id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.23866 | Jan Tobias Muehlberg | Jason Kayembe, Iness Ben Guirat, Jan Tobias M\"uhlberg | Exploring Privacy and Security as Drivers for Environmental Sustainability in Cloud-Based Office Solutions | Post-proceedings paper presented at LOCO '24: 1st International Workshop on Low Carbon Computing, 2024-12-03, in Glasgow, UK/Online. This paper is based on https://doi.org/10.48550/arXiv.2411.16340 but presents novel insights and additional technical content; this paper has been reviewed again by the LOCO PC | null | null | null | cs.CR cs.CY cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we explore the intersection of privacy, security, and environmental sustainability in cloud-based office solutions, focusing on quantifying user- and network-side energy use and associated carbon emissions. We hypothesise that privacy-focused services are typically more energy-efficient than those funded through data collection and advertising. To evaluate this, we propose a framework that systematically measures environmental costs based on energy usage and network data traffic during well-defined, automated usage scenarios. To test our hypothesis, we first analyse how underlying architectures and business models, such as monetisation through personalised advertising, contribute to the environmental footprint of these services. We then explore existing methodologies and tools for software environmental impact assessment. We apply our framework to three mainstream email services selected to reflect different privacy policies, from ad-supported tracking-intensive models to privacy-focused designs: Microsoft Outlook, Google Mail (Gmail), and Proton Mail. We extend this comparison to a self-hosted email solution, evaluated with and without end-to-end encryption. We show that the self-hosted solution, even with 14% of device energy and 15% of emissions overheads from PGP encryption, remains the most energy-efficient, saving up to 33% of emissions per session compared to Gmail. Among commercial providers, Proton Mail is the most efficient, saving up to 0.1 gCO2 e per session compared to Outlook, whose emissions can be further reduced by 2% through ad-blocking. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 13:58:22 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 13:40:13 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kayembe",
"Jason",
""
],
[
"Guirat",
"Iness Ben",
""
],
[
"Mühlberg",
"Jan Tobias",
""
]
] |
2506.23882 | Pei Gao | Pei Gao and Qiyu Yang | An improved upper bound for the distribution of iterated Euler totient functions | null | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | Let $\phi(n)$ be the Euler totient function and $\phi_k(n)$ its $k$-fold iterate. In this note, we improve the upper bound for the number of positive $n\leqslant x$ such that $\phi_{k+1}(n)\geqslant cn$. Comparing with the upper bound which was obtained from Pollack's asymptotic formula of the summation of $\phi_{k+1}(n)$ for $n\leqslant x$, we have successfully increased the denominator exponent of the main term of the upper bound from $k$ to $k+1$. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 14:11:37 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 08:15:29 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gao",
"Pei",
""
],
[
"Yang",
"Qiyu",
""
]
] |
2506.23943 | Johannes Zink | Emilio Di Giacomo and Walter Didimo and Henry F\"orster and Torsten Ueckerdt and Johannes Zink | Linear Layouts of Graphs with Priority Queues | Appears in Proc. 19th Algorithms and Data Structures Symposium (WADS 2025) | null | null | null | cs.DM cs.CG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A linear layout of a graph consists of a linear ordering of its vertices and a partition of its edges into pages such that the edges assigned to the same page obey some constraint. The two most prominent and widely studied types of linear layouts are stack and queue layouts, in which any two edges assigned to the same page are forbidden to cross and nest, respectively. The names of these two layouts derive from the fact that, when parsing the graph according to the linear vertex ordering, the edges in a single page can be stored using a single stack or queue, respectively. Recently, the concepts of stack and queue layouts have been extended by using a double-ended queue or a restricted-input queue for storing the edges of a page. We extend this line of study to edge-weighted graphs by introducing priority queue layouts, that is, the edges on each page are stored in a priority queue whose keys are the edge weights. First, we show that there are edge-weighted graphs that require a linear number of priority queues. Second, we characterize the graphs that admit a priority queue layout with a single queue, regardless of the edge-weight function, and we provide an efficient recognition algorithm. Third, we show that the number of priority queues required independently of the edge-weight function is bounded by the pathwidth of the graph, but can be arbitrarily large already for graphs of treewidth two. Finally, we prove that determining the minimum number of priority queues is NP-complete if the linear ordering of the vertices is fixed. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 15:09:14 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 09:49:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Di Giacomo",
"Emilio",
""
],
[
"Didimo",
"Walter",
""
],
[
"Förster",
"Henry",
""
],
[
"Ueckerdt",
"Torsten",
""
],
[
"Zink",
"Johannes",
""
]
] |
2506.24031 | Grant Moles | Grant Moles | Multiplicative Relationships of Subrings and their Applications to Factorization | 25 pages, 0 figures | null | null | null | math.AC | http://creativecommons.org/licenses/by/4.0/ | When studying the properties of a ring $R$, it is often useful to compare $R$ to other rings whose properties are already known. In this paper, we define three ways in which a subring $R$ might be compared to a larger ring $T$: being associated, being ideal-preserving, or being locally associated. We then explore how these properties of a subring might be leveraged to give information about $R$, including applications to the field of factorization. Of particular interest is the result that an order in a number field is associated if and only if it is both ideal-preserving and locally associated. We conclude with a discussion of how these properties are realized in the case of orders in a number field and how such orders might be found. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 16:36:18 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 14:50:37 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Moles",
"Grant",
""
]
] |
2507.00109 | Ben Kain | Ben Kain | Dynamical evolution and stability of quantum corrected Schwarzschild black holes in semiclassical gravity | 12 pages, 8 figures | Phys. Rev. D 111, 124057 (2025) | 10.1103/ms6s-92tx | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Schwarzschild solution describes a classical static black hole in general relativity. When general relativity is extended by including semiclassical corrections in the form of a renormalized energy-momentum tensor, the horizon of the Schwarzschild black hole disappears and is replaced by a wormhole. We study the stability of this quantum corrected static Schwarzschild solution in semiclassical gravity by using it as the initial data of a dynamical evolution. We find that the quantum corrected solution is unstable and that the wormhole can expand or collapse when perturbed. In vacuum, the wormhole expands, but in the presence of even a small amount of classical matter, the wormhole collapses, forming a horizon and evolving to an evaporating black hole. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 18:00:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kain",
"Ben",
""
]
] |
2507.00182 | Jesus Ivan Ruiz Martinez | J. I. Ruiz-Martinez, A. Mendez-Vazquez, E. Rodriguez-Tello | Graph-Based Deep Learning for Component Segmentation of Maize Plants | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In precision agriculture, one of the most important tasks when exploring crop production is identifying individual plant components. There are several attempts to accomplish this task by the use of traditional 2D imaging, 3D reconstructions, and Convolutional Neural Networks (CNN). However, they have several drawbacks when processing 3D data and identifying individual plant components. Therefore, in this work, we propose a novel Deep Learning architecture to detect components of individual plants on Light Detection and Ranging (LiDAR) 3D Point Cloud (PC) data sets. This architecture is based on the concept of Graph Neural Networks (GNN), and feature enhancing with Principal Component Analysis (PCA). For this, each point is taken as a vertex and by the use of a K-Nearest Neighbors (KNN) layer, the edges are established, thus representing the 3D PC data set. Subsequently, Edge-Conv layers are used to further increase the features of each point. Finally, Graph Attention Networks (GAT) are applied to classify visible phenotypic components of the plant, such as the leaf, stem, and soil. This study demonstrates that our graph-based deep learning approach enhances segmentation accuracy for identifying individual plant components, achieving percentages above 80% in the IoU average, thus outperforming other existing models based on point clouds. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 18:44:27 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 17:15:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ruiz-Martinez",
"J. I.",
""
],
[
"Mendez-Vazquez",
"A.",
""
],
[
"Rodriguez-Tello",
"E.",
""
]
] |
2507.00219 | Yahya Alnashri | Hasan Alzubaidi and Yahya Alnashri | Error Estimates for Non Conforming Discretisation of Time-dependent Convection-Diffusion-Reaction Model | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We use a generic framework, namely the gradient discretisation method (GDM), to propose a unified numerical analysis for general time-dependent convection-diffusion-reaction models. We establish novel results for convergence rates of numerical approximations of such models under reasonable assumptions on exact solutions, and prove the existence and uniqueness of the approximate solution for suitably small time steps. The main interest of our results lies in covering several approximation methods and various applications of the considered model such as the generalised Burgers-Fisher (GBF) and the generalised Burgers-Huxley (GBH) models. Numerical tests based on the hybrid mimetic mixed (HMM) method for the GBF model are performed on various types of general meshes to examine the accuracy of the proposed gradient scheme. The results confirm our theoretical rates of convergence, even on mesh with extreme distortions. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 19:42:30 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 02:23:29 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Alzubaidi",
"Hasan",
""
],
[
"Alnashri",
"Yahya",
""
]
] |
2507.00235 | Bubai Manna | Bubai Manna | Minimum Selective Subset on Some Graph Classes | This work has been accepted in CCCG 2025 | null | null | null | cs.CG cs.CC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a connected simple graph G = (V(G),E(G)), each vertex is assigned a color from the set of colors C={1, 2,..., c}. The set of vertices V(G) is partitioned as V_1, V_2, ... ,V_c, where all vertices in V_j share the same color j. A subset S of V(G) is called Selective Subset if, for every vertex v in V(G), and if v is in V_j, at least one of its nearest neighbors in (S union (V(G)\ V_j)) has the same color as v. The Minimum Selective Subset (MSS) problem seeks to find a selective subset of minimum size. The problem was first introduced by Wilfong in 1991 for a set of points in the Euclidean plane, where two major problems, MCS (Minimum Consistent Subset) and MSS, were proposed.
In graph algorithms, the only known result is that the MSS problem is NP-complete, as shown in 2018. Beyond this, no further progress has been made to date. In contrast, the MCS problem has been widely studied in various graph classes over the years. Therefore, in this work, we also extend the algorithmic study of MSS on various graph classes. We first present a log(n)-approximation algorithm for general graphs with n vertices and regardless of the number of colors. We also show that the problem remains NP-complete in planar graphs when restricted to just two colors.. Finally, we provide polynomial-time algorithms for computing optimal solutions in trees and unit interval graphs for any number of colors. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 20:04:56 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 06:03:49 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Manna",
"Bubai",
""
]
] |
2507.00241 | Shahab Shahidi | Shahab Shahidi and Shiva Kayedi | Cosmological implications of hybrid metric-Palatini $f(\mathcal{R},\mathcal{R}_{\mu\nu}\mathcal{R}^{\mu\nu})$ gravity | 13 pages, 12 figures; accepted for publication in EPJC | null | null | null | gr-qc | http://creativecommons.org/publicdomain/zero/1.0/ | The hybrid metric-Palatini gravity with Lagrangian density $L =R+f(\mathcal{R},\mathcal{R}_{\mu\nu}\mathcal{R}^{\mu\nu})$ is considered, where $R$ is the metric Ricci scalar and $\mathcal{R}_{\mu\nu}$ is the Palatini Ricci tensor. Contrary to the standard hybrid metric-Palatini theory, because of the term $\mathcal{R}_{\mu\nu}\mathcal{R}^{\mu\nu}$ in the action, the model can not be analytically transformed to a scalar-tensor theory. However, on top of a maximally symmetric space-times like the FRW universe, there is a way to solve for a metric compatible connection which we will follow in this paper. The cosmological implications of the resulting model will then be fully considered. The best fit values of the model and cosmological parameters will be obtained by confronting the model with the recent observational data on the Hubble parameter. We will see that the observational data can be explained very good in this model, but, significant deviations from the standard $\Lambda$CDM model could be seen in derivatives of the Hubble parameter $q$, $j$ and $s$. We will perform a statefinder analysis for the model and show that its behavior differs from that of the $\Lambda$CDM model. Also, we will consider the recently proposed $Om$ diagnostics to categorize the dark energy type of the model and obtain the $\omega$-varying alternative of the model that mimics the Hubble flow. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 20:16:09 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 13:35:20 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Shahidi",
"Shahab",
""
],
[
"Kayedi",
"Shiva",
""
]
] |
2507.00259 | Amr Abourayya | Amr Abourayya, Jens Kleesiek, Bharat Rao, Michael Kamp | Who Should I Listen To? Adaptive Collaboration in Personalized Federated Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data heterogeneity is a central challenge in federated learning, and personalized federated learning (PFL) aims to address it by tailoring models to each client's distribution. Yet many PFL methods fail to outperform local or centralized baselines, suggesting a mismatch between the collaboration they enforce and the structure of the data. We propose an approach based on adaptive collaboration, where clients decide adaptively not only how much to rely on others, but also whom to trust at the level of individual examples. We instantiate this principle in FEDMOSAIC, a federated co-training method in which clients exchange predictions over a shared unlabeled dataset. This enables fine-grained trust decisions that are difficult to achieve with parameter sharing alone. Each client adjusts its loss weighting based on the agreement between private and public data, and contributes to global pseudo-labels in proportion to its estimated per-example confidence. Empirically, FEDMOSAIC improves upon state-of-the-art PFL methods across diverse non-IID settings, and we provide convergence guarantees under standard assumptions. Our results demonstrate the potential of data-aware collaboration for robust and effective personalization. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 20:53:01 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 10:26:16 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Abourayya",
"Amr",
""
],
[
"Kleesiek",
"Jens",
""
],
[
"Rao",
"Bharat",
""
],
[
"Kamp",
"Michael",
""
]
] |
2507.00292 | Pietro Gori | Ali Mammadov, Lo\"ic Le Folgoc, Guillaume Hocquet, Pietro Gori | Reducing Variability of Multiple Instance Learning Methods for Digital Pathology | MICCAI 2025 - This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this article is published in LNCS, Springer | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Digital pathology has revolutionized the field by enabling the digitization of tissue samples into whole slide images (WSIs). However, the high resolution and large size of WSIs present significant challenges when it comes to applying Deep Learning models. As a solution, WSIs are often divided into smaller patches with a global label (\textit{i.e., diagnostic}) per slide, instead of a (too) costly pixel-wise annotation. By treating each slide as a bag of patches, Multiple Instance Learning (MIL) methods have emerged as a suitable solution for WSI classification. A major drawback of MIL methods is their high variability in performance across different runs, which can reach up to 10-15 AUC points on the test set, making it difficult to compare different MIL methods reliably. This variability mainly comes from three factors: i) weight initialization, ii) batch (shuffling) ordering, iii) and learning rate. To address that, we introduce a Multi-Fidelity, Model Fusion strategy for MIL methods. We first train multiple models for a few epochs and average the most stable and promising ones based on validation scores. This approach can be applied to any existing MIL model to reduce performance variability. It also simplifies hyperparameter tuning and improves reproducibility while maintaining computational efficiency. We extensively validate our approach on WSI classification tasks using 2 different datasets, 3 initialization strategies and 5 MIL methods, for a total of more than 2000 experiments. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 22:10:24 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 12:37:04 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mammadov",
"Ali",
""
],
[
"Folgoc",
"Loïc Le",
""
],
[
"Hocquet",
"Guillaume",
""
],
[
"Gori",
"Pietro",
""
]
] |
2507.00307 | Joseph Fry | Joseph Fry | Robust Inference when Nuisance Parameters may be Partially Identified with Applications to Synthetic Controls | null | null | null | null | econ.EM | http://creativecommons.org/licenses/by/4.0/ | When conducting inference for the average treatment effect on the treated with a Synthetic Control Estimator, the vector of control weights is a nuisance parameter which is often constrained, high-dimensional, and may be only partially identified even when the average treatment effect on the treated is point-identified. All three of these features of a nuisance parameter can lead to failure of asymptotic normality for the estimate of the parameter of interest when using standard methods. I provide a new method yielding asymptotic normality for an estimate of the parameter of interest, even when all three of these complications are present. This is accomplished by first estimating the nuisance parameter using a regularization penalty to achieve a form of identification, and then estimating the parameter of interest using moment conditions that have been orthogonalized with respect to the nuisance parameter. I present high-level sufficient conditions for the estimator and verify these conditions in an example involving Synthetic Controls. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 22:48:00 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 17:42:34 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Fry",
"Joseph",
""
]
] |
2507.00316 | Siyou Li | Siyou Li, Pengyao Qin, Huanan Wu, Dong Nie, Arun J. Thirunavukarasu, Juntao Yu, Le Zhang | ${\mu}^2$Tokenizer: Differentiable Multi-Scale Multi-Modal Tokenizer for Radiology Report Generation | Accepted by MICCAI 2025 | null | null | null | cs.LG cs.CL eess.IV | http://creativecommons.org/licenses/by/4.0/ | Automated radiology report generation (RRG) aims to produce detailed textual reports from clinical imaging, such as computed tomography (CT) scans, to improve the accuracy and efficiency of diagnosis and provision of management advice. RRG is complicated by two key challenges: (1) inherent complexity in extracting relevant information from imaging data under resource constraints, and (2) difficulty in objectively evaluating discrepancies between model-generated and expert-written reports. To address these challenges, we propose $\mu^2$LLM, a $\underline{\textbf{mu}}$ltiscale $\underline{\textbf{mu}}$ltimodal large language models for RRG tasks. The novel ${\mu}^2$Tokenizer, as an intermediate layer, integrates multi-modal features from the multiscale visual tokenizer and the text tokenizer, then enhances report generation quality through direct preference optimization (DPO), guided by GREEN-RedLlama. Experimental results on four large CT image-report medical datasets demonstrate that our method outperforms existing approaches, highlighting the potential of our fine-tuned $\mu^2$LLMs on limited data for RRG tasks. At the same time, for prompt engineering, we introduce a five-stage, LLM-driven pipeline that converts routine CT reports into paired visual-question-answer triples and citation-linked reasoning narratives, creating a scalable, high-quality supervisory corpus for explainable multimodal radiology LLM. All code, datasets, and models will be publicly available in our official repository. https://github.com/Siyou-Li/u2Tokenizer | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 23:14:49 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 01:08:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Li",
"Siyou",
""
],
[
"Qin",
"Pengyao",
""
],
[
"Wu",
"Huanan",
""
],
[
"Nie",
"Dong",
""
],
[
"Thirunavukarasu",
"Arun J.",
""
],
[
"Yu",
"Juntao",
""
],
[
"Zhang",
"Le",
""
]
] |
2507.00348 | Numan Guldemir | Numan Halit Guldemir, Oluwafemi Olukoya, Jes\'us Mart\'inez-del-Rinc\'on | Addressing malware family concept drift with triplet autoencoder | null | Proceedings of the Eighteenth International Conference on Emerging Security Information, Systems and Technologies (SECURWARE 2024) | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Machine learning is increasingly vital in cybersecurity, especially in malware detection. However, concept drift, where the characteristics of malware change over time, poses a challenge for maintaining the efficacy of these detection systems. Concept drift can occur in two forms: the emergence of entirely new malware families and the evolution of existing ones. This paper proposes an innovative method to address the former, focusing on effectively identifying new malware families. Our approach leverages a supervised autoencoder combined with triplet loss to differentiate between known and new malware families. We create clear and robust clusters that enhance the accuracy and resilience of malware family classification by utilizing this metric learning technique and the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. The effectiveness of our method is validated using an Android malware dataset and a Windows portable executable (PE) malware dataset, showcasing its capability to sustain model performance within the dynamic landscape of emerging malware threats. Our results demonstrate a significant improvement in detecting new malware families, offering a reliable solution for ongoing cybersecurity challenges. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 00:55:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Guldemir",
"Numan Halit",
""
],
[
"Olukoya",
"Oluwafemi",
""
],
[
"Martínez-del-Rincón",
"Jesús",
""
]
] |
2507.00487 | Jianghao Lin | Jianghao Lin, Xinyuan Wang, Xinyi Dai, Menghui Zhu, Bo Chen, Ruiming Tang, Yong Yu, Weinan Zhang | MassTool: A Multi-Task Search-Based Tool Retrieval Framework for Large Language Models | null | null | null | null | cs.IR cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tool retrieval is a critical component in enabling large language models (LLMs) to interact effectively with external tools. It aims to precisely filter the massive tools into a small set of candidates for the downstream tool-augmented LLMs. However, most existing approaches primarily focus on optimizing tool representations, often neglecting the importance of precise query comprehension. To address this gap, we introduce MassTool, a multi-task search-based framework designed to enhance both query representation and tool retrieval accuracy. MassTool employs a two-tower architecture: a tool usage detection tower that predicts the need for function calls, and a tool retrieval tower that leverages a query-centric graph convolution network (QC-GCN) for effective query-tool matching. It also incorporates search-based user intent modeling (SUIM) to handle diverse and out-of-distribution queries, alongside an adaptive knowledge transfer (AdaKT) module for efficient multi-task learning. By jointly optimizing tool usage detection loss, list-wise retrieval loss, and contrastive regularization loss, MassTool establishes a robust dual-step sequential decision-making pipeline for precise query understanding. Extensive experiments demonstrate its effectiveness in improving retrieval accuracy. Our code is available at https://github.com/wxydada/MassTool. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:02:26 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 04:35:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lin",
"Jianghao",
""
],
[
"Wang",
"Xinyuan",
""
],
[
"Dai",
"Xinyi",
""
],
[
"Zhu",
"Menghui",
""
],
[
"Chen",
"Bo",
""
],
[
"Tang",
"Ruiming",
""
],
[
"Yu",
"Yong",
""
],
[
"Zhang",
"Weinan",
""
]
] |
