on careful measurements of animal subject's gaze direction. Rodents are the most widely studied animal subjects for such research because of their economic advantage and hardiness. Recently, video based eye trackers that use image processing techniques have become a popular option for gaze tracking because they are easy to use and are completely noninvasive. Although significant progress has been made in improving the accuracy and robustness of eye tracking algorithms, unfortunately, almost all of the techniques have focused on human eyes, which does not account for the unique characteristics of the rodent eye images, e.g., variability in eye parameters, abundance of surrounding hair, and their small size. To overcome these unique challenges, this work presents a flexible, robust, and highly accurate model for pupil and corneal reflection identification in rodent gaze determination that can be incrementally trained to account for variability in eye parameters encountered in the field. To the best of our knowledge, this is the first paper that demonstrates a highly accurate and practical biomedical image segmentation based convolutional neural network architecture for pupil and corneal reflection identification in eye images. This new method, in conjunction with our automated infrared videobased eye recording system, offers the state of the art technology in eye tracking for neuroscience and vision science research for rodents. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:48:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Puri",
"Isha",
""
],
[
"Cox",
"David",
""
]
] |
2506.08185 | Huixin Zhan | Huixin Zhan, Jason H. Moore | Surgeon Style Fingerprinting and Privacy Risk Quantification via Discrete Diffusion Models in a Vision-Language-Action Framework | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Surgeons exhibit distinct operating styles due to differences in training, experience, and motor behavior - yet current AI systems often ignore this personalization signal. We propose a novel approach to model fine-grained, surgeon-specific fingerprinting in robotic surgery using a discrete diffusion framework integrated with a vision-language-action (VLA) pipeline. Our method formulates gesture prediction as a structured sequence denoising task, conditioned on multimodal inputs including endoscopic video, surgical intent language, and a privacy-aware embedding of surgeon identity and skill. Personalized surgeon fingerprinting is encoded through natural language prompts using third-party language models, allowing the model to retain individual behavioral style without exposing explicit identity. We evaluate our method on the JIGSAWS dataset and demonstrate that it accurately reconstructs gesture sequences while learning meaningful motion fingerprints unique to each surgeon. To quantify the privacy implications of personalization, we perform membership inference attacks and find that more expressive embeddings improve task performance but simultaneously increase susceptibility to identity leakage. These findings demonstrate that while personalized embeddings improve performance, they also increase vulnerability to identity leakage, revealing the importance of balancing personalization with privacy risk in surgical modeling. Code is available at: https://github.com/huixin-zhan-ai/Surgeon_style_fingerprinting. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:49:55 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zhan",
"Huixin",
""
],
[
"Moore",
"Jason H.",
""
]
] |
2506.08186 | Ilya Beskin | Ilya M Beskin, Jordan Zesch, Emma Hunt, Alexander Weinstein, Ernst-Ludwig Florin | Optical tweezers as a tool for quantitative imaging | Main Text: 21 pages, 7 figures. Supplemental Materials: 8 pages, 5 figures Submitted for publication in Optics Express | null | null | null | physics.optics physics.bio-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Optical tweezers equipped with position detection allow for application of piconewton-scale forces and high-temporal-resolution measurements of nanometer-scale motion. While typically used for trapping microscopic objects, the optical tweezer detection pathway can also be used for a microscopy technique sensitive to nanometer-sized structures. Optical tweezers most commonly use back-focal-plane detection to determine the position of the trapped object. This technique involves analyzing the interference pattern between scattered and unscattered light. Despite the reliance on interference, we show that an image of an extended object can be understood as a sum of the images of individual point-like particles that make up the object. This allows for optical tweezers with back-focal-plane detection to be used as an imaging tool capable of determining the mass distribution of scanned structures. Furthermore, the sample-orientation-dependent detector response allows for a unique method of background subtraction. We demonstrate the quantitative imaging capabilities of optical tweezer microscopy by determining the size distributions multiple nearby particles and measuring the changing diameter of collagen fibrils. The background subtraction technique is demonstrated by imaging surface-bound microtubules with a strong background from protein aggregate co-adsorption. Optical tweezer microscopy allows for quantitative imaging of objects far from the coverslip surface. This makes it an excellent tool for studying the link between the structure and mechanics of microscopic systems. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:51:39 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Beskin",
"Ilya M",
""
],
[
"Zesch",
"Jordan",
""
],
[
"Hunt",
"Emma",
""
],
[
"Weinstein",
"Alexander",
""
],
[
"Florin",
"Ernst-Ludwig",
""
]
] |
2506.08187 | Mateusz Dembny | Mateusz Dembny and Miko{\l}aj Sier\.z\k{e}ga | Geometric aspects of the Harnack Inequality for a nonlocal heat equation | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We establish a connection between a sharp double-sided Harnack bound for positive solutions of a fractional heat equation and the circular geometry in higher dimensions. The present work extends and generalizes the results obtained in the preceding paper. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:52:43 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Dembny",
"Mateusz",
""
],
[
"Sierżęga",
"Mikołaj",
""
]
] |
2506.08188 | Wenlong Meng | Wenlong Meng, Shuguo Fan, Chengkun Wei, Min Chen, Yuwei Li, Yuanchao Zhang, Zhikun Zhang, Wenzhi Chen | GradEscape: A Gradient-Based Evader Against AI-Generated Text Detectors | Accepted to USENIX Security'25 | null | null | null | cs.CR cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we introduce GradEscape, the first gradient-based evader designed to attack AI-generated text (AIGT) detectors. GradEscape overcomes the undifferentiable computation problem, caused by the discrete nature of text, by introducing a novel approach to construct weighted embeddings for the detector input. It then updates the evader model parameters using feedback from victim detectors, achieving high attack success with minimal text modification. To address the issue of tokenizer mismatch between the evader and the detector, we introduce a warm-started evader method, enabling GradEscape to adapt to detectors across any language model architecture. Moreover, we employ novel tokenizer inference and model extraction techniques, facilitating effective evasion even in query-only access.
We evaluate GradEscape on four datasets and three widely-used language models, benchmarking it against four state-of-the-art AIGT evaders. Experimental results demonstrate that GradEscape outperforms existing evaders in various scenarios, including with an 11B paraphrase model, while utilizing only 139M parameters. We have successfully applied GradEscape to two real-world commercial AIGT detectors. Our analysis reveals that the primary vulnerability stems from disparity in text expression styles within the training data. We also propose a potential defense strategy to mitigate the threat of AIGT evaders. We open-source our GradEscape for developing more robust AIGT detectors. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:56:42 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Meng",
"Wenlong",
""
],
[
"Fan",
"Shuguo",
""
],
[
"Wei",
"Chengkun",
""
],
[
"Chen",
"Min",
""
],
[
"Li",
"Yuwei",
""
],
[
"Zhang",
"Yuanchao",
""
],
[
"Zhang",
"Zhikun",
""
],
[
"Chen",
"Wenzhi",
""
]
] |
2506.08189 | Amartya Dutta | Amartya Dutta, Kazi Sajeed Mehrab, Medha Sawhney, Abhilash Neog, Mridul Khurana, Sepideh Fatemi, Aanish Pradhan, M. Maruf, Ismini Lourentzou, Arka Daw, Anuj Karpatne | Open World Scene Graph Generation using Vision Language Models | Accepted in CVPR 2025 Workshop (CVinW) | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by/4.0/ | Scene-Graph Generation (SGG) seeks to recognize objects in an image and distill their salient pairwise relationships. Most methods depend on dataset-specific supervision to learn the variety of interactions, restricting their usefulness in open-world settings, involving novel objects and/or relations. Even methods that leverage large Vision Language Models (VLMs) typically require benchmark-specific fine-tuning. We introduce Open-World SGG, a training-free, efficient, model-agnostic framework that taps directly into the pretrained knowledge of VLMs to produce scene graphs with zero additional learning. Casting SGG as a zero-shot structured-reasoning problem, our method combines multimodal prompting, embedding alignment, and a lightweight pair-refinement strategy, enabling inference over unseen object vocabularies and relation sets. To assess this setting, we formalize an Open-World evaluation protocol that measures performance when no SGG-specific data have been observed either in terms of objects and relations. Experiments on Visual Genome, Open Images V6, and the Panoptic Scene Graph (PSG) dataset demonstrate the capacity of pretrained VLMs to perform relational understanding without task-level training. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 19:59:05 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Dutta",
"Amartya",
""
],
[
"Mehrab",
"Kazi Sajeed",
""
],
[
"Sawhney",
"Medha",
""
],
[
"Neog",
"Abhilash",
""
],
[
"Khurana",
"Mridul",
""
],
[
"Fatemi",
"Sepideh",
""
],
[
"Pradhan",
"Aanish",
""
],
[
"Maruf",
"M.",
""
],
[
"Lourentzou",
"Ismini",
""
],
[
"Daw",
"Arka",
""
],
[
"Karpatne",
"Anuj",
""
]
] |
