id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.12673 | Hadi Mehrabpour | Hadi Mehrabpour | Imprint of $\alpha$-clustering on nucleon-nucleon correlations in relativistic light ion collisions | 14 pages, 7 figures | null | null | null | nucl-th hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | This study investigates the influence of $\alpha$-cluster structures in relativistic light nuclear collisions. Using a cluster framework, the characteristics of the nucleonic configurations of $^{16}$ O and $^{20}$ Ne are extracted as predicted by various \textit{ab initio} models, including Nuclear Lattice Effective Field Theory (NLEFT), Variational Monte Carlo (VMC), and the Projected Generator Coordinate Method (PGCM). Additionally, I analyze configurations derived from a three-parameter Fermi (3pF) density function. The investigation focuses on the effects of cluster parameters on two-point correlators using a rotor model for symmetric collisions ($^{16}$O+$^{16}$O and $^{20}$Ne+$^{20}$Ne) and asymmetric collisions ($^{208}$Pb+$^{16}$O and $^{208}$Pb+$^{20}$Ne). The cluster parameters are determined by minimizing the \textit{chi-square} statistic to align the nucleon distributions with those predicted by the mentioned theories. The results reveal that perturbative calculations effectively capture the structural features of these nuclei, while comparisons with Monte Carlo simulations validate these findings. Furthermore, The analysis reveals distinct cluster geometries: VMC suggests tetrahedral shapes, while NLEFT, PGCM, and 3pF indicate irregular triangular pyramids. Notably, NLEFT shows a bowling pin-like $\alpha$ cluster structure for $^{20}$Ne. The study also identifies constraints on cluster parameters in the ground and excited oxygen states, with a gradual increase in $\varepsilon_2\{2\}$ for excited states. Accurate modeling of asymmetric collisions requires a range of nucleons from heavy spherical nuclei, leading to weighted correlators in perturbative calculations. I demonstrate consistency between perturbative calculations and Monte Carlo models, with analytical calculations providing more insights into asymmetric than symmetric collisions. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 00:40:20 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:05:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mehrabpour",
"Hadi",
""
]
] |
2506.12674 | Paul Landes | Paul Landes, Aaron J Chaise, Tarak Nath Nandi, Ravi K Madduri | Enhancing Clinical Models with Pseudo Data for De-identification | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Many models are pretrained on redacted text for privacy reasons. Clinical foundation models are often trained on de-identified text, which uses special syntax (masked) text in place of protected health information. Even though these models have increased in popularity, there has been little effort in understanding the effects of training them on redacted text. In this work, we pretrain several encoder-only models on a dataset that contains redacted text and a version with replaced realistic pseudo text. We then fine-tuned models for the protected health information de-identification task and show how our methods significantly outperform previous baselines. The contributions of this work include: a) our novel, and yet surprising findings with training recommendations, b) redacted text replacements used to produce the pseudo dataset, c) pretrained embeddings and fine-tuned task specific models, and d) freely available pseudo training dataset generation and model source code used in our experiments. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 00:57:48 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 01:12:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Landes",
"Paul",
""
],
[
"Chaise",
"Aaron J",
""
],
[
"Nandi",
"Tarak Nath",
""
],
[
"Madduri",
"Ravi K",
""
]
] |
2506.12680 | Chen-Bin Feng | Chen-Bin Feng, Kangdao Liu, Jian Sun, Jiping Jin, Yiguo Jiang, and Chi-Man Vong | 3D Hand Mesh-Guided AI-Generated Malformed Hand Refinement with Hand Pose Transformation via Diffusion Model | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The malformed hands in the AI-generated images seriously affect the authenticity of the images. To refine malformed hands, existing depth-based approaches use a hand depth estimator to guide the refinement of malformed hands. Due to the performance limitations of the hand depth estimator, many hand details cannot be represented, resulting in errors in the generated hands, such as confusing the palm and the back of the hand. To solve this problem, we propose a 3D mesh-guided refinement framework using a diffusion pipeline. We use a state-of-the-art 3D hand mesh estimator, which provides more details of the hands. For training, we collect and reannotate a dataset consisting of RGB images and 3D hand mesh. Then we design a diffusion inpainting model to generate refined outputs guided by 3D hand meshes. For inference, we propose a double check algorithm to facilitate the 3D hand mesh estimator to obtain robust hand mesh guidance to obtain our refined results. Beyond malformed hand refinement, we propose a novel hand pose transformation method. It increases the flexibility and diversity of the malformed hand refinement task. We made the restored images mimic the hand poses of the reference images. The pose transformation requires no additional training. Extensive experimental results demonstrate the superior performance of our proposed method. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 01:30:22 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 02:48:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Feng",
"Chen-Bin",
""
],
[
"Liu",
"Kangdao",
""
],
[
"Sun",
"Jian",
""
],
[
"Jin",
"Jiping",
""
],
[
"Jiang",
"Yiguo",
""
],
[
"Vong",
"Chi-Man",
""
]
] |
2506.12727 | Minhyuk Choi | Minhyuk Choi, Injae Kim, Hyunwoo J. Kim | Efficient multi-view training for 3D Gaussian Splatting | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | 3D Gaussian Splatting (3DGS) has emerged as a preferred choice alongside Neural Radiance Fields (NeRF) in inverse rendering due to its superior rendering speed. Currently, the common approach in 3DGS is to utilize "single-view" mini-batch training, where only one image is processed per iteration, in contrast to NeRF's "multi-view" mini-batch training, which leverages multiple images. We observe that such single-view training can lead to suboptimal optimization due to increased variance in mini-batch stochastic gradients, highlighting the necessity for multi-view training. However, implementing multi-view training in 3DGS poses challenges. Simply rendering multiple images per iteration incurs considerable overhead and may result in suboptimal Gaussian densification due to its reliance on single-view assumptions. To address these issues, we modify the rasterization process to minimize the overhead associated with multi-view training and propose a 3D distance-aware D-SSIM loss and multi-view adaptive density control that better suits multi-view scenarios. Our experiments demonstrate that the proposed methods significantly enhance the performance of 3DGS and its variants, freeing 3DGS from the constraints of single-view training. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 05:40:50 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 02:12:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Choi",
"Minhyuk",
""
],
[
"Kim",
"Injae",
""
],
[
"Kim",
"Hyunwoo J.",
""
]
] |
2506.12732 | Takeru Matsuda | Hayato Nishimori, Takeru Matsuda | On the attainment of the Wasserstein--Cramer--Rao lower bound | null | null | null | null | math.ST stat.ML stat.TH | http://creativecommons.org/licenses/by/4.0/ | Recently, a Wasserstein analogue of the Cramer--Rao inequality has been developed using the Wasserstein information matrix (Otto metric). This inequality provides a lower bound on the Wasserstein variance of an estimator, which quantifies its robustness against additive noise. In this study, we investigate conditions for an estimator to attain the Wasserstein--Cramer--Rao lower bound (asymptotically), which we call the (asymptotic) Wasserstein efficiency. We show a condition under which Wasserstein efficient estimators exist for one-parameter statistical models. This condition corresponds to a recently proposed Wasserstein analogue of one-parameter exponential families (e-geodesics). We also show that the Wasserstein estimator, a Wasserstein analogue of the maximum likelihood estimator based on the Wasserstein score function, is asymptotically Wasserstein efficient in location-scale families. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 05:56:12 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:04:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nishimori",
"Hayato",
""
],
[
"Matsuda",
"Takeru",
""
]
] |
2506.12796 | Zhihang Tan | Zhihang Tan, Jingrui Hou, Ping Wang, Qibiao Hu, Peng Zhu | Surprise Calibration for Better In-Context Learning | 16 pages, 11 figures | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In-context learning (ICL) has emerged as a powerful paradigm for task adaptation in large language models (LLMs), where models infer underlying task structures from a few demonstrations. However, ICL remains susceptible to biases that arise from prior knowledge and contextual demonstrations, which can degrade the performance of LLMs. Existing bias calibration methods typically apply fixed class priors across all inputs, limiting their efficacy in dynamic ICL settings where the context for each query differs. To address these limitations, we adopt implicit sequential Bayesian inference as a framework for interpreting ICL, identify "surprise" as an informative signal for class prior shift, and introduce a novel method--Surprise Calibration (SC). SC leverages the notion of surprise to capture the temporal dynamics of class priors, providing a more adaptive and computationally efficient solution for in-context learning. We empirically demonstrate the superiority of SC over existing bias calibration techniques across a range of benchmark natural language processing tasks. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 10:04:42 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 07:46:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tan",
"Zhihang",
""
],
[
"Hou",
"Jingrui",
""
],
[
"Wang",
"Ping",
""
],
[
"Hu",
"Qibiao",
""
],
[
"Zhu",
"Peng",
""
]
] |
2506.12863 | Priyanka Saha | Priyanka Saha, Dipanjan Dey, Kaushik Bhattacharya | Cosmological effect of coherent oscillation of ultralight scalar fields in a multicomponent universe | 19 Pages, 3 figures | null | null | null | gr-qc astro-ph.CO hep-th | http://creativecommons.org/licenses/by/4.0/ | The idea that coherent oscillations of a scalar field, oscillating over a time period that is much shorter than the cosmological timescale, can exhibit cold dark matter (CDM) like behavior was previously established. In our work we first show that this equivalence between the oscillating scalar field model and the CDM sector is exact only in a flat Friedmann-Lemaitre-Robertson-Walker (FLRW) spacetime in the absence of cosmological constant and any other possible matter components in the universe when the mass of the scalar field is very large compared to the Hubble parameter. Then we show how to generalize the equivalence between the coherently oscillating scalar field model and the CDM sector in a spatially curved universe with multiple matter components. Using our general method, we will show how a coherently oscillating scalar field model can represent the CDM sector in the presence of non-minimal coupling of the CDM sector with radiation. Our method is powerful enough to work out the dynamics of gravitational collapse in a closed FLRW spacetime where the coherently oscillating scalar field model represents the CDM sector. We have, for the first time, presented a consistent method which specifies how a coherently oscillating scalar field model, where the scalar field is ultralight, acts like the CDM sector in a multicomponent universe. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 14:26:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Saha",