2507.00490 | Zijian Chen | Zijian Chen, Yuan Tian, Yuze Sun, Wei Sun, Zicheng Zhang, Weisi Lin, Guangtao Zhai, Wenjun Zhang | Just Noticeable Difference for Large Multimodal Models | 19 pages, 19 figures | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Just noticeable difference (JND), the minimum change that the human visual system (HVS) can perceive, has been studied for decades. Although recent work has extended this line of research into machine vision, there has been a scarcity of studies systematically exploring its perceptual boundaries across multiple tasks and stimulus types, particularly in the current era of rapidly advancing large multimodal models (LMMs), where studying the multifaceted capabilities of models has become a mainstream focus. Moreover, the perceptual defects of LMMs are not investigated thoroughly, resulting in potential security issues and suboptimal response efficiency. In this paper, we take an initial attempt and demonstrate that there exist significant visual blind spots in current LMMs. To systemically quantify this characteristic, we propose a new concept, {\bf LMM-JND}, together with its determination pipeline. Targeting uncovering the behavior commonalities in HVS-aligned visual perception tasks, we delve into several LMM families and construct a large-scale dataset, named VPA-JND, which contains 21.5k reference images with over 489k stimuli across 12 distortion types, to facilitate LMM-JND studies. VPA-JND exposes areas where state-of-the-art LMMs, including GPT-4o and the InternVL2.5 series, struggle with basic comparison queries and fall significantly short of human-level visual performance. We further explore the effects of vision and language backbones and find a notable correlation between their design philosophy that may instruct the future refinement of LMMs for their visual acuity. Together, our research underscores the significance of LMM-JND as a unique perspective for studying LMMs, and predictable LMM-JND is crucial for security concerns. This work will be available at https://github.com/zijianchen98/LMM-JND. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:06:32 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 13:58:48 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Chen",
"Zijian",
""
],
[
"Tian",
"Yuan",
""
],
[
"Sun",
"Yuze",
""
],
[
"Sun",
"Wei",
""
],
[
"Zhang",
"Zicheng",
""
],
[
"Lin",
"Weisi",
""
],
[
"Zhai",
"Guangtao",
""
],
[
"Zhang",
"Wenjun",
""
]
] |
2507.00502 | Jianchao Zhao | JianChao Zhao, Chenhao Ding, Songlin Dong, Yuhang He, Yihong Gong | ExPaMoE: An Expandable Parallel Mixture of Experts for Continual Test-Time Adaptation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Continual Test-Time Adaptation (CTTA) aims to enable models to adapt on-the-fly to a stream of unlabeled data under evolving distribution shifts. However, existing CTTA methods typically rely on shared model parameters across all domains, making them vulnerable to feature entanglement and catastrophic forgetting in the presence of large or non-stationary domain shifts. To address this limitation, we propose ExPaMoE, a novel framework based on an Expandable Parallel Mixture-of-Experts architecture. ExPaMoE decouples domain-general and domain-specific knowledge via a dual-branch expert design with token-guided feature separation, and dynamically expands its expert pool based on a Spectral-Aware Online Domain Discriminator (SODD) that detects distribution changes in real-time using frequency-domain cues. Extensive experiments demonstrate the superiority of ExPaMoE across diverse CTTA scenarios. We evaluate our method on standard benchmarks including CIFAR-10C, CIFAR-100C, ImageNet-C, and Cityscapes-to-ACDC for semantic segmentation. Additionally, we introduce ImageNet++, a large-scale and realistic CTTA benchmark built from multiple ImageNet-derived datasets, to better reflect long-term adaptation under complex domain evolution. ExPaMoE consistently outperforms prior arts, showing strong robustness, scalability, and resistance to forgetting. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:17:33 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 11:38:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhao",
"JianChao",
""
],
[
"Ding",
"Chenhao",
""
],
[
"Dong",
"Songlin",
""
],
[
"He",
"Yuhang",
""
],
[
"Gong",
"Yihong",
""
]
] |
2507.00521 | Mugeng Liu | Mugeng Liu, Siqi Zhong, Qi Yang, Yudong Han, Xuanzhe Liu, and Yun Ma | WebANNS: Fast and Efficient Approximate Nearest Neighbor Search in Web Browsers | SIGIR 2025 | null | 10.1145/3726302.3730115 | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Approximate nearest neighbor search (ANNS) has become vital to modern AI infrastructure, particularly in retrieval-augmented generation (RAG) applications. Numerous in-browser ANNS engines have emerged to seamlessly integrate with popular LLM-based web applications, while addressing privacy protection and challenges of heterogeneous device deployments. However, web browsers present unique challenges for ANNS, including computational limitations, external storage access issues, and memory utilization constraints, which state-of-the-art (SOTA) solutions fail to address comprehensively. We propose WebANNS, a novel ANNS engine specifically designed for web browsers. WebANNS leverages WebAssembly to overcome computational bottlenecks, designs a lazy loading strategy to optimize data retrieval from external storage, and applies a heuristic approach to reduce memory usage. Experiments show that WebANNS is fast and memory efficient, achieving up to $743.8\times$ improvement in 99th percentile query latency over the SOTA engine, while reducing memory usage by up to 39\%. Note that WebANNS decreases query time from 10 seconds to the 10-millisecond range in browsers, making in-browser ANNS practical with user-acceptable latency. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:37:18 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 02:20:54 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Liu",
"Mugeng",
""
],
[
"Zhong",
"Siqi",
""
],
[
"Yang",
"Qi",
""
],
[
"Han",
"Yudong",
""
],
[
"Liu",
"Xuanzhe",
""
],
[
"Ma",
"Yun",
""
]
] |
2507.00539 | Paul Jeanney | Paul Jeanney, Ashton Hetherington, Shady E. Ahmed, David Lanceta, Susana Saiz, Jos\'e Miguel Perez and Soledad Le Clainche | Ensemble Kalman Filter for Data Assimilation coupled with low-resolution computations techniques applied in Fluid Dynamics | article, 49 pages, 29 figures, 4 tables | null | null | null | cs.CE physics.flu-dyn | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper presents an innovative Reduced-Order Model (ROM) for merging experimental and simulation data using Data Assimilation (DA) to estimate the "True" state of a fluid dynamics system, leading to more accurate predictions. Our methodology introduces a novel approach implementing the Ensemble Kalman Filter (EnKF) within a reduced-dimensional framework, grounded in a robust theoretical foundation and applied to fluid dynamics. To address the substantial computational demands of DA, the proposed ROM employs low-resolution (LR) techniques to drastically reduce computational costs. This approach involves downsampling datasets for DA computations, followed by an advanced reconstruction technique based on low-cost Singular Value Decomposition (lcSVD). The lcSVD method, a key innovation in this paper, has never been applied to DA before and offers a highly efficient way to enhance resolution with minimal computational resources. Our results demonstrate significant reductions in both computation time and RAM usage through the LR techniques without compromising the accuracy of the estimations. For instance, in a turbulent test case, the LR approach with a compression rate of 15.9 can achieve a speed-up of 13.7 and a RAM compression of 90.9% while maintaining a low Relative Root Mean Square Error (RRMSE) of 2.6%, compared to 0.8% in the high-resolution (HR) reference. Furthermore, we highlight the effectiveness of the EnKF in estimating and predicting the state of fluid flow systems based on limited observations and low-fidelity numerical data. This paper highlights the potential of the proposed DA method in fluid dynamics applications, particularly for improving computational efficiency in CFD and related fields. Its ability to balance accuracy with low computational and memory costs makes it suitable for large-scale and real-time applications, such as environmental monitoring or aerospace. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:59:16 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 01:54:36 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Jeanney",
"Paul",
""
],
[
"Hetherington",
"Ashton",
""
],
[
"Ahmed",
"Shady E.",
""
],
[
"Lanceta",
"David",
""
],
[
"Saiz",
"Susana",
""
],
[
"Perez",
"José Miguel",
""
],
[
"Clainche",
"Soledad Le",
""
]
] |
2507.00587 | Immo Christopher Reis | Immo C. Reis, Joanna E. Sobczyk, Sonia Bacca | Optimized binning for response function reconstruction via Chebyshev expansions | 9 pages, 4 figures | null | null | MITP-25-050 | nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an optimized histogram binning strategy to reconstruct nuclear response functions via the Chebyshev expansion bound-state method. Our approach employs a stochastic regularization of the density of states to define adaptive, equal-area bins. Using the deuteron solved in a harmonic-oscillator basis with a chiral interaction, we benchmark on dipole and longitudinal responses, obtaining excellent agreement with exact theory and experiment. This general framework readily extends to other many-body systems and opens the door to new ab initio calculations of lepton-nucleus cross sections in medium-mass nuclei. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 09:08:12 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Reis",
"Immo C.",
""
],
[
"Sobczyk",
"Joanna E.",
""
],
[
"Bacca",
"Sonia",
""
]
] |
2507.00601 | Ruotong Wang | Shuangquan Lyu, Yingnan Deng, Guiran Liu, Zhen Qi, Ruotong Wang | Transferable Modeling Strategies for Low-Resource LLM Tasks: A Prompt and Alignment-Based Approach | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses the limited transfer and adaptation capabilities of large language models in low-resource language scenarios. It proposes a unified framework that combines a knowledge transfer module with parameter-efficient fine-tuning strategies. The method introduces knowledge alignment loss and soft prompt tuning to guide the model in effectively absorbing the structural features of target languages or tasks under minimal annotation. This enhances both generalization performance and training stability. The framework includes lightweight adaptation modules to reduce computational costs. During training, it integrates freezing strategies and prompt injection to preserve the model's original knowledge while enabling quick adaptation to new tasks. The study also conducts stability analysis experiments and synthetic pseudo-data transfer experiments to systematically evaluate the method's applicability and robustness across different low-resource tasks. Experimental results show that compared with existing multilingual pre-trained models and mainstream transfer methods, the proposed approach achieves higher performance and stability on cross-lingual tasks such as MLQA, XQuAD, and PAWS-X. It demonstrates particularly strong advantages under extremely data-scarce conditions. The proposed method offers strong generality and scalability. It enhances task-specific adaptability while preserving the general capabilities of large language models. This makes it well-suited for complex semantic modeling and multilingual processing tasks. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 09:34:49 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 06:39:55 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lyu",
"Shuangquan",
""
],
[
"Deng",
"Yingnan",
""
],
[
"Liu",
"Guiran",
""
],
[
"Qi",
"Zhen",
""
],
[
"Wang",
"Ruotong",
""
]
] |
2507.00621 | Lars Eric Hientzsch | Matteo Caggio, Donatella Donatelli, Lars Eric Hientzsch | Inviscid incompressible limit for capillary fluids with density dependent viscosity | null | null | null | null | math.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The asymptotic limit of the 2D and 3D Navier-Stokes-Korteweg system for barotropic capillary fluids with density dependent viscosities in the low Mach number and vanishing viscosity regime is established. In the relative energy framework, we prove the convergence of weak solutions of the Navier-Stokes-Korteweg system to the strong solution of the incompressible Euler system. The convergence is obtained through the use of suitable dispersive estimates for an acoustic system altered by the presence of the Korteweg tensor. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 10:05:07 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 12:15:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Caggio",
"Matteo",
""
],
[
"Donatelli",
"Donatella",
""
],
[
"Hientzsch",
"Lars Eric",
""
]
] |
2507.00629 | Enrico Maria Malatesta | Jean Barbier, Federica Gerace, Alessandro Ingrosso, Clarissa Lauditi, Enrico M. Malatesta, Gibbs Nwemadji, Rodrigo P\'erez Ortiz | Generalization performance of narrow one-hidden layer networks in the teacher-student setting | 34 pages, figures | null | null | null | cond-mat.dis-nn cs.LG math.PR math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding the generalization abilities of neural networks for simple input-output distributions is crucial to account for their learning performance on real datasets. The classical teacher-student setting, where a network is trained from data obtained thanks to a label-generating teacher model, serves as a perfect theoretical test bed. In this context, a complete theoretical account of the performance of fully connected one-hidden layer networks in the presence of generic activation functions is lacking. In this work, we develop such a general theory for narrow networks, i.e. networks with a large number of hidden units, yet much smaller than the input dimension. Using methods from statistical physics, we provide closed-form expressions for the typical performance of both finite temperature (Bayesian) and empirical risk minimization estimators, in terms of a small number of weight statistics. In doing so, we highlight the presence of a transition where hidden neurons specialize when the number of samples is sufficiently large and proportional to the number of parameters of the network. Our theory accurately predicts the generalization error of neural networks trained on regression or classification tasks with either noisy full-batch gradient descent (Langevin dynamics) or full-batch gradient descent. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 10:18:20 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 15:49:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Barbier",
"Jean",
""
],
[
"Gerace",
"Federica",
""
],
[
"Ingrosso",
"Alessandro",
""
],
[
"Lauditi",
"Clarissa",
""
],
[
"Malatesta",
"Enrico M.",
""
],
[
"Nwemadji",
"Gibbs",
""
],
[
"Ortiz",
"Rodrigo Pérez",
""
]
] |
2507.00680 | Brendah Nansereko | Brendah Nansereko, Marcel Wolbers, James Carpenter, Jonathan Bartlett | Bayesian analysis of the causal reference-based model for missing data in clinical trials | Inadvertently missed Jonathan Bartlett as an author | null | null | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The statistical analysis of clinical trials is often complicated by missing data. Patients sometimes experience intercurrent events (ICEs), which usually (although not always) lead to missing subsequent outcome measurements for such individuals. The reference-based imputation methods were proposed by Carpenter et al. (2013) and have been commonly adopted for handling missing data due to ICEs when estimating treatment policy strategy estimands. Conventionally, the variance for reference-based estimators was obtained using Rubin's rules. However, Rubin's rules variance estimator is biased compared to the repeated sampling variance of the point estimator, due to uncongeniality. Repeated sampling variance estimators were proposed as an alternative to variance estimation for reference-based estimators. However, these have the property that they decrease as the proportion of ICEs increases. White et al. (2019) introduced a causal model incorporating the concept of a 'maintained treatment effect' following the occurrence of ICEs and showed that this causal model included common reference-based estimators as special cases. Building on this framework, we propose introducing a prior distribution for the maintained effect parameter to account for uncertainty in this assumption. Our approach provides inference for reference-based estimators that explicitly reflects our uncertainty about how much treatment effects are maintained after the occurrence of ICEs. In trials where no or little post-ICE data are observed, our proposed Bayesian reference-based causal model approach can be used to estimate the treatment policy treatment effect, incorporating uncertainty about the reference-based assumption. We compare the frequentist properties of this approach with existing reference-based methods through simulations and by application to an antidepressant trial. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 11:24:51 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 07:55:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Nansereko",
"Brendah",
""
],
[
"Wolbers",
"Marcel",
""
],
[
"Carpenter",
"James",
""
],
[
"Bartlett",
"Jonathan",
""
]
] |
2507.00726 | Jongho Park | Dongyoon Hwang, Hojoon Lee, Jaegul Choo, Dongmin Park, Jongho Park | Can Large Language Models Develop Strategic Reasoning? Post-training Insights from Learning Chess | 27 pages | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | While reinforcement learning (RL) for large language models (LLMs) has shown promise in mathematical reasoning, strategic reasoning for LLMs using RL remains largely unexplored. We investigate whether LLMs can develop strategic reasoning capabilities through RL in chess. To this end, we leverage a chess-pretrained action-value network to provide dense reward on the LLM's output move quality, which can be seen as a form of knowledge distillation. Our experiments show that our distillation-based dense rewards often outperform sparse binary rewards. However, surprisingly, all models plateau far below expert levels. We provide SFT and RL ablations on chess reasoning training and find evidence that this limitation stems from a deficit in the pretrained models' internal understanding of chess--a deficit which RL alone may not be able to fully overcome. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 13:16:34 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 05:31:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Hwang",
"Dongyoon",
""
],
[
"Lee",
"Hojoon",
""
],
[
"Choo",
"Jaegul",
""
],
[
"Park",
"Dongmin",
""
],
[
"Park",
"Jongho",
""
]
] |
2507.00762 | Tom Maus | Tom Maus, Asma Atamna and Tobias Glasmachers | Leveraging Genetic Algorithms for Efficient Demonstration Generation in Real-World Reinforcement Learning Environments | This manuscript corresponds to the submitted version to LOD 2025. The final Version of Record will appear in the official conference proceedings | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Reinforcement Learning (RL) has demonstrated significant potential in certain real-world industrial applications, yet its broader deployment remains limited by inherent challenges such as sample inefficiency and unstable learning dynamics. This study investigates the utilization of Genetic Algorithms (GAs) as a mechanism for improving RL performance in an industrially inspired sorting environment. We propose a novel approach in which GA-generated expert demonstrations are used to enhance policy learning. These demonstrations are incorporated into a Deep Q-Network (DQN) replay buffer for experience-based learning and utilized as warm-start trajectories for Proximal Policy Optimization (PPO) agents to accelerate training convergence. Our experiments compare standard RL training with rule-based heuristics, brute-force optimization, and demonstration data, revealing that GA-derived demonstrations significantly improve RL performance. Notably, PPO agents initialized with GA-generated data achieved superior cumulative rewards, highlighting the potential of hybrid learning paradigms, where heuristic search methods complement data-driven RL. The utilized framework is publicly available and enables further research into adaptive RL strategies for real-world applications. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:04:17 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 14:06:48 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Maus",
"Tom",
""
],
[
"Atamna",
"Asma",
""
],
[
"Glasmachers",
"Tobias",
""
]
] |
2507.00808 | Kenichi Fujita | Hiroki Kanagawa, Kenichi Fujita, Aya Watanabe, Yusuke Ijima | Multi-interaction TTS toward professional recording reproduction | 7 pages,6 figures, Accepted to Speech Synthesis Workshop 2025 (SSW13) | null | null | null | cs.SD cs.CL eess.AS | http://creativecommons.org/licenses/by-sa/4.0/ | Voice directors often iteratively refine voice actors' performances by providing feedback to achieve the desired outcome. While this iterative feedback-based refinement process is important in actual recordings, it has been overlooked in text-to-speech synthesis (TTS). As a result, fine-grained style refinement after the initial synthesis is not possible, even though the synthesized speech often deviates from the user's intended style. To address this issue, we propose a TTS method with multi-step interaction that allows users to intuitively and rapidly refine synthesized speech. Our approach models the interaction between the TTS model and its user to emulate the relationship between voice actors and voice directors. Experiments show that the proposed model with its corresponding dataset enables iterative style refinements in accordance with users' directions, thus demonstrating its multi-interaction capability. Sample audios are available: https://ntt-hilab-gensp.github.io/ssw13multiinteractiontts/ | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:40:51 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 05:42:47 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kanagawa",
"Hiroki",
""
],
[
"Fujita",
"Kenichi",
""
],
[
"Watanabe",
"Aya",
""
],
[
"Ijima",
"Yusuke",
""
]
] |
2507.00840 | Jing Song | Jing Song, Yi-Yao Li, Eulogio Oset | Molecular states with bottom mesons and multistrange baryons systems | Typo: $g_i G_i$ in MeV | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate molecular states formed by bottom mesons and multistrange baryons from both octet and decuplet flavor representations, using the local hidden gauge approach combined with coupled-channel Bethe-Salpeter equations. Focusing on strangeness sectors \(S = -1\) to \(-4\), we predict several bound states in the \(S=-1, I=1/2,~3/2\) and \(S=-2, I=0\) sectors. No bound states are found in other isospin channels, where the interaction is repulsive, nor in higher strangeness sectors. The binding energies are analyzed under different values of the cutoff regularization parameters, providing estimates of theoretical uncertainties. This study provides concrete predictions to support future experimental investigations and improve understanding of heavy-flavor multistrange exotic hadrons. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 15:09:38 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 16:48:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Song",
"Jing",
""
],
[
"Li",
"Yi-Yao",
""
],
[
"Oset",
"Eulogio",
""
]
] |
2507.00846 | Rishal Aggarwal | Rishal Aggarwal, Jacky Chen, Nicholas M. Boffi, David Ryan Koes | BoltzNCE: Learning Likelihoods for Boltzmann Generation with Stochastic Interpolants and Noise Contrastive Estimation | 19 pages, 25 figures, submitted to NeurIPS 2025 | null | null | null | cs.LG physics.bio-ph | http://creativecommons.org/licenses/by/4.0/ | Efficient sampling from the Boltzmann distribution defined by an energy function is a key challenge in modeling physical systems such as molecules. Boltzmann Generators tackle this by leveraging Continuous Normalizing Flows that transform a simple prior into a distribution that can be reweighted to match the Boltzmann distribution using sample likelihoods. However, obtaining likelihoods requires computing costly Jacobians during integration, making it impractical for large molecular systems. To overcome this, we propose learning the likelihood of the generated distribution via an energy-based model trained with noise contrastive estimation and score matching. By using stochastic interpolants to anneal between the prior and generated distributions, we combine both the objective functions to efficiently learn the density function. On the alanine dipeptide system, we demonstrate that our method yields free energy profiles and energy distributions comparable to those obtained with exact likelihoods. Additionally, we show that free energy differences between metastable states can be estimated accurately with orders-of-magnitude speedup. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 15:18:28 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 01:16:47 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Aggarwal",
"Rishal",
""
],
[
"Chen",
"Jacky",
""
],
[
"Boffi",
"Nicholas M.",
""
],
[
"Koes",
"David Ryan",
""
]
] |
2507.00868 | Simon Rei{\ss} | Simon Rei{\ss}, Zdravko Marinov, Alexander Jaus, Constantin Seibold, M. Saquib Sarfraz, Erik Rodner, Rainer Stiefelhagen | Is Visual in-Context Learning for Compositional Medical Tasks within Reach? | Accepted to ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we explore the potential of visual in-context learning to enable a single model to handle multiple tasks and adapt to new tasks during test time without re-training. Unlike previous approaches, our focus is on training in-context learners to adapt to sequences of tasks, rather than individual tasks. Our goal is to solve complex tasks that involve multiple intermediate steps using a single model, allowing users to define entire vision pipelines flexibly at test time. To achieve this, we first examine the properties and limitations of visual in-context learning architectures, with a particular focus on the role of codebooks. We then introduce a novel method for training in-context learners using a synthetic compositional task generation engine. This engine bootstraps task sequences from arbitrary segmentation datasets, enabling the training of visual in-context learners for compositional tasks. Additionally, we investigate different masking-based training objectives to gather insights into how to train models better for solving complex, compositional tasks. Our exploration not only provides important insights especially for multi-modal medical task sequences but also highlights challenges that need to be addressed. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 15:32:23 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 09:16:31 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Reiß",
"Simon",
""
],
[
"Marinov",
"Zdravko",
""
],
[
"Jaus",
"Alexander",
""
],
[
"Seibold",
"Constantin",
""
],
[
"Sarfraz",
"M. Saquib",
""
],
[
"Rodner",
"Erik",
""
],
[
"Stiefelhagen",
"Rainer",
""
]
] |
2507.00930 | Krist\'of B\'erczi | Krist\'of B\'erczi, Lydia Mirabel Mendoza-Cadena, Jos\'e Soto | Inverse matroid optimization under subset constraints | 20 pages | null | null | null | cs.DS cs.DM | http://creativecommons.org/licenses/by/4.0/ | In the Inverse Matroid problem, we are given a matroid, a fixed basis $B$, and an initial weight function, and the goal is to minimally modify the weights -- measured by some function -- so that $B$ becomes a maximum-weight basis. The problem arises naturally in settings where one wishes to explain or enforce a given solution by minimally perturbing the input.