2506.08190 | Aleksandra Sierant | Aleksandra Sierant, Diana M\'endez-Avalos, Santiago Tabares Giraldo, Morgan W. Mitchell | Quantum noise in a squeezed-light-enhanced multiparameter quantum sensor | 6 pages, 4 figures, comments welcome | null | null | null | quant-ph physics.app-ph physics.atom-ph physics.ins-det | http://creativecommons.org/licenses/by/4.0/ | We study quantum enhancement of sensitivity using squeezed light in a multi-parameter quantum sensor, the hybrid rf-dc optically pumped magnetometer (hOPM) [Phys. Rev. Applied 21, 034054, (2024)]. Using a single spin ensemble, the hOPM acquires both the dc field strength (scalar magnetometry), and resonantly detects one quadrature of the ac magnetic field at a chosen frequency (rf magnetometry). In contrast to the Bell-Bloom scalar magnetometer (BBOPM) [Phys. Rev. Lett. 127, 193601 (2021)], the back-action evasion in the hOPM is incomplete, leading to a complex interplay of the three quantum noise sources in this system: photon shot noise, spin projection noise, and measurement back-action noise. We observe these interactions using squeezed light as a tool to control the distribution of optical quantum noise between $S_2$ and $S_3$ polarization Stokes components, and the resulting effect on readout quantum noise and measurement back-action. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:00:51 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Sierant",
"Aleksandra",
""
],
[
"Méndez-Avalos",
"Diana",
""
],
[
"Giraldo",
"Santiago Tabares",
""
],
[
"Mitchell",
"Morgan W.",
""
]
] |
2506.08191 | Antoni Nowinowski | Antoni Nowinowski, Krzysztof Krawiec | Generative Learning of Differentiable Object Models for Compositional Interpretation of Complex Scenes | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study builds on the architecture of the Disentangler of Visual Priors (DVP), a type of autoencoder that learns to interpret scenes by decomposing the perceived objects into independent visual aspects of shape, size, orientation, and color appearance. These aspects are expressed as latent parameters which control a differentiable renderer that performs image reconstruction, so that the model can be trained end-to-end with gradient using reconstruction loss. In this study, we extend the original DVP so that it can handle multiple objects in a scene. We also exploit the interpretability of its latent by using the decoder to sample additional training examples and devising alternative training modes that rely on loss functions defined not only in the image space, but also in the latent space. This significantly facilitates training, which is otherwise challenging due to the presence of extensive plateaus in the image-space reconstruction loss. To examine the performance of this approach, we propose a new benchmark featuring multiple 2D objects, which subsumes the previously proposed Multi-dSprites dataset while being more parameterizable. We compare the DVP extended in these ways with two baselines (MONet and LIVE) and demonstrate its superiority in terms of reconstruction quality and capacity to decompose overlapping objects. We also analyze the gradients induced by the considered loss functions, explain how they impact the efficacy of training, and discuss the limitations of differentiable rendering in autoencoders and the ways in which they can be addressed. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:06:42 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nowinowski",
"Antoni",
""
],
[
"Krawiec",
"Krzysztof",
""
]
] |
2506.08192 | Steven Cheung | Jared Claypoole, Steven Cheung, Ashish Gehani, Vinod Yegneswaran, and Ahmad Ridley | Interpreting Agent Behaviors in Reinforcement-Learning-Based Cyber-Battle Simulation Platforms | null | The AAAI-25 Workshop on Artificial Intelligence for Cyber Security (AICS), March 2025 | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze two open source deep reinforcement learning agents submitted to the CAGE Challenge 2 cyber defense challenge, where each competitor submitted an agent to defend a simulated network against each of several provided rules-based attack agents. We demonstrate that one can gain interpretability of agent successes and failures by simplifying the complex state and action spaces and by tracking important events, shedding light on the fine-grained behavior of both the defense and attack agents in each experimental scenario. By analyzing important events within an evaluation episode, we identify patterns in infiltration and clearing events that tell us how well the attacker and defender played their respective roles; for example, defenders were generally able to clear infiltrations within one or two timesteps of a host being exploited. By examining transitions in the environment's state caused by the various possible actions, we determine which actions tended to be effective and which did not, showing that certain important actions are between 40% and 99% ineffective. We examine how decoy services affect exploit success, concluding for instance that decoys block up to 94% of exploits that would directly grant privileged access to a host. Finally, we discuss the realism of the challenge and ways that the CAGE Challenge 4 has addressed some of our concerns. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:07:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Claypoole",
"Jared",
""
],
[
"Cheung",
"Steven",
""
],
[
"Gehani",
"Ashish",
""
],
[
"Yegneswaran",
"Vinod",
""
],
[
"Ridley",
"Ahmad",
""
]
] |
2506.08193 | Vasilis Oikonomou | S.D. Odintsov, V.K. Oikonomou | GW170817 Viable Einstein-Gauss-Bonnet Inflation Compatible with the Atacama Cosmology Telescope Data | null | null | null | null | gr-qc astro-ph.CO hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we investigate several Einstein-Gauss-Bonnet models that are compatible with the GW170817 event, the Atacama Cosmology Telescope data and the BICEP/Keck updated Planck constraints on the tensor-to-scalar ratio. We consider two distinct classes of Einstein-Gauss-Bonnet theories, which are equally successful for GW170817-compatible model building and we examine their viability against the Atacama Cosmology Telescope data and the updated Planck constraints on the tensor-to-scalar ratio. The two models are distinct since the first class relates directly the non-minimal Gauss-Bonnet scalar coupling function with the scalar potential and yields $c_T^2=1$ while in the second class, the non-minimal Gauss-Bonnet scalar coupling function and the scalar potential are freely but conveniently chosen and the class of models respects the constraint $\left| c_T^2 - 1 \right| < 6 \times 10^{-15}$. We provide several examples of models belonging to both the two classes of GW170817-compatible Einstein-Gauss-Bonnet theories and we demonstrate that Einstein-Gauss-Bonnet theories provide a promising theoretical framework for inflationary dynamics. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:08:22 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Odintsov",
"S. D.",
""
],
[
"Oikonomou",
"V. K.",
""
]
] |
2506.08195 | Jack Lubin | Jack Lubin, Erik A. Petigura, Velibor V. Mi\v{s}i\'c, Judah Van Zandt, Luke B. Handley | AstroQ: Automated Scheduling of Cadenced Astronomical Observations | Submitted to AJ, welcoming comments | null | null | null | astro-ph.IM astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Astronomy relies heavily on time domain observations. To maximize the scientific yield of such observations, astronomers must carefully match the observational cadence to the phenomena of interest. This presents significant scheduling challenges for observatories with multiple large programs, each with different cadence needs. To address this challenge, we developed AstroQ, an automated framework for scheduling cadenced observations. We tested this on a suite of Doppler exoplanet programs at Keck Observatory, where the algorithm powers the KPF-Community Cadence project. As a point of reference, AstroQ can determine the provably optimal ordering of 3680 observations of 200 targets -- each with its own cadence needs and accessibility constraints -- over a six month period to five minute time resolution. Schedules of this size may be constructed in ~120 seconds on modern workstation, enabling dynamic rescheduling due to weather changes, target-of-opportunity interrupts, and other needs. A key advantage of AstroQ over manual scheduling is realistic projections of program completion, savings in human effort, and elimination of human bias in balancing many programs. AstroQ is open source and may be applied to other scheduling needs, both in astronomy and beyond. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:12:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Lubin",
"Jack",
""
],
[
"Petigura",
"Erik A.",
""
],
[
"Mišić",
"Velibor V.",
""
],
[
"Van Zandt",
"Judah",
""
],
[
"Handley",
"Luke B.",
""
]
] |
2506.08196 | Brian Zimmerman | Brian D. Zimmerman, Julien Aubert-B\'educhaud, Florian Boudin, Akiko Aizawa, Olga Vechtomova | No Stupid Questions: An Analysis of Question Query Generation for Citation Recommendation | 6 pages, 5 figures, 2 tables | null | null | null | cs.IR cs.DL | http://creativecommons.org/licenses/by/4.0/ | Existing techniques for citation recommendation are constrained by their adherence to article contents and metadata. We leverage GPT-4o-mini's latent expertise as an inquisitive assistant by instructing it to ask questions which, when answered, could expose new insights about an excerpt from a scientific article. We evaluate the utility of these questions as retrieval queries, measuring their effectiveness in retrieving and ranking masked target documents. In some cases, generated questions ended up being better queries than extractive keyword queries generated by the same model. We additionally propose MMR-RBO, a variation of Maximal Marginal Relevance (MMR) using Rank-Biased Overlap (RBO) to identify which questions will perform competitively with the keyword baseline. As all question queries yield unique result sets, we contend that there are no stupid questions. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:13:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zimmerman",
"Brian D.",
""
],
[
"Aubert-Béduchaud",
"Julien",
""
],
[
"Boudin",
"Florian",
""
],
[
"Aizawa",
"Akiko",
""
],
[
"Vechtomova",
"Olga",
""
]
] |
2506.08197 | Anton Kuncinas | Anton Kun\v{c}inas | Dark Matter and CP Violation in Some Symmetry-Constrained 3HDMs | PhD thesis based on arXiv:2001.01994 , arXiv:2108.07026 , arXiv:2204.05684 , arXiv:2301.12194 , arXiv:2302.07210 , arXiv:2408.02728 | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | Multi-Higgs-doublet models can accommodate a dark matter candidate. An underlying symmetry not only ensures the candidate's stability but also helps control the number of free parameters. We consider two scenarios. First, we explore dark matter candidates in the $S_3$-symmetric three-Higgs-doublet model. Notably, our findings reveal that the cases we discuss allow for different dark matter mass ranges compared to other candidates within three-Higgs-doublet models. Additionally, we investigate an alternative to conventional stabilisation via discrete symmetries by examining stabilisation through continuous symmetries. This alternative approach exhibits a unique characteristic-the emergence of mass-degenerate states. We also explore the impact of CP violation in the $S_3$-symmetric three-Higgs-doublet model. Our analysis investigates how different representations of quarks under the $S_3$ group, combined with various vacuum configurations, lead to distinct phenomenological consequences, including the possibility of very light neutral scalars. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:16:17 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Kunčinas",