"Priyanka",
""
],
[
"Dey",
"Dipanjan",
""
],
[
"Bhattacharya",
"Kaushik",
""
]
] |
2506.13022 | Suo Liu | Liu Suo, Shahid Idrees, Liu Dian, Zeng Shuguang | The Properties of Non-Potential Magnetic Field Parameters in a Super-Active Region with Complex Structures and Strong Solar Flares | 14 pages 10 figures solar physics 2025 | null | 10.1007/s11207-025-02456-6 | Solar Physics 2025 | astro-ph.SR astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, the non-potential magnetic field parameters of active region NOAA 9077 are investigated; this AR experienced a super-strong X5.7 solar flare. Using advanced extrapolation techniques, the 3D magnetic field structure from vector magnetograms is obtained from the Solar Magnetic Field Telescope (SMFT) at Huairou Solar Observing Station (HSOS). Then various non-potential parameters are calculated, including current density, shear angle, quasi-separatrix layers (QSLs), twist, and field line helicity. By analyzing the spatial and temporal distributions of these parameters, we aim to shed light on the relationship between magnetic field properties and solar flare occurrence. Our findings reveal that high twist and complex magnetic field configurations are prevalent before flares, while these features tend to weaken after the eruption. Additionally, we observe decreases in helicity and free energy after the flare, while the free energy peaks approximately 1.5 days prior to the onset of the flare. Furthermore, we investigate the distribution of quasi-separatrix layers and twist, finding high degrees of complexity before flares. Multiple patterns of high current density regions suggest unstable magnetic structures prone to flaring, coinciding with shear angle distribution. Relative field line helicity patterns exhibit distinct characteristics compared to current density, concentrating before flares and diverging afterward. Overall, our results highlight the contrasting nature of current density and relative field line helicity patterns in relation to solar flares, in addition to the aforementioned feature in the set of commonly derived non-potential parameters for this particular event. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 01:17:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 02:19:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Suo",
"Liu",
""
],
[
"Idrees",
"Shahid",
""
],
[
"Dian",
"Liu",
""
],
[
"Shuguang",
"Zeng",
""
]
] |
2506.13038 | Xuecheng Wu | Zijian Zhang, Xuecheng Wu, Danlei Huang, Siyu Yan, Chong Peng, Xuezhi Cao | HKD4VLM: A Progressive Hybrid Knowledge Distillation Framework for Robust Multimodal Hallucination and Factuality Detection in VLMs | null | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Driven by the rapid progress in vision-language models (VLMs), the responsible behavior of large-scale multimodal models has become a prominent research area, particularly focusing on hallucination detection and factuality checking. In this paper, we present the solution for the two tracks of Responsible AI challenge. Inspirations from the general domain demonstrate that a smaller distilled VLM can often outperform a larger VLM that is directly tuned on downstream tasks, while achieving higher efficiency. We thus jointly tackle two tasks from the perspective of knowledge distillation and propose a progressive hybrid knowledge distillation framework termed HKD4VLM. Specifically, the overall framework can be decomposed into Pyramid-like Progressive Online Distillation and Ternary-Coupled Refinement Distillation, hierarchically moving from coarse-grained knowledge alignment to fine-grained refinement. Besides, we further introduce the mapping shift-enhanced inference and diverse augmentation strategies to enhance model performance and robustness. Extensive experimental results demonstrate the effectiveness of our HKD4VLM. Ablation studies provide insights into the critical design choices driving performance gains. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 02:03:41 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:31:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Zijian",
""
],
[
"Wu",
"Xuecheng",
""
],
[
"Huang",
"Danlei",
""
],
[
"Yan",
"Siyu",
""
],
[
"Peng",
"Chong",
""
],
[
"Cao",
"Xuezhi",
""
]
] |
2506.13042 | Shaul Hanany | Shaul Hanany, Scott Cray, Samuel Dietterich, Jan Dusing, Calvin Firth, Jurgen Koch, Rex Lam, Tomotake Matsumura, Haruyuki Sakurai, Yuki Sakurai, Aritoki Suzuki, Ryota Takaku, Qi Wen, Alexander Wienke, Andrew Y. Yan | Laser ablated sub-wavelength structure anti-reflection coating on an alumina lens | 9 pages, 7 figures, updated and improved version of paper presented at SPIE's Astronomical Telescopes and Instrumentation 2024 | null | null | null | astro-ph.IM physics.optics | http://creativecommons.org/licenses/by/4.0/ | We used laser ablation to fabricate sub-wavelength structure anti-reflection coating (SWS-ARC) on a 5 cm diameter alumina lens. With an aspect ratio of 2.5, the SWS-ARC are designed to give a broad-band low reflectance response between 110 and 290 GHz. SWS shape measurements conducted on both sides of the lens give 303 $\mu$m pitch and total height between 750 and 790 $\mu$m, matching or exceeding the aspect ratio design values. Millimeter-wave transmittance measurements in a band between 140 and 260 GHz show the increase in transmittance expected with the ARC when compared to finite element analysis electromagnetic simulations. To our knowledge, this is the first demonstration of SWS-ARC on an alumina lens, opening the path for implementing the technique for larger diameter lenses. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 02:08:45 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 16:46:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hanany",
"Shaul",
""
],
[
"Cray",
"Scott",
""
],
[
"Dietterich",
"Samuel",
""
],
[
"Dusing",
"Jan",
""
],
[
"Firth",
"Calvin",
""
],
[
"Koch",
"Jurgen",
""
],
[
"Lam",
"Rex",
""
],
[
"Matsumura",
"Tomotake",
""
],
[
"Sakurai",
"Haruyuki",
""
],
[
"Sakurai",
"Yuki",
""
],
[
"Suzuki",
"Aritoki",
""
],
[
"Takaku",
"Ryota",
""
],
[
"Wen",
"Qi",
""
],
[
"Wienke",
"Alexander",
""
],
[
"Yan",
"Andrew Y.",
""
]
] |
2506.13068 | Xueping Li | Haowen Xu, Yulin Sun, Jose Tupayachi, Olufemi Omitaomu, Sisi Zlatanova, Xueping Li | Towards the Autonomous Optimization of Urban Logistics: Training Generative AI with Scientific Tools via Agentic Digital Twins and Model Context Protocol | null | null | null | null | cs.MA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Optimizing urban freight logistics is critical for developing sustainable, low-carbon cities. Traditional methods often rely on manual coordination of simulation tools, optimization solvers, and expert-driven workflows, limiting their efficiency and scalability. This paper presents an agentic system architecture that leverages the model context protocol (MCP) to orchestrate multi-agent collaboration among scientific tools for autonomous, simulation-informed optimization in urban logistics. The system integrates generative AI agents with domain-specific engines - such as Gurobi for optimization and AnyLogic for agent-based simulation - forming a generative digital twin capable of reasoning, planning, and acting across multimodal freight networks. By incorporating integrated chatbots, retrieval-augmented generation, and structured memory, the framework enables agents to interpret user intent from natural language conversations, retrieve relevant datasets and models, coordinate solvers and simulators, and execute complex workflows. We demonstrate this approach through a freight decarbonization case study, showcasing how MCP enables modular, interoperable, and adaptive agent behavior across diverse toolchains. The results reveal that our system transforms digital twins from static visualizations into autonomous, decision-capable systems, advancing the frontiers of urban operations research. By enabling context-aware, generative agents to operate scientific tools automatically and collaboratively, this framework supports more intelligent, accessible, and dynamic decision-making in transportation planning and smart city management. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 03:23:27 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:06:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Haowen",
""
],
[
"Sun",
"Yulin",
""
],
[
"Tupayachi",
"Jose",
""
],
[
"Omitaomu",
"Olufemi",
""
],
[
"Zlatanova",
"Sisi",
""
],
[
"Li",
"Xueping",
""
]
] |
2506.13069 | Mohamed Nouidha | Mohamed Nouidha | An Isomorphism Theorem for Some Linear Elliptic Differential Operators on Quasi-Asymptotically Conical Manifolds | null | null | null | null | math.DG math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a linear elliptic differential operator of the form $\mathcal{P}=\Delta + V - \lambda$ on a quasi-asymptotically conical manifold $(M, g)$, where $g$ is a polyhomogeneous metric and $V$ is a $b$-vector field that is unbounded with respect to the metric $g$. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 03:24:57 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 03:33:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nouidha",
"Mohamed",
""
]
] |
2506.13085 | Yiqiu Ma | Yubao Liu, Yanbei Chen, Kentaro Somiya, and Yiqiu Ma | Testing the quantum nature of gravity through interferometry | 30 pages, 15 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We propose a Michelson-type interferometric protocol for testing the quantum nature of gravity through testing the phenomenology of semi-classical gravity theory, which predicts a state-dependent Schrodinger-Newton (SN) evolution of the test mass. The protocol's feature lies in utilizing the asymmetry of two interferometric arms induced by SN self-gravity to create cross-talk between the common and differential motion of the test masses. This cross-talk is imprinted as a clean binary signature in the correlation measurements of the interferometer's output light fields. Our results demonstrate that, when assisted by 10 dB squeezed input states, 3 hours of aggregated measurement data can provide sufficient signal-to-noise ratio to conclusively test the SN theory in 1 Kelvin environment. This shows the strong feasibility of using such interferometric protocols to test if gravity operates quantum-mechanically. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 04:05:41 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:42:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Yubao",
""
],
[
"Chen",
"Yanbei",
""
],
[
"Somiya",
"Kentaro",
""
],
[
"Ma",
"Yiqiu",
""
]
] |