We extend this classical problem by replacing the fixed basis with a subset $S_0$ of the ground set and imposing various structural constraints on the set of maximum-weight bases relative to $S_0$. Specifically, we study six variants: (A) Inverse Matroid Exists, where $S_0$ must contain at least one maximum-weight basis; (B) Inverse Matroid All, where all bases contained in $S_0$ are maximum-weight; and (C) Inverse Matroid Only, where $S_0$ contains exactly the maximum-weight bases, along with their natural negated counterparts.
For all variants, we develop combinatorial polynomial-time algorithms under the $\ell_\infty$-norm. A key ingredient is a refined min-max theorem for Inverse Matroid under the $\ell_\infty$-norm, which enables simpler and faster algorithms than previous approaches and may be of independent combinatorial interest. Our work significantly broadens the range of inverse optimization problems on matroids that can be solved efficiently, especially those that constrain the structure of optimal solutions through subset inclusion or exclusion. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 16:36:38 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 06:36:04 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bérczi",
"Kristóf",
""
],
[
"Mendoza-Cadena",
"Lydia Mirabel",
""
],
[
"Soto",
"José",
""
]
] |
2507.00992 | Yuanrui Wang | Yuanrui Wang, Cong Han, Yafei Li, Zhipeng Jin, Xiawei Li, SiNan Du, Wen Tao, Yi Yang, Shuanglong Li, Chun Yuan, Liu Lin | UniGlyph: Unified Segmentation-Conditioned Diffusion for Precise Visual Text Synthesis | Accepted by ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-to-image generation has greatly advanced content creation, yet accurately rendering visual text remains a key challenge due to blurred glyphs, semantic drift, and limited style control. Existing methods often rely on pre-rendered glyph images as conditions, but these struggle to retain original font styles and color cues, necessitating complex multi-branch designs that increase model overhead and reduce flexibility. To address these issues, we propose a segmentation-guided framework that uses pixel-level visual text masks -- rich in glyph shape, color, and spatial detail -- as unified conditional inputs. Our method introduces two core components: (1) a fine-tuned bilingual segmentation model for precise text mask extraction, and (2) a streamlined diffusion model augmented with adaptive glyph conditioning and a region-specific loss to preserve textual fidelity in both content and style. Our approach achieves state-of-the-art performance on the AnyText benchmark, significantly surpassing prior methods in both Chinese and English settings. To enable more rigorous evaluation, we also introduce two new benchmarks: GlyphMM-benchmark for testing layout and glyph consistency in complex typesetting, and MiniText-benchmark for assessing generation quality in small-scale text regions. Experimental results show that our model outperforms existing methods by a large margin in both scenarios, particularly excelling at small text rendering and complex layout preservation, validating its strong generalization and deployment readiness. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 17:42:19 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 04:08:46 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Wang",
"Yuanrui",
""
],
[
"Han",
"Cong",
""
],
[
"Li",
"Yafei",
""
],
[
"Jin",
"Zhipeng",
""
],
[
"Li",
"Xiawei",
""
],
[
"Du",
"SiNan",
""
],
[
"Tao",
"Wen",
""
],
[
"Yang",
"Yi",
""
],
[
"Li",
"Shuanglong",
""
],
[
"Yuan",
"Chun",
""
],
[
"Lin",
"Liu",
""
]
] |
2507.01004 | Yuhong Chou | Yuhong Chou, Zehao Liu, Ruijie Zhu, Xinyi Wan, Tianjian Li, Congying Chu, Qian Liu, Jibin Wu, Zejun Ma | ZeCO: Zero Communication Overhead Sequence Parallelism for Linear Attention | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Linear attention mechanisms deliver significant advantages for Large Language Models (LLMs) by providing linear computational complexity, enabling efficient processing of ultra-long sequences (e.g., 1M context). However, existing Sequence Parallelism (SP) methods, essential for distributing these workloads across devices, become the primary bottleneck due to substantial communication overhead. In this paper, we introduce ZeCO (Zero Communication Overhead) sequence parallelism for linear attention models, a new SP method designed to overcome these limitations and achieve end-to-end near-linear scalability for long sequence training. For example, training a model with a 1M sequence length across 64 devices using ZeCO takes roughly the same time as training with an 16k sequence on a single device. At the heart of ZeCO lies All-Scan, a new collective communication primitive. All-Scan provides each SP rank with precisely the initial operator state it requires while maintaining a minimal communication footprint, effectively eliminating communication overhead. Theoretically, we prove the optimaity of ZeCO, showing that it introduces only negligible time and space overhead. Empirically, we compare the communication costs of different sequence parallelism strategies and demonstrate that All-Scan achieves the fastest communication in SP scenarios. Specifically, on 256 GPUs with an 8M sequence length, ZeCO achieves a 60\% speedup compared to the current state-of-the-art (SOTA) SP method. We believe ZeCO establishes a clear path toward efficiently training next-generation LLMs on previously intractable sequence lengths. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 17:54:53 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 10:04:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Chou",
"Yuhong",
""
],
[
"Liu",
"Zehao",
""
],
[
"Zhu",
"Ruijie",
""
],
[
"Wan",
"Xinyi",
""
],
[
"Li",
"Tianjian",
""
],
[
"Chu",
"Congying",
""
],
[
"Liu",
"Qian",
""
],
[
"Wu",
"Jibin",
""
],
[
"Ma",
"Zejun",
""
]
] |
2507.01006 | Jiale Cheng | GLM-V Team: Wenyi Hong, Wenmeng Yu, Xiaotao Gu, Guo Wang, Guobing Gan, Haomiao Tang, Jiale Cheng, Ji Qi, Junhui Ji, Lihang Pan, Shuaiqi Duan, Weihan Wang, Yan Wang, Yean Cheng, Zehai He, Zhe Su, Zhen Yang, Ziyang Pan, Aohan Zeng, Baoxu Wang, Boyan Shi, Changyu Pang, Chenhui Zhang, Da Yin, Fan Yang, Guoqing Chen, Jiazheng Xu, Jiali Chen, Jing Chen, Jinhao Chen, Jinghao Lin, Jinjiang Wang, Junjie Chen, Leqi Lei, Letian Gong, Leyi Pan, Mingzhi Zhang, Qinkai Zheng, Sheng Yang, Shi Zhong, Shiyu Huang, Shuyuan Zhao, Siyan Xue, Shangqin Tu, Shengbiao Meng, Tianshu Zhang, Tianwei Luo, Tianxiang Hao, Wenkai Li, Wei Jia, Xin Lyu, Xuancheng Huang, Yanling Wang, Yadong Xue, Yanfeng Wang, Yifan An, Yifan Du, Yiming Shi, Yiheng Huang, Yilin Niu, Yuan Wang, Yuanchang Yue, Yuchen Li, Yutao Zhang, Yuxuan Zhang, Zhanxiao Du, Zhenyu Hou, Zhao Xue, Zhengxiao Du, Zihan Wang, Peng Zhang, Debing Liu, Bin Xu, Juanzi Li, Minlie Huang, Yuxiao Dong, Jie Tang | GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present GLM-4.1V-Thinking, a vision-language model (VLM) designed to advance general-purpose multimodal understanding and reasoning. In this report, we share our key findings in the development of the reasoning-centric training framework. We first develop a capable vision foundation model with significant potential through large-scale pre-training, which arguably sets the upper bound for the final performance. We then propose Reinforcement Learning with Curriculum Sampling (RLCS) to unlock the full potential of the model, leading to comprehensive capability enhancement across a diverse range of tasks, including STEM problem solving, video understanding, content recognition, coding, grounding, GUI-based agents, and long document understanding. We open-source GLM-4.1V-9B-Thinking, which achieves state-of-the-art performance among models of comparable size. In a comprehensive evaluation across 28 public benchmarks, our model outperforms Qwen2.5-VL-7B on nearly all tasks and achieves comparable or even superior performance on 18 benchmarks relative to the significantly larger Qwen2.5-VL-72B. Notably, GLM-4.1V-9B-Thinking also demonstrates competitive or superior performance compared to closed-source models such as GPT-4o on challenging tasks including long document understanding and STEM reasoning, further underscoring its strong capabilities. Code, models and more information are released at https://github.com/THUDM/GLM-4.1V-Thinking. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 17:55:04 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 15:53:43 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"V Team",
"",
""
],
[
"Hong",
"Wenyi",
""
],
[
"Yu",
"Wenmeng",
""
],
[
"Gu",
"Xiaotao",
""
],
[
"Wang",
"Guo",
""
],
[
"Gan",
"Guobing",
""
],
[
"Tang",
"Haomiao",
""
],
[
"Cheng",
"Jiale",
""
],
[
"Qi",
"Ji",
""
],
[
"Ji",
"Junhui",
""
],
[
"Pan",
"Lihang",
""
],
[
"Duan",
"Shuaiqi",
""
],
[
"Wang",
"Weihan",
""
],
[
"Wang",
"Yan",
""
],
[
"Cheng",
"Yean",
""
],
[
"He",
"Zehai",
""
],
[
"Su",
"Zhe",
""
],
[
"Yang",
"Zhen",
""
],
[
"Pan",
"Ziyang",
""
],
[
"Zeng",
"Aohan",
""
],
[
"Wang",
"Baoxu",
""
],
[
"Shi",
"Boyan",
""
],
[
"Pang",
"Changyu",
""
],
[
"Zhang",
"Chenhui",
""
],
[
"Yin",
"Da",
""
],
[
"Yang",
"Fan",
""
],
[
"Chen",
"Guoqing",
""
],
[
"Xu",
"Jiazheng",
""
],
[
"Chen",
"Jiali",
""
],
[
"Chen",
"Jing",
""
],
[
"Chen",
"Jinhao",
""
],
[
"Lin",
"Jinghao",
""
],
[
"Wang",
"Jinjiang",
""
],
[
"Chen",
"Junjie",
""
],
[
"Lei",
"Leqi",
""
],
[
"Gong",
"Letian",
""
],
[
"Pan",
"Leyi",
""
],
[
"Zhang",
"Mingzhi",
""
],
[
"Zheng",
"Qinkai",
""
],
[
"Yang",
"Sheng",
""
],
[
"Zhong",
"Shi",
""
],
[
"Huang",
"Shiyu",
""
],
[
"Zhao",
"Shuyuan",
""
],
[
"Xue",
"Siyan",
""
],
[
"Tu",
"Shangqin",
""
],
[
"Meng",
"Shengbiao",
""
],
[
"Zhang",
"Tianshu",
""
],
[
"Luo",
"Tianwei",
""
],
[
"Hao",
"Tianxiang",
""
],
[
"Li",
"Wenkai",
""
],
[
"Jia",
"Wei",
""
],
[
"Lyu",
"Xin",
""
],
[
"Huang",
"Xuancheng",
""
],
[
"Wang",
"Yanling",
""
],
[
"Xue",
"Yadong",
""
],
[
"Wang",
"Yanfeng",
""
],
[
"An",
"Yifan",
""
],
[
"Du",
"Yifan",
""
],
[
"Shi",
"Yiming",
""
],
[
"Huang",
"Yiheng",
""
],
[
"Niu",
"Yilin",
""
],
[
"Wang",
"Yuan",
""
],
[
"Yue",
"Yuanchang",
""
],
[
"Li",
"Yuchen",
""
],
[
"Zhang",
"Yutao",
""
],
[
"Zhang",
"Yuxuan",
""
],
[
"Du",
"Zhanxiao",
""
],
[
"Hou",
"Zhenyu",
""
],
[
"Xue",
"Zhao",
""
],
[
"Du",
"Zhengxiao",
""
],
[
"Wang",
"Zihan",
""
],
[
"Zhang",
"Peng",
""
],
[
"Liu",
"Debing",
""
],
[
"Xu",
"Bin",
""
],
[
"Li",
"Juanzi",
""
],
[
"Huang",
"Minlie",
""
],
[
"Dong",
"Yuxiao",
""
],
[
"Tang",
"Jie",
""
]
] |
2507.01011 | Francesco Scazza | Alessandro Muzi Falconi, Riccardo Panza, Sara Sbernardori, Riccardo Forti, Ralf Klemt, Omar Abdel Karim, Matteo Marinelli, Francesco Scazza | Microsecond-scale high-survival and number-resolved detection of ytterbium atom arrays | 8+10 pages, 4+6 figures; typos corrected | null | null | null | cond-mat.quant-gas physics.atom-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scalable atom-based quantum platforms for simulation, computing, and metrology require fast high-fidelity, low-loss imaging of individual atoms. Standard fluorescence detection methods rely on continuous cooling, limiting the detection range to one atom and imposing long imaging times that constrain the experimental cycle and mid-circuit conditional operations. Here, we demonstrate fast and low-loss single-atom imaging in optical tweezers without active cooling, enabled by the favorable properties of ytterbium. Collecting fluorescence over microsecond timescales, we reach single-atom discrimination fidelities above 99.9% and single-shot survival probabilities above 99.5%. Through interleaved recooling pulses, as short as a few hundred microseconds for atoms in magic traps, we perform tens of consecutive detections with constant atom-retention probability per image - an essential step toward fast atom re-use in tweezer-based processors and clocks. Our scheme does not induce parity projection in multiply-occupied traps, enabling number-resolved single-shot detection of several atoms per site. This allows us to study the near-deterministic preparation of single atoms in tweezers driven by blue-detuned light-assisted collisions. Moreover, the near-diffraction-limited spatial resolution of our low-loss imaging enables number-resolved microscopy in dense arrays, opening the way to direct site-occupancy readout in optical lattices for density fluctuation and correlation measurements in quantum simulators. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 17:58:23 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 11:14:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Falconi",
"Alessandro Muzi",
""
],
[
"Panza",
"Riccardo",
""
],
[
"Sbernardori",
"Sara",
""
],
[
"Forti",
"Riccardo",
""
],
[
"Klemt",
"Ralf",
""
],
[
"Karim",
"Omar Abdel",
""
],
[
"Marinelli",
"Matteo",
""
],
[
"Scazza",
"Francesco",
""
]
] |
2507.01018 | Mohammed K Alzaylaee Dr | Mohammed K. Alzaylaee | A Systematic Review of Security Vulnerabilities in Smart Home Devices and Mitigation Techniques | null | IJCSNS Vol.25, No.3 (2025) | 10.22937/IJCSNS.2025.25.3.24 | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Smart homes that integrate Internet of Things (IoT) devices face increasing cybersecurity risks, posing significant challenges to these environments. The study explores security threats in smart homes ecosystems, categorizing them into vulnerabilities at the network layer, device level, and those from cloud-based and AI-driven systems. Research findings indicate that post-quantum encryption, coupled with AI-driven anomaly detection, is highly effective in enhancing security; however, computational resource demands present significant challenges. Blockchain authentication together with zero-trust structures builds security resilience, although they need changes to existing infrastructure. The specific security strategies show their effectiveness through ANOVA, Chi-square tests, and Monte Carlo simulations yet lack sufficient scalability according to the results. The research demonstrates the requirement for improvement in cryptographic techniques, alongside AI-enhanced threat detection and adaptive security models which must achieve a balance between performance and efficiency and real-time applicability within smart home ecosystems. | [
{
"version": "v1",
"created": "Thu, 03 Apr 2025 00:03:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Alzaylaee",
"Mohammed K.",
""
]
] |
2507.01019 | Kevin Zhu | Imran Mirza, Cole Huang, Ishwara Vasista, Rohan Patil, Asli Akalin, Sean O'Brien, Kevin Zhu | MALIBU Benchmark: Multi-Agent LLM Implicit Bias Uncovered | Accepted to Building Trust in LLMs @ ICLR 2025 and NAACL SRW 2025 | null | null | null | cs.CL cs.CY | http://creativecommons.org/licenses/by/4.0/ | Multi-agent systems, which consist of multiple AI models interacting within a shared environment, are increasingly used for persona-based interactions. However, if not carefully designed, these systems can reinforce implicit biases in large language models (LLMs), raising concerns about fairness and equitable representation. We present MALIBU, a novel benchmark developed to assess the degree to which LLM-based multi-agent systems implicitly reinforce social biases and stereotypes. MALIBU evaluates bias in LLM-based multi-agent systems through scenario-based assessments. AI models complete tasks within predefined contexts, and their responses undergo evaluation by an LLM-based multi-agent judging system in two phases. In the first phase, judges score responses labeled with specific demographic personas (e.g., gender, race, religion) across four metrics. In the second phase, judges compare paired responses assigned to different personas, scoring them and selecting the superior response. Our study quantifies biases in LLM-generated outputs, revealing that bias mitigation may favor marginalized personas over true neutrality, emphasizing the need for nuanced detection, balanced fairness strategies, and transparent evaluation benchmarks in multi-agent systems. | [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 19:16:40 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mirza",
"Imran",
""
],
[
"Huang",
"Cole",
""
],
[
"Vasista",
"Ishwara",
""
],
[
"Patil",
"Rohan",
""
],
[
"Akalin",
"Asli",
""
],
[
"O'Brien",
"Sean",
""
],
[
"Zhu",
"Kevin",
""
]
] |
2507.01020 | Aashray Reddy | Aashray Reddy, Andrew Zagula, Nicholas Saban | AutoAdv: Automated Adversarial Prompting for Multi-Turn Jailbreaking of Large Language Models | 16 pages, 4 figures, submitted to LLMSEC | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) continue to exhibit vulnerabilities to jailbreaking attacks: carefully crafted malicious inputs intended to circumvent safety guardrails and elicit harmful responses. As such, we present AutoAdv, a novel framework that automates adversarial prompt generation to systematically evaluate and expose vulnerabilities in LLM safety mechanisms. Our approach leverages a parametric attacker LLM to produce semantically disguised malicious prompts through strategic rewriting techniques, specialized system prompts, and optimized hyperparameter configurations. The primary contribution of our work is a dynamic, multi-turn attack methodology that analyzes failed jailbreak attempts and iteratively generates refined follow-up prompts, leveraging techniques such as roleplaying, misdirection, and contextual manipulation. We quantitatively evaluate attack success rate (ASR) using the StrongREJECT (arXiv:2402.10260 [cs.CL]) framework across sequential interaction turns. Through extensive empirical evaluation of state-of-the-art models--including ChatGPT, Llama, and DeepSeek--we reveal significant vulnerabilities, with our automated attacks achieving jailbreak success rates of up to 86% for harmful content generation. Our findings reveal that current safety mechanisms remain susceptible to sophisticated multi-turn attacks, emphasizing the urgent need for more robust defense strategies. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 08:38:56 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Reddy",