"Anton",
""
]
] |
2506.08198 | Naim Goksel Karacayli | N. G. Kara\c{c}ayl{\i}, P. L. Taylor | Unified reconstruction of the Lyman-alpha power spectrum with Hamiltonian Monte Carlo | 13 pages, 10 figures | null | null | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The complex geometry of the Ly$\alpha$ forest data has motivated the use of various two-point statistics as alternatives to the three-dimensional power spectrum ($P_{\mathrm{3D}}$), which carries cosmological information in Fourier space. On large scales, the three-dimensional correlation function ($\xi_\mathrm{3D}$) has provided robust measurements of the baryon acoustic oscillation (BAO) scale at 150 Mpc. On smaller scales, the one-dimensional power spectrum, $P_{\mathrm{1D}}(k_\|)$, has been the primary tool for extracting information. At the same time, the cross-spectrum, $P_\times(\theta, k_\|)$, has been introduced to incorporate angular information without the complications caused by survey window functions. We propose an analytical forward-modeling framework to reconstruct $P_{\mathrm{3D}}$ from all these observables in a nearly model-independent way. We demonstrate the performance of our method using a hypothetical mock data vector representative of future Dark Energy Spectroscopic Instrument (DESI) measurements and show that the monopole of $P_{\mathrm{3D}}$ can be reconstructed in 25 $k$ bins between $0.07~\mathrm{Mpc}^{-1}$ and $1.8~\mathrm{Mpc}^{-1}$, achieving a median precision of $8\%$ and a mean precision of $13\%$. Our method can serve as an intermediary for consistency checks, though it is not intended to replace direct $P_{\mathrm{3D}}$ estimation. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:18:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Karaçaylı",
"N. G.",
""
],
[
"Taylor",
"P. L.",
""
]
] |
2506.08199 | Brian Zimmerman | Brian D. Zimmerman, Joshua Folkins, Olga Vechtomova | Extracting Information About Publication Venues Using Citation-Informed Transformers | null | null | null | null | cs.DL | http://creativecommons.org/licenses/by/4.0/ | Scientific document embeddings contain a variety of rich features which can be harnessed for downstream tasks such as recommendation, ranking, and clustering. We explore which tangible insights can be drawn from scientific document embeddings to understand trends in computer science research featured across nine well-known venues. We collect approximately 60,000 scientific documents published between 2015 and 2023 and analyze their embeddings, which we produce with the SPECTER pre-trained language model. In particular, we examine whether similarity between two venues can be measured using the embeddings of the scientific documents they admit for publication. Our findings indicate that some venues within computer science are indistinguishable when only considering the distributions of their document embeddings. We additionally examine whether any two venues are becoming increasingly similar over time and identify a trend of convergence within some venues in our analysis. We discuss the implications of these results and the potential impact on new scientific contributions. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:20:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zimmerman",
"Brian D.",
""
],
[
"Folkins",
"Joshua",
""
],
[
"Vechtomova",
"Olga",
""
]
] |
2506.08200 | Phoebe Chua Ms | Kat R. Agres, Adyasha Dash, Phoebe Chua, Stefan K. Ehrlich | AffectMachine-Pop: A controllable expert system for real-time pop music generation | null | null | null | null | cs.HC cs.MM | http://creativecommons.org/licenses/by/4.0/ | Music is a powerful medium for influencing listeners' emotional states, and this capacity has driven a surge of research interest in AI-based affective music generation in recent years. Many existing systems, however, are a black box which are not directly controllable, thus making these systems less flexible and adaptive to users. We present \textit{AffectMachine-Pop}, an expert system capable of generating retro-pop music according to arousal and valence values, which can either be pre-determined or based on a listener's real-time emotion states. To validate the efficacy of the system, we conducted a listening study demonstrating that AffectMachine-Pop is capable of generating affective music at target levels of arousal and valence. The system is tailored for use either as a tool for generating interactive affective music based on user input, or for incorporation into biofeedback or neurofeedback systems to assist users with emotion self-regulation. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:21:21 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Agres",
"Kat R.",
""
],
[
"Dash",
"Adyasha",
""
],
[
"Chua",
"Phoebe",
""
],
[
"Ehrlich",
"Stefan K.",
""
]
] |
2506.08201 | Jalaj Upadhyay | Krishna Pillutla, Jalaj Upadhyay, Christopher A. Choquette-Choo, Krishnamurthy Dvijotham, Arun Ganesh, Monika Henzinger, Jonathan Katz, Ryan McKenna, H. Brendan McMahan, Keith Rush, Thomas Steinke, Abhradeep Thakurta | Correlated Noise Mechanisms for Differentially Private Learning | 212 pages | null | null | null | cs.LG cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This monograph explores the design and analysis of correlated noise mechanisms for differential privacy (DP), focusing on their application to private training of AI and machine learning models via the core primitive of estimation of weighted prefix sums. While typical DP mechanisms inject independent noise into each step of a stochastic gradient (SGD) learning algorithm in order to protect the privacy of the training data, a growing body of recent research demonstrates that introducing (anti-)correlations in the noise can significantly improve privacy-utility trade-offs by carefully canceling out some of the noise added on earlier steps in subsequent steps. Such correlated noise mechanisms, known variously as matrix mechanisms, factorization mechanisms, and DP-Follow-the-Regularized-Leader (DP-FTRL) when applied to learning algorithms, have also been influential in practice, with industrial deployment at a global scale. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:21:50 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Pillutla",
"Krishna",
""
],
[
"Upadhyay",
"Jalaj",
""
],
[
"Choquette-Choo",
"Christopher A.",
""
],
[
"Dvijotham",
"Krishnamurthy",
""
],
[
"Ganesh",
"Arun",
""
],
[
"Henzinger",
"Monika",
""
],
[
"Katz",
"Jonathan",
""
],
[
"McKenna",
"Ryan",
""
],
[
"McMahan",
"H. Brendan",
""
],
[
"Rush",
"Keith",
""
],
[
"Steinke",
"Thomas",
""
],
[
"Thakurta",
"Abhradeep",
""
]
] |
2506.08202 | Davide Augusto Bignamini | Davide A. Bignamini and Enrico Priola | Stochastic dissipative systems in Banach spaces driven by L\'evy noise | null | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we are interested in the well-posedness of stochastic reaction diffusion equations like \begin{equation} \begin{cases} dX(t)(\xi)=\big(\Delta_\xi X(t)(\xi)-p(X(t)(\xi))\big)dt+RdW(t)+dL(t) , \quad t\in [0,T];\\ X(0)=x\in L^2(\mathcal{O}) \end{cases} \end{equation} where $\mathcal{O}$ is a bounded open domain of $\mathbb{R}^d$ with regular boundary, $d\in\mathbb{N}$, $p:\mathbb{R}\rightarrow\mathbb{R}$ is a polynomial of odd degree with positive leading coefficient, $R$ is a linear bounded operator on $L^2(\mathcal{O})$, $\{W(t)\}_{t\geq 0}$ is a $L^2(\mathcal{O})$-cylindrical Wiener process, $\{L(t)\}_{t\geq 0}$ is a pure-jump L\'evy process on $L^2(\mathcal{O})$. We complement the equation with suitable boundary conditions on $\partial \mathcal{O}.$ Some papers in literature analize existence and uniqueness of mild solutions for every $x\in L^p(\mathcal{O})$, for some suitable $p\geq 2$. The results of this paper allow to study reaction diffusion equations also on the space of continuous function $C(\overline{O})$. This seems to be new in the L\'evy case (it is already done in the Wiener case).\\ We also discuss and review the previous cited works with the aim of unifying the different frameworks. We underline that when $R=0$ for every $x\in C(\overline{O})$ (or $x\in L^p(\mathcal{O})$) the mild solution to the equation has a c\`adl\`ag modifications in $C(\overline{O})$ (or $\in L^p(\mathcal{O})$), even if $\{L(t)\}_{t \geq 0}$ is not a L\'evy process taking values in $C(\overline{O})$ (or $\in L^p(\mathcal{O})$). This phenomenon for the linear problem (i.e., $F\equiv 0$ in the SPDE) has been investigated in other papers. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:22:03 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bignamini",
"Davide A.",
""
],
[
"Priola",
"Enrico",
""
]
] |
2506.08203 | Marina Chadeeva | M. Chadeeva and A. Stifutkin | Response to cosmic muons of scintillator-SiPM assemblies measured at different temperatures | 11 pages, 6 figures | null | null | null | physics.ins-det hep-ex | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Highly segmented calorimeters represent a modern trend in experimental particle physics aimed at improving the energy resolution with the particle flow reconstruction. The widely used and cost-effective solution is the structure assembled from small scintillator elements readout by tiny photosensors. The improvement of signal to noise ratio can be achieved by operating at low temperatures. The paper presents studies of temperature dependence of response to muons for elements comprised of plastic scintillator tiles with different options of reflective coverage and read out by silicon photomultipliers. No temperature dependence of response for bare tiles has been detected. With reflective film, the reduction of response with decreasing temperature is observed. The measured effect amounts to 10% in the range from +30oC to -30oC. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:22:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Chadeeva",
"M.",
""
],
[
"Stifutkin",
"A.",
""
]
] |
2506.08204 | Alexander Von Hoegen | Alexander von Hoegen, Tommy Tai, Clifford J. Allington, Matthew Yeung, Jacob Pettine, Marios H. Michael, Emil Vi\~nas Bostr\"om, Xiaomeng Cui, Kierstin Torres, Alexander E. Kossak, Byunghun Lee, Geoffrey S. D. Beach, Angel Rubio, Philip Kim, Nuh Gedik | Visualizing a Terahertz Superfluid Plasmon in a Two-Dimensional Superconductor | 23 pages, 4 figures | null | null | null | cond-mat.supr-con cond-mat.mes-hall cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The superconducting gap defines the fundamental energy scale for the emergence of dissipationless transport and collective phenomena in a superconductor. In layered high-temperature cuprate superconductors, where the Cooper pairs are confined to weakly coupled two-dimensional copper-oxygen planes, terahertz (THz) spectroscopy at sub-gap millielectronvolt energies has provided crucial insights into the collective superfluid response perpendicular to the superconducting layers. However, within the copper-oxygen planes the collective superfluid response manifests as plasmonic charge oscillations at energies far exceeding the superconducting gap, obscured by strong dissipation. Here, we present spectroscopic evidence of a below-gap, two-dimensional superfluid plasmon in few-layer Bi2Sr2CaCu2O8+x and spatially resolve its deeply sub-diffractive THz electrodynamics. By placing the superconductor in the near-field of a spintronic THz emitter, we reveal this distinct resonance-absent in bulk samples and observed only in the superconducting phase-and determine its plasmonic nature by mapping the geometric anisotropy and dispersion. Crucially, these measurements offer a direct view of the momentum- and frequency dependent superconducting transition in two dimensions. These results establish a new platform for investigating superfluid phenomena at finite momenta and THz frequencies, highlighting the potential to engineer and visualize superconducting devices operating at ultrafast THz rates. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:26:49 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"von Hoegen",