2506.13087 | Zeyu Zhang | Zeyu Zhang, Ziyuan Jiao | IKDiffuser: Fast and Diverse Inverse Kinematics Solution Generation for Multi-arm Robotic Systems | under review | null | null | null | cs.RO cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Solving Inverse Kinematics (IK) problems is fundamental to robotics, but has primarily been successful with single serial manipulators. For multi-arm robotic systems, IK remains challenging due to complex self-collisions, coupled joints, and high-dimensional redundancy. These complexities make traditional IK solvers slow, prone to failure, and lacking in solution diversity. In this paper, we present IKDiffuser, a diffusion-based model designed for fast and diverse IK solution generation for multi-arm robotic systems. IKDiffuser learns the joint distribution over the configuration space, capturing complex dependencies and enabling seamless generalization to multi-arm robotic systems of different structures. In addition, IKDiffuser can incorporate additional objectives during inference without retraining, offering versatility and adaptability for task-specific requirements. In experiments on 6 different multi-arm systems, the proposed IKDiffuser achieves superior solution accuracy, precision, diversity, and computational efficiency compared to existing solvers. The proposed IKDiffuser framework offers a scalable, unified approach to solving multi-arm IK problems, facilitating the potential of multi-arm robotic systems in real-time manipulation tasks. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 04:12:04 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:43:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Zeyu",
""
],
[
"Jiao",
"Ziyuan",
""
]
] |
2506.13108 | Jinwon Yi | Jinwon Yi, Soo-Chang Rey, Suk Kim, Youngdae Lee | Formation Pathways of Compact Elliptical Galaxies: Perspective of Internal Structures | 13 pages, 5 figures, Accepted for publication in ApJ | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a structural analysis of 138 compact elliptical galaxies (cEs) in the redshift range of $z < 0.05$ using the Sloan Digital Sky Survey (SDSS) DR12 data. We perform single- and double-component S\'ersic model fitting to their SDSS $r$-band surface brightness profiles. By dividing cEs into those with [cE(w)] and without [cE(w/o)] a bright host galaxy, we find a significant structural dichotomy: the majority ($\sim$85\%) of cE(w)s exhibit single-component profiles, while a similar proportion ($\sim$85\%) of isolated cE(w/o)s display double-component profiles, characterized by a compact, inner component and a diffuse, disk-like outer component. These results suggest that host-associated cE(w)s primarily form through the tidal stripping of larger progenitors, resulting in a compact bulge-like core. In contrast, isolated cE(w/o)s appear to form intrinsically at early epochs, likely through gas-rich mergers, and retain disk-like outer structures. The S\'ersic index distribution of cE(w)s with single-component structure indicates progenitor types ranging from pseudo-bulge to classical bulge, supported by differences in stellar populations. A small fraction of cEs, including double-component cE(w)s and single-component cE(w/o)s, suggests complex evolutionary channels involving environmental capture or ejection. Our results emphasize that the structural characteristics of cEs, specifically the presence or absence of an extended outer envelope, serve as a crucial diagnostic tool to distinguish tidally stripped remnants from intrinsically formed low-mass cEs in isolation. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 05:35:05 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:16:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yi",
"Jinwon",
""
],
[
"Rey",
"Soo-Chang",
""
],
[
"Kim",
"Suk",
""
],
[
"Lee",
"Youngdae",
""
]
] |
2506.13147 | Yusuke Tanimura | Moemi Matsumoto, Yusuke Tanimura, and Kouichi Hagino | Application of the optimized-basis generator coordinate method to low-lying excited states of sd-shell nuclei | 7 pages, 2 figures | null | null | KUNS-3058 | nucl-th | http://creativecommons.org/licenses/by/4.0/ | We apply the optimized-basis generator coordinate method (OptGCM) to sd-shell nuclei, $^{20}$Ne, $^{24}$Mg, and $^{28}$Si. This method variationally optimizes both the basis Slater determinants in the generator coordinate method (GCM) and the corresponding weight coefficients. To analyze the low-lying excited states of those nuclei, we implement the angular momentum projection. With the Skyrme interaction, we show that the simultaneous optimzation of the basis functions and the weight factors lowers the energy of the excited states and at the same time leads to an appreciable effect on transition probabilities. These results highlight the effectiveness of the OptGCM method. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 06:59:58 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Matsumoto",
"Moemi",
""
],
[
"Tanimura",
"Yusuke",
""
],
[
"Hagino",
"Kouichi",
""
]
] |
2506.13167 | Zhe Wang | Zhenxin Liu, Benoit Saussol, Sandro Vaienti, Zhe Wang | Quenched invariance principle with a rate for random dynamical systems | 31 pages | null | null | null | math.DS math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider the quenched invariance principle for random Young towers driven by an ergodic system. In particular, we obtain the Wassertein convergence rate in the quenched invariance principle. As a key ingredient, we derive a new martingale-coboundary decomposition for the random tower map, which provides a good control over sums of squares of the approximating martingale. We apply our results to a class of random dynamical systems that admit a random Young tower, such as independent and identically distributed (i.i.d.) translations of Viana maps, intermittent maps of the interval and small random perturbations of Anosov maps with an ergodic driving system. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 07:22:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Zhenxin",
""
],
[
"Saussol",
"Benoit",
""
],
[
"Vaienti",
"Sandro",
""
],
[
"Wang",
"Zhe",
""
]
] |
2506.13172 | Evgeny Markhasin | Evgeny Markhasin | AI-Facilitated Analysis of Abstracts and Conclusions: Flagging Unsubstantiated Claims and Ambiguous Pronouns | 13 pages | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | We present and evaluate a suite of proof-of-concept (PoC), structured workflow prompts designed to elicit human-like hierarchical reasoning while guiding Large Language Models (LLMs) in the high-level semantic and linguistic analysis of scholarly manuscripts. The prompts target two non-trivial analytical tasks within academic summaries (abstracts and conclusions): identifying unsubstantiated claims (informational integrity) and flagging semantically confusing ambiguous pronoun references (linguistic clarity). We conducted a systematic, multi-run evaluation on two frontier models (Gemini Pro 2.5 Pro and ChatGPT Plus o3) under varied context conditions. Our results for the informational integrity task reveal a significant divergence in model performance: while both models successfully identified an unsubstantiated head of a noun phrase (95% success), ChatGPT consistently failed (0% success) to identify an unsubstantiated adjectival modifier that Gemini correctly flagged (95% success), raising a question regarding the potential influence of the target's syntactic role. For the linguistic analysis task, both models performed well (80-90% success) with full manuscript context. Surprisingly, in a summary-only setting, Gemini's performance was substantially degraded, while ChatGPT achieved a perfect (100%) success rate. Our findings suggest that while structured prompting is a viable methodology for complex textual analysis, prompt performance may be highly dependent on the interplay between the model, task type, and context, highlighting the need for rigorous, model-specific testing. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 07:34:31 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:24:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Markhasin",
"Evgeny",
""
]
] |
2506.13242 | Alexandre Sedoglavic | Jean-Guillaume Dumas (UGA, LJK, CASC), Cl\'ement Pernet (CASC, LJK, UGA), Alexandre Sedoglavic (CRIStAL) | A non-commutative algorithm for multiplying 4x4 matrices using 48 non-complex multiplications | null | null | null | null | cs.SC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The quest for non-commutative matrix multiplication algorithms in small dimensions has seen a lot of recent improvements recently. In particular, the number of scalar multiplications required to multiply two $4\times4$ matrices was first reduced in \cite{Fawzi:2022aa} from 49 (two recursion levels of Strassen's algorithm) to 47 but only in characteristic 2 or more recently to 48 in\cite{alphaevolve} but over complex numbers. We propose an algorithm in 48 multiplications with only rational coefficients, hence removing the complex number requirement. It was derived from the latter one, under the action of an isotropy which happen to project the algorithm on the field of rational numbers. We also produce a straight line program of this algorithm, reducing the leading constant in the complexity, as well as an alternative basis variant of it, leading to an algorithm running in ${\frac{19}{16}n^{2+\frac{\log_2 3 }{2}} + o\left({n^{2+\frac{\log_2 3}{2}}}\right)}$ operations over any ring containing an inverse of 2. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 08:42:15 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:51:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dumas",
"Jean-Guillaume",
"",
"UGA, LJK, CASC"
],
[
"Pernet",
"Clément",
"",
"CASC, LJK, UGA"
],
[
"Sedoglavic",
"Alexandre",
"",
"CRIStAL"
]
] |
2506.13277 | Huayang Li | Huayang Li, Yahui Liu, Hongyu Sun, Deng Cai, Leyang Cui, Wei Bi, Peilin Zhao, Taro Watanabe | SeqPE: Transformer with Sequential Position Encoding | null | null | null | null | cs.LG cs.AI cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | Since self-attention layers in Transformers are permutation invariant by design, positional encodings must be explicitly incorporated to enable spatial understanding. However, fixed-size lookup tables used in traditional learnable position embeddings (PEs) limit extrapolation capabilities beyond pre-trained sequence lengths. Expert-designed methods such as ALiBi and RoPE, mitigate this limitation but demand extensive modifications for adapting to new modalities, underscoring fundamental challenges in adaptability and scalability. In this work, we present SeqPE, a unified and fully learnable position encoding framework that represents each $n$-dimensional position index as a symbolic sequence and employs a lightweight sequential position encoder to learn their embeddings in an end-to-end manner. To regularize SeqPE's embedding space, we introduce two complementary objectives: a contrastive objective that aligns embedding distances with a predefined position-distance function, and a knowledge distillation loss that anchors out-of-distribution position embeddings to in-distribution teacher representations, further enhancing extrapolation performance. Experiments across language modeling, long-context question answering, and 2D image classification demonstrate that SeqPE not only surpasses strong baselines in perplexity, exact match (EM), and accuracy--particularly under context length extrapolation--but also enables seamless generalization to multi-dimensional inputs without requiring manual architectural redesign. We release our code, data, and checkpoints at https://github.com/ghrua/seqpe. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 09:16:40 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 10:12:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Huayang",
""
],
[
"Liu",
"Yahui",
""
],
[
"Sun",
"Hongyu",
""
],
[
"Cai",
"Deng",
""
],
[
"Cui",
"Leyang",