"Aashray",
""
],
[
"Zagula",
"Andrew",
""
],
[
"Saban",
"Nicholas",
""
]
] |
2507.01021 | Kumarmanas Nethil | Kumarmanas Nethil, Vaibhav Mishra, Kriti Anandan, Kavya Manohar | Scalable Offline ASR for Command-Style Dictation in Courtrooms | Accepted to Interspeech 2025 Show & Tell | null | null | null | eess.AS cs.CL cs.SD | http://creativecommons.org/licenses/by/4.0/ | We propose an open-source framework for Command-style dictation that addresses the gap between resource-intensive Online systems and high-latency Batch processing. Our approach uses Voice Activity Detection (VAD) to segment audio and transcribes these segments in parallel using Whisper models, enabling efficient multiplexing across audios. Unlike proprietary systems like SuperWhisper, this framework is also compatible with most ASR architectures, including widely used CTC-based models. Our multiplexing technique maximizes compute utilization in real-world settings, as demonstrated by its deployment in around 15% of India's courtrooms. Evaluations on live data show consistent latency reduction as user concurrency increases, compared to sequential batch processing. The live demonstration will showcase our open-sourced implementation and allow attendees to interact with it in real-time. | [
{
"version": "v1",
"created": "Sat, 07 Jun 2025 17:26:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Nethil",
"Kumarmanas",
""
],
[
"Mishra",
"Vaibhav",
""
],
[
"Anandan",
"Kriti",
""
],
[
"Manohar",
"Kavya",
""
]
] |
2507.01022 | Shayan Dadman | Shayan Dadman, Bernt Arild Bremdal, Andreas Bergsland | Workflow-Based Evaluation of Music Generation Systems | 54 pages, 3 figures, 6 tables, 5 appendices | null | null | null | eess.AS cs.HC cs.LG cs.MM cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study presents an exploratory evaluation of Music Generation Systems (MGS) within contemporary music production workflows by examining eight open-source systems. The evaluation framework combines technical insights with practical experimentation through criteria specifically designed to investigate the practical and creative affordances of the systems within the iterative, non-linear nature of music production. Employing a single-evaluator methodology as a preliminary phase, this research adopts a mixed approach utilizing qualitative methods to form hypotheses subsequently assessed through quantitative metrics. The selected systems represent architectural diversity across both symbolic and audio-based music generation approaches, spanning composition, arrangement, and sound design tasks. The investigation addresses limitations of current MGS in music production, challenges and opportunities for workflow integration, and development potential as collaborative tools while maintaining artistic authenticity. Findings reveal these systems function primarily as complementary tools enhancing rather than replacing human expertise. They exhibit limitations in maintaining thematic and structural coherence that emphasize the indispensable role of human creativity in tasks demanding emotional depth and complex decision-making. This study contributes a structured evaluation framework that considers the iterative nature of music creation. It identifies methodological refinements necessary for subsequent comprehensive evaluations and determines viable areas for AI integration as collaborative tools in creative workflows. The research provides empirically-grounded insights to guide future development in the field. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 13:07:47 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Dadman",
"Shayan",
""
],
[
"Bremdal",
"Bernt Arild",
""
],
[
"Bergsland",
"Andreas",
""
]
] |
2507.01023 | Mizuki Tani | Mizuki Tani and Kenichi L. Ishikawa and Tomohito Otobe | Electron Excitation Probability in Dielectrics under Two-color Intense Laser Fields | null | null | null | null | physics.optics physics.atom-ph physics.plasm-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Two-color laser fields offer significantly enhanced control over electron excitation dynamics under ultrashort intense laser pulses compared to monochromatic fields. However, their strong nonlinearity necessitates computationally expensive first-principles calculations to accurately predict ionization dynamics. To overcome this challenge, we derive an analytical expression for the ionization rate in dielectrics subjected to intense two-color laser fields, refining the theoretical framework introduced in JPSJ {\bf 88}, 024706 (2019). By benchmarking our formula against first-principles calculations based on time-dependent density functional theory (TDDFT) for $\alpha$-quartz, we demonstrate that our model captures the essential physics of ionization dynamics with remarkable qualitative accuracy, despite employing certain approximations. This analytical approach not only provides deeper physical insight but also offers a computationally efficient alternative for predicting strong-field interactions in dielectrics. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 18:46:03 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Tani",
"Mizuki",
""
],
[
"Ishikawa",
"Kenichi L.",
""
],
[
"Otobe",
"Tomohito",
""
]
] |
2507.01024 | Ifeanyi George Igwegbe | George Igwegbe, Martins Awojide, Mboh Bless, Nirel Kadzo | Hello Afrika: Speech Commands in Kinyarwanda | Data Science Africa, 2024 | null | null | null | eess.AS cs.AI cs.SD | http://creativecommons.org/licenses/by/4.0/ | Voice or Speech Commands are a subset of the broader Spoken Word Corpus of a language which are essential for non-contact control of and activation of larger AI systems in devices used in everyday life especially for persons with disabilities. Currently, there is a dearth of speech command models for African languages. The Hello Afrika project aims to address this issue and its first iteration is focused on the Kinyarwanda language since the country has shown interest in developing speech recognition technologies culminating in one of the largest datasets on Mozilla Common Voice. The model was built off a custom speech command corpus made up of general directives, numbers, and a wake word. The final model was deployed on multiple devices (PC, Mobile Phone and Edge Devices) and the performance was assessed using suitable metrics. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:30:19 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Igwegbe",
"George",
""
],
[
"Awojide",
"Martins",
""
],
[
"Bless",
"Mboh",
""
],
[
"Kadzo",
"Nirel",
""
]
] |
2507.01025 | Pin Chen | Yutong Lu, Dan Huang and Pin Chen | HPC-AI Coupling Methodology for Scientific Applications | 14 pages, 11 figures | null | null | null | cs.CE cs.AI physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Artificial intelligence (AI) technologies have fundamentally transformed numerical-based high-performance computing (HPC) applications with data-driven approaches and endeavored to address existing challenges, e.g. high computational intensity, in various scientific domains. In this study, we explore the scenarios of coupling HPC and AI (HPC-AI) in the context of emerging scientific applications, presenting a novel methodology that incorporates three patterns of coupling: surrogate, directive, and coordinate. Each pattern exemplifies a distinct coupling strategy, AI-driven prerequisite, and typical HPC-AI ensembles. Through case studies in materials science, we demonstrate the application and effectiveness of these patterns. The study highlights technical challenges, performance improvements, and implementation details, providing insight into promising perspectives of HPC-AI coupling. The proposed coupling patterns are applicable not only to materials science but also to other scientific domains, offering valuable guidance for future HPC-AI ensembles in scientific discovery. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 09:09:03 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lu",
"Yutong",
""
],
[
"Huang",
"Dan",
""
],
[
"Chen",
"Pin",
""
]
] |
2507.01026 | Md Shakil Ahamed Shohag | Md Shakil Ahamed Shohag, Q. M. Jonathan Wu, Farhad Pourpanah | Few-Shot Inspired Generative Zero-Shot Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative zero-shot learning (ZSL) methods typically synthesize visual features for unseen classes using predefined semantic attributes, followed by training a fully supervised classification model. While effective, these methods require substantial computational resources and extensive synthetic data, thereby relaxing the original ZSL assumptions. In this paper, we propose FSIGenZ, a few-shot-inspired generative ZSL framework that reduces reliance on large-scale feature synthesis. Our key insight is that class-level attributes exhibit instance-level variability, i.e., some attributes may be absent or partially visible, yet conventional ZSL methods treat them as uniformly present. To address this, we introduce Model-Specific Attribute Scoring (MSAS), which dynamically re-scores class attributes based on model-specific optimization to approximate instance-level variability without access to unseen data. We further estimate group-level prototypes as clusters of instances based on MSAS-adjusted attribute scores, which serve as representative synthetic features for each unseen class. To mitigate the resulting data imbalance, we introduce a Dual-Purpose Semantic Regularization (DPSR) strategy while training a semantic-aware contrastive classifier (SCC) using these prototypes. Experiments on SUN, AwA2, and CUB benchmarks demonstrate that FSIGenZ achieves competitive performance using far fewer synthetic features. | [
{
"version": "v1",
"created": "Wed, 18 Jun 2025 02:39:36 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Shohag",
"Md Shakil Ahamed",
""
],
[
"Wu",
"Q. M. Jonathan",
""
],
[
"Pourpanah",
"Farhad",
""
]
] |
2507.01027 | Zijian Ye | Zijian Ye, Wei Huang, Yifei Yu, Tianhe Ren, Zhongrui Wang, Xiaojuan Qi | DBellQuant: Breaking the Bell with Double-Bell Transformation for LLMs Post Training Binarization | 19 pages; Appendix added | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) demonstrate remarkable performance but face substantial computational and memory challenges that limit their practical deployment. Quantization has emerged as a promising solution; however, its effectiveness is often limited by quantization errors arising from weight distributions that are not quantization-friendly and the presence of activation outliers. To address these challenges, we introduce DBellQuant, an innovative post-training quantization (PTQ) framework that achieves nearly 1-bit weight compression and 6-bit activation quantization with minimal performance degradation. DBellQuant uses Learnable Transformation for Dual-Bell (LTDB) algorithm, which transforms single-bell weight distributions into dual-bell forms to reduce binarization errors and applies inverse transformations to smooth activations. DBellQuant sets a new state-of-the-art by preserving superior model performance under aggressive weight and activation quantization. For example, on the Wikitext2 dataset, DBellQuant achieves a perplexity of 14.39 on LLaMA2-13B with 6-bit activation quantization, significantly outperforming BiLLM's 21.35 without activation quantization, underscoring its potential in compressing LLMs for real-world applications. | [
{
"version": "v1",
"created": "Wed, 18 Jun 2025 06:41:03 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ye",
"Zijian",
""
],
[
"Huang",
"Wei",
""
],
[
"Yu",
"Yifei",
""
],
[
"Ren",
"Tianhe",
""
],
[
"Wang",
"Zhongrui",
""
],
[
"Qi",
"Xiaojuan",
""
]
] |
2507.01028 | Basile Terver | Jean Ponce (WILLOW), Martial Hebert (CMU), Basile Terver (FAIR, WILLOW) | Dual Perspectives on Non-Contrastive Self-Supervised Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The objective of non-contrastive approaches to self-supervised learning is to train on pairs of different views of the data an encoder and a predictor that minimize the mean discrepancy between the code predicted from the embedding of the first view and the embedding of the second one. In this setting, the stop gradient and exponential moving average iterative procedures are commonly used to avoid representation collapse, with excellent performance in downstream supervised applications. This presentation investigates these procedures from the dual theoretical viewpoints of optimization and dynamical systems. We first show that, in general, although they do not optimize the original objective, or for that matter, any other smooth function, they do avoid collapse. Following Tian et al. [2021], but without any of the extra assumptions used in their proofs, we then show using a dynamical system perspective that, in the linear case, minimizing the original objective function without the use of a stop gradient or exponential moving average always leads to collapse. Conversely, we finally show that the limit points of the dynamical systems associated with these two procedures are, in general, asymptotically stable equilibria, with no risk of degenerating to trivial solutions. | [
{
"version": "v1",
"created": "Wed, 18 Jun 2025 07:46:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ponce",
"Jean",
"",
"WILLOW"
],
[
"Hebert",
"Martial",
"",
"CMU"
],
[
"Terver",
"Basile",
"",
"FAIR, WILLOW"
]
] |
2507.01029 | Junjie Zhou | Junjie Zhou, Yingli Zuo, Shichang Feng, Peng Wan, Qi Zhu, Daoqiang Zhang, and Wei Shao | PathCoT: Chain-of-Thought Prompting for Zero-shot Pathology Visual Reasoning | null | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | With the development of generative artificial intelligence and instruction tuning techniques, multimodal large language models (MLLMs) have made impressive progress on general reasoning tasks. Benefiting from the chain-of-thought (CoT) methodology, MLLMs can solve the visual reasoning problem step-by-step. However, existing MLLMs still face significant challenges when applied to pathology visual reasoning tasks: (1) LLMs often underperforms because they lack domain-specific information, which can lead to model hallucinations. (2) The additional reasoning steps in CoT may introduce errors, leading to the divergence of answers. To address these limitations, we propose PathCoT, a novel zero-shot CoT prompting method which integrates the pathology expert-knowledge into the reasoning process of MLLMs and incorporates self-evaluation to mitigate divergence of answers. Specifically, PathCoT guides the MLLM with prior knowledge to perform as pathology experts, and provides comprehensive analysis of the image with their domain-specific knowledge. By incorporating the experts' knowledge, PathCoT can obtain the answers with CoT reasoning. Furthermore, PathCoT incorporates a self-evaluation step that assesses both the results generated directly by MLLMs and those derived through CoT, finally determining the reliable answer. The experimental results on the PathMMU dataset demonstrate the effectiveness of our method on pathology visual understanding and reasoning. | [
{
"version": "v1",
"created": "Wed, 18 Jun 2025 09:20:23 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhou",
"Junjie",
""
],
[
"Zuo",
"Yingli",
""
],
[
"Feng",
"Shichang",
""
],
[
"Wan",
"Peng",
""
],
[
"Zhu",
"Qi",
""
],
[
"Zhang",
"Daoqiang",
""
],
[
"Shao",
"Wei",
""
]
] |
2507.01030 | Seyed Mohammad Jafar Sobhani | Reza Lotfi Navaei, Mohammad Safarzadeh, Seyed Mohammad Jafar Sobhani | Optimizing Flamelet Generated Manifold Models: A Machine Learning Performance Study | It has been submitted to ASME Journal of Heat and Mass Transfer | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In chemistry tabulations and Flamelet combustion models, the Flamelet Generated Manifold (FGM) is recognized for its precision and physical representation. The practical implementation of FGM requires a significant allocation of memory resources. FGM libraries are developed specifically for a specific fuel and subsequently utilized for all numerical problems using machine learning techniques. This research aims to develop libraries of Laminar FGM utilizing machine learning algorithms for application in combustion simulations of methane fuel. This study employs four Machine Learning algorithms to regenerate Flamelet libraries, based on an understanding of data sources, techniques, and data-driven concepts. 1. Multi-Layer Perceptron; 2. Random Forest; 3. Linear Regression; 4. Support Vector Machine. Seven libraries were identified as appropriate for constructing a database for training machine learning models, giving an error rate of 2.30%. The default architectures of each method were evaluated to determine the optimal approach, leading to the selection of the MLP method as the primary choice. The method was enhanced through hyperparameter tuning to improve accuracy. The quantity of hidden layers and neurons significantly influences method performance. The optimal model, comprising four hidden layers with 10, 15, 20, and 25 neurons respectively, achieved an accuracy of 99.81%. | [