"Alexander",
""
],
[
"Tai",
"Tommy",
""
],
[
"Allington",
"Clifford J.",
""
],
[
"Yeung",
"Matthew",
""
],
[
"Pettine",
"Jacob",
""
],
[
"Michael",
"Marios H.",
""
],
[
"Boström",
"Emil Viñas",
""
],
[
"Cui",
"Xiaomeng",
""
],
[
"Torres",
"Kierstin",
""
],
[
"Kossak",
"Alexander E.",
""
],
[
"Lee",
"Byunghun",
""
],
[
"Beach",
"Geoffrey S. D.",
""
],
[
"Rubio",
"Angel",
""
],
[
"Kim",
"Philip",
""
],
[
"Gedik",
"Nuh",
""
]
] |
2506.08205 | Shadab Anwar Shaikh | Shadab Anwar Shaikh, Kranthi Balusu, Ayoub Soulami | A Machine Learning Approach to Generate Residual Stress Distributions using Sparse Characterization Data in Friction-Stir Processed Parts | null | null | null | null | cs.LG cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Residual stresses, which remain within a component after processing, can deteriorate performance. Accurately determining their full-field distributions is essential for optimizing the structural integrity and longevity. However, the experimental effort required for full-field characterization is impractical. Given these challenges, this work proposes a machine learning (ML) based Residual Stress Generator (RSG) to infer full-field stresses from limited measurements. An extensive dataset was initially constructed by performing numerous process simulations with a diverse parameter set. A ML model based on U-Net architecture was then trained to learn the underlying structure through systematic hyperparameter tuning. Then, the model's ability to generate simulated stresses was evaluated, and it was ultimately tested on actual characterization data to validate its effectiveness. The model's prediction of simulated stresses shows that it achieved excellent predictive accuracy and exhibited a significant degree of generalization, indicating that it successfully learnt the latent structure of residual stress distribution. The RSG's performance in predicting experimentally characterized data highlights the feasibility of the proposed approach in providing a comprehensive understanding of residual stress distributions from limited measurements, thereby significantly reducing experimental efforts. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:26:57 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Shaikh",
"Shadab Anwar",
""
],
[
"Balusu",
"Kranthi",
""
],
[
"Soulami",
"Ayoub",
""
]
] |
2506.08206 | Satyanarayan Kothe | Sharmishtha Raghuvanshi, Supriya Sanjay Nikam, Manisha Karne, Satyanarayan Kishan Kothe | Unmasking inequility: socio-economic determinants and gender disparities in Maharashtra and India's health outcomes -- Insights from NFHS-5 | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | This research examines the persistent challenge of health inequalities in India, departing from the conventional focus on aggregate improvements in mortality rates. While India has achieved progress in overall health indicators since independence, the distribution of health outcomes remains uneven, a fact starkly highlighted by the COVID-19 pandemic. This study investigates the socio-economic determinants of health disparities using the National Family and Health Survey (NFHS)-5 data from 2019-20, focusing on both national and state-level analyses, specifically for Maharashtra. Employing a health economics framework, the analysis delves into individual-level data, population shares, self-reported morbidity prevalence, and treatment patterns across diverse socio-economic groups. Regression analyses, stratified by gender, are conducted to quantify the impact of socio-economic factors on reported morbidity. Furthermore, a Fairlie decomposition, an extension of the Oaxaca decomposition, is utilised to dissect the gender gap in morbidity, assessing the extent to which observed differences are attributable to explanatory variables. The findings reveal a significant burden of self-reported morbidity, with approximately one in nine individuals in India and one in eight in Maharashtra reporting morbidity. Notably, women exhibit nearly double the morbidity rate compared to men. The decomposition analysis identifies key drivers of gender disparities. In India, marital status exacerbates these differences, while insurance coverage, caste, urban residence, and wealth mitigate them. In Maharashtra, urban residence and marital status widen the gap, whereas religion, caste, and insurance coverage narrow it. This research underscores the importance of targeted policy interventions to address the complex interplay of socio-economic factors driving health inequalities in India. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:27:43 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Raghuvanshi",
"Sharmishtha",
""
],
[
"Nikam",
"Supriya Sanjay",
""
],
[
"Karne",
"Manisha",
""
],
[
"Kothe",
"Satyanarayan Kishan",
""
]
] |
2506.08207 | Alejandro Pe\~na-Torres PhD. | Alejandro Pe\~na-Torres and Michail Stamatakis and Hannes J\'onsson | Dimerization of Ag adatoms on Si(100) at surprisingly low temperature due to adatom migration on top of Si-dimer rows | 10 pages, 5 figures | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | The puzzling experimental observations of Ag addimer formation on the Si(100) surface upon vapor deposition at low temperature, even as low as 140 K, reported by Huang et al. [Phys. Chem. Chem. Phys. 2021, 23, 4161], is explained by facile thermal migration on top of the Si dimer rows, while diffusion is inactive once an Ag adatom lands in one of the optimal binding sites in between dimer rows. The previously hypothesized transient mobility due to the hot spot formed as an Ag atom lands on the Si surface is found not to contribute significantly to mobility and dimer formation. The experimental conditions are simulated by a combination of classical dynamics calculations of the deposition events, systematic searches of saddle points representing transition states for thermally activated events and kinetic Monte Carlo simulations of long timescale evolution of the system over a temperature range from 100 K to 300 K. While the calculated energy barriers for diffusion parallel and perpendicular to the Si dimer rows are found to be nearly equal, indicating isotropic diffusion, the simulations show highly anisotropic diffusion, as the optimal migration mechanism involves multiple hops of Ag adatoms on top of Si dimer rows. Impinging Ag atoms are found to have a 90% chance of landing on top of a dimer row and while the hot spot created cools down too fast for transient mobility to play an important role, the energy barrier for thermally activated hops along the dimer row is low enough for migration to be active even at 100 K. A migrating Ag adatom can dimerise with another Ag adatom sitting at a stable binding site in between rows or another adatom on top of the same row. The simulations for deposition at 140 K show that most of the deposited Ag atoms end up forming dimers. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:28:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Peña-Torres",
"Alejandro",
""
],
[
"Stamatakis",
"Michail",
""
],
[
"Jónsson",
"Hannes",
""
]
] |
2506.08208 | Teresa Sicignano | Teresa Sicignano, Vincenzo Ripepi, Marina Rejkuba, Martino Romaniello, Marcella Marconi, Roberto Molinaro, Anupam Bhardwaj, Giulia De Somma, Maria-Rosa Cioni, Felice Cusano, Gisella Clementini, Richard de Grijs, Valentin Ivanov, Jesper Storm, Martin Groenewegen | The VMC Survey -- LIV. Anomalous Cepheids in the Magellanic Clouds Period-Luminosity relations in the near-infrared bands | 28 pages, 14 figures, Accepted for publication in A&A | null | null | null | astro-ph.SR astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomalous Cepheids (ACs) are less studied metal-poor pulsating stars ([Fe/H]<-1.5) compared to Classical Cepheids (CCs) and RR Lyrae stars. They follow distinct Period-Luminosity (PL) and Period-Wesenheit (PW) relations and pulsate in either the fundamental (F) or first overtone (1O) mode. Our goal is to assess the precision and accuracy of AC-based distances and evaluate their potential for establishing an independent distance scale. We derive new PL and PW relations for F-mode, 1O-mode, and, for the first time, combined F+1O ACs in the Magellanic Clouds. We study their wavelength dependence and apply the relations to estimate distances to Local Group stellar systems hosting ACs, while also confirming AC classifications. Our analysis is based on near-infrared time-series photometry in the Y, J, and Ks bands for about 200 ACs in the Magellanic Clouds from the VISTA survey of the Magellanic Clouds system (VMC, 2009-2018). VMC data are complemented with optical photometry from Gaia DR3 and OGLE-IV, which also provide periods and pulsation modes. Custom light-curve templates were used to derive precise intensity-averaged magnitudes for 118 ACs in the Large Magellanic Cloud (LMC) and 75 in the Small Magellanic Cloud. These data were used to derive multi-band PL and PW relations, calibrated using the geometric LMC distance from eclipsing binaries. We find that PL relation slopes increase and dispersions decrease with wavelength. Using Gaia parallaxes, we determine the LMC distance modulus and the LMC-SMC relative distance. We also confirm the AC nature of several new candidates in Galactic Globular Clusters and derive a distance modulus for the Draco dSph galaxy of 19.425+/-0.048 mag. A 0.1 mag discrepancy with RR Lyrae-based distances may reflect metallicity effects. Future spectroscopic surveys and Gaia DR4 will help refine the AC distance scale and quantify metallicity impacts. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:29:05 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Sicignano",
"Teresa",
""
],
[
"Ripepi",
"Vincenzo",
""
],
[
"Rejkuba",
"Marina",
""
],
[
"Romaniello",
"Martino",
""
],
[
"Marconi",
"Marcella",
""
],
[
"Molinaro",
"Roberto",
""
],
[
"Bhardwaj",
"Anupam",
""
],
[
"De Somma",
"Giulia",
""
],
[
"Cioni",
"Maria-Rosa",
""
],
[
"Cusano",
"Felice",
""
],
[
"Clementini",
"Gisella",
""
],
[
"de Grijs",
"Richard",
""
],
[
"Ivanov",
"Valentin",
""
],
[
"Storm",
"Jesper",
""
],
[
"Groenewegen",
"Martin",
""
]
] |