""
],
[
"Bi",
"Wei",
""
],
[
"Zhao",
"Peilin",
""
],
[
"Watanabe",
"Taro",
""
]
] |
2506.13304 | Minze Chen | Minze Chen, Tianqi Mao, Yang Zhao, Wei Xiao, Dezhi Zheng, Zhaocheng Wang, Jun Zhang, and Sheng Chen | New Paradigm for Integrated Sensing and Communication with Rydberg Atomic Receiver | null | null | null | null | quant-ph physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | The RYDberg Atomic Receiver (RYDAR) has been demonstrated to surmount the limitation on both the sensitivity and operating bandwidth of the classical electronic counterpart, which can theoretically detect indiscernible electric signals below -174 dBm/Hz with optical measurement through Rydberg-state atoms. Such miracle has established a new quantum-based paradigm for communications and sensing, which motivates a revolution of the transceiver design philosophies to fully unleash the potential of RYDAR towards next-generation networks. Against this background, this article provides a thorough investigation of Rydberg atomic communications and sensing from theory to hardware implementations. Specifically, we highlight the great opportunities from the hybridization between the RYDAR and the cutting-edge integrated sensing and communication (ISAC), followed by essential preliminaries of the quantum-based receiver. Then we propose a theoretical framework for broadband ISAC based on RYDAR, demonstrated by the proof-of-concept experiments. Afterwards, the enabling technologies for the ISAC framework are explored ranging from channel characterization, waveform design to array-based receiver configurations, where the open problems are also summarized. Finally, the future applications of RYDAR-based ISAC are envisioned, indicating its significant potential for both civilian and military purposes. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 09:45:43 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:42:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chen",
"Minze",
""
],
[
"Mao",
"Tianqi",
""
],
[
"Zhao",
"Yang",
""
],
[
"Xiao",
"Wei",
""
],
[
"Zheng",
"Dezhi",
""
],
[
"Wang",
"Zhaocheng",
""
],
[
"Zhang",
"Jun",
""
],
[
"Chen",
"Sheng",
""
]
] |
2506.13338 | Peter Zeiler Skands | Peter Skands | Theoretical Summary: Moriond QCD and High-Energy Interactions 2025 | Contribution to Moriond QCD and High-Energy Interactions 2025 | null | null | null | hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | The theory talks at Moriond QCD and High-Energy Interactions 2025 covered the full range of scales from BSM, top, Higgs, EW, and hard QCD physics, through resummation, factorisation, and PDFs, to hadronic, heavy-ion, nonperturbative, and lattice QCD. A few talks also touched on methodologies. We here summarise main points of most of these contributions. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 10:27:50 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:42:33 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Skands",
"Peter",
""
]
] |
2506.13369 | Sajad Salavatidezfouli | Sajad Salavatidezfouli, Henrik Karstoft, Alexandros Iosifidis and Mahdi Abkar | Dual guidance: ROM-informed field reconstruction with generative models | null | null | null | null | physics.flu-dyn cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | We present a dual-guided framework for reconstructing unsteady incompressible flow fields using sparse observations. The approach combines optimized sensor placement with a physics-informed guided generative model. Sensor locations are selected using mutual information theory applied to a reduced-order model of the flow, enabling efficient identification of high-information observation points with minimal computational cost. These sensors, once selected, provide targeted observations that guide a denoising diffusion probabilistic model conditioned by physical constraints. Extensive experiments on 2D laminar cylinder wake flows demonstrate that under sparse sensing conditions, the structured sensor layouts fail to capture key flow dynamics, yielding high reconstruction errors. In contrast, our optimized sensor placement strategy achieves accurate reconstructions with L2 errors as low as 0.05, even with a limited number of sensors, confirming the effectiveness of the proposed approach in data-limited regimes. When the number of sensors is higher than a threshold, however, both methods perform comparably. Our dual-guided approach bridges reduced order model-based sensor position optimization with modern generative modeling, providing accurate, physics-consistent reconstruction from sparse data for scientific machine-learning problems. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 11:24:28 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:36:20 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Salavatidezfouli",
"Sajad",
""
],
[
"Karstoft",
"Henrik",
""
],
[
"Iosifidis",
"Alexandros",
""
],
[
"Abkar",
"Mahdi",
""
]
] |
2506.13390 | Seok-Jin Kim | Seok-Jin Kim, Gi-Soo Kim, Min-hwan Oh | Experimental Design for Semiparametric Bandits | Accepted at COLT 2025 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study finite-armed semiparametric bandits, where each arm's reward combines a linear component with an unknown, potentially adversarial shift. This model strictly generalizes classical linear bandits and reflects complexities common in practice. We propose the first experimental-design approach that simultaneously offers a sharp regret bound, a PAC bound, and a best-arm identification guarantee. Our method attains the minimax regret $\tilde{O}(\sqrt{dT})$, matching the known lower bound for finite-armed linear bandits, and further achieves logarithmic regret under a positive suboptimality gap condition. These guarantees follow from our refined non-asymptotic analysis of orthogonalized regression that attains the optimal $\sqrt{d}$ rate, paving the way for robust and efficient learning across a broad class of semiparametric bandit problems. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 11:53:00 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 08:20:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kim",
"Seok-Jin",
""
],
[
"Kim",
"Gi-Soo",
""
],
[
"Oh",
"Min-hwan",
""
]
] |
2506.13422 | Andrew Sullivan Mr | Andrew Sullivan, Ross J. Turner, Stanislav S. Shabala, Chris Power, Sophie A. Young | Jet outbursts, non-thermal pressure and the AGN jet duty cycle | Added further acknowledgements | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We predict the non-thermal pressure (NTP) induced in the cores of galaxy clusters by kinetic jet feedback from an active galactic nucleus (AGN). We model a population of Fanaroff-Riley type I jets when sampling power-law distributions in jet power and age, which we evolve in time with a two-phase jet-lobe model. We couple the energy of each jet outburst to the surrounding gas inside spherical shells, allowing us to estimate the fraction of NTP to total pressure induced in the cluster. We predict the mean profile for this NTP fraction over the source population in a variety of cluster environments and for different AGN jet duty cycles. For typical gas and dark matter profiles, the mean NTP fraction peaks at ~4-6% when the AGN jets are active for 10-30% of the total AGN lifecycle. These predictions are in good agreement with observational constraints, suggesting that AGN feedback imparts only small non-thermal contributions to the cluster's core. Furthermore, we find a relationship between the peak in the mean NTP fraction and the AGN jet duty cycle in a given cluster environment. Applying this to Hitomi measurements of the NTP in the Perseus cluster, we infer an AGN jet duty cycle that is consistent with independent evidence of Perseus' AGN jet activity. We propose this as a novel approach for observationally inferring the past AGN activity of real clusters from their observed NTP fraction and environmental profiles. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 12:37:49 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 13:31:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sullivan",
"Andrew",
""
],
[
"Turner",
"Ross J.",
""
],
[
"Shabala",
"Stanislav S.",
""
],
[
"Power",
"Chris",
""
],
[
"Young",
"Sophie A.",
""
]
] |
2506.13444 | Laiyan Ding | Laiyan Ding, Hualie Jiang, Jiwei Chen, Rui Huang | Self-Supervised Enhancement for Depth from a Lightweight ToF Sensor with Monocular Images | accepted by IROS 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Depth map enhancement using paired high-resolution RGB images offers a cost-effective solution for improving low-resolution depth data from lightweight ToF sensors. Nevertheless, naively adopting a depth estimation pipeline to fuse the two modalities requires groundtruth depth maps for supervision. To address this, we propose a self-supervised learning framework, SelfToF, which generates detailed and scale-aware depth maps. Starting from an image-based self-supervised depth estimation pipeline, we add low-resolution depth as inputs, design a new depth consistency loss, propose a scale-recovery module, and finally obtain a large performance boost. Furthermore, since the ToF signal sparsity varies in real-world applications, we upgrade SelfToF to SelfToF* with submanifold convolution and guided feature fusion. Consequently, SelfToF* maintain robust performance across varying sparsity levels in ToF data. Overall, our proposed method is both efficient and effective, as verified by extensive experiments on the NYU and ScanNet datasets. The code is available at \href{https://github.com/denyingmxd/selftof}{https://github.com/denyingmxd/selftof}. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 12:57:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 06:43:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ding",
"Laiyan",
""
],
[
"Jiang",
"Hualie",
""
],
[
"Chen",
"Jiwei",
""
],
[
"Huang",
"Rui",
""
]
] |
2506.13460 | Lorenzo Mercolli | Lorenzo Mercolli, William M. Steinberger, Pascal V. Grundler, Anzhelika Moiseeva, Saverio Braccini, Maurizio Conti, Pawe{\l} Moskal, Narendra Rathod, Axel Rominger, Hasan Sari, Roger Schibli, Robert Seifert, Kuangyu Shi, Ewa {\L}. St\k{e}pie\'n, Nicholas P. van der Meulen | First Positronium Lifetime Imaging with Scandium-44 on a Long Axial Field-of-view PET/CT | null | null | null | null | physics.med-ph | http://creativecommons.org/licenses/by/4.0/ | Purpose: 44Sc has been successfully produced, synthesized, labeled and first-in-human studies were conducted some years ago. The decay properties of 44Sc, together with being close to a clinical implementation, make it an ideal candidate for in vivo positronium lifetime measurements. In this study, we investigate the count statistics for ortho-positronium (oPs) measurements with 44Sc.
Method: A NEMA image quality phantom was filled with 41.7 MBq of 44Sc dissolved in water and scanned on a commercial long-axial field-of-view PET/CT. Three-photon events were identified using a prototype feature of the scanner and dedicated software. The lifetime of oPs was determined in the phantom spheres and in 4x4x4 mm^3 voxels.
Results: All measured oPs lifetimes are compatible, within the uncertainties, with the literature values for water. The oPs lifetime is 2.65+-0.50, 1.39+-0.20 and 1.76+-0.18 ns in the three smallest spheres of the phantom and 1.79+-0.57 ns for a single voxel in the central region of the largest sphere. The relative standard deviation in the background regions of the time difference distributions, i.e., for time differences smaller than -2.7 ns, is above 20% - even for voxels inside the phantom spheres.