{
"version": "v1",
"created": "Wed, 18 Jun 2025 16:27:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Navaei",
"Reza Lotfi",
""
],
[
"Safarzadeh",
"Mohammad",
""
],
[
"Sobhani",
"Seyed Mohammad Jafar",
""
]
] |
2507.01031 | Fanchen Bu | Fanchen Bu, Kijung Shin | PyTorch-based Geometric Learning with Non-CUDA Processing Units: Experiences from Intel Gaudi-v2 HPUs | Conference paper: Accepted in Korea Computer Congress (KCC) 2025. The library is available at https://github.com/bokveizen/gaudi-geometric-learning | null | null | null | cs.LG cs.SE | http://creativecommons.org/licenses/by/4.0/ | Geometric learning has emerged as a powerful paradigm for modeling non-Euclidean data, especially graph-structured ones, with applications spanning social networks, molecular structures, knowledge graphs, and recommender systems. While Nvidia's CUDA-enabled graphics processing units (GPUs) largely dominate the hardware landscape, emerging accelerators such as Intel's Gaudi Habana Processing Units (HPUs) offer competitive performance and energy efficiency. However, the usage of such non-CUDA processing units requires significant engineering effort and novel software adaptations. In this work, we present our experiences porting PyTorch-based geometric learning frameworks to Gaudi-v2 HPUs. We introduce a collection of core utilities that restore essential operations (e.g., scatter, sparse indexing, k-nearest neighbors) on Gaudi-v2 HPUs, and we consolidate sixteen guided tutorials and eleven real-world examples with diagnostic analyses of encountered failures and detailed workarounds. We collect all our experiences into a publicly accessible GitHub repository. Our contributions lower the barrier for researchers to experiment with geometric-learning algorithms and models on non-CUDA hardware, providing a foundation for further optimization and cross-platform portability. | [
{
"version": "v1",
"created": "Fri, 20 Jun 2025 04:10:48 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bu",
"Fanchen",
""
],
[
"Shin",
"Kijung",
""
]
] |
2507.01032 | Chen Zhao | Nan Mu, Hongbo Yang, Chen Zhao | An Uncertainty-Aware Dynamic Decision Framework for Progressive Multi-Omics Integration in Classification Tasks | null | null | null | null | cs.LG cs.AI q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Background and Objective: High-throughput multi-omics technologies have proven invaluable for elucidating disease mechanisms and enabling early diagnosis. However, the high cost of multi-omics profiling imposes a significant economic burden, with over reliance on full omics data potentially leading to unnecessary resource consumption. To address these issues, we propose an uncertainty-aware, multi-view dynamic decision framework for omics data classification that aims to achieve high diagnostic accuracy while minimizing testing costs. Methodology: At the single-omics level, we refine the activation functions of neural networks to generate Dirichlet distribution parameters, utilizing subjective logic to quantify both the belief masses and uncertainty mass of classification results. Belief mass reflects the support of a specific omics modality for a disease class, while the uncertainty parameter captures limitations in data quality and model discriminability, providing a more trustworthy basis for decision-making. At the multi omics level, we employ a fusion strategy based on Dempster-Shafer theory to integrate heterogeneous modalities, leveraging their complementarity to boost diagnostic accuracy and robustness. A dynamic decision mechanism is then applied that omics data are incrementally introduced for each patient until either all data sources are utilized or the model confidence exceeds a predefined threshold, potentially before all data sources are utilized. Results and Conclusion: We evaluate our approach on four benchmark multi-omics datasets, ROSMAP, LGG, BRCA, and KIPAN. In three datasets, over 50% of cases achieved accurate classification using a single omics modality, effectively reducing redundant testing. Meanwhile, our method maintains diagnostic performance comparable to full-omics models and preserves essential biological insights. | [
{
"version": "v1",
"created": "Fri, 20 Jun 2025 13:44:14 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mu",
"Nan",
""
],
[
"Yang",
"Hongbo",
""
],
[
"Zhao",
"Chen",
""
]
] |
2507.01033 | Changyoung Kim | Younsik Kim, Suk Bum Chung, Changyoung Kim | Pedagogical approach to anomalous position and velocity | null | null | 10.1016/j.cap.2025.03.003 | null | cond-mat.mes-hall cond-mat.str-el quant-ph | http://creativecommons.org/licenses/by/4.0/ | In this work, we discuss a pedagogical method in deriving the expressions for anomalous position and velocity. While we follow the steps used in optics in the derivation of the group velocity, we use Bloch wave functions instead of plane wave states. In comparison to the plane wave case, application of Bloch wave functions results in two additional terms in the expression of the group velocity: the Berry phase factor and anomalous position contributions. These two new terms with distinct origins eventually lead to the known anomalous velocity. Aiming for an intuitive understanding, we simulate the situation under an electric field using linear-combination-of-atomic-orbital states and visually demonstrate that the envelope function exhibits the transverse motion expected from an anomalous velocity. | [
{
"version": "v1",
"created": "Fri, 20 Jun 2025 22:59:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kim",
"Younsik",
""
],
[
"Chung",
"Suk Bum",
""
],
[
"Kim",
"Changyoung",
""
]
] |
2507.01034 | Mansour Essgaer Dr | Asma Agaal, Mansour Essgaer, Hend M. Farkash, Zulaiha Ali Othman | Data-driven Insights for Informed Decision-Making: Applying LSTM Networks for Robust Electricity Forecasting in Libya | This article was published in International Journal of Intelligent Systems and Applications (IJISA) (MECS Press), Vol. 17, No. 3, 8 Jun. 2025, DOI: https://doi.org/10.5815/ijisa.2025.03.05 | International Journal of Intelligent Systems and Applications (IJISA) (MECS Press), Vol. 17, No. 3, 8 Jun. 2025 | 10.5815/ijisa.2025.03.05 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Accurate electricity forecasting is crucial for grid stability and energy planning, especially in Benghazi, Libya, where frequent load shedding, generation deficits, and infrastructure limitations persist. This study proposes a data-driven approach to forecast electricity load, generation, and deficits for 2025 using historical data from 2019 (a year marked by instability) and 2023 (a more stable year). Multiple time series models were applied, including ARIMA, seasonal ARIMA, dynamic regression ARIMA, exponential smoothing, extreme gradient boosting, and Long Short-Term Memory (LSTM) neural networks. The dataset was enhanced through missing value imputation, outlier smoothing, and log transformation. Performance was assessed using mean squared error, root mean squared error, mean absolute error, and mean absolute percentage error. LSTM outperformed all other models, showing strong capabilities in modeling non-stationary and seasonal patterns. A key contribution of this work is an optimized LSTM framework that integrates exogenous factors such as temperature and humidity, offering robust performance in forecasting multiple electricity indicators. These results provide practical insights for policymakers and grid operators to enable proactive load management and resource planning in data-scarce, volatile regions. | [
{
"version": "v1",
"created": "Fri, 20 Jun 2025 23:41:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Agaal",
"Asma",
""
],
[
"Essgaer",
"Mansour",
""
],
[
"Farkash",
"Hend M.",
""
],
[
"Othman",
"Zulaiha Ali",
""
]
] |
2507.01035 | Yushang Zhao | Yushang Zhao, Haotian Lyu, Yike Peng, Aijia Sun, Feng Jiang, Xinyue Han | Research on Low-Latency Inference and Training Efficiency Optimization for Graph Neural Network and Large Language Model-Based Recommendation Systems | null | null | null | null | cs.LG cs.AI cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The incessant advent of online services demands high speed and efficient recommender systems (ReS) that can maintain real-time performance along with processing very complex user-item interactions. The present study, therefore, considers computational bottlenecks involved in hybrid Graph Neural Network (GNN) and Large Language Model (LLM)-based ReS with the aim optimizing their inference latency and training efficiency. An extensive methodology was used: hybrid GNN-LLM integrated architecture-optimization strategies(quantization, LoRA, distillation)-hardware acceleration (FPGA, DeepSpeed)-all under R 4.4.2. Experimental improvements were significant, with the optimal Hybrid + FPGA + DeepSpeed configuration reaching 13.6% more accuracy (NDCG@10: 0.75) at 40-60ms of latency, while LoRA brought down training time by 66% (3.8 hours) in comparison to the non-optimized baseline. Irrespective of domain, such as accuracy or efficiency, it can be established that hardware-software co-design and parameter-efficient tuning permit hybrid models to outperform GNN or LLM approaches implemented independently. It recommends the use of FPGA as well as LoRA for real-time deployment. Future work should involve federated learning along with advanced fusion architectures for better scalability and privacy preservation. Thus, this research marks the fundamental groundwork concerning next-generation ReS balancing low-latency response with cutting-edge personalization. | [
{
"version": "v1",
"created": "Sat, 21 Jun 2025 03:10:50 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhao",
"Yushang",
""
],
[
"Lyu",
"Haotian",
""
],
[
"Peng",
"Yike",
""
],
[
"Sun",
"Aijia",
""
],
[
"Jiang",
"Feng",
""
],
[
"Han",
"Xinyue",
""
]
] |
2507.01036 | Seth Bulin | Seth Bulin | Systemic Constraints of Undecidability | Submitted version; includes appendices with formal definitions and structural embeddings. Prepared in Nature Computational Science format. Keywords: computability theory, undecidability, causal systems, structural closure, recursion theory, Turing machines, hypercomputation, metaundecidability, epistemic limits, consciousness, modeling limits | null | null | null | cs.FL cs.AI math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a theory of systemic undecidability, reframing incomputability as a structural property of systems rather than a localized feature of specific functions or problems. We define a notion of causal embedding and prove a closure principle: any subsystem that participates functionally in the computation of an undecidable system inherits its undecidability. This result positions undecidability as a pervasive constraint on prediction, modeling, and epistemic access in both natural and artificial systems. Our framework disarms oracle mimicry and challenges the view that computational limits can be circumvented through architectural innovation. By generalizing classical results into a dynamic systems context, this work augments the logical trajectory of G\"odel, Turing, and Chaitin, offering a new perspective of the topology of computability and its interrelation to the boundaries of scientific knowledge. | [
{
"version": "v1",
"created": "Sat, 21 Jun 2025 22:56:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bulin",
"Seth",
""
]
] |
2507.01037 | Wenbin Ouyang | Wenbin Ouyang, Sirui Li, Yining Ma, Cathy Wu | Learning to Segment for Vehicle Routing Problems | null | null | null | null | cs.LG cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Iterative search heuristics are widely recognized as state-of-the-art for solving Vehicle Routing Problems (VRPs). In this work, we identify and exploit a critical observation: within these solvers, a large portion of the solution remains stable, i.e., unchanged across search iterations, causing redundant computations, especially for large-scale VRPs with long subtours. To address this, we pioneer the formal study of the First-Segment-Then-Aggregate (FSTA) decomposition technique to accelerate iterative solvers. Specifically, FSTA preserves stable solution segments during the search, aggregates nodes within each segment into fixed hypernodes, and focuses the search only on unstable portions. Yet, a key challenge lies in identifying which segments should be aggregated by FSTA. To this end, we then introduce Learning-to-Segment (L2Seg), a novel neural framework to intelligently differentiate potentially stable and unstable portions for FSTA decomposition. We present three L2Seg variants: non-autoregressive (globally comprehensive but locally indiscriminate), autoregressive (locally refined but globally deficient), and their synergy, with bespoke training and inference strategies. Empirical results on CVRP and VRPTW suggest that L2Seg accelerates state-of-the-art iterative solvers by up to 7x. Additionally, we provide in-depth analysis showing NAR and AR synergy achieves best performance by combining their complementary strengths. Notably, L2Seg is a flexible framework that is compatible with traditional, learning-based, and hybrid solvers, while supporting a broad class of VRPs. | [
{
"version": "v1",
"created": "Sun, 22 Jun 2025 05:38:15 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ouyang",
"Wenbin",
""
],
[
"Li",
"Sirui",
""
],
[
"Ma",
"Yining",
""
],
[
"Wu",
"Cathy",
""
]
] |
2507.01038 | Seong-Joon Park | Seong-Joon Park, Hee-Youl Kwak, Sang-Hyo Kim, Yongjune Kim, Jong-Seon No | Cross-Attention Message-Passing Transformers for Code-Agnostic Decoding in 6G Networks | null | null | null | null | cs.IT cs.LG eess.SP math.IT | http://creativecommons.org/licenses/by/4.0/ | Channel coding for 6G networks is expected to support a wide range of requirements arising from heterogeneous communication scenarios. These demands challenge traditional code-specific decoders, which lack the flexibility and scalability required for next-generation systems. To tackle this problem, we propose an AI-native foundation model for unified and code-agnostic decoding based on the transformer architecture. We first introduce a cross-attention message-passing transformer (CrossMPT). CrossMPT employs two masked cross-attention blocks that iteratively update two distinct input representations-magnitude and syndrome vectors-allowing the model to effectively learn the decoding problem. Notably, our CrossMPT has achieved state-of-the-art decoding performance among single neural decoders. Building on this, we develop foundation CrossMPT (FCrossMPT) by making the architecture invariant to code length, rate, and class, allowing a single trained model to decode a broad range of codes without retraining. To further enhance decoding performance, particularly for short blocklength codes, we propose CrossMPT ensemble decoder (CrossED), an ensemble decoder composed of multiple parallel CrossMPT blocks employing different parity-check matrices. This architecture can also serve as a foundation model, showing strong generalization across diverse code types. Overall, the proposed AI-native code-agnostic decoder offers flexibility, scalability, and high performance, presenting a promising direction to channel coding for 6G networks. | [
{
"version": "v1",
"created": "Sun, 22 Jun 2025 16:08:42 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Park",
"Seong-Joon",
""
],
[
"Kwak",
"Hee-Youl",
""
],
[
"Kim",
"Sang-Hyo",
""
],
[
"Kim",
"Yongjune",
""
],
[
"No",
"Jong-Seon",
""
]
] |
2507.01039 | Kaaustaaub Shankar | Kaaustaaub Shankar, Wilhelm Louw, Kelly Cohen | On-Policy Optimization of ANFIS Policies Using Proximal Policy Optimization | Submitted to NAFIPS 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | We propose a reinforcement learning (RL) approach for training neuro-fuzzy controllers using Proximal Policy Optimization (PPO). Building on prior work that applied Deep Q-Learning to Adaptive Neuro-Fuzzy Inference Systems (ANFIS), our method replaces the off-policy value-based framework with a stable on-policy actor-critic loop. We evaluate this approach in the CartPole-v1 environment using multiple random seeds and compare its learning performance against ANFIS-Deep Q-Network (DQN) baselines. It was found that PPO-trained fuzzy agents achieved a mean return of 500 +/- 0 on CartPole-v1 after 20000 updates, showcasing less variance than prior DQN-based methods during training and overall faster convergence. These findings suggest that PPO offers a promising pathway for training explainable neuro-fuzzy controllers in reinforcement learning tasks. | [
{
"version": "v1",
"created": "Sun, 22 Jun 2025 17:49:49 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Shankar",
"Kaaustaaub",
""
],
[
"Louw",
"Wilhelm",
""
],
[
"Cohen",
"Kelly",
""
]
] |
2507.01040 | Tianxiang Xia | Tianxiang Xia, Max Neuwinger, Lin Xiao | Fast Clifford Neural Layers | 7 pages content-wise | null | null | null | cs.LG cs.AI cs.NE cs.PF | http://creativecommons.org/licenses/by/4.0/ | Clifford Neural Layers improve PDE modeling by introducing Clifford Algebra into neural networks. In this project we focus on optimizing the inference of 2/3D Clifford convolutional layers and multivector activation layers for one core CPU performance.
Overall, by testing on a real network block involving Clifford convolutional layers and multivector activation layers, we observe that our implementation is 30% faster than standard PyTorch implementation in relatively large data + network size (>L2 cache).