2506.08209 | Lev Ricanati | Lev Ricanati | Valuing Diffuse Global Public Goods from Satellite Constellations: Evidence from GPS and Airline Delays | 53 pages, 6 figures | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | This paper studies the welfare impact of discrete improvements to global public goods in the context of the Global Positioning System (GPS). Specifically, I find that by disabling Selective Availability in May, 2000, and thus significantly increasing the accuracy of GPS, the United States generated at least \$268 million (2000 dollars) of additional welfare gains. To quantify this welfare impact, I apply a difference-in-differences model to the Bureau of Transportation Statistics's Airline On-Time Performance Data in the years 1999 and 2000. I use this model to estimate the time saved per flight attributable to the improved GPS and multiply these time savings by the number of passengers in the ensuing year and their values of time. I conclude by estimating the economic loss from current threats to the provision of satellite-based global public goods. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:29:45 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ricanati",
"Lev",
""
]
] |
2506.08210 | Andrew Wang | Andrew Z. Wang, Songwei Ge, Tero Karras, Ming-Yu Liu, Yogesh Balaji | A Comprehensive Study of Decoder-Only LLMs for Text-to-Image Generation | CVPR 2025 | null | null | null | cs.CV cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Both text-to-image generation and large language models (LLMs) have made significant advancements. However, many text-to-image models still employ the somewhat outdated T5 and CLIP as their text encoders. In this work, we investigate the effectiveness of using modern decoder-only LLMs as text encoders for text-to-image diffusion models. We build a standardized training and evaluation pipeline that allows us to isolate and evaluate the effect of different text embeddings. We train a total of 27 text-to-image models with 12 different text encoders to analyze the critical aspects of LLMs that could impact text-to-image generation, including the approaches to extract embeddings, different LLMs variants, and model sizes. Our experiments reveal that the de facto way of using last-layer embeddings as conditioning leads to inferior performance. Instead, we explore embeddings from various layers and find that using layer-normalized averaging across all layers significantly improves alignment with complex prompts. Most LLMs with this conditioning outperform the baseline T5 model, showing enhanced performance in advanced visio-linguistic reasoning skills. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:29:53 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Andrew Z.",
""
],
[
"Ge",
"Songwei",
""
],
[
"Karras",
"Tero",
""
],
[
"Liu",
"Ming-Yu",
""
],
[
"Balaji",
"Yogesh",
""
]
] |
2506.08211 | Jose Guadalupe Romero Velazquez | Romeo Ortega, Jose Guadalupe Romero, Stanislav Aranovskiy, Gang Tao | Standard LSParameter Estimators Ensure Finite Convergence Time for Linear Regression Equations Under an Interval Excitation Assumption | null | null | null | null | eess.SY cs.SY math.ST stat.TH | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this brief note we recall the little-known fact that, for linear regression equations (LRE) with intervally excited (IE) regressors, standard Least Square (LS) parameter estimators ensure finite convergence time (FCT) of the estimated parameters. The convergence time being equal to the time length needed to comply with the IE assumption. As is well-known, IE is necessary and sufficient for the identifiability of the LRE-hence, it is the weakest assumption for the on-or off-line solution of the parameter estimation problem. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:30:30 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ortega",
"Romeo",
""
],
[
"Romero",
"Jose Guadalupe",
""
],
[
"Aranovskiy",
"Stanislav",
""
],
[
"Tao",
"Gang",
""
]
] |
2506.08212 | Zahra Shomali | Seyedeh Ameneh Bahadori, Zahra Shomali, Reza Asgari | Heat Dissipation and Thermoelectric Performance of InSe-Based Monolayers: A Monte Carlo Simulation Study | 9 pages, 10 figures | null | null | null | physics.app-ph cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Using nonequilibrium Monte Carlo simulations of the phonon Boltzmann transport equation, we study transient heat transfer in five indium-based two-dimensional monolayers: Janus monolayers In$_2$SeTe and In$_2$SSe, pristine InSe, and InSe under 4$\%$ and 6$\%$ tensile strain. In this work, the potential of these materials for energy conversion in thermoelectric generators and hotspot control in metal-oxide-semiconductor field-effect transistors is investigated. A promising option for an effective heat dissipation and enhanced transistor reliability is found to be a strained InSe, which shows the lowest peak temperature during the heating among the studied materials. On the other hand, with a high Seebeck coefficient, low thermal conductivity, and an improved figure of merit, the Janus In$_2$SeTe monolayer, compensates for its increased phonon scattering to reach the maximum temperature, making it a potent thermoelectric material. Our findings emphasis the importance of strain engineering and structural asymmetry in tuning phonon transport, enabling material optimization for next-generation nanoelectronic and energy-harvesting devices. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:31:23 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bahadori",
"Seyedeh Ameneh",
""
],
[
"Shomali",
"Zahra",
""
],
[
"Asgari",
"Reza",
""
]
] |
2506.08214 | Ioannis Iakovidis | Ioannis Iakovidis, Zahra Kalantari, Amir Hossein Payberah, Fernando Jaramillo, Francisco Pena Escobar | Using Satellite Images And Self-supervised Machine Learning Networks To Detect Water Hidden Under Vegetation | 16 pages, 9 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In recent years the wide availability of high-resolution radar satellite images along with the advancement of computer vision models have enabled the remote monitoring of the surface area of wetlands. However, these models require large amounts of manually annotated satellite images, which are slow and expensive to produce. To overcome this problem, self-supervised training methods have been deployed to train models without using annotated data. In this paper we use a combination of deep clustering and negative sampling to train a model to segment radar satellite images into areas that separate water from land without the use of any manual annotations. Furthermore, we implement an ensemble version of the model to reduce variance and improve performance. Compared to a single fully-supervised model using the same architecture, our ensemble of self-supervised models achieves a 0.02 improvement in the Intersection Over Union metric over our test dataset. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:35:31 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Iakovidis",
"Ioannis",
""
],
[
"Kalantari",
"Zahra",
""
],
[
"Payberah",
"Amir Hossein",
""
],
[
"Jaramillo",
"Fernando",
""
],
[
"Escobar",
"Francisco Pena",
""
]
] |
2506.08215 | Marin Bu\v{z}an\v{c}i\'c | Marin Bu\v{z}an\v{c}i\'c, Igor Vel\v{c}i\'c, Josip \v{Z}ubrini\'c | Homogenization of elasto-plastic plate equations with vanishing hardening | 44 pages. arXiv admin note: substantial text overlap with arXiv:2302.14758, arXiv:2212.02116 | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | Building on recent work analyzing the interaction between homogenization and dimension reduction in perfectly plastic materials, we focus on the extreme regime where the plate thickness vanishes much faster than the period of oscillation of the heterogeneous material. Our investigation is devoted to the derivation of the existence of a quasistatic evolution for a multiphase elasto-plastic composite without imposing any kind of ordering constraints on admissible yield surfaces of the various phases. The analysis is carried out in two steps: we first derive a heterogeneous plate model with hardening via dimension reduction using evolutionary $\Gamma$-convergence techniques, and then perform two-scale homogenization while simultaneously letting the hardening parameters tend to zero using a stress-strain approach. This approach allows us to reconcile the Kirchhoff-Love structure of limiting displacements with the expected requirement that the effective dissipation potential at each point of the interface between two phases needs to be the pointwise-in-space inf-convolution of that in either phase. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:35:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bužančić",
"Marin",
""
],
[
"Velčić",
"Igor",
""
],
[
"Žubrinić",
"Josip",
""
]
] |
2506.08216 | Shahaf Bassan | Shahaf Bassan, Guy Amir, Meirav Zehavi, Guy Katz | What makes an Ensemble (Un) Interpretable? | To appear in ICML 2025 | null | null | null | cs.LG cs.CC cs.LO | http://creativecommons.org/licenses/by/4.0/ | Ensemble models are widely recognized in the ML community for their limited interpretability. For instance, while a single decision tree is considered interpretable, ensembles of trees (e.g., boosted trees) are often treated as black-boxes. Despite this folklore recognition, there remains a lack of rigorous mathematical understanding of what particularly makes an ensemble (un)-interpretable, including how fundamental factors like the (1) *number*, (2) *size*, and (3) *type* of base models influence its interpretability. In this work, we seek to bridge this gap by applying concepts from computational complexity theory to study the challenges of generating explanations for various ensemble configurations. Our analysis uncovers nuanced complexity patterns influenced by various factors. For example, we demonstrate that under standard complexity assumptions like P$\neq$NP, interpreting ensembles remains intractable even when base models are of constant size. Surprisingly, the complexity changes drastically with the number of base models: small ensembles of decision trees are efficiently interpretable, whereas interpreting ensembles with even a constant number of linear models remains intractable. We believe that our findings provide a more robust foundation for understanding the interpretability of ensembles, emphasizing the benefits of examining it through a computational complexity lens. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:38:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bassan",
"Shahaf",
""
],
[
"Amir",
"Guy",
""
],
[
"Zehavi",
"Meirav",
""
],
[
"Katz",
"Guy",
""
]
] |
2506.08217 | Filippo Calderoni | Filippo Calderoni and Luca Motto Ros | Structural results on idealistic equivalence relations | 26 pages | null | null | null | math.LO | http://creativecommons.org/licenses/by/4.0/ | We address some fundamental problems concerning the structure of idealistic equivalence relations. In particular, we show that, under analytic determinacy, there are continuum many idealistic analytic equivalence relations that are not classwise Borel isomorphic to an orbit equivalence relation. Moreover, we also discuss alternative formulations of the E1 conjecture, and present an elementary counterexample to one of its earliest variants proposed by Hjorth and Kechris in 1997. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:38:17 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Calderoni",