Conclusions: Despite the favorable physical properties of 44Sc, the count statistics of three-photon events remains a challenge. The high prompt-photon energy causes a significant amount of random three-photon coincidences with the given methodology and, therefore, increases the statistical uncertainties on the measured oPs lifetime. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 13:15:35 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:11:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mercolli",
"Lorenzo",
""
],
[
"Steinberger",
"William M.",
""
],
[
"Grundler",
"Pascal V.",
""
],
[
"Moiseeva",
"Anzhelika",
""
],
[
"Braccini",
"Saverio",
""
],
[
"Conti",
"Maurizio",
""
],
[
"Moskal",
"Paweł",
""
],
[
"Rathod",
"Narendra",
""
],
[
"Rominger",
"Axel",
""
],
[
"Sari",
"Hasan",
""
],
[
"Schibli",
"Roger",
""
],
[
"Seifert",
"Robert",
""
],
[
"Shi",
"Kuangyu",
""
],
[
"Stępień",
"Ewa Ł.",
""
],
[
"van der Meulen",
"Nicholas P.",
""
]
] |
2506.13472 | Geom Lee | Junho Yoon, Geom Lee, Donghyeon Jeon, Inho Kang, Seung-Hoon Na | ROSAQ: Rotation-based Saliency-Aware Weight Quantization for Efficiently Compressing Large Language Models | 10 pages, 2 figures | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Quantization has been widely studied as an effective technique for reducing the memory requirement of large language models (LLMs), potentially improving the latency time as well. Utilizing the characteristic of rotational invariance of transformer, we propose the rotation-based saliency-aware weight quantization (ROSAQ), which identifies salient channels in the projection feature space, not in the original feature space, where the projected "principal" dimensions are naturally considered as "salient" features. The proposed ROSAQ consists of 1) PCA-based projection, which first performs principal component analysis (PCA) on a calibration set and transforms via the PCA projection, 2) Salient channel dentification, which selects dimensions corresponding to the K-largest eigenvalues as salient channels, and 3) Saliency-aware quantization with mixed-precision, which uses FP16 for salient dimensions and INT3/4 for other dimensions. Experiment results show that ROSAQ shows improvements over the baseline saliency-aware quantization on the original feature space and other existing quantization methods. With kernel fusion, ROSAQ presents about 2.3x speed up over FP16 implementation in generating 256 tokens with a batch size of 64. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 13:30:33 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 09:13:54 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yoon",
"Junho",
""
],
[
"Lee",
"Geom",
""
],
[
"Jeon",
"Donghyeon",
""
],
[
"Kang",
"Inho",
""
],
[
"Na",
"Seung-Hoon",
""
]
] |
2506.13482 | Irene Urso | I. Urso, F. H. Vincent, M. Wielgus, T. Paumard, G. Perrin | Gravity versus astrophysics in black hole images and photon rings: equatorial emissions and spherically symmetric spacetimes | Submitted to A&A, 16 pages + 8 pages of appendix, 9 figures + 10 figures in appendix, 3 tables | null | null | null | astro-ph.HE gr-qc | http://creativecommons.org/licenses/by/4.0/ | The EHT collaboration released in 2019 the first horizon-scale images of a black hole accretion flow, opening a novel route for plasma physics comprehension and gravitational tests. Although the present unresolved images deeply depend on the astrophysical properties of the accreted matter, GR predicts that they contain highly lensed observables, the "photon rings", embodying the effects of strong-field gravity. Focusing on the supermassive black hole M87* and adopting a geometrically thin, equatorial disc as a phenomenological configuration for the accreting matter, our goal is to study the degeneracy of spacetime curvature and of physically-motivated emission processes on EHT-like images observed at 230 and 345 GHz. In a parametric framework, we simulate adaptively ray-traced images using GYOTO in various spherically-symmetric spacetime geometries, for a comprehensive class of disc velocities and realistic synchrotron emission profiles. We then extract the width and the peak position of 1D intensity cross sections on the direct image and the first photon ring. We show that, among the investigated quantities, the most appropriate observables to probe the geometry are the peak positions of the first photon ring. Small geometric deviations can be unequivocally detected, regardless of the motion of the disc, ranging from a Keplerian to a radially infalling one, if the black hole mass-to-distance estimate is accurate up to around 2%; the current uncertainty of 11% being sufficient just to access extreme deviations. The equatorial set-up of this paper, favoured by present EHT observations of M87*, is adapted to model future measurements at higher observing frequencies, where absorption effects are negligible, and with higher resolution, indispensable to resolve the photon rings. Additional work is needed to investigate if our conclusions hold for more realistic disc configurations. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 13:39:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Urso",
"I.",
""
],
[
"Vincent",
"F. H.",
""
],
[
"Wielgus",
"M.",
""
],
[
"Paumard",
"T.",
""
],
[
"Perrin",
"G.",
""
]
] |
2506.13510 | Saleh Afroogh | Junfeng Jiao, Saleh Afroogh, Kevin Chen, Abhejay Murali, David Atkinson, Amit Dhurandhar | Safe-Child-LLM: A Developmental Benchmark for Evaluating LLM Safety in Child-LLM Interactions | null | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As Large Language Models (LLMs) increasingly power applications used by children and adolescents, ensuring safe and age-appropriate interactions has become an urgent ethical imperative. Despite progress in AI safety, current evaluations predominantly focus on adults, neglecting the unique vulnerabilities of minors engaging with generative AI. We introduce Safe-Child-LLM, a comprehensive benchmark and dataset for systematically assessing LLM safety across two developmental stages: children (7-12) and adolescents (13-17). Our framework includes a novel multi-part dataset of 200 adversarial prompts, curated from red-teaming corpora (e.g., SG-Bench, HarmBench), with human-annotated labels for jailbreak success and a standardized 0-5 ethical refusal scale. Evaluating leading LLMs -- including ChatGPT, Claude, Gemini, LLaMA, DeepSeek, Grok, Vicuna, and Mistral -- we uncover critical safety deficiencies in child-facing scenarios. This work highlights the need for community-driven benchmarks to protect young users in LLM interactions. To promote transparency and collaborative advancement in ethical AI development, we are publicly releasing both our benchmark datasets and evaluation codebase at https://github.com/The-Responsible-AI-Initiative/Safe_Child_LLM_Benchmark.git | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 14:04:54 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 02:13:08 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jiao",
"Junfeng",
""
],
[
"Afroogh",
"Saleh",
""
],
[
"Chen",
"Kevin",
""
],
[
"Murali",
"Abhejay",
""
],
[
"Atkinson",
"David",
""
],
[
"Dhurandhar",
"Amit",
""
]
] |
2506.13528 | Meir Shimon | Meir Shimon | Microphysical Origins of Gravitational Potentials in Astrophysical Systems | Submitted for publication | null | null | null | gr-qc astro-ph.CO hep-ph | http://creativecommons.org/licenses/by/4.0/ | Most astrophysical systems (except for very compact objects such as, e.g., black holes and neutron stars) in our universe are characterized by shallow gravitational potentials, with dimensionless compactness parameter. While the existence and characteristic scales of such virialized systems depend on gravity, we demonstrate that the value of the compactness -- and thus the non-relativistic nature of most astrophysical objects -- arises from microphysical parameters, specifically the fine structure constant and the electron-to-proton mass ratio, and is fundamentally independent of the gravitational constant G. It then follows that large-scale astrophysical velocities and escape velocities associated with virialized systems are determined by electromagnetic and atomic physics, not by gravitation, and that the compactness is always set by microphysical scales -- even for the most compact objects, such as neutron stars, where the compactness is determined by quantities like the pion-to-proton mass ratio. This observation, largely overlooked in the literature, explains why the universe is not dominated by relativistic, compact objects and connects the low entropy of the observable universe to underlying microphysics. Our results emphasize the central but underappreciated role played by dimensionless microphysical constants in shaping the macroscopic gravitational landscape of the universe. In particular, we clarify that this independence of compactness from G applies specifically to entire, virialized, naturally formed astrophysical systems -- such as stars, galaxies, and planets -- that have reached equilibrium through self-gravity and microphysical processes. In contrast, arbitrary non-virialized subsystems (e.g., a piece cut from a planet) do not exhibit this property. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 14:19:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shimon",
"Meir",
""
]
] |
2506.13566 | Jonathan Hoss | Jonathan Hoss, Felix Schelling and Noah Klarmann | A Production Scheduling Framework for Reinforcement Learning Under Real-World Constraints | This paper has been accepted for presentation at the IEEE 21st International Conference on Automation Science and Engineering (CASE 2025) | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The classical Job Shop Scheduling Problem (JSSP) focuses on optimizing makespan under deterministic constraints. Real-world production environments introduce additional complexities that cause traditional scheduling approaches to be less effective. Reinforcement learning (RL) holds potential in addressing these challenges, as it allows agents to learn adaptive scheduling strategies. However, there is a lack of a comprehensive, general-purpose frameworks for effectively training and evaluating RL agents under real-world constraints. To address this gap, we propose a modular framework that extends classical JSSP formulations by incorporating key real-world constraints inherent to the shopfloor, including transport logistics, buffer management, machine breakdowns, setup times, and stochastic processing conditions, while also supporting multi-objective optimization. The framework is a customizable solution that offers flexibility in defining problem instances and configuring simulation parameters, enabling adaptation to diverse production scenarios. A standardized interface ensures compatibility with various RL approaches, providing a robust environment for training RL agents and facilitating the standardized comparison of different scheduling methods under dynamic and uncertain conditions. We release JobShopLab as an open-source tool for both research and industrial applications, accessible at: https://github.com/proto-lab-ro/jobshoplab | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 14:50:26 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:27:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hoss",
"Jonathan",
""
],
[
"Schelling",
"Felix",
""
],
[
"Klarmann",
"Noah",
""
]
] |
2506.13601 | Biao Yang | Yiran Hao, Dong Liu, Liyou Luo, Jialu Mu, Hanyu Wang, Zibo Liu, Jensen Li, Zhihong Zhu, Qinghua Guo and Biao Yang | Searching for topological semi-complete bandgap in elastic truss lattices | null | null | null | null | cond-mat.mes-hall physics.app-ph | http://creativecommons.org/licenses/by/4.0/ | Gapless topological phases have attracted significant interest across both quantum and classical systems owing to their novel physics and promising applications. However, the search for ideal gapless topological nodes inside a clear bandgap is still lacking in elastic systems. The degenerate points are always hidden in the trivial bulk bands due to the intricate elastic modes involved. Here, we find a topological semi-complete bandgap in a three-dimensional elastic truss lattice by tuning a supporting rod, which exhibits a complete bandgap except for the inevitable topological degenerate points. Furthermore, we experimentally map the topological semi-complete bandgap and the inside nontrivial surface state arcs with a scanning laser vibrometer. The introduced scheme provides a systematic approach for the idealization of semi-complete bandgaps and thus may significantly advance the practical utility of topological phases in mechanical engineering domains. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 15:25:42 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:30:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hao",
"Yiran",
""
],
[
"Liu",
"Dong",
""
],
[
"Luo",
"Liyou",
""
],
[
"Mu",
"Jialu",
""
],
[
"Wang",
"Hanyu",
""
],
[
"Liu",
"Zibo",
""
],
[
"Li",
"Jensen",
""
],
[
"Zhu",
"Zhihong",
""
],
[
"Guo",
"Qinghua",
""
],
[
"Yang",
"Biao",
""
]
] |
2506.13615 | Arsenii Onuchin Andreevich | Bulat Batuev, Arsenii Onuchin, Sergey Sukhov | Effective Stimulus Propagation in Neural Circuits: Driver Node Selection | null | null | null | null | q-bio.NC cs.NE q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Precise control of signal propagation in modular neural networks represents a fundamental challenge in computational neuroscience. We establish a framework for identifying optimal control nodes that maximize stimulus transmission between weakly coupled neural populations. Using spiking stochastic block model networks, we systematically compare driver node selection strategies - including random sampling and topology-based centrality measures (degree, betweenness, closeness, eigenvector, harmonic, and percolation centrality) - to determine minimal control inputs for achieving inter-population synchronization. Targeted stimulation of just 10-20% of the most central neurons in the source population significantly enhances spiking propagation fidelity compared to random selection. This approach yields a 64-fold increase in signal transfer efficiency at critical inter-module connection densities. These findings establish a theoretical foundation for precision neuromodulation in biological neural systems and neurotechnology applications. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 15:43:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 11:35:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Batuev",
"Bulat",
""
],
[
"Onuchin",
"Arsenii",
""
],
[
"Sukhov",
"Sergey",
""
]
] |
2506.13628 | Francesco Fabbri | Francesco Fabbri, Martino Andrea Scarpolini, Angelo Iollo, Francesco Viola, Francesco Tudisco | Graph-Convolutional-Beta-VAE for Synthetic Abdominal Aorta Aneurysm Generation | Typo in the title | null | null | null | cs.LG cs.AI q-bio.TO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Synthetic data generation plays a crucial role in medical research by mitigating privacy concerns and enabling large-scale patient data analysis. This study presents a beta-Variational Autoencoder Graph Convolutional Neural Network framework for generating synthetic Abdominal Aorta Aneurysms (AAA). Using a small real-world dataset, our approach extracts key anatomical features and captures complex statistical relationships within a compact disentangled latent space. To address data limitations, low-impact data augmentation based on Procrustes analysis was employed, preserving anatomical integrity. The generation strategies, both deterministic and stochastic, manage to enhance data diversity while ensuring realism. Compared to PCA-based approaches, our model performs more robustly on unseen data by capturing complex, nonlinear anatomical variations. This enables more comprehensive clinical and statistical analyses than the original dataset alone. The resulting synthetic AAA dataset preserves patient privacy while providing a scalable foundation for medical research, device testing, and computational modeling. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 15:55:56 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:36:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fabbri",
"Francesco",
""
],
[
"Scarpolini",
"Martino Andrea",
""
],
[
"Iollo",
"Angelo",
""
],
[
"Viola",
"Francesco",
""
],
[
"Tudisco",
"Francesco",
""
]
] |
2506.13657 | Dipayan Biswas | Dipayan Biswas, Shishir Shah, Jaspal Subhlok | Lecture Video Visual Objects (LVVO) Dataset: A Benchmark for Visual Object Detection in Educational Videos | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | We introduce the Lecture Video Visual Objects (LVVO) dataset, a new benchmark for visual object detection in educational video content. The dataset consists of 4,000 frames extracted from 245 lecture videos spanning biology, computer science, and geosciences. A subset of 1,000 frames, referred to as LVVO_1k, has been manually annotated with bounding boxes for four visual categories: Table, Chart-Graph, Photographic-image, and Visual-illustration. Each frame was labeled independently by two annotators, resulting in an inter-annotator F1 score of 83.41%, indicating strong agreement. To ensure high-quality consensus annotations, a third expert reviewed and resolved all cases of disagreement through a conflict resolution process. To expand the dataset, a semi-supervised approach was employed to automatically annotate the remaining 3,000 frames, forming LVVO_3k. The complete dataset offers a valuable resource for developing and evaluating both supervised and semi-supervised methods for visual content detection in educational videos. The LVVO dataset is publicly available to support further research in this domain. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:18:21 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 04:05:44 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Biswas",
"Dipayan",
""
],
[
"Shah",
"Shishir",
""
],
[
"Subhlok",
"Jaspal",
""
]
] |
2506.13659 | Jaehyeon Seo | Joonkyung Lee, Jaeseong Oh, Jaehyeon Seo | Counting homomorphisms in antiferromagnetic graphs via Lorentzian polynomials | 30 pages, 6 figures. Extended abstract accepted to FPSAC 2025 | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | An edge-weighted graph $G$, possibly with loops, is said to be antiferromagnetic if it has nonnegative weights and at most one positive eigenvalue, counting multiplicities. The number of graph homomorphisms from a graph $H$ to an antiferromagnetic graph $G$ generalises various important parameters in graph theory, including the number of independent sets and proper vertex-colourings, as well as their relaxations in statistical physics.