We open source our code base at https://github.com/egretwAlker/c-opt-clifford-layers | [
{
"version": "v1",
"created": "Sun, 22 Jun 2025 20:43:42 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Xia",
"Tianxiang",
""
],
[
"Neuwinger",
"Max",
""
],
[
"Xiao",
"Lin",
""
]
] |
2507.01042 | Gautam Siddharth Kashyap | Harsh Joshi, Gautam Siddharth Kashyap, Rafiq Ali, Ebad Shabbir, Niharika Jain, Sarthak Jain, Jiechao Gao, Usman Naseem | Can Argus Judge Them All? Comparing VLMs Across Domains | null | null | null | null | cs.IR cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Vision-Language Models (VLMs) are advancing multimodal AI, yet their performance consistency across tasks is underexamined. We benchmark CLIP, BLIP, and LXMERT across diverse datasets spanning retrieval, captioning, and reasoning. Our evaluation includes task accuracy, generation quality, efficiency, and a novel Cross-Dataset Consistency (CDC) metric. CLIP shows strongest generalization (CDC: 0.92), BLIP excels on curated data, and LXMERT leads in structured reasoning. These results expose trade-offs between generalization and specialization, informing industrial deployment of VLMs and guiding development toward robust, task-flexible architectures. | [
{
"version": "v1",
"created": "Mon, 23 Jun 2025 09:58:35 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Joshi",
"Harsh",
""
],
[
"Kashyap",
"Gautam Siddharth",
""
],
[
"Ali",
"Rafiq",
""
],
[
"Shabbir",
"Ebad",
""
],
[
"Jain",
"Niharika",
""
],
[
"Jain",
"Sarthak",
""
],
[
"Gao",
"Jiechao",
""
],
[
"Naseem",
"Usman",
""
]
] |
2507.01043 | Agnieszka Jastrzebska | Szymon \'Swiderski and Agnieszka Jastrz\k{e}bska | Data Classification with Dynamically Growing and Shrinking Neural Networks | Paper submitted to Journal of Computational Science | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The issue of data-driven neural network model construction is one of the core problems in the domain of Artificial Intelligence. A standard approach assumes a fixed architecture with trainable weights. A conceptually more advanced assumption is that we not only train the weights, but also find out the optimal model architecture. We present a new method that realizes just that. This article is an extended version of our conference paper titled "Dynamic Growing and Shrinking of Neural Networks with Monte Carlo Tree Search [26]". In the paper, we show in detail how to create a neural network with a procedure that allows dynamic shrinking and growing of the model while it is being trained. The decision-making mechanism for the architectural design is governed by a Monte Carlo tree search procedure which simulates network behavior and allows to compare several candidate architecture changes to choose the best one. The proposed method was validated using both visual and time series datasets, demonstrating its particular effectiveness in multivariate time series classification. This is attributed to the architecture's ability to adapt dynamically, allowing independent modifications for each time series. The approach is supplemented by Python source code for reproducibility. Experimental evaluations in visual pattern and multivariate time series classification tasks revealed highly promising performance, underscoring the method's robustness and adaptability. | [
{
"version": "v1",
"created": "Mon, 23 Jun 2025 19:52:01 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Świderski",
"Szymon",
""
],
[
"Jastrzębska",
"Agnieszka",
""
]
] |
2507.01044 | Vivek Borkar | Vivek Borkar and Parthe Pandit | Asymptotic convexity of wide and shallow neural networks | 5 pages | null | null | null | stat.ML cs.LG math.PR | http://creativecommons.org/licenses/by/4.0/ | For a simple model of shallow and wide neural networks, we show that the epigraph of its input-output map as a function of the network parameters approximates epigraph of a. convex function in a precise sense. This leads to a plausible explanation of their observed good performance. | [
{
"version": "v1",
"created": "Mon, 23 Jun 2025 20:26:14 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Borkar",
"Vivek",
""
],
[
"Pandit",
"Parthe",
""
]
] |
2507.01045 | Xiao Gu | Xiao Gu, Wei Tang, Jinpei Han, Veer Sangha, Fenglin Liu, Shreyank N Gowda, Antonio H. Ribeiro, Patrick Schwab, Kim Branson, Lei Clifton, Antonio Luiz P. Ribeiro, Zhangdaihong Liu, and David A. Clifton | Sensing Cardiac Health Across Scenarios and Devices: A Multi-Modal Foundation Model Pretrained on Heterogeneous Data from 1.7 Million Individuals | null | null | null | null | cs.LG cs.AI eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cardiac biosignals, such as electrocardiograms (ECG) and photoplethysmograms (PPG), are of paramount importance for the diagnosis, prevention, and management of cardiovascular diseases, and have been extensively used in a variety of clinical tasks. Conventional deep learning approaches for analyzing these signals typically rely on homogeneous datasets and static bespoke models, limiting their robustness and generalizability across diverse clinical settings and acquisition protocols. In this study, we present a cardiac sensing foundation model (CSFM) that leverages advanced transformer architectures and a generative, masked pretraining strategy to learn unified representations from vast, heterogeneous health records. Our model is pretrained on an innovative multi-modal integration of data from multiple large-scale datasets (including MIMIC-III-WDB, MIMIC-IV-ECG, and CODE), comprising cardiac signals and the corresponding clinical or machine-generated text reports from approximately 1.7 million individuals. We demonstrate that the embeddings derived from our CSFM not only serve as effective feature extractors across diverse cardiac sensing scenarios, but also enable seamless transfer learning across varying input configurations and sensor modalities. Extensive evaluations across diagnostic tasks, demographic information recognition, vital sign measurement, clinical outcome prediction, and ECG question answering reveal that CSFM consistently outperforms traditional one-modal-one-task approaches. Notably, CSFM exhibits robust performance across multiple ECG lead configurations from standard 12-lead systems to single-lead setups, and in scenarios where only ECG, only PPG, or a combination thereof is available. These findings highlight the potential of CSFM as a versatile and scalable solution, for comprehensive cardiac monitoring. | [
{
"version": "v1",
"created": "Mon, 23 Jun 2025 20:58:12 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gu",
"Xiao",
""
],
[
"Tang",
"Wei",
""
],
[
"Han",
"Jinpei",
""
],
[
"Sangha",
"Veer",
""
],
[
"Liu",
"Fenglin",
""
],
[
"Gowda",
"Shreyank N",
""
],
[
"Ribeiro",
"Antonio H.",
""
],
[
"Schwab",
"Patrick",
""
],
[
"Branson",
"Kim",
""
],
[
"Clifton",
"Lei",
""
],
[
"Ribeiro",
"Antonio Luiz P.",
""
],
[
"Liu",
"Zhangdaihong",
""
],
[
"Clifton",
"David A.",
""
]
] |
2507.01046 | Christian Parkinson | Christian Parkinson, Weinan Wang | A Compartmental Model for Epidemiology with Human Behavior and Stochastic Effects | 28 pages, 5 figures | null | null | null | math.DS q-bio.PE | http://creativecommons.org/licenses/by/4.0/ | We propose a compartmental model for epidemiology wherein the population is split into groups with either comply or refuse to comply with protocols designed to slow the spread of a disease. Parallel to the disease spread, we assume that noncompliance with protocols spreads as a social contagion. We begin by deriving the reproductive ratio for a deterministic version of the model, and use this to fully characterize the local stability of disease free equilibrium points. We then append the deterministic model with stochastic effects, specifically assuming that the transmission rate of the disease and the transmission rate of the social contagion are uncertain. We prove global existence and nonnegativity for our stochastic model. Then using suitably constructed stochastic Lyapunov functions, we analyze the behavior of the stochastic system with respect to certain disease free states. We demonstrate all of our results with numerical simulations. | [
{
"version": "v1",
"created": "Tue, 24 Jun 2025 20:51:47 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Parkinson",
"Christian",
""
],
[
"Wang",
"Weinan",
""
]
] |
2507.01047 | Majdi Radaideh | Logan A. Burnett, Umme Mahbuba Nabila, Majdi I. Radaideh | Variational Digital Twins | 33 pages, 14 figures, and 7 tables | null | null | null | cs.LG cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | While digital twins (DT) hold promise for providing real-time insights into complex energy assets, much of the current literature either does not offer a clear framework for information exchange between the model and the asset, lacks key features needed for real-time implementation, or gives limited attention to model uncertainty. Here, we aim to solve these gaps by proposing a variational digital twin (VDT) framework that augments standard neural architectures with a single Bayesian output layer. This lightweight addition, along with a novel VDT updating algorithm, lets a twin update in seconds on commodity GPUs while producing calibrated uncertainty bounds that can inform experiment design, control algorithms, and model reliability. The VDT is evaluated on four energy-sector problems. For critical-heat-flux prediction, uncertainty-driven active learning reaches R2 = 0.98 using 47 % fewer experiments and one-third the training time of random sampling. A three-year renewable-generation twin maintains R2 > 0.95 for solar output and curbs error growth for volatile wind forecasts via monthly updates that process only one month of data at a time. A nuclear reactor transient cooldown twin reconstructs thermocouple signals with R2 > 0.99 and preserves accuracy after 50 % sensor loss, demonstrating robustness to degraded instrumentation. Finally, a physics-informed Li-ion battery twin, retrained after every ten discharges, lowers voltage mean-squared error by an order of magnitude relative to the best static model while adapting its credible intervals as the cell approaches end-of-life. These results demonstrate that combining modest Bayesian augmentation with efficient update schemes turns conventional surrogates into uncertainty-aware, data-efficient, and computationally tractable DTs, paving the way for dependable models across industrial and scientific energy systems. | [
{
"version": "v1",
"created": "Wed, 25 Jun 2025 02:05:30 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Burnett",
"Logan A.",
""
],
[
"Nabila",
"Umme Mahbuba",
""
],
[
"Radaideh",
"Majdi I.",
""
]
] |
2507.01048 | Ricardo Vargas | Ricardo Emanuel Vaz Vargas, Afr\^anio Jos\'e de Melo Junior, Celso Jos\'e Munaro, Cl\'audio Benevenuto de Campos Lima, Eduardo Toledo de Lima Junior, Felipe Muntzberg Barrocas, Fl\'avio Miguel Varej\~ao, Guilherme Fidelis Peixer, Igor de Melo Nery Oliveira, Jader Riso Barbosa Jr., Jaime Andr\'es Lozano Cadena, Jean Carlos Dias de Ara\'ujo, Jo\~ao Neuenschwander Escosteguy Carneiro, Lucas Gouveia Omena Lopes, Lucas Pereira de Gouveia, Mateus de Araujo Fernandes, Matheus Lima Scramignon, Patrick Marques Ciarelli, Rodrigo Castello Branco, Rog\'erio Leite Alves Pinto | 3W Dataset 2.0.0: a realistic and public dataset with rare undesirable real events in oil wells | 21 pages, 10 figures, and 7 tables | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In the oil industry, undesirable events in oil wells can cause economic losses, environmental accidents, and human casualties. Solutions based on Artificial Intelligence and Machine Learning for Early Detection of such events have proven valuable for diverse applications across industries. In 2019, recognizing the importance and the lack of public datasets related to undesirable events in oil wells, Petrobras developed and publicly released the first version of the 3W Dataset, which is essentially a set of Multivariate Time Series labeled by experts. Since then, the 3W Dataset has been developed collaboratively and has become a foundational reference for numerous works in the field. This data article describes the current publicly available version of the 3W Dataset, which contains structural modifications and additional labeled data. The detailed description provided encourages and supports the 3W community and new 3W users to improve previous published results and to develop new robust methodologies, digital products and services capable of detecting undesirable events in oil wells with enough anticipation to enable corrective or mitigating actions. | [
{
"version": "v1",
"created": "Wed, 25 Jun 2025 23:42:32 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Vargas",
"Ricardo Emanuel Vaz",
""
],
[
"Junior",
"Afrânio José de Melo",
""
],
[
"Munaro",
"Celso José",
""
],
[
"Lima",
"Cláudio Benevenuto de Campos",
""
],
[
"Junior",
"Eduardo Toledo de Lima",
""
],
[
"Barrocas",
"Felipe Muntzberg",
""
],
[
"Varejão",
"Flávio Miguel",
""
],
[
"Peixer",
"Guilherme Fidelis",
""
],
[
"Oliveira",
"Igor de Melo Nery",
""
],
[
"Barbosa",
"Jader Riso",
"Jr."
],
[
"Cadena",
"Jaime Andrés Lozano",
""
],
[
"de Araújo",
"Jean Carlos Dias",
""
],
[
"Carneiro",
"João Neuenschwander Escosteguy",
""
],
[
"Lopes",
"Lucas Gouveia Omena",
""
],
[
"de Gouveia",
"Lucas Pereira",
""
],
[
"Fernandes",
"Mateus de Araujo",
""
],
[
"Scramignon",
"Matheus Lima",
""
],
[
"Ciarelli",
"Patrick Marques",
""
],
[
"Branco",
"Rodrigo Castello",
""
],
[
"Pinto",
"Rogério Leite Alves",
""
]
] |
2507.01049 | Pranav Jadhav | Pranav Jadhav | Cohort Retrieval using Dense Passage Retrieval | null | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Patient cohort retrieval is a pivotal task in medical research and clinical practice, enabling the identification of specific patient groups from extensive electronic health records (EHRs). In this work, we address the challenge of cohort retrieval in the echocardiography domain by applying Dense Passage Retrieval (DPR), a prominent methodology in semantic search. We propose a systematic approach to transform an echocardiographic EHR dataset of unstructured nature into a Query-Passage dataset, framing the problem as a Cohort Retrieval task. Additionally, we design and implement evaluation metrics inspired by real-world clinical scenarios to rigorously test the models across diverse retrieval tasks. Furthermore, we present a custom-trained DPR embedding model that demonstrates superior performance compared to traditional and off-the-shelf SOTA methods.To our knowledge, this is the first work to apply DPR for patient cohort retrieval in the echocardiography domain, establishing a framework that can be adapted to other medical domains. | [
{
"version": "v1",
"created": "Thu, 26 Jun 2025 18:11:25 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Jadhav",
"Pranav",
""
]
] |
2507.01050 | Jing Yu | Jing Yu, Yibo Zhao, Jiapeng Zhu, Wenming Shao, Bo Pang, Zhao Zhang, Xiang Li | Text Detoxification: Data Efficiency, Semantic Preservation and Model Generalization | null | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | The widespread dissemination of toxic content on social media poses a serious threat to both online environments and public discourse, highlighting the urgent need for detoxification methods that effectively remove toxicity while preserving the original semantics. However, existing approaches often struggle to simultaneously achieve strong detoxification performance, semantic preservation, and robustness to out-of-distribution data. Moreover, they typically rely on costly, manually annotated parallel corpora while showing poor data efficiency. To address these challenges, we propose a two-stage training framework that jointly optimizes for data efficiency, semantic preservation, and model generalization. We first perform supervised fine-tuning on a small set of high-quality, filtered parallel data to establish a strong initialization. Then, we leverage unlabeled toxic inputs and a custom-designed reward model to train the LLM using Group Relative Policy Optimization. Experimental results demonstrate that our method effectively mitigates the trade-offs faced by previous work, achieving state-of-the-art performance with improved generalization and significantly reduced dependence on annotated data. Our code is available at: https://anonymous.4open.science/r/Detoxification-of-Text-725F/ | [
{
"version": "v1",
"created": "Mon, 23 Jun 2025 05:48:10 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Yu",
"Jing",
""
],
[
"Zhao",
"Yibo",
""
],
[
"Zhu",
"Jiapeng",
""
],
[
"Shao",
"Wenming",
""
],
[
"Pang",
"Bo",
""
],
[
"Zhang",
"Zhao",
""
],
[
"Li",
"Xiang",
""
]
] |
2507.01051 | Giada Pistilli | Giada Pistilli, Bruna Trevelin | Can AI be Consentful? | null | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | The evolution of generative AI systems exposes the challenges of traditional legal and ethical frameworks built around consent. This chapter examines how the conventional notion of consent, while fundamental to data protection and privacy rights, proves insufficient in addressing the implications of AI-generated content derived from personal data. Through legal and ethical analysis, we show that while individuals can consent to the initial use of their data for AI training, they cannot meaningfully consent to the numerous potential outputs their data might enable or the extent to which the output is used or distributed. We identify three fundamental challenges: the scope problem, the temporality problem, and the autonomy trap, which collectively create what we term a ''consent gap'' in AI systems and their surrounding ecosystem. We argue that current legal frameworks inadequately address these emerging challenges, particularly regarding individual autonomy, identity rights, and social responsibility, especially in cases where AI-generated content creates new forms of personal representation beyond the scope of the original consent. By examining how these consent limitations intersect with broader principles of responsible AI (including fairness, transparency, accountability, and autonomy) we demonstrate the need to evolve ethical and legal approaches to consent. | [
{
"version": "v1",
"created": "Fri, 27 Jun 2025 15:32:16 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Pistilli",
"Giada",
""
],
[
"Trevelin",
"Bruna",
""
]
] |
2507.01052 | Ahmed Farooq | Ahmed Farooq | Long-Sequence Memory with Temporal Kernels and Dense Hopfield Functionals | null | null | null | null | cs.LG cs.AI cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study we introduce a novel energy functional for long-sequence memory, building upon the framework of dense Hopfield networks which achieves exponential storage capacity through higher-order interactions. Building upon earlier work on long-sequence Hopfield memory models, we propose a temporal kernal $K(m, k)$ to incorporate temporal dependencies, enabling efficient sequential retrieval of patterns over extended sequences. We demonstrate the successful application of this technique for the storage and sequential retrieval of movies frames which are well suited for this because of the high dimensional vectors that make up each frame creating enough variation between even sequential frames in the high dimensional space. The technique has applications in modern transformer architectures, including efficient long-sequence modeling, memory augmentation, improved attention with temporal bias, and enhanced handling of long-term dependencies in time-series data. Our model offers a promising approach to address the limitations of transformers in long-context tasks, with potential implications for natural language processing, forecasting, and beyond. | [
{
"version": "v1",
"created": "Fri, 27 Jun 2025 15:57:58 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Farooq",
"Ahmed",
""
]
] |
2507.01053 | Pedro Moreira | Rafi Al Attrach, Pedro Moreira, Rajna Fani, Renato Umeton, Leo Anthony Celi | Conversational LLMs Simplify Secure Clinical Data Access, Understanding, and Analysis | 10 pages, 4 figures | null | null | null | cs.IR cs.AI cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As ever-larger clinical datasets become available, they have the potential to unlock unprecedented opportunities for medical research. Foremost among them is Medical Information Mart for Intensive Care (MIMIC-IV), the world's largest open-source EHR database. However, the inherent complexity of these datasets, particularly the need for sophisticated querying skills and the need to understand the underlying clinical settings, often presents a significant barrier to their effective use. M3 lowers the technical barrier to understanding and querying MIMIC-IV data. With a single command it retrieves MIMIC-IV from PhysioNet, launches a local SQLite instance (or hooks into the hosted BigQuery), and-via the Model Context Protocol (MCP)-lets researchers converse with the database in plain English. Ask a clinical question in natural language; M3 uses a language model to translate it into SQL, executes the query against the MIMIC-IV dataset, and returns structured results alongside the underlying query for verifiability and reproducibility. Demonstrations show that minutes of dialogue with M3 yield the kind of nuanced cohort analyses that once demanded hours of handcrafted SQL and relied on understanding the complexities of clinical workflows. By simplifying access, M3 invites the broader research community to mine clinical critical-care data and accelerates the translation of raw records into actionable insight. | [
{
"version": "v1",
"created": "Fri, 27 Jun 2025 16:24:17 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Attrach",
"Rafi Al",
""
],
[
"Moreira",
"Pedro",
""
],
[
"Fani",
"Rajna",
""
],
[
"Umeton",
"Renato",
""
],
[
"Celi",
"Leo Anthony",
""
]
] |
2507.01054 | Jithendaraa Subramanian | Jithendaraa Subramanian, Linda Hung, Daniel Schweigert, Santosh Suram, Weike Ye | XxaCT-NN: Structure Agnostic Multimodal Learning for Materials Science | 10 pages, 6 figures | null | null | null | cs.LG cond-mat.mtrl-sci cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Recent advances in materials discovery have been driven by structure-based models, particularly those using crystal graphs. While effective for computational datasets, these models are impractical for real-world applications where atomic structures are often unknown or difficult to obtain. We propose a scalable multimodal framework that learns directly from elemental composition and X-ray diffraction (XRD) -- two of the more available modalities in experimental workflows without requiring crystal structure input. Our architecture integrates modality-specific encoders with a cross-attention fusion module and is trained on the 5-million-sample Alexandria dataset. We present masked XRD modeling (MXM), and apply MXM and contrastive alignment as self-supervised pretraining strategies. Pretraining yields faster convergence (up to 4.2x speedup) and improves both accuracy and representation quality. We further demonstrate that multimodal performance scales more favorably with dataset size than unimodal baselines, with gains compounding at larger data regimes. Our results establish a path toward structure-free, experimentally grounded foundation models for materials science. | [
{
"version": "v1",
"created": "Fri, 27 Jun 2025 21:45:56 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Subramanian",
"Jithendaraa",
""
],
[
"Hung",
"Linda",
""
],
[
"Schweigert",
"Daniel",
""
],
[
"Suram",
"Santosh",
""
],
[
"Ye",
"Weike",
""
]
] |
2507.01055 | Hao Yang | Hao Yang, Xinlong Liang, Zhang Li, Yue Sun, Zheyu Hu, Xinghe Xie, Behdad Dashtbozorg, Jincheng Huang, Shiwei Zhu, Luyi Han, Jiong Zhang, Shanshan Wang, Ritse Mann, Qifeng Yu, Tao Tan | Prompt Mechanisms in Medical Imaging: A Comprehensive Survey | null | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | Deep learning offers transformative potential in medical imaging, yet its clinical adoption is frequently hampered by challenges such as data scarcity, distribution shifts, and the need for robust task generalization. Prompt-based methodologies have emerged as a pivotal strategy to guide deep learning models, providing flexible, domain-specific adaptations that significantly enhance model performance and adaptability without extensive retraining. This systematic review critically examines the burgeoning landscape of prompt engineering in medical imaging. We dissect diverse prompt modalities, including textual instructions, visual prompts, and learnable embeddings, and analyze their integration for core tasks such as image generation, segmentation, and classification. Our synthesis reveals how these mechanisms improve task-specific outcomes by enhancing accuracy, robustness, and data efficiency and reducing reliance on manual feature engineering while fostering greater model interpretability by making the model's guidance explicit. Despite substantial advancements, we identify persistent challenges, particularly in prompt design optimization, data heterogeneity, and ensuring scalability for clinical deployment. Finally, this review outlines promising future trajectories, including advanced multimodal prompting and robust clinical integration, underscoring the critical role of prompt-driven AI in accelerating the revolution of diagnostics and personalized treatment planning in medicine. | [