"Filippo",
""
],
[
"Ros",
"Luca Motto",
""
]
] |
2506.08218 | Alan Mills | Alan Mills, Jonathan White, Phil Legg | gh0stEdit: Exploiting Layer-Based Access Vulnerability Within Docker Container Images | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Containerisation is a popular deployment process for application-level virtualisation using a layer-based approach. Docker is a leading provider of containerisation, and through the Docker Hub, users can supply Docker images for sharing and re-purposing popular software application containers. Using a combination of in-built inspection commands, publicly displayed image layer content, and static image scanning, Docker images are designed to ensure end users can clearly assess the content of the image before running them. In this paper we present \textbf{\textit{gh0stEdit}}, a vulnerability that fundamentally undermines the integrity of Docker images and subverts the assumed trust and transparency they utilise. The use of gh0stEdit allows an attacker to maliciously edit Docker images, in a way that is not shown within the image history, hierarchy or commands. This attack can also be carried out against signed images (Docker Content Trust) without invalidating the image signature. We present two use case studies for this vulnerability, and showcase how gh0stEdit is able to poison an image in a way that is not picked up through static or dynamic scanning tools. Our attack case studies highlight the issues in the current approach to Docker image security and trust, and expose an attack method which could potentially be exploited in the wild without being detected. To the best of our knowledge we are the first to provide detailed discussion on the exploit of this vulnerability. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:38:17 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Mills",
"Alan",
""
],
[
"White",
"Jonathan",
""
],
[
"Legg",
"Phil",
""
]
] |
2506.08219 | Vibhuti Rai | Vibhuti N. Rai, Junyoung Sim, Florian Faaber, Nils Bogdanoff, Sergey Trishin, Paul Wiechers, Tom S. Seifert, Tobias Kampfrath, Christian Lotze and Katharina J. Franke | Influence of atomic-scale defects on coherent phonon excitations by THz near fields in an STM | 12 pages, 10 figures | null | null | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Coherent phonons describe the collective, ultrafast motion of atoms and play a central role in light-induced structural dynamics. Here, we employ terahertz scanning tunneling microscopy (THz-STM) to excite and detect coherent phonons in semiconducting 2H-MoTe$_{2}$ and resolve how their excitation is influenced by atomic-scale defects. In a THz pump-probe scheme, we observe long-lived oscillatory signals that we assign to out-of-plane breathing and in-plane shear modes, which are both forbidden in the bulk. Remarkably, the relative excitation strength of these modes varies near defects, indicating that local band bending modulates the coupling to the THz field. This defect-tunable coupling offers new opportunities to control material properties via selective excitation of vibrational modes at the nanoscale. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:39:35 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Rai",
"Vibhuti N.",
""
],
[
"Sim",
"Junyoung",
""
],
[
"Faaber",
"Florian",
""
],
[
"Bogdanoff",
"Nils",
""
],
[
"Trishin",
"Sergey",
""
],
[
"Wiechers",
"Paul",
""
],
[
"Seifert",
"Tom S.",
""
],
[
"Kampfrath",
"Tobias",
""
],
[
"Lotze",
"Christian",
""
],
[
"Franke",
"Katharina J.",
""
]
] |
2506.08220 | Octave Mariotti | Octave Mariotti, Zhipeng Du, Yash Bhalgat, Oisin Mac Aodha, Hakan Bilen | Jamais Vu: Exposing the Generalization Gap in Supervised Semantic Correspondence | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | Semantic correspondence (SC) aims to establish semantically meaningful matches across different instances of an object category. We illustrate how recent supervised SC methods remain limited in their ability to generalize beyond sparsely annotated training keypoints, effectively acting as keypoint detectors. To address this, we propose a novel approach for learning dense correspondences by lifting 2D keypoints into a canonical 3D space using monocular depth estimation. Our method constructs a continuous canonical manifold that captures object geometry without requiring explicit 3D supervision or camera annotations. Additionally, we introduce SPair-U, an extension of SPair-71k with novel keypoint annotations, to better assess generalization. Experiments not only demonstrate that our model significantly outperforms supervised baselines on unseen keypoints, highlighting its effectiveness in learning robust correspondences, but that unsupervised baselines outperform supervised counterparts when generalized across different datasets. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:40:47 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Mariotti",
"Octave",
""
],
[
"Du",
"Zhipeng",
""
],
[
"Bhalgat",
"Yash",
""
],
[
"Mac Aodha",
"Oisin",
""
],
[
"Bilen",
"Hakan",
""
]
] |
2506.08221 | Samra Zafar Ms | Samra Zafar, Shaheer Minhas, Syed Ali Hassan Zaidi, Arfa Naeem, Zahra Ali | "I Wrote, I Paused, I Rewrote" Teaching LLMs to Read Between the Lines of Student Writing | 7 pages, 6 figures, 2 tables | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models(LLMs) like Gemini are becoming common tools for supporting student writing. But most of their feedback is based only on the final essay missing important context about how that text was written. In this paper, we explore whether using writing process data, collected through keystroke logging and periodic snapshots, can help LLMs give feedback that better reflects how learners think and revise while writing. We built a digital writing tool that captures both what students type and how their essays evolve over time. Twenty students used this tool to write timed essays, which were then evaluated in two ways: (i) LLM generated feedback using both the final essay and the full writing trace, and (ii) After the task, students completed surveys about how useful and relatable they found the feedback. Early results show that learners preferred the process-aware LLM feedback, finding it more in tune with their own thinking. We also found that certain types of edits, like adding new content or reorganizing paragraphs, aligned closely with higher scores in areas like coherence and elaboration. Our findings suggest that making LLMs more aware of the writing process can lead to feedback that feels more meaningful, personal, and supportive. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:42:02 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zafar",
"Samra",
""
],
[
"Minhas",
"Shaheer",
""
],
[
"Zaidi",
"Syed Ali Hassan",
""
],
[
"Naeem",
"Arfa",
""
],
[
"Ali",
"Zahra",
""
]
] |
2506.08222 | Elise Chavez | Elise Chavez, Maria Acosta-Flechas, Christophe Bonnaud, Burt Holzman, Tulika Bose | The Elastic Analysis Facility's (EAF's) Contribution to the Future of Analysis at Multi-Experiment Institutions and Future Colliders | 11 pages, 6 figures, 1 table, accepted as contribution for the European Strategy for Particle Physics - 2026 update, Matches published version | null | null | FERMILAB-PUB-25-0152-CMS | physics.data-an hep-ex | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The Elastic Analysis Facility (EAF) hosted at Fermi National Accelerator Laboratory (Fermilab) is a platform being developed with the goal of providing a fast and efficient facility for physics analysis. As high-energy physics moves towards collecting larger datasets, such as those from the High-Luminosity LHC, the EAF strives to provide a powerful and adaptable framework for future colliders and multi-experiment institutions. Currently, the EAF supports several experiments including CMS, NOvA, and DUNE as well as serving accelerator physicists and beam line operations through integrated software and secure connections to Fermilab's computing resources. In addition, the EAF was designed with a user-friendly interface, intended to be more intuitive for emerging generations of physicists, that is still accessible for established styles of analysis. The EAF can also achieve better analysis efficiency due to the modernization of software and tools that can better utilize Fermilab's computing power. Furthermore, its design incorporates industry standards whenever possible, enhancing its sustainability and making it a possible template for other national or international laboratories and research facilities. Overall, the EAF is a forward-looking solution that will meet the evolving needs of particle physics, ensuring readiness for future colliders and multi-experiment research institutions. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:43:47 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Chavez",
"Elise",
""
],
[
"Acosta-Flechas",
"Maria",
""
],
[
"Bonnaud",
"Christophe",
""
],
[
"Holzman",
"Burt",
""
],
[
"Bose",
"Tulika",
""
]
] |
2506.08223 | Mikhail Kats | Jin-Woo Cho, Jonathan L. King, Martin Hafermann, Karla Paz, Shenwei Yin, Tanuj Kumar, Hongyan Mei, Joseph A. Andrade, Dung. T. Quach, Colin Hessel, Chenghao Wan, Carsten Ronning, David Woolf, and Mikhail A. Kats | Precise control of the phase-transition temperature of VO2 from 70 to -35 {\deg}C for tunable optics | Main text + supplementary | null | null | null | physics.optics cond-mat.mtrl-sci cond-mat.str-el physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We demonstrate the combination of metal doping and ion bombardment as a hybrid technique to precisely engineer the insulator-to-metal transition (IMT) properties of vanadium dioxide (VO$_2$) films. This approach enables (i) setting the phase-transition temperature ($T_c$) from $-35$ to $70$ {\deg}C, (ii) near-zero hysteresis in highly doped and irradiated films, and (iii) robust optical contrast across a broad range of doping and irradiation conditions, down to $T_c = -35$ {\deg}C. While previous work has focused on engineering the IMT via one of metal doping or ion irradiation, our hybrid method utilizes both. In doing so, we address several lingering challenges in VO$_2$ engineering, including the trade-off between $T_c$ reduction and optical contrast, and limitations in the achievable value of T$_c$. Strain$-T_c$ correlation from XRD and Raman measurements suggests a common mechanism behind doping- and irradiation-induced $T_c$ reduction. The ability to reduce the $T_c$ and hysteresis of VO$_2$ while maintaining large optical contrast will extend the functionality of VO$_2$-based optical and optoelectronic devices, including detectors, limiters, and modulators. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:45:06 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Cho",