We obtain homomorphism inequalities for various graphs $H$ and antiferromagnetic graphs~$G$ of the form \[ \lvert\operatorname{Hom}(H,G)\rvert^2 \leq \lvert\operatorname{Hom}(H\times K_2,G)\rvert, \] where $H\times K_2$ denotes the tensor product of $H$ and $K_2$. Firstly, we show that the inequality holds for any $H$ obtained by blowing up vertices of a bipartite graph into complete graphs and any antiferromagnetic $G$. In particular, one can take $H=K_{d+1}$, which already implies a new result for the Sah--Sawhney--Stoner--Zhao conjecture on the maximum number of $d$-regular graphs in antiferromagnetic graphs. Secondly, the inequality also holds for $G=K_q$ and those $H$ obtained by blowing up vertices of a bipartite graph into complete multipartite graphs, paths or even cycles.
Both results can be seen as the first progress towards Zhao's conjecture on $q$-colourings, which states that the inequality holds for any $H$ and $G=K_q$, after his own work. Our method leverages on the emerging theory of Lorentzian polynomials due to Br\"and\'en and Huh and log-concavity of the list colourings of bipartite graphs, which may be of independent interest. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:18:42 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 14:36:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lee",
"Joonkyung",
""
],
[
"Oh",
"Jaeseong",
""
],
[
"Seo",
"Jaehyeon",
""
]
] |
2506.13674 | Haonan Wang | Haonan Wang, Brian Chen, Siquan Li, Xinhe Liang, Hwee Kuan Lee, Kenji Kawaguchi, Tianyang Hu | Prefix-Tuning+: Modernizing Prefix-Tuning by Decoupling the Prefix from Attention | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Parameter-Efficient Fine-Tuning (PEFT) methods have become crucial for rapidly adapting large language models (LLMs) to downstream tasks. Prefix-Tuning, an early and effective PEFT technique, demonstrated the ability to achieve performance comparable to full fine-tuning with significantly reduced computational and memory overhead. However, despite its earlier success, its effectiveness in training modern state-of-the-art LLMs has been very limited. In this work, we demonstrate empirically that Prefix-Tuning underperforms on LLMs because of an inherent tradeoff between input and prefix significance within the attention head. This motivates us to introduce Prefix-Tuning+, a novel architecture that generalizes the principles of Prefix-Tuning while addressing its shortcomings by shifting the prefix module out of the attention head itself. We further provide an overview of our construction process to guide future users when constructing their own context-based methods. Our experiments show that, across a diverse set of benchmarks, Prefix-Tuning+ consistently outperforms existing Prefix-Tuning methods. Notably, it achieves performance on par with the widely adopted LoRA method on several general benchmarks, highlighting the potential modern extension of Prefix-Tuning approaches. Our findings suggest that by overcoming its inherent limitations, Prefix-Tuning can remain a competitive and relevant research direction in the landscape of parameter-efficient LLM adaptation. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:30:26 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 15:25:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Haonan",
""
],
[
"Chen",
"Brian",
""
],
[
"Li",
"Siquan",
""
],
[
"Liang",
"Xinhe",
""
],
[
"Lee",
"Hwee Kuan",
""
],
[
"Kawaguchi",
"Kenji",
""
],
[
"Hu",
"Tianyang",
""
]
] |
2506.13686 | Hadi Kazemi | Hadi Kazemi, Ankit Pensia, Varun Jog | The Sample Complexity of Distributed Simple Binary Hypothesis Testing under Information Constraints | 1-page extended abstract appears in COLT2025 | null | null | null | cs.IT math.IT math.ST stat.ML stat.TH | http://creativecommons.org/licenses/by/4.0/ | This paper resolves two open problems from a recent paper, arXiv:2403.16981, concerning the sample complexity of distributed simple binary hypothesis testing under information constraints. The first open problem asks whether interaction reduces the sample complexity of distributed simple binary hypothesis testing. In this paper, we show that sequential interaction does not help. The second problem suggests tightening existing sample complexity bounds for communication-constrained simple binary hypothesis testing. We derive optimally tight bounds for this setting and resolve this problem. Our main technical contributions are: (i) a one-shot lower bound on the Bayes error in simple binary hypothesis testing that satisfies a crucial tensorisation property; (ii) a streamlined proof of the formula for the sample complexity of simple binary hypothesis testing without constraints, first established in arXiv:2403.16981; and (iii) a reverse data-processing inequality for Hellinger-$\lambda$ divergences, generalising the results from arXiv:1812.03031 and arXiv:2206.02765. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:50:27 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kazemi",
"Hadi",
""
],
[
"Pensia",
"Ankit",
""
],
[
"Jog",
"Varun",
""
]
] |
2506.13693 | Zijun Zhao | Zijun Zhao, Shuhang Li, Francesco A. Evangelista | Multireference equation-of-motion driven similarity renormalization group: theoretical foundations and applications to ionized states | null | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | We present a formulation and implementation of an equation-of-motion (EOM) extension of the multireference driven similarity renormalization group (MR-DSRG) formalism for ionization potentials (IP-EOM-DSRG). The IP-EOM-DSRG formalism results in a Hermitian generalized eigenvalue problem, delivering accurate ionization potentials for systems with strongly correlated ground and excited states. The EOM step scales as $O(N^5)$ with the basis set size $N$, allowing for efficient calculation of spectroscopic properties, such as transition energies and intensities. The IP-EOM-DSRG formalism is combined with three truncation schemes of the parent MR-DSRG theory: an iterative nonperturbative method with up to two-body excitations [MR-LDSRG(2)] and second- and third-order perturbative approximations [DSRG-MRPT2/3]. We benchmark these variants by computing 1) the vertical valence ionization potentials of a series of small molecules at both equilibrium and stretched geometries; 2) the spectroscopic constants of several low-lying electronic states of the OH, CN, N2+, and CO+ radicals; and 3) the binding curves of low-lying electronic states of the CN radical. A comparison with experimental data and theoretical results shows that all three IP-EOM-DSRG methods accurately reproduce the vertical ionization potentials and spectroscopic constants of these systems. Notably, the DSRG-MRPT3 and MR-LDSRG(2) versions outperform several state-of-the-art multireference methods of comparable or higher cost. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 16:55:43 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 12:06:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhao",
"Zijun",
""
],
[
"Li",
"Shuhang",
""
],
[
"Evangelista",
"Francesco A.",
""
]
] |
2506.13754 | Edward Li | Edward Li, Zichen Wang, Jiahe Huang, Jeong Joon Park | VideoPDE: Unified Generative PDE Solving via Video Inpainting Diffusion Models | Project page: https://videopde.github.io/ | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present a unified framework for solving partial differential equations (PDEs) using video-inpainting diffusion transformer models. Unlike existing methods that devise specialized strategies for either forward or inverse problems under full or partial observation, our approach unifies these tasks under a single, flexible generative framework. Specifically, we recast PDE-solving as a generalized inpainting problem, e.g., treating forward prediction as inferring missing spatiotemporal information of future states from initial conditions. To this end, we design a transformer-based architecture that conditions on arbitrary patterns of known data to infer missing values across time and space. Our method proposes pixel-space video diffusion models for fine-grained, high-fidelity inpainting and conditioning, while enhancing computational efficiency through hierarchical modeling. Extensive experiments show that our video inpainting-based diffusion model offers an accurate and versatile solution across a wide range of PDEs and problem setups, outperforming state-of-the-art baselines. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 17:58:00 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2025 02:15:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Edward",
""
],
[
"Wang",
"Zichen",
""
],
[
"Huang",
"Jiahe",
""
],
[
"Park",
"Jeong Joon",
""
]
] |
2506.13767 | Chenpeng Feng | Chenpeng Feng | Hodge structures of the surface of planes in a cubic 5-fold | null | null | null | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | We study the geometry of the moduli space of planes in a general cubic 5-fold and its deformation. We show that this moduli space is a smooth projective surface whose canonical bundle is ample. We also show that the variation of degree 1 Hodge structures of a particular family of such surfaces is maximal. The main technical input is the hyper-Kahler geometry and an elaborated calculation of the Hodge numbers of such surfaces Keywords: Cubic hypersurfaces, Hyper-Kahler manifolds, Schubert calculus, Variation of Hodge structures | [
{
"version": "v1",
"created": "Thu, 20 Mar 2025 15:31:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Feng",
"Chenpeng",
""
]
] |
2506.13768 | Reimann Stefan | Stefan Reimann | 'Memory States' from Almost Nothing: Representing and Computing in a Non-associative Algebra | 27 pages, 6 figures, journal article (accepted) | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This note presents a non-associative algebraic framework for the representation and computation of information items in high-dimensional space. This framework is consistent with the principles of spatial computing and with the empirical findings in cognitive science about memory. Computations are performed through a process of multiplication-like binding and non-associative interference-like bundling. Models that rely on associative bundling typically lose order information, which necessitates the use of auxiliary order structures, such as position markers, to represent sequential information that is important for cognitive tasks. In contrast, the non-associative bundling proposed allows the construction of sparse representations of arbitrarily long sequences that maintain their temporal structure across arbitrary lengths. In this operation, noise is a constituent element of the representation of order information, rather than a means of obscuring it. The non-associative nature of the proposed framework results in the representation of a single sequence by two distinct states. The L-state, generated through left-associative bundling, continuously updates and emphasises a recency effect, while the R-state, formed through right-associative bundling, encodes finite sequences or chunks, capturing a primacy effect. The construction of these states may be associated with activity in the prefrontal cortex in relation to short-term memory and hippocampal encoding in long-term memory, respectively. The accuracy of retrieval is contingent upon a decision-making process that is based on the mutual information between the memory states and the cue. The model is able to replicate the Serial Position Curve, which reflects the empirical recency and primacy effects observed in cognitive experiments. | [