{
"version": "v1",
"created": "Sat, 28 Jun 2025 03:06:25 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Yang",
"Hao",
""
],
[
"Liang",
"Xinlong",
""
],
[
"Li",
"Zhang",
""
],
[
"Sun",
"Yue",
""
],
[
"Hu",
"Zheyu",
""
],
[
"Xie",
"Xinghe",
""
],
[
"Dashtbozorg",
"Behdad",
""
],
[
"Huang",
"Jincheng",
""
],
[
"Zhu",
"Shiwei",
""
],
[
"Han",
"Luyi",
""
],
[
"Zhang",
"Jiong",
""
],
[
"Wang",
"Shanshan",
""
],
[
"Mann",
"Ritse",
""
],
[
"Yu",
"Qifeng",
""
],
[
"Tan",
"Tao",
""
]
] |
2507.01056 | Lu Gao | Lidan Peng, Lu Gao, Feng Hong, Jingran Sun | Evaluating Pavement Deterioration Rates Due to Flooding Events Using Explainable AI | null | Buildings 15.9 (2025): 1452 | 10.3390/buildings15091452 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Flooding can damage pavement infrastructure significantly, causing both immediate and long-term structural and functional issues. This research investigates how flooding events affect pavement deterioration, specifically focusing on measuring pavement roughness by the International Roughness Index (IRI). To quantify these effects, we utilized 20 years of pavement condition data from TxDOT's PMIS database, which is integrated with flood event data, including duration and spatial extent. Statistical analyses were performed to compare IRI values before and after flooding and to calculate the deterioration rates influenced by flood exposure. Moreover, we applied Explainable Artificial Intelligence (XAI) techniques, such as SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME), to assess the impact of flooding on pavement performance. The results demonstrate that flood-affected pavements experience a more rapid increase in roughness compared to non-flooded sections. These findings emphasize the need for proactive flood mitigation strategies, including improved drainage systems, flood-resistant materials, and preventative maintenance, to enhance pavement resilience in vulnerable regions. | [
{
"version": "v1",
"created": "Sat, 28 Jun 2025 04:13:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Peng",
"Lidan",
""
],
[
"Gao",
"Lu",
""
],
[
"Hong",
"Feng",
""
],
[
"Sun",
"Jingran",
""
]
] |
2507.01057 | Lushun Fan | Lushun Fan, Yuqin Xia, Jun Li, Karl Jenkins | Loop2Net: Data-Driven Generation and Optimization of Airfoil CFD Meshes from Sparse Boundary Coordinates | null | null | null | null | cs.LG physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, an innovative intelligent optimization system for mesh quality is proposed, which is based on a deep convolutional neural network architecture, to achieve mesh generation and optimization. The core of the study is the Loop2Net generator and loss function, it predicts the mesh based on the given wing coordinates. And the model's performance is continuously optimised by two key loss functions during the training. Then discipline by adding penalties, the goal of mesh generation was finally reached. | [
{
"version": "v1",
"created": "Sat, 28 Jun 2025 13:27:07 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Fan",
"Lushun",
""
],
[
"Xia",
"Yuqin",
""
],
[
"Li",
"Jun",
""
],
[
"Jenkins",
"Karl",
""
]
] |
2507.01058 | Puspendu Banerjee | Puspendu Banerjee, Aritra Mazumdar, Wazib Ansar, Saptarsi Goswami and Amlan Chakrabarti | A Data Science Approach to Calcutta High Court Judgments: An Efficient LLM and RAG-powered Framework for Summarization and Similar Cases Retrieval | 12 pages, 6 figures | null | null | null | cs.IR cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | The judiciary, as one of democracy's three pillars, is dealing with a rising amount of legal issues, needing careful use of judicial resources. This research presents a complex framework that leverages Data Science methodologies, notably Large Language Models (LLM) and Retrieval-Augmented Generation (RAG) techniques, to improve the efficiency of analyzing Calcutta High Court verdicts. Our framework focuses on two key aspects: first, the creation of a robust summarization mechanism that distills complex legal texts into concise and coherent summaries; and second, the development of an intelligent system for retrieving similar cases, which will assist legal professionals in research and decision making. By fine-tuning the Pegasus model using case head note summaries, we achieve significant improvements in the summarization of legal cases. Our two-step summarizing technique preserves crucial legal contexts, allowing for the production of a comprehensive vector database for RAG. The RAG-powered framework efficiently retrieves similar cases in response to user queries, offering thorough overviews and summaries. This technique not only improves legal research efficiency, but it also helps legal professionals and students easily acquire and grasp key legal information, benefiting the overall legal scenario. | [
{
"version": "v1",
"created": "Sat, 28 Jun 2025 20:24:34 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Banerjee",
"Puspendu",
""
],
[
"Mazumdar",
"Aritra",
""
],
[
"Ansar",
"Wazib",
""
],
[
"Goswami",
"Saptarsi",
""
],
[
"Chakrabarti",
"Amlan",
""
]
] |
2507.01059 | Xiangbo Gao | Xiangbo Gao, Keshu Wu, Hao Zhang, Kexin Tian, Yang Zhou, Zhengzhong Tu | Automated Vehicles Should be Connected with Natural Language | null | null | null | null | cs.MA cs.AI cs.CL cs.CV cs.RO | http://creativecommons.org/licenses/by/4.0/ | Multi-agent collaborative driving promises improvements in traffic safety and efficiency through collective perception and decision making. However, existing communication media -- including raw sensor data, neural network features, and perception results -- suffer limitations in bandwidth efficiency, information completeness, and agent interoperability. Moreover, traditional approaches have largely ignored decision-level fusion, neglecting critical dimensions of collaborative driving. In this paper we argue that addressing these challenges requires a transition from purely perception-oriented data exchanges to explicit intent and reasoning communication using natural language. Natural language balances semantic density and communication bandwidth, adapts flexibly to real-time conditions, and bridges heterogeneous agent platforms. By enabling the direct communication of intentions, rationales, and decisions, it transforms collaborative driving from reactive perception-data sharing into proactive coordination, advancing safety, efficiency, and transparency in intelligent transportation systems. | [
{
"version": "v1",
"created": "Sun, 29 Jun 2025 16:41:19 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gao",
"Xiangbo",
""
],
[
"Wu",
"Keshu",
""
],
[
"Zhang",
"Hao",
""
],
[
"Tian",
"Kexin",
""
],
[
"Zhou",
"Yang",
""
],
[
"Tu",
"Zhengzhong",
""
]
] |
2507.01060 | Kang Liu | Kang Liu | Optimizing Conversational Product Recommendation via Reinforcement Learning | null | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a reinforcement learning-based approach to optimize conversational strategies for product recommendation across diverse industries. As organizations increasingly adopt intelligent agents to support sales and service operations, the effectiveness of a conversation hinges not only on what is recommended but how and when recommendations are delivered. We explore a methodology where agentic systems learn optimal dialogue policies through feedback-driven reinforcement learning. By mining aggregate behavioral patterns and conversion outcomes, our approach enables agents to refine talk tracks that drive higher engagement and product uptake, while adhering to contextual and regulatory constraints. We outline the conceptual framework, highlight key innovations, and discuss the implications for scalable, personalized recommendation in enterprise environments. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 00:59:58 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Liu",
"Kang",
""
]
] |
2507.01061 | Yulei Ye | Jingjing Qu, Kejia Hu, Jun Zhu, Wenhao Li, Teng Wang, Zhiyun Chen, Yulei Ye, Chaochao Lu, Aimin Zhou, Xiangfeng Wang, James Evan | Epitome: Pioneering an Experimental Platform for AI-Social Science Integration | 18 pages, 5figures | null | null | null | cs.CY cs.AI cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The integration of Large Language Models (LLMs) into social science experiments represents a transformative approach to understanding human-AI interactions and their societal impacts. We introduce Epitome, the world's first open experimental platform dedicated to the deep integration of artificial intelligence and social science. Rooted in theoretical foundations from management, communication studies, sociology, psychology, and ethics, Epitome focuses on the interactive impacts of AI on individuals, organizations, and society during its real-world deployment. It constructs a theoretical support system through cross-disciplinary experiments. The platform offers a one-stop comprehensive experimental solution spanning "foundation models-complex application development-user feedback" through seven core modules, while embedding the classical "control-comparison-comparative causal logic" of social science experiments into multilevel human-computer interaction environments, including dialogues, group chats, and multi-agent virtual scenarios. With its canvas-style, user-friendly interface, Epitome enables researchers to easily design and run complex experimental scenarios, facilitating systematic investigations into the social impacts of AI and exploration of integrated solutions.To demonstrate its capabilities, we replicated three seminal social science experiments involving LLMs, showcasing Epitome's potential to streamline complex experimental designs and produce robust results, suitable for publishing in the top selective journals. Our findings highlight the platform's utility in enhancing the efficiency and quality of human-AI interactions, providing valuable insights into the societal implications of AI technologies. Epitome thus offers a powerful tool for advancing interdisciplinary research at the intersection of AI and social science, with potential applications in policy-making, ... | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 09:06:16 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Qu",
"Jingjing",
""
],
[
"Hu",
"Kejia",
""
],
[
"Zhu",
"Jun",
""
],
[
"Li",
"Wenhao",
""
],
[
"Wang",
"Teng",
""
],
[
"Chen",
"Zhiyun",
""
],
[
"Ye",
"Yulei",
""
],
[
"Lu",
"Chaochao",
""
],
[
"Zhou",
"Aimin",
""
],
[
"Wang",
"Xiangfeng",
""
],
[
"Evan",
"James",
""
]
] |
2507.01062 | Seyma Yaman Kayadibi | Seyma Yaman Kayadibi | Quantifying Student Success with Generative AI: A Monte Carlo Simulation Informed by Systematic Review | 35 pages, 4 figures. All figures are image-based: one Python code screenshot, one regression model output, one success score distribution chart, and one PRISMA diagram. This article presents a standalone segment from the author's master's thesis at Victoria University | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | The exponential development of generative artificial intelligence (GenAI) technologies like ChatGPT has raised increasing curiosity about their use in higher education, specifically with respect to how students view them, make use of them, and the implications for learning outcomes. This paper employs a hybrid methodological approach involving a systematic literature review and simulation-based modeling to explore student perceptions of GenAI use in the context of higher education. A total of nineteen empirical articles from 2023 through 2025 were selected from the PRISMA-based search targeting the Scopus database. Synthesis of emerging patterns from the literature was achieved by thematic categorization. Six of these had enough quantitative information, i.e., item-level means and standard deviations, to permit probabilistic modeling. One dataset, from the resulting subset, was itself selected as a representative case with which to illustrate inverse-variance weighting by Monte Carlo simulation, by virtue of its well-designed Likert scale format and thematic alignment with the use of computing systems by the researcher.
The simulation provided a composite "Success Score" forecasting the strength of the relationship between student perceptions and learning achievements. Findings reveal that attitude factors concerned with usability and real-world usefulness are significantly better predictors of positive learning achievement than affective or trust-based factors. Such an interdisciplinary perspective provides a unique means of linking thematic results with predictive modelling, resonating with longstanding controversies about the proper use of GenAI tools within the university. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 09:50:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kayadibi",
"Seyma Yaman",
""
]
] |
2507.01063 | Madhav Kotecha | Madhav Kotecha | FAIR-MATCH: A Multi-Objective Framework for Bias Mitigation in Reciprocal Dating Recommendations | null | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Online dating platforms have fundamentally transformed the formation of romantic relationships, with millions of users worldwide relying on algorithmic matching systems to find compatible partners. However, current recommendation systems in dating applications suffer from significant algorithmic deficiencies, including but not limited to popularity bias, filter bubble effects, and inadequate reciprocity modeling that limit effectiveness and introduce harmful biases. This research integrates foundational work with recent empirical findings to deliver a detailed analysis of dating app recommendation systems, highlighting key issues and suggesting research-backed solutions. Through analysis of reciprocal recommendation frameworks, fairness evaluation metrics, and industry implementations, we demonstrate that current systems achieve modest performance with collaborative filtering reaching 25.1\% while reciprocal methods achieve 28.7\%. Our proposed mathematical framework addresses these limitations through enhanced similarity measures, multi-objective optimization, and fairness-aware algorithms that maintain competitive accuracy while improving demographic representation to reduce algorithmic bias. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 10:36:57 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kotecha",
"Madhav",
""
]
] |
2507.01064 | Riccardo Finotello | Riccardo Finotello, Vincent Lahoche, Dine Ousmane Samary | Functional Renormalization for Signal Detection: Dimensional Analysis and Dimensional Phase Transition for Nearly Continuous Spectra Effective Field Theory | 34 pages | null | null | COMETA-2025-23 | physics.data-an cond-mat.stat-mech cs.IT hep-th math.IT stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Signal detection is one of the main challenges of data science. According to the nature of the data, the presence of noise may corrupt measurements and hinder the discovery of significant patterns. A wide range of techniques aiming at extracting the relevant degrees of freedom from data has been thus developed over the years. However, signal detection in almost continuous spectra, for small signal-to-noise ratios, remains a known difficult issue. This paper develops over recent advancements proposing to tackle this issue by analysing the properties of the underlying effective field theory arising as a sort of maximal entropy distribution in the vicinity of universal random matrix distributions. Nearly continuous spectra provide an intrinsic and non-conventional scaling law for field and couplings, the scaling dimensions depending on the energy scale. The coarse-graining over small eigenvalues of the empirical spectrum defines a specific renormalization group, whose characteristics change when the collective behaviour of "informational" modes become significant, that is, stronger than the intrinsic fluctuations of noise. This paper pursues three different goals. First, we propose to quantify the real effects of fluctuations relative to what can be called "signal", while improving the robustness of the results obtained in our previous work. Second, we show that quantitative changes in the presence of a signal result in a counterintuitive modification of the distribution of eigenvectors. Finally, we propose a method for estimating the number of noise components and define a limit of detection in a general nearly continuous spectrum using the renormalization group. The main statements of this paper are essentially numeric, and their reproducibility can be checked using the associated code. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 18:00:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Finotello",
"Riccardo",
""
],
[
"Lahoche",
"Vincent",
""
],
[
"Samary",
"Dine Ousmane",
""
]
] |
2507.01065 | Daniel Russo | Christiaan Verwijs, Evelien Acun-Roos, Daniel Russo | Is It Safe To Learn And Share? On Psychological Safety and Social Learning in (Agile) Communities of Practice | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As hybrid, distributed, and asynchronous work models become more prevalent, continuous learning in Agile Software Development (ASD) gains renewed importance. Communities of Practice (CoPs) are increasingly adopted to support social learning beyond formal education, often relying on virtual communication. Psychological safety, a prerequisite for effective learning, remains insufficiently understood in these settings. This mixed-methods study investigates psychological safety within Agile CoPs through survey data from 143 participants. Results indicate that psychological safety is significantly lower in online interactions compared to face-to-face settings. Moreover, low psychological safety reduces participants' intent to continue contributing and avoidance of interpersonal risk. No significant differences emerged based on gender, community seniority, or content creation activity. However, differences by role and age group suggest potential generational or role-related effects. Thematic analysis revealed exclusionary behavior, negative interaction patterns, and hostility as primary threats to psychological safety, often reinforced by tribalism and specific community dynamics. Suggested interventions include establishing explicit norms, structured facilitation, and active moderation. The findings were validated through member checking with 30 participants. This study provides a comparative perspective on interaction modalities and offers practical guidance for organizers seeking to cultivate inclusive, high-impact CoPs and similarly structured virtual or hybrid work environments. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 19:06:43 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Verwijs",
"Christiaan",
""
],
[
"Acun-Roos",
"Evelien",
""
],
[
"Russo",
"Daniel",
""
]
] |
2507.01066 | Jinghao Shi | Hanzhong Liang, Jinghao Shi, Xiang Shen, Zixuan Wang, Vera Wen, Ardalan Mehrani, Zhiqian Chen, Yifan Wu, Zhixin Zhang | Embedding-based Retrieval in Multimodal Content Moderation | Camera ready for SIGIR 2025 | null | 10.1145/3726302.3731945 | null | cs.IR cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video understanding plays a fundamental role for content moderation on short video platforms, enabling the detection of inappropriate content. While classification remains the dominant approach for content moderation, it often struggles in scenarios requiring rapid and cost-efficient responses, such as trend adaptation and urgent escalations. To address this issue, we introduce an Embedding-Based Retrieval (EBR) method designed to complement traditional classification approaches. We first leverage a Supervised Contrastive Learning (SCL) framework to train a suite of foundation embedding models, including both single-modal and multi-modal architectures. Our models demonstrate superior performance over established contrastive learning methods such as CLIP and MoCo. Building on these embedding models, we design and implement the embedding-based retrieval system that integrates embedding generation and video retrieval to enable efficient and effective trend handling. Comprehensive offline experiments on 25 diverse emerging trends show that EBR improves ROC-AUC from 0.85 to 0.99 and PR-AUC from 0.35 to 0.95. Further online experiments reveal that EBR increases action rates by 10.32% and reduces operational costs by over 80%, while also enhancing interpretability and flexibility compared to classification-based solutions. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 19:11:25 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Liang",
"Hanzhong",
""
],
[
"Shi",
"Jinghao",
""
],
[
"Shen",
"Xiang",
""
],
[
"Wang",
"Zixuan",
""
],
[
"Wen",
"Vera",
""
],
[
"Mehrani",
"Ardalan",
""
],
[
"Chen",
"Zhiqian",
""
],
[
"Wu",
"Yifan",
""
],
[
"Zhang",
"Zhixin",
""
]
] |
2507.01067 | Keun Soo Yim | Keun Soo Yim | Evaluation of a Foundational Model and Stochastic Models for Forecasting Sporadic or Spiky Production Outages of High-Performance Machine Learning Services | null | null | null | null | cs.LG cs.AI cs.DC cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | Time series forecasting models have diverse real world applications (e.g., from electricity metrics to software workload). Latest foundational models trained for time series forecasting show strengths (e.g., for long sequences and in zero-shot settings). However, foundational model was not yet used for forecasting rare, spiky events, i.e., a challenging target because those are a corner case of extreme events. In this paper, we optimize a state-of-the-art foundational model to forecast sporadic or spiky production outages of high-performance machine learning services powering billions of client devices. We evaluate the forecasting errors of the foundational model compared with classical stochastic forecasting models (e.g., moving average and autoregressive). The analysis helps us understand how each of the evaluated models performs for the sporadic or spiky events. For example, it identifies the key patterns in the target data that are well tracked by the foundational model vs. each of the stochastic models. We use the models with optimal parameters to estimate a year-long outage statistics of a particular root cause with less than 6% value errors. | [
{
"version": "v1",
"created": "Mon, 30 Jun 2025 23:59:12 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Yim",
"Keun Soo",
""
]
] |
2507.01068 | Biplov Paneru | Biplov Paneru | Prediction of Freezing of Gait in Parkinsons Disease using Explainable AI and Federated Deep Learning for Wearable Sensors | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | This study leverages an Inertial Measurement Unit (IMU) dataset to develop explainable AI methods for the early detection and prediction of Freezing of Gait (FOG), a common symptom in Parkinson's disease. Machine learning models, including CatBoost, XGBoost, and Extra Trees classifiers, are employed to accurately categorize FOG episodes based on relevant clinical features. A Stacking Ensemble model achieves superior performance, surpassing a hybrid bidirectional GRU model and reaching nearly 99% classification accuracy. SHAP interpretability analysis reveals that time (seconds) is the most influential factor in distinguishing gait patterns. Additionally, the proposed FOG prediction framework incorporates federated learning, where models are trained locally on individual devices and aggregated on a central server using a federated averaging approach, utilizing a hybrid Conv1D + LSTM architecture for enhanced predictive capability. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 01:22:12 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Paneru",
"Biplov",
""
]
] |