"Jin-Woo",
""
],
[
"King",
"Jonathan L.",
""
],
[
"Hafermann",
"Martin",
""
],
[
"Paz",
"Karla",
""
],
[
"Yin",
"Shenwei",
""
],
[
"Kumar",
"Tanuj",
""
],
[
"Mei",
"Hongyan",
""
],
[
"Andrade",
"Joseph A.",
""
],
[
"Quach",
"Dung. T.",
""
],
[
"Hessel",
"Colin",
""
],
[
"Wan",
"Chenghao",
""
],
[
"Ronning",
"Carsten",
""
],
[
"Woolf",
"David",
""
],
[
"Kats",
"Mikhail A.",
""
]
] |
2506.08224 | Qiang Zhu | Osman Goni Ridwan, Sylvain Piti\'e, Monish Soundar Raj, Dong Dai, Gilles Frapper, Hongfei Xue and Qiang Zhu | AI-Assisted Rapid Crystal Structure Generation Towards a Target Local Environment | 14 pages, 7 figures | null | null | null | cond-mat.mtrl-sci cs.AI physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | In the field of material design, traditional crystal structure prediction approaches require extensive structural sampling through computationally expensive energy minimization methods using either force fields or quantum mechanical simulations. While emerging artificial intelligence (AI) generative models have shown great promise in generating realistic crystal structures more rapidly, most existing models fail to account for the unique symmetries and periodicity of crystalline materials, and they are limited to handling structures with only a few tens of atoms per unit cell. Here, we present a symmetry-informed AI generative approach called Local Environment Geometry-Oriented Crystal Generator (LEGO-xtal) that overcomes these limitations. Our method generates initial structures using AI models trained on an augmented small dataset, and then optimizes them using machine learning structure descriptors rather than traditional energy-based optimization. We demonstrate the effectiveness of LEGO-xtal by expanding from 25 known low-energy sp2 carbon allotropes to over 1,700, all within 0.5 eV/atom of the ground-state energy of graphite. This framework offers a generalizable strategy for the targeted design of materials with modular building blocks, such as metal-organic frameworks and next-generation battery materials. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:47:36 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ridwan",
"Osman Goni",
""
],
[
"Pitié",
"Sylvain",
""
],
[
"Raj",
"Monish Soundar",
""
],
[
"Dai",
"Dong",
""
],
[
"Frapper",
"Gilles",
""
],
[
"Xue",
"Hongfei",
""
],
[
"Zhu",
"Qiang",
""
]
] |
2506.08225 | Francisco Olivares | Davide Cenzato, Francisco Olivares and Nicola Prezza | Testing Suffixient Sets | null | null | null | null | cs.DS | http://creativecommons.org/licenses/by/4.0/ | Suffixient sets are a novel prefix array (PA) compression technique based on subsampling PA (rather than compressing the entire array like previous techniques used to do): by storing very few entries of PA (in fact, a compressed number of entries), one can prove that pattern matching via binary search is still possible provided that random access is available on the text. In this paper, we tackle the problems of determining whether a given subset of text positions is (1) a suffixient set or (2) a suffixient set of minimum cardinality. We provide linear-time algorithms solving these problems. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:49:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Cenzato",
"Davide",
""
],
[
"Olivares",
"Francisco",
""
],
[
"Prezza",
"Nicola",
""
]
] |
2506.08226 | Arthur Feeney | Arthur Feeney, Kuei-Hsiang Huang, Aparna Chandramowlishwaran | Mondrian: Transformer Operators via Domain Decomposition | 26 pages, 7 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Operator learning enables data-driven modeling of partial differential equations (PDEs) by learning mappings between function spaces. However, scaling transformer-based operator models to high-resolution, multiscale domains remains a challenge due to the quadratic cost of attention and its coupling to discretization. We introduce \textbf{Mondrian}, transformer operators that decompose a domain into non-overlapping subdomains and apply attention over sequences of subdomain-restricted functions. Leveraging principles from domain decomposition, Mondrian decouples attention from discretization. Within each subdomain, it replaces standard layers with expressive neural operators, and attention across subdomains is computed via softmax-based inner products over functions. The formulation naturally extends to hierarchical windowed and neighborhood attention, supporting both local and global interactions. Mondrian achieves strong performance on Allen-Cahn and Navier-Stokes PDEs, demonstrating resolution scaling without retraining. These results highlight the promise of domain-decomposed attention for scalable and general-purpose neural operators. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:52:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Feeney",
"Arthur",
""
],
[
"Huang",
"Kuei-Hsiang",
""
],
[
"Chandramowlishwaran",
"Aparna",
""
]
] |
2506.08227 | Shyamgopal Karthik | Vishaal Udandarao, Mehdi Cherti, Shyamgopal Karthik, Jenia Jitsev, Samuel Albanie, Matthias Bethge | A Good CREPE needs more than just Sugar: Investigating Biases in Compositional Vision-Language Benchmarks | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We investigate 17 benchmarks (e.g. SugarCREPE, VALSE) commonly used for measuring compositional understanding capabilities of vision-language models (VLMs). We scrutinize design choices in their construction, including data source (e.g. MS-COCO) and curation procedures (e.g. constructing negative images/captions), uncovering several inherent biases across most benchmarks. We find that blind heuristics (e.g. token-length, log-likelihood under a language model) perform on par with CLIP models, indicating that these benchmarks do not effectively measure compositional understanding. We demonstrate that the underlying factor is a distribution asymmetry between positive and negative images/captions, induced by the benchmark construction procedures. To mitigate these issues, we provide a few key recommendations for constructing more robust vision-language compositional understanding benchmarks, that would be less prone to such simple attacks. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:53:43 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Udandarao",
"Vishaal",
""
],
[
"Cherti",
"Mehdi",
""
],
[
"Karthik",
"Shyamgopal",
""
],
[
"Jitsev",
"Jenia",
""
],
[
"Albanie",
"Samuel",
""
],
[
"Bethge",
"Matthias",
""
]
] |
2506.08228 | Sergio Casas | Mustafa Baniodeh, Kratarth Goel, Scott Ettinger, Carlos Fuertes, Ari Seff, Tim Shen, Cole Gulino, Chenjie Yang, Ghassen Jerfel, Dokook Choe, Rui Wang, Vinutha Kallem, Sergio Casas, Rami Al-Rfou, Benjamin Sapp, Dragomir Anguelov | Scaling Laws of Motion Forecasting and Planning -- A Technical Report | null | null | null | null | cs.LG cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the empirical scaling laws of a family of encoder-decoder autoregressive transformer models on the task of joint motion forecasting and planning in the autonomous driving domain. Using a 500 thousand hours driving dataset, we demonstrate that, similar to language modeling, model performance improves as a power-law function of the total compute budget, and we observe a strong correlation between model training loss and model evaluation metrics. Most interestingly, closed-loop metrics also improve with scaling, which has important implications for the suitability of open-loop metrics for model development and hill climbing. We also study the optimal scaling of the number of transformer parameters and the training data size for a training compute-optimal model. We find that as the training compute budget grows, optimal scaling requires increasing the model size 1.5x as fast as the dataset size. We also study inference-time compute scaling, where we observe that sampling and clustering the output of smaller models makes them competitive with larger models, up to a crossover point beyond which a larger models becomes more inference-compute efficient. Overall, our experimental results demonstrate that optimizing the training and inference-time scaling properties of motion forecasting and planning models is a key lever for improving their performance to address a wide variety of driving scenarios. Finally, we briefly study the utility of training on general logged driving data of other agents to improve the performance of the ego-agent, an important research area to address the scarcity of robotics data for large capacity models training. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:54:23 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Baniodeh",
"Mustafa",
""
],
[
"Goel",
"Kratarth",
""
],
[
"Ettinger",
"Scott",
""
],
[
"Fuertes",
"Carlos",
""
],
[
"Seff",
"Ari",
""
],
[
"Shen",
"Tim",
""
],
[
"Gulino",
"Cole",
""
],
[
"Yang",
"Chenjie",
""
],
[
"Jerfel",
"Ghassen",
""
],
[
"Choe",
"Dokook",
""
],
[
"Wang",
"Rui",
""
],
[
"Kallem",
"Vinutha",
""
],
[
"Casas",
"Sergio",
""
],
[
"Al-Rfou",
"Rami",
""
],
[
"Sapp",
"Benjamin",
""
],
[
"Anguelov",
"Dragomir",
""
]
] |
2506.08230 | Xinchun Wu | Xinchun Wu, Xuezhi Bian, Jonathan Rawlinson, Robert G. Littlejohn, and Joseph E. Subotnik | Recovering Exact Vibrational Energies Within a Phase Space Electronic Structure Framework | null | null | null | null | physics.chem-ph quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In recent years, there has been a push to go beyond Born-Oppenheimer theory and build electronic states from a phase space perspective, i.e. parameterize electronic states by both nuclear position(R) and nuclear momentum(P). Previous empirical studies have demonstrated that such approaches can yield improved single-surface observables, including vibrational energies, electronic momenta, and vibrational circular dichroism spectra. That being said, unlike the case of BO theory, there is no unique phase space electronic Hamiltonian, nor any theory for using phase space eigenvectors (as opposed to BO eigenvectors) so as to recover exact quantum vibrational eigenvalues. As such, one might consider such phase space approaches ad hoc. To that end, here we show how to formally extract exact quantum energies from a coupled nuclear-electronic Hamiltonian using perturbation theory on top of a phase space electronic framework. Thus, while we cannot isolate an "optimal" phase space electronic Hamiltonian, this work does justify a phase space electronic structure approach by offering a rigorous framework for correcting the zeroth order phase space electronic states. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:57:33 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wu",
"Xinchun",
""
],
[
"Bian",
"Xuezhi",
""
],
[
"Rawlinson",
"Jonathan",
""
],
[
"Littlejohn",
"Robert G.",
""
],
[
"Subotnik",
"Joseph E.",
""
]
] |