{
"version": "v1",
"created": "Tue, 13 May 2025 08:43:02 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Reimann",
"Stefan",
""
]
] |
2506.13769 | Filippo Leveni | Filippo Leveni | Non-planar Object Detection and Identification by Features Matching and Triangulation Growth | Master's thesis at Politecnico di Milano | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Object detection and identification is surely a fundamental topic in the computer vision field; it plays a crucial role in many applications such as object tracking, industrial robots control, image retrieval, etc. We propose a feature-based approach for detecting and identifying distorted occurrences of a given template in a scene image by incremental grouping of feature matches between the image and the template. For this purpose, we consider the Delaunay triangulation of template features as an useful tool through which to be guided in this iterative approach. The triangulation is treated as a graph and, starting from a single triangle, neighboring nodes are considered and the corresponding features are identified; then matches related to them are evaluated to determine if they are worthy to be grouped. This evaluation is based on local consistency criteria derived from geometric and photometric properties of local features. Our solution allows the identification of the object in situations where geometric models (e.g. homography) does not hold, thus enable the detection of objects such that the template is non planar or when it is planar but appears distorted in the image. We show that our approach performs just as well or better than application of homography-based RANSAC in scenarios in which distortion is nearly absent, while when the deformation becomes relevant our method shows better description performance. | [
{
"version": "v1",
"created": "Mon, 19 May 2025 06:20:07 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Leveni",
"Filippo",
""
]
] |
2506.13770 | Shiwen Zhang | Shiwen Zhang, Zhuowei Chen, Lang Chen, Yanze Wu | CDST: Color Disentangled Style Transfer for Universal Style Reference Customization | codes and models will be released if the paper is accepted | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce Color Disentangled Style Transfer (CDST), a novel and efficient two-stream style transfer training paradigm which completely isolates color from style and forces the style stream to be color-blinded. With one same model, CDST unlocks universal style transfer capabilities in a tuning-free manner during inference. Especially, the characteristics-preserved style transfer with style and content references is solved in the tuning-free way for the first time. CDST significantly improves the style similarity by multi-feature image embeddings compression and preserves strong editing capability via our new CDST style definition inspired by Diffusion UNet disentanglement law. By conducting thorough qualitative and quantitative experiments and human evaluations, we demonstrate that CDST achieves state-of-the-art results on various style transfer tasks. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 18:44:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Shiwen",
""
],
[
"Chen",
"Zhuowei",
""
],
[
"Chen",
"Lang",
""
],
[
"Wu",
"Yanze",
""
]
] |
2506.13771 | Dongkyu Kim | Banseok Lee, Dongkyu Kim, Youngcheon You, Youngmin Kim | LittleBit: Ultra Low-Bit Quantization via Latent Factorization | null | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Deploying large language models (LLMs) often faces challenges from substantial memory and computational costs. Quantization offers a solution, yet performance degradation in the sub-1-bit regime remains particularly difficult. This paper introduces LittleBit, a novel method for extreme LLM compression. It targets levels like 0.1 bits per weight (BPW), achieving nearly 31$\times$ memory reduction, e.g., Llama2-13B to under 0.9 GB. LittleBit represents weights in a low-rank form using latent matrix factorization, subsequently binarizing these factors. To counteract information loss from this extreme precision, it integrates a multi-scale compensation mechanism. This includes row, column, and an additional latent dimension that learns per-rank importance. Two key contributions enable effective training: Dual Sign-Value-Independent Decomposition (Dual-SVID) for stable quantization-aware training (QAT) initialization, and integrated Residual Compensation to mitigate errors. Extensive experiments confirm LittleBit's superiority in sub-1-bit quantization: e.g., its 0.1 BPW performance on Llama2-7B surpasses the leading method's 0.7 BPW. This establishes a superior size-performance trade-off, with kernel-level benchmarks indicating potential for a 5$\times$ speedup compared to FP16. LittleBit paves the way for deploying powerful LLMs in resource-constrained environments. | [
{
"version": "v1",
"created": "Fri, 30 May 2025 06:43:03 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lee",
"Banseok",
""
],
[
"Kim",
"Dongkyu",
""
],
[
"You",
"Youngcheon",
""
],
[
"Kim",
"Youngmin",
""
]
] |
2506.13772 | Zhenyan Lu | Zhenyan Lu, Daliang Xu, Dongqi Cai, Zexi Li, Wei Liu, Fangming Liu, Shangguang Wang, Mengwei Xu | MobiEdit: Resource-efficient Knowledge Editing for Personalized On-device LLMs | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are deployed on mobile devices to power killer applications such as intelligent assistants. LLMs pre-trained on general corpora often hallucinate when handling personalized or unseen queries, leading to incorrect or outdated responses. Knowledge editing addresses this by identifying and adjusting a small crucial portion of model weights, without compromising the general knowledge. However, prior knowledge editing methods are impractical to run on local devices due to the resource-heavy backpropagation (BP) needed for updates. We present MobiEdit, the first mobile knowledge editing framework that enables efficient LLM personalization on commercial off-the-shelf (COTS) mobile devices. MobiEdit replaces full-precision BP with quantized forward-only gradient estimation, thus compatible with the energy-efficient mobile neural processing units (NPUs). MobiEdit replaces full-precision backpropagation with quantized forward-only gradient estimation, making it compatible with energy-efficient mobile NPUs. To further improve gradient estimation efficiency, we introduce two optimizations: an early stoping mechanism that adaptively terminates editing upon success and a prefix cache that reuses computation across steps. Our approach enables real-time editing of a 3B-parameter model (Qwen2.5-3B-Instruct) on COTS mobile devices with 7.6$\times$ less memory, 14.7 $\times$ less energy and 3.6$\times$ less latency compared to previous knowledge editing methods. | [
{
"version": "v1",
"created": "Thu, 05 Jun 2025 03:02:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lu",
"Zhenyan",
""
],
[
"Xu",
"Daliang",
""
],
[
"Cai",
"Dongqi",
""
],
[
"Li",
"Zexi",
""
],
[
"Liu",
"Wei",
""
],
[
"Liu",
"Fangming",
""
],
[
"Wang",
"Shangguang",
""
],
[
"Xu",
"Mengwei",
""
]
] |
2506.13773 | Milapji Singh Gill | Milapji Singh Gill, Tom Jeleniewski, Felix Gehlhoff, Alexander Fay | Representing Time-Continuous Behavior of Cyber-Physical Systems in Knowledge Graphs | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time-continuous dynamic models are essential for various Cyber-Physical System (CPS) applications. To ensure effective usability in different lifecycle phases, such behavioral information in the form of differential equations must be contextualized and integrated with further CPS information. While knowledge graphs provide a formal description and structuring mechanism for this task, there is a lack of reusable ontological artifacts and methods to reduce manual instantiation effort. Hence, this contribution introduces two artifacts: Firstly, a modular semantic model based on standards is introduced to represent differential equations directly within knowledge graphs and to enrich them semantically. Secondly, a method for efficient knowledge graph generation is presented. A validation of these artifacts was conducted in the domain of aviation maintenance. Results show that differential equations of a complex Electro-Hydraulic Servoactuator can be formally represented in a knowledge graph and be contextualized with other lifecycle data, proving the artifacts' practical applicability. | [
{
"version": "v1",
"created": "Sun, 08 Jun 2025 13:54:23 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Gill",
"Milapji Singh",
""
],
[
"Jeleniewski",
"Tom",
""
],
[
"Gehlhoff",
"Felix",
""
],
[
"Fay",
"Alexander",
""
]
] |
2506.13774 | Nell Watson | Nell Watson, Ahmed Amer, Evan Harris, Preeti Ravindra, Shujun Zhang | Personalized Constitutionally-Aligned Agentic Superego: Secure AI Behavior Aligned to Diverse Human Values | 39 pages, 5 figures | null | null | null | cs.AI cs.CY cs.MA | http://creativecommons.org/licenses/by/4.0/ | Agentic AI systems, possessing capabilities for autonomous planning and action, exhibit immense potential across diverse domains. However, their practical deployment is significantly hampered by challenges in aligning their behavior with varied human values, complex safety requirements, and specific compliance needs. Existing alignment methodologies often falter when faced with the intricate task of providing deep, personalized contextual information without inducing confabulation or operational inefficiencies. This paper introduces a novel solution: a 'superego' agent, designed as a personalized oversight mechanism for agentic AI. This system dynamically steers AI planning by referencing user-selected "Creed Constitutions"-encapsulating diverse rule sets-with adjustable adherence levels to fit non-negotiable values. A real-time compliance enforcer validates plans against these constitutions and a universal ethical floor before execution. We present a functional system, including a demonstration interface (www.Creed.Space) with a prototypical constitution-sharing portal, and successful integration with third-party models via the Model Context Protocol (MCP). Comprehensive benchmark evaluations (HarmBench, AgentHarm) demonstrate that our Superego agent dramatically reduces harmful outputs, achieving up to a 98.3% harm score reduction and near-perfect refusal rates (e.g., 100% with Claude Sonnet 4 on AgentHarm's harmful set) for leading LLMs like Gemini 2.5 Flash and GPT-4o. This approach substantially simplifies personalized AI alignment, rendering agentic systems more reliably attuned to individual and cultural contexts, while also enabling substantial safety improvements. An overview on this research with examples is available at https://superego.creed.space. | [
{
"version": "v1",
"created": "Sun, 08 Jun 2025 20:31:26 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Watson",
"Nell",
""
],
[
"Amer",
"Ahmed",
""
],
[
"Harris",
"Evan",
""
],
[
"Ravindra",
"Preeti",
""
],
[
"Zhang",
"Shujun",
""
]
] |
2506.13775 | Slimane Zaim | Slimane Zaim | On the scalar tachyons creation in noncommutative de Sitter space-time | 13 pages | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We investigate tachyon production in noncommutative de Sitter space by solving the deformed Klein-Gordon equation. Using the Bogoliubov transformation method, we compute the number density of created scalar tachyons and demonstrate that their spectrum follows a thermal distribution. Our analysis reveals that noncommutativity reduces the effective temperature of tachyon production, analogous to its effect on ordinary particle creation. Furthermore, we establish that the role of the noncommutative parameter in tachyon production is functionally equivalent to that of an electric field in standard particle production in de Sitter space. | [
{
"version": "v1",
"created": "Sun, 08 Jun 2025 22:18:58 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zaim",
"Slimane",
""
]
] |
2506.13777 | En Xu | En Xu, Huandong Wang, Yunke Zhang, Sibo Li, Yinzhou Tang, Zhilun Zhou, Yuming Lin, Yuan Yuan, Xiaochen Fan, Jingtao Ding, Yong Li | A Survey of Physics-Informed AI for Complex Urban Systems | null | null | null | null | physics.soc-ph cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Urban systems are typical examples of complex systems, where the integration of physics-based modeling with artificial intelligence (AI) presents a promising paradigm for enhancing predictive accuracy, interpretability, and decision-making. In this context, AI excels at capturing complex, nonlinear relationships, while physics-based models ensure consistency with real-world laws and provide interpretable insights. We provide a comprehensive review of physics-informed AI methods in urban applications. The proposed taxonomy categorizes existing approaches into three paradigms - Physics-Integrated AI, Physics-AI Hybrid Ensemble, and AI-Integrated Physics - and further details seven representative methods. This classification clarifies the varying degrees and directions of physics-AI integration, guiding the selection and development of appropriate methods based on application needs and data availability. We systematically examine their applications across eight key urban domains: energy, environment, economy, transportation, information, public services, emergency management, and the urban system as a whole. Our analysis highlights how these methodologies leverage physical laws and data-driven models to address urban challenges, enhancing system reliability, efficiency, and adaptability. By synthesizing existing methodologies and their urban applications, we identify critical gaps and outline future research directions, paving the way toward next-generation intelligent urban system modeling. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 06:19:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"En",
""
],
[
"Wang",
"Huandong",
""
],
[
"Zhang",
"Yunke",
""
],
[
"Li",
"Sibo",
""
],
[
"Tang",
"Yinzhou",
""
],
[
"Zhou",
"Zhilun",
""
],
[
"Lin",
"Yuming",
""
],
[
"Yuan",
"Yuan",
""
],
[
"Fan",
"Xiaochen",
""
],
[
"Ding",
"Jingtao",
""
],
[
"Li",
"Yong",
""
]
] |
2506.13778 | Anvi Alex Eponon | Anvi Alex Eponon, Moein Shahiki-Tash, Ildar Batyrshin, Christian E. Maldonado-Sifuentes, Grigori Sidorov, Alexander Gelbukh | Knowledge Compression via Question Generation: Enhancing Multihop Document Retrieval without Fine-tuning | null | null | null | null | cs.IR cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | This study presents a question-based knowledge encoding approach that improves retrieval-augmented generation (RAG) systems without requiring fine-tuning or traditional chunking. We encode textual content using generated questions that span the lexical and semantic space, creating targeted retrieval cues combined with a custom syntactic reranking method.
In single-hop retrieval over 109 scientific papers, our approach achieves a Recall@3 of 0.84, outperforming traditional chunking methods by 60 percent. We also introduce "paper-cards", concise paper summaries under 300 characters, which enhance BM25 retrieval, increasing MRR@3 from 0.56 to 0.85 on simplified technical queries.
For multihop tasks, our reranking method reaches an F1 score of 0.52 with LLaMA2-Chat-7B on the LongBench 2WikiMultihopQA dataset, surpassing chunking and fine-tuned baselines which score 0.328 and 0.412 respectively.
This method eliminates fine-tuning requirements, reduces retrieval latency, enables intuitive question-driven knowledge access, and decreases vector storage demands by 80%, positioning it as a scalable and efficient RAG alternative. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 16:15:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Eponon",
"Anvi Alex",
""
],
[
"Shahiki-Tash",
"Moein",
""
],
[
"Batyrshin",
"Ildar",
""
],
[
"Maldonado-Sifuentes",
"Christian E.",
""
],
[
"Sidorov",
"Grigori",
""
],
[
"Gelbukh",
"Alexander",
""
]
] |
2506.13779 | Scott Powers | Ruoqian Zhu, Dave Love, Scott Powers | Ball path curvature and in-game free throw shooting proficiency in the National Basketball Association | 12 pages, 6 figures | null | null | null | physics.soc-ph stat.AP | http://creativecommons.org/licenses/by/4.0/ | Basketball shooting coaches agree that smoother shooting motions are better, but there is less agreement about what "smooth" means quantitatively or what part of the shooting motion needs to be smooth. Using ball tracking data from the 2023-2024 National Basketball Association regular season, we explore the relationship between ball path curvature and free throw shooting performance. We fit B\'ezier curves to the ball tracking data in the sagittal plane and test different methods of calculating path curvature. We find that both max curvature and terminal curvature are negatively associated with shooting performance, but terminal curvature explains much more of the between-player variance in free throw shooting performance. This suggests that shooting coaches would be better off focusing on the smoothness at the end of the shot rather than at the beginning of the forward motion of the ball. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:15:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhu",
"Ruoqian",
""
],
[
"Love",
"Dave",
""
],
[
"Powers",
"Scott",
""
]
] |
2506.13780 | Sedat Porikli | Sedat Porikli, Vedat Porikli | Hidden Bias in the Machine: Stereotypes in Text-to-Image Models | Equal contribution by both authors, Published at CVPR 2025 Workshop on Experimental Model Auditing via Controllable Synthesis (EMACS) and Workshop on Demographic Diversity in Computer Vision (DemoDiv) | null | null | null | cs.CV cs.AI cs.CY cs.LG | http://creativecommons.org/licenses/by/4.0/ | Text-to-Image (T2I) models have transformed visual content creation, producing highly realistic images from natural language prompts. However, concerns persist around their potential to replicate and magnify existing societal biases. To investigate these issues, we curated a diverse set of prompts spanning thematic categories such as occupations, traits, actions, ideologies, emotions, family roles, place descriptions, spirituality, and life events. For each of the 160 unique topics, we crafted multiple prompt variations to reflect a wide range of meanings and perspectives. Using Stable Diffusion 1.5 (UNet-based) and Flux-1 (DiT-based) models with original checkpoints, we generated over 16,000 images under consistent settings. Additionally, we collected 8,000 comparison images from Google Image Search. All outputs were filtered to exclude abstract, distorted, or nonsensical results. Our analysis reveals significant disparities in the representation of gender, race, age, somatotype, and other human-centric factors across generated images. These disparities often mirror and reinforce harmful stereotypes embedded in societal narratives. We discuss the implications of these findings and emphasize the need for more inclusive datasets and development practices to foster fairness in generative visual systems. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:06:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Porikli",
"Sedat",
""
],
[
"Porikli",
"Vedat",
""
]
] |
2506.13781 | Pablo Ari\~no Fern\'andez | Pablo Ari\~no Fern\'andez | Solving the Job Shop Scheduling Problem with Graph Neural Networks: A Customizable Reinforcement Learning Environment | Bachelor's thesis, Universidad Polit\'ecnica de Madrid, 2025. 150 pages, 23 figures | null | null | null | cs.LG cs.AI cs.DM | http://creativecommons.org/licenses/by/4.0/ | The job shop scheduling problem is an NP-hard combinatorial optimization problem relevant to manufacturing and timetabling. Traditional approaches use priority dispatching rules based on simple heuristics. Recent work has attempted to replace these with deep learning models, particularly graph neural networks (GNNs), that learn to assign priorities from data. However, training such models requires customizing numerous factors: graph representation, node features, action space, and reward functions. The lack of modular libraries for experimentation makes this research time-consuming. This work introduces JobShopLib, a modular library that allows customizing these factors and creating new components with its reinforcement learning environment. We trained several dispatchers through imitation learning to demonstrate the environment's utility. One model outperformed various graph-based dispatchers using only individual operation features, highlighting the importance of feature customization. Our GNN model achieved near state-of-the-art results on large-scale problems. These results suggest significant room for improvement in developing such models. JobShopLib provides the necessary tools for future experimentation. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 08:09:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fernández",
"Pablo Ariño",
""
]
] |
2506.13782 | Ke Wang | Ke Wang, Bo Pan, Yingchaojie Feng, Yuwei Wu, Jieyi Chen, Minfeng Zhu, and Wei Chen | XGraphRAG: Interactive Visual Analysis for Graph-based Retrieval-Augmented Generation | Accepted to IEEE Pacific Visualization Conference 2025 | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Graph-based Retrieval-Augmented Generation (RAG) has shown great capability in enhancing Large Language Model (LLM)'s answer with an external knowledge base. Compared to traditional RAG, it introduces a graph as an intermediate representation to capture better structured relational knowledge in the corpus, elevating the precision and comprehensiveness of generation results. However, developers usually face challenges in analyzing the effectiveness of GraphRAG on their dataset due to GraphRAG's complex information processing pipeline and the overwhelming amount of LLM invocations involved during graph construction and query, which limits GraphRAG interpretability and accessibility. This research proposes a visual analysis framework that helps RAG developers identify critical recalls of GraphRAG and trace these recalls through the GraphRAG pipeline. Based on this framework, we develop XGraphRAG, a prototype system incorporating a set of interactive visualizations to facilitate users' analysis process, boosting failure cases collection and improvement opportunities