2507.01069 | Nishant A. Parikh | Nishant A. Parikh | Agentic AI in Product Management: A Co-Evolutionary Model | 41 pages, 2 figures | null | null | null | cs.CE cs.SE | http://creativecommons.org/licenses/by/4.0/ | This study explores agentic AI's transformative role in product management, proposing a conceptual co-evolutionary framework to guide its integration across the product lifecycle. Agentic AI, characterized by autonomy, goal-driven behavior, and multi-agent collaboration, redefines product managers (PMs) as orchestrators of socio-technical ecosystems. Using systems theory, co-evolutionary theory, and human-AI interaction theory, the framework maps agentic AI capabilities in discovery, scoping, business case development, development, testing, and launch. An integrative review of 70+ sources, including case studies from leading tech firms, highlights PMs' evolving roles in AI orchestration, supervision, and strategic alignment. Findings emphasize mutual adaptation between PMs and AI, requiring skills in AI literacy, governance, and systems thinking. Addressing gaps in traditional frameworks, this study provides a foundation for future research and practical implementation to ensure responsible, effective agentic AI integration in software organizations. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 02:32:32 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Parikh",
"Nishant A.",
""
]
] |
2507.01070 | Gopinath Guin | Gopinath Guin, Souvik Paul and Sunandan Gangopadhyay | Interacting Barrow holographic dark energy model in the presence of radiation and matter | 26 pages, 6 figures, This version will be updated soon, comments are welcome | null | null | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We have studied the effect of radiation in the interacting barrow holographic dark energy model for a non-flat universe. For both open and closed universes, we have obtained the evolution equation for the density parameters for dark energy, dark matter and radiation for a specific kind of interaction among the seven possible linear interactions; the analysis with the remaining six interactions will be updated soon in this manuscript. We have then numerically solved those coupled differential equations to show their behaviour with the redshift parameter. We have also found different epochs corresponding to dark energy-dark matter, dark energy-radiation and dark matter-radiation crossings. We have also obtained various observational constraints for our interacting Barrow holographic dark matter model using the Cosmic chronometer data and BAO data. In the upcoming version of the manuscript, we will add the observational constraints of our model using Pantheon data, also several calculations of other interaction scenarios will be added. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 07:51:56 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Guin",
"Gopinath",
""
],
[
"Paul",
"Souvik",
""
],
[
"Gangopadhyay",
"Sunandan",
""
]
] |
2507.01071 | Alessia Ferrara | Alessia Ferrara and Falk K. Wittel | Orthotropic Viscoelastic Creep in Cellular Scaffolds | null | null | null | null | q-bio.CB cond-mat.mtrl-sci physics.bio-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Recent measurements of Norway spruce have revealed stress-state-dependent normalized creep behavior, highlighting a gap in our fundamental understanding. This study examines whether the anisotropic response originates from the micro-structural, cellular nature of composite cell walls with varying tracheid types. Cell wall creep parameters are identified via surrogate-based inverse parameter identification, applied to hierarchical micro-mechanical and FEM models of increasing topological complexity up to the growth ring scale. Despite microstructural disorder, simulated creep curves converge toward a universal set of proportionality factors. The results indicate that directional creep behavior cannot be attributed solely to tissue-scale topology, and that realistic predictions require the inclusion of non-linear material responses at stress concentration sites. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 08:35:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ferrara",
"Alessia",
""
],
[
"Wittel",
"Falk K.",
""
]
] |
2507.01072 | Adeel Riaz | Adeel Riaz, Alexander Stangl, M\'onica Burriel, Michel Mermoux | In situ Study of Phase Transitions in La$_2$NiO$_{4+\delta}$ using Raman Spectroscopy | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | La$_2$NiO$_{4+\delta}$ has attracted increasing interest in recent years, both as oxygen electrode in solid oxide fuel cells and electrolysers due to its high electrochemical activity at intermediate to high temperatures, and as key component of memristive devices for neuromorphic computing, owing to its variable oxygen stoichiometry. The integration of La$_2$NiO$_{4+\delta}$ into devices operating at different temperatures and oxygen partial pressures requires knowledge of the effects of hyper-stoichiometry ($\delta$) on its crystalline structure. La$_2$NiO$_{4+\delta}$ is known to accommodate oxygen at interstitial sites allowing for large delta values, up to ~ 0.16. In addition, the O-doping - temperature phase diagram is known to be complex, exhibiting several phase transitions with increasing delta. Herein, we use Raman spectroscopy to monitor the effects of O-doping in the phase diagram and the various structures it contains. Throughout this work, we studied this material in its usual ceramic form, as well as in the form of thin films. Results are discussed in terms of phase transitions, chemical expansion, and some of the possible consequences of the low mean grain size inherent to such thin films. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 08:42:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Riaz",
"Adeel",
""
],
[
"Stangl",
"Alexander",
""
],
[
"Burriel",
"Mónica",
""
],
[
"Mermoux",
"Michel",
""
]
] |
2507.01073 | Dian Jin | Dian Jin | Rotational Sampling: A Plug-and-Play Encoder for Rotation-Invariant 3D Molecular GNNs | null | null | null | null | cs.LG q-bio.BM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph neural networks (GNNs) have achieved remarkable success in molecular property prediction. However, traditional graph representations struggle to effectively encode the inherent 3D spatial structures of molecules, as molecular orientations in 3D space introduce significant variability, severely limiting model generalization and robustness. Existing approaches primarily focus on rotation-invariant and rotation-equivariant methods. Invariant methods often rely heavily on prior knowledge and lack sufficient generalizability, while equivariant methods suffer from high computational costs. To address these limitations, this paper proposes a novel plug-and-play 3D encoding module leveraging rotational sampling. By computing the expectation over the SO(3) rotational group, the method naturally achieves approximate rotational invariance. Furthermore, by introducing a carefully designed post-alignment strategy, strict invariance can be achieved without compromising performance. Experimental evaluations on the QM9 and C10 Datasets demonstrate superior predictive accuracy, robustness, and generalization performance compared to existing methods. Moreover, the proposed approach maintains low computational complexity and enhanced interpretability, providing a promising direction for efficient and effective handling of 3D molecular information in drug discovery and material design. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 08:58:12 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Jin",
"Dian",
""
]
] |
2507.01074 | Fernando Garc\'ia-Torres | N. P. Garc\'ia-de-la-Puente, Roc\'io del Amor, Fernando Garc\'ia-Torres, Niels M{\o}ller Israelsen, Coraline Lapre, Christian Rosenberg Petersen, Ole Bang, Dominik Brouczek, Martin Schwentenwein, Kevin Neumann, Niels Benson, and Valery Naranjo | MID-INFRARED (MIR) OCT-based inspection in industry | Paper accepted at i-ESA 2024 12th International Conference on Interoperability for Enterprise Systems and Applications 6 pages, 2 figures, 2 tables | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper aims to evaluate mid-infrared (MIR) Optical Coherence Tomography (OCT) systems as a tool to penetrate different materials and detect sub-surface irregularities. This is useful for monitoring production processes, allowing Non-Destructive Inspection Techniques of great value to the industry. In this exploratory study, several acquisitions are made on composite and ceramics to know the capabilities of the system. In addition, it is assessed which preprocessing and AI-enhanced vision algorithms can be anomaly-detection methodologies capable of detecting abnormal zones in the analyzed objects. Limitations and criteria for the selection of optimal parameters will be discussed, as well as strengths and weaknesses will be highlighted. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 11:25:42 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"García-de-la-Puente",
"N. P.",
""
],
[
"del Amor",
"Rocío",
""
],
[
"García-Torres",
"Fernando",
""
],
[
"Israelsen",
"Niels Møller",
""
],
[
"Lapre",
"Coraline",
""
],
[
"Petersen",
"Christian Rosenberg",
""
],
[
"Bang",
"Ole",
""
],
[
"Brouczek",
"Dominik",
""
],
[
"Schwentenwein",
"Martin",
""
],
[
"Neumann",
"Kevin",
""
],
[
"Benson",
"Niels",
""
],
[
"Naranjo",
"Valery",
""
]
] |
2507.01075 | Gabriele Padovani | Gabriele Padovani, Valentine Anantharaj, Sandro Fiore | Provenance Tracking in Large-Scale Machine Learning Systems | null | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As the demand for large scale AI models continues to grow, the optimization of their training to balance computational efficiency, execution time, accuracy and energy consumption represents a critical multidimensional challenge. Achieving this balance requires not only innovative algorithmic techniques and hardware architectures but also comprehensive tools for monitoring, analyzing, and understanding the underlying processes involved in model training and deployment. Provenance data information about the origins, context, and transformations of data and processes has become a key component in this pursuit. By leveraging provenance, researchers and engineers can gain insights into resource usage patterns, identify inefficiencies, and ensure reproducibility and accountability in AI development workflows. For this reason, the question of how distributed resources can be optimally utilized to scale large AI models in an energy efficient manner is a fundamental one. To support this effort, we introduce the yProv4ML library, a tool designed to collect provenance data in JSON format, compliant with the W3C PROV and ProvML standards. yProv4ML focuses on flexibility and extensibility, and enables users to integrate additional data collection tools via plugins. The library is fully integrated with the yProv framework, allowing for higher level pairing in tasks run also through workflow management systems. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:10:02 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Padovani",
"Gabriele",
""
],
[
"Anantharaj",
"Valentine",
""
],
[
"Fiore",
"Sandro",
""
]
] |
2507.01076 | Vanja Stojanovi\'c | Vanja Stojanovi\'c, Bor Panger\v{s}i\v{c} | Empirical Analysis Of Heuristic and Approximation Algorithms for the The Mutual-Visibility Problem | null | null | null | null | cs.CG cs.AI cs.PF math.CO | http://creativecommons.org/licenses/by/4.0/ | The NP-complete mutual-visibility (MV) problem currently lacks empirical analysis on its practical behaviour despite theoretical studies. This paper addresses this gap by implementing and evaluating three distinct algorithms - a direct greedy heuristic, a hypergraph-based approximation, and a genetic algorithm - on diverse synthetic graph datasets, including those with analytically known $\mu(G)$ values and general graph models. Our results demonstrate that for smaller graphs, the algorithms consistently achieve MV set sizes aligning with theoretical bounds. However, for larger instances, achieved solution sizes notably diverge from theoretical limits; this, combined with the absence of tight bounds, complicates absolute quality assessment. Nevertheless, validation on known optimal graphs showed the Genetic Algorithm and other heuristics empirically performing best among tested methods. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:35:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Stojanović",
"Vanja",
""
],
[
"Pangeršič",
"Bor",
""
]
] |
2507.01077 | Bogdan-Mihai Bogdan | Bogdan Bogdan, Arina Cazacu, Laura Vasilie | Good Enough to Learn: LLM-based Anomaly Detection in ECU Logs without Reliable Labels | 6 pages, 7 figures, 4 tables, accepted to IEEE Intelligent Vehicles Symposium (IV) 2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomaly detection often relies on supervised or clustering approaches, with limited success in specialized domains like automotive communication systems where scalable solutions are essential. We propose a novel decoder-only Large Language Model (LLM) to detect anomalies in Electronic Control Unit (ECU) communication logs. Our approach addresses two key challenges: the lack of LLMs tailored for ECU communication and the complexity of inconsistent ground truth data. By learning from UDP communication logs, we formulate anomaly detection simply as identifying deviations in time from normal behavior. We introduce an entropy regularization technique that increases model's uncertainty in known anomalies while maintaining consistency in similar scenarios. Our solution offers three novelties: a decoder-only anomaly detection architecture, a way to handle inconsistent labeling, and an adaptable LLM for different ECU communication use cases. By leveraging the generative capabilities of decoder-only models, we present a new technique that addresses the high cost and error-prone nature of manual labeling through a more scalable system that is able to learn from a minimal set of examples, while improving detection accuracy in complex communication environments. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:56:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bogdan",
"Bogdan",
""
],
[
"Cazacu",
"Arina",
""
],
[
"Vasilie",
"Laura",
""
]
] |
2507.01078 | Gabriele Padovani | Gabriele Padovani, Valentine Anantharaj, Sandro Fiore | yProv4ML: Effortless Provenance Tracking for Machine Learning Systems | null | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of interest in large language models (LLMs) reflects their potential for flexibility and generalization, and attracted the attention of a diverse range of researchers. However, the advent of these techniques has also brought to light the lack of transparency and rigor with which development is pursued. In particular, the inability to determine the number of epochs and other hyperparameters in advance presents challenges in identifying the best model. To address this challenge, machine learning frameworks such as MLFlow can automate the collection of this type of information. However, these tools capture data using proprietary formats and pose little attention to lineage. This paper proposes yProv4ML, a framework to capture provenance information generated during machine learning processes in PROV-JSON format, with minimal code modifications. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 14:59:52 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Padovani",
"Gabriele",
""
],
[
"Anantharaj",
"Valentine",
""
],
[
"Fiore",
"Sandro",
""
]
] |
2507.01079 | Taehwan Park | Taehwan Park, Geonho Lee, Min-Soo Kim | MobileRAG: A Fast, Memory-Efficient, and Energy-Efficient Method for On-Device RAG | 14 pages | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Retrieval-Augmented Generation (RAG) has proven effective on server infrastructures, but its application on mobile devices is still underexplored due to limited memory and power resources. Existing vector search and RAG solutions largely assume abundant computation resources, making them impractical for on-device scenarios. In this paper, we propose MobileRAG, a fully on-device pipeline that overcomes these limitations by combining a mobile-friendly vector search algorithm, \textit{EcoVector}, with a lightweight \textit{Selective Content Reduction} (SCR) method. By partitioning and partially loading index data, EcoVector drastically reduces both memory footprint and CPU usage, while the SCR method filters out irrelevant text to diminish Language Model (LM) input size without degrading accuracy. Extensive experiments demonstrated that MobileRAG significantly outperforms conventional vector search and RAG methods in terms of latency, memory usage, and power consumption, while maintaining accuracy and enabling offline operation to safeguard privacy in resource-constrained environments. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 15:12:14 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Park",
"Taehwan",
""
],
[
"Lee",
"Geonho",
""
],
[
"Kim",
"Min-Soo",
""
]
] |
2507.01080 | Edouard Lansiaux | Edouard Lansiaux, Ramy Azzouz, Emmanuel Chazard, Am\'elie Vromant, Eric Wiel | Development and Comparative Evaluation of Three Artificial Intelligence Models (NLP, LLM, JEPA) for Predicting Triage in Emergency Departments: A 7-Month Retrospective Proof-of-Concept | 15 pages, 6 figures | null | null | null | cs.LG cs.PF | http://creativecommons.org/licenses/by/4.0/ | Triage errors, including undertriage and overtriage, are persistent challenges in emergency departments (EDs). With increasing patient influx and staff shortages, the integration of artificial intelligence (AI) into triage protocols has gained attention. This study compares the performance of three AI models [Natural Language Processing (NLP), Large Language Models (LLM), and Joint Embedding Predictive Architecture (JEPA)] in predicting triage outcomes against the FRENCH scale and clinical practice.We conducted a retrospective analysis of a prospectively recruited cohort gathering adult patient triage data over a 7-month period at the Roger Salengro Hospital ED (Lille, France). Three AI models were trained and validated : (1) TRIAGEMASTER (NLP), (2) URGENTIAPARSE (LLM), and (3) EMERGINET (JEPA). Data included demographic details, verbatim chief complaints, vital signs, and triage outcomes based on the FRENCH scale and GEMSA coding. The primary outcome was the concordance of AI-predicted triage level with the FRENCH gold-standard. It was assessed thanks to various indicators : F1-Score, Weighted Kappa, Spearman, MAE, RMSE. The LLM model (URGENTIAPARSE) showed higher accuracy (composite score: 2.514) compared to JEPA (EMERGINET, 0.438) and NLP (TRIAGEMASTER, -3.511), outperforming nurse triage (-4.343). Secondary analyses highlighted the effectiveness of URGENTIAPARSE in predicting hospitalization needs (GEMSA) and its robustness with structured data versus raw transcripts (either for GEMSA prediction or for FRENCH prediction). LLM architecture, through abstraction of patient representations, offers the most accurate triage predictions among tested models. Integrating AI into ED workflows could enhance patient safety and operational efficiency, though integration into clinical workflows requires addressing model limitations and ensuring ethical transparency. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 16:37:55 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lansiaux",
"Edouard",
""
],
[
"Azzouz",
"Ramy",
""
],
[
"Chazard",
"Emmanuel",
""
],
[
"Vromant",
"Amélie",
""
],
[
"Wiel",
"Eric",
""
]
] |
2507.01081 | Megan DeBettencourt | Megan T. deBettencourt, Sruthi Sakthivel, Emily A. Holmes, Mark Chevillet | AI-guided digital intervention with physiological monitoring reduces intrusive memories after experimental trauma | null | null | null | null | cs.HC cs.AI | http://creativecommons.org/licenses/by/4.0/ | Trauma prevalence is vast globally. Evidence-based digital treatments can help, but most require human guidance. Human guides provide tailored instructions and responsiveness to internal cognitive states, but limit scalability. Can generative AI and neurotechnology provide a scalable alternative? Here we test ANTIDOTE, combining AI guidance and pupillometry to automatically deliver and monitor an evidence-based digital treatment, specifically the Imagery Competing Task Intervention (ICTI), to reduce intrusive memories after psychological trauma. One hundred healthy volunteers were exposed to videos of traumatic events and randomly assigned to an intervention or active control condition. As predicted, intervention participants reported significantly fewer intrusive memories over the following week. Post-hoc assessment against clinical rubrics confirmed the AI guide delivered the intervention successfully. Additionally, pupil size tracked intervention engagement and predicted symptom reduction, providing a candidate biomarker of intervention effectiveness. These findings open a path toward rigorous AI-guided digital interventions that can scale to trauma prevalence. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 17:59:01 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"deBettencourt",
"Megan T.",
""
],
[
"Sakthivel",
"Sruthi",
""
],
[
"Holmes",
"Emily A.",
""
],
[
"Chevillet",
"Mark",
""
]
] |
2507.01082 | Masaki Yamada | Kazunori Nakayama and Masaki Yamada | PQ-ball and its Real Scalar Analogue in an Expanding Universe | 12 pages, 6 figures | null | null | TU-1266; KEK-QUP-2025-0014 | hep-ph astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We demonstrate the formation of quasi-stable localized scalar configurations in spontaneously symmetry breaking U(1) model by 3+1-dimensional classical lattice simulations. Such configurations are called PQ-balls, as the primary motivation of this kind of configuration is Peccei-Quinn theory under the kinetic misalignment mechanism. Our numerical simulations demonstrate that they can form if the PQ charge is generated through the coherent rotation of a complex scalar field in the complex plane, via dynamics analogous to the Affleck-Dine mechanism. These configurations subsequently decay due to the U(1)-breaking effect induced by spontaneous symmetry breaking. We also demonstrate the formation and decay of oscillons in a similar setup in a real scalar field theory. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 18:00:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Nakayama",
"Kazunori",
""
],
[
"Yamada",
"Masaki",
""
]
] |
2507.01083 | Mark Ho-Yeuk Cheung | Mark Ho-Yeuk Cheung, Digvijay Wadekar, Ajit Kumar Mehta, Tousif Islam, Javier Roulet, Emanuele Berti, Tejaswi Venumadhav, Barak Zackay, Matias Zaldarriaga | Searching for intermediate mass ratio binary black hole mergers in the third observing run of LIGO-Virgo-KAGRA | 23 pages, 10 figures, 2 tables | null | null | null | gr-qc astro-ph.CO astro-ph.HE astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Intermediate mass ratio inspirals (IMRIs) of binary black holes with mass ratios $10^{-4}\lesssim q \lesssim 0.1$ are astrophysically interesting sources of gravitational waves. Mergers of intermediate-mass black holes (IMBHs) with stellar-mass black holes would be IMRIs, so their detection can help us probe the formation mechanisms of IMBHs. They can also help us perform precise tests of general relativity due to the presence of strong higher-order mode emission. We perform a search for aligned-spin IMRIs within the data of the two LIGO detectors in the third observing run (O3) of the LIGO-Virgo-KAGRA (LVK) collaboration, including higher modes in the template banks for the first time. We use the IAS-HM pipeline for our search and construct template banks in the range $1/100 < q<1/18$ using the SEOBNRv5HM waveform model. Our banks retain a similar level of effectualness for IMRPhenomXHM and BHPTNRSur2dq1e3 waveforms, making our search results relatively robust against waveform systematics. We show that the sensitivity volume of the search increases by up to $\sim 500\%$ upon inclusion of higher modes. We do not find any significant candidates with inverse false alarm rate (IFAR) $> 1$ year in the O3 data. This gives us upper limits on the IMRI merger rate in the local Universe, ranging from $\sim 30$ to $10^3$ Gpc$^{-3}$ yr$^{-1}$ depending on the masses of the black holes in the binary. These constraints are consistent with rate predictions in the literature. Our projections indicate that we would be able to detect IMRIs or constrain some of their proposed formation channels in the fourth (O4) and fifth (O5) observing runs. | [
{
"version": "v1",
"created": "Tue, 01 Jul 2025 18:00:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Cheung",
"Mark Ho-Yeuk",
""
],
[
"Wadekar",
"Digvijay",
""
],
[
"Mehta",
"Ajit Kumar",
""
],
[
"Islam",
"Tousif",
""
],
[
"Roulet",
"Javier",
""
],
[
"Berti",
"Emanuele",
""
],
[
"Venumadhav",
"Tejaswi",
""
],
[
"Zackay",
"Barak",
""
],
[
"Zaldarriaga",
"Matias",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.