2506.08231 | Aaron Cohen | Melissa Estevez, Nisha Singh, Lauren Dyson, Blythe Adamson, Qianyu Yuan, Megan W. Hildner, Erin Fidyk, Olive Mbah, Farhad Khan, Kathi Seidl-Rathkopf, Aaron B. Cohen | Ensuring Reliability of Curated EHR-Derived Data: The Validation of Accuracy for LLM/ML-Extracted Information and Data (VALID) Framework | 18 pages, 3 tables, 1 figure | null | null | null | cs.LG cs.AI cs.PF | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Large language models (LLMs) are increasingly used to extract clinical data from electronic health records (EHRs), offering significant improvements in scalability and efficiency for real-world data (RWD) curation in oncology. However, the adoption of LLMs introduces new challenges in ensuring the reliability, accuracy, and fairness of extracted data, which are essential for research, regulatory, and clinical applications. Existing quality assurance frameworks for RWD and artificial intelligence do not fully address the unique error modes and complexities associated with LLM-extracted data. In this paper, we propose a comprehensive framework for evaluating the quality of clinical data extracted by LLMs. The framework integrates variable-level performance benchmarking against expert human abstraction, automated verification checks for internal consistency and plausibility, and replication analyses comparing LLM-extracted data to human-abstracted datasets or external standards. This multidimensional approach enables the identification of variables most in need of improvement, systematic detection of latent errors, and confirmation of dataset fitness-for-purpose in real-world research. Additionally, the framework supports bias assessment by stratifying metrics across demographic subgroups. By providing a rigorous and transparent method for assessing LLM-extracted RWD, this framework advances industry standards and supports the trustworthy use of AI-powered evidence generation in oncology research and practice. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 20:59:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Estevez",
"Melissa",
""
],
[
"Singh",
"Nisha",
""
],
[
"Dyson",
"Lauren",
""
],
[
"Adamson",
"Blythe",
""
],
[
"Yuan",
"Qianyu",
""
],
[
"Hildner",
"Megan W.",
""
],
[
"Fidyk",
"Erin",
""
],
[
"Mbah",
"Olive",
""
],
[
"Khan",
"Farhad",
""
],
[
"Seidl-Rathkopf",
"Kathi",
""
],
[
"Cohen",
"Aaron B.",
""
]
] |
2506.08232 | Anna Taylor | Anna Ruth Taylor, Tommi T. Koskinen, Luca Argenti, Nicholas Lewis, Chenliang Huang, Anthony Arfaux, and Panayotis Lavvas | A Multi-Species Atmospheric Escape Model with Excited Hydrogen and Helium: Application to HD209458b | 31 Pages, 20 Figures, 3 Tables. Accepted to ApJ on June 9 2025 | null | null | null | astro-ph.EP astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Atmospheric escape shapes exoplanet evolution and star-planet interactions, with He I 10830 \AA\ absorption serving as a key tracer of mass loss in hot gas giants. However, transit depths vary significantly across observed systems for reasons that remain poorly understood. HD209458b, the archetypal hot-Jupiter, exhibits relatively weak He I 10830 \AA\ and H$\alpha$ absorption, which has been interpreted as evidence for a high H/He ratio (98/2), possibly due to diffusive separation. To investigate this possibility and other processes that control these transit depths, we reassess excitation and de-excitation rates for metastable helium and explore the impact of diffusion processes, stellar activity, and tidal forces on the upper atmosphere and transit depths using a model framework spanning the whole atmosphere. Our model reproduces the observed He I transit depth and H$\alpha$ upper limit, showing strong diffusive separation. We match the observations assuming a photoelectron efficiency of 20-40\%, depending on the composition of the atmosphere, corresponding to mass-loss rates of $1.9-3\times10^{10}$ g/s. We find that the He I 10830 \AA\ transit depth is sensitive to both stellar activity and diffusion processes, while H$\alpha$ is largely unaffected due to its strong dependence on Lyman-$\alpha$ excitation. These differences may help explain the system-to-system scatter seen in population-level studies of the He I line. While He I data alone may not tightly constrain mass-loss rates or temperatures, they do confirm atmospheric escape and help narrow the viable parameter space when interpreted with physically motivated models. Simultaneous observations of He I, H$\alpha$, and stellar activity indicators provide powerful constraints on upper atmosphere dynamics and composition, even in the absence of full transmission spectra. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:01:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Taylor",
"Anna Ruth",
""
],
[
"Koskinen",
"Tommi T.",
""
],
[
"Argenti",
"Luca",
""
],
[
"Lewis",
"Nicholas",
""
],
[
"Huang",
"Chenliang",
""
],
[
"Arfaux",
"Anthony",
""
],
[
"Lavvas",
"Panayotis",
""
]
] |
2506.08233 | Long Qian | Andr\'e Platzer, Long Qian | Approximate Axiomatization for Differentially-Defined Functions | null | null | null | null | cs.LO math.LO | http://creativecommons.org/licenses/by/4.0/ | This article establishes a complete approximate axiomatization for the real-closed field $\mathbb{R}$ expanded with all differentially-defined functions, including special functions such as $\sin(x), \cos(x), e^x, \dots$. Every true sentence is provable up to some numerical approximation, and the truth of such approximations converge under mild conditions. Such an axiomatization is a fragment of the axiomatization for differential dynamic logic, and is therefore a finite extension of the axiomatization of real-closed fields. Furthermore, the numerical approximations approximate formulas containing special function symbols by $\text{FOL}_{\mathbb{R}}$ formulas, improving upon earlier decidability results only concerning closed sentences. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:03:38 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Platzer",
"André",
""
],
[
"Qian",
"Long",
""
]
] |
2506.08234 | Yu-Ang Lee | Yu-Ang Lee, Guan-Ting Yi, Mei-Yi Liu, Jui-Chao Lu, Guan-Bo Yang, Yun-Nung Chen | Compound AI Systems Optimization: A Survey of Methods, Challenges, and Future Directions | 15 pages, 4 figures, 1 table | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in large language models (LLMs) and AI systems have led to a paradigm shift in the design and optimization of complex AI workflows. By integrating multiple components, compound AI systems have become increasingly adept at performing sophisticated tasks. However, as these systems grow in complexity, new challenges arise in optimizing not only individual components but also their interactions. While traditional optimization methods such as supervised fine-tuning (SFT) and reinforcement learning (RL) remain foundational, the rise of natural language feedback introduces promising new approaches, especially for optimizing non-differentiable systems. This paper provides a systematic review of recent progress in optimizing compound AI systems, encompassing both numerical and language-based techniques. We formalize the notion of compound AI system optimization, classify existing methods along several key dimensions, and highlight open research challenges and future directions in this rapidly evolving field. A list of surveyed papers is publicly available at https://github.com/MiuLab/AISysOpt-Survey. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:04:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Lee",
"Yu-Ang",
""
],
[
"Yi",
"Guan-Ting",
""
],
[
"Liu",
"Mei-Yi",
""
],
[
"Lu",
"Jui-Chao",
""
],
[
"Yang",
"Guan-Bo",
""
],
[
"Chen",
"Yun-Nung",
""
]
] |
2506.08235 | Shashidhar Reddy Javaji | Shashidhar Reddy Javaji, Yupeng Cao, Haohang Li, Yangyang Yu, Nikhil Muralidhar, Zining Zhu | Can AI Validate Science? Benchmarking LLMs for Accurate Scientific Claim $\rightarrow$ Evidence Reasoning | 21 pages, 6 figures, Under review | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) are increasingly being used for complex research tasks such as literature review, idea generation, and scientific paper analysis, yet their ability to truly understand and process the intricate relationships within complex research papers, such as the logical links between claims and supporting evidence remains largely unexplored. In this study, we present CLAIM-BENCH, a comprehensive benchmark for evaluating LLMs' capabilities in scientific claim-evidence extraction and validation, a task that reflects deeper comprehension of scientific argumentation. We systematically compare three approaches which are inspired by divide and conquer approaches, across six diverse LLMs, highlighting model-specific strengths and weaknesses in scientific comprehension. Through evaluation involving over 300 claim-evidence pairs across multiple research domains, we reveal significant limitations in LLMs' ability to process complex scientific content. Our results demonstrate that closed-source models like GPT-4 and Claude consistently outperform open-source counterparts in precision and recall across claim-evidence identification tasks. Furthermore, strategically designed three-pass and one-by-one prompting approaches significantly improve LLMs' abilities to accurately link dispersed evidence with claims, although this comes at increased computational cost. CLAIM-BENCH sets a new standard for evaluating scientific comprehension in LLMs, offering both a diagnostic tool and a path forward for building systems capable of deeper, more reliable reasoning across full-length papers. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:04:39 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Javaji",
"Shashidhar Reddy",
""
],
[
"Cao",
"Yupeng",
""
],
[
"Li",
"Haohang",
""
],
[
"Yu",
"Yangyang",
""
],
[
"Muralidhar",
"Nikhil",
""
],
[
"Zhu",
"Zining",
""
]
] |
2506.08236 | Adam Brandenburger | Adam Brandenburger and Pierfrancesco La Mura | Delayed Arrow-of-Time Detection in Signed Laplacian Dynamics | 7 pages | null | null | null | quant-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study how rapidly the direction of time becomes operationally detectable from mesoscopic data when state-weights may be positive or negative. In contrast with classical Markov processes -- where forward evolution is instantly distinguishable from its reverse -- signed dynamics can render the arrow of time undetectable during an initial interval. Assume the generator of the signed dynamics is a symmetric signed Laplacian (with a single linear invariant) and a phase-space Second Law holds in the form of non-decreasing R\'enyi-$2$ entropy. Drawing on recent results on eventual exponential positivity of signed Laplacians (Chen et al., 2021), we define a test that correctly identifies the direction of thermodynamic time if conducted over a time interval of length at least $\tau$. We go on to prove that the test cannot deliver an incorrect conclusion if conducted over a shorter interval. Dropping symmetry, we exhibit a no-signaling example where $\tau = +\infty$, so that the arrow of time remains permanently undetectable under our test. We comment briefly on quantum realizability of our framework. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:08:59 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Brandenburger",
"Adam",
""
],
[
"La Mura",
"Pierfrancesco",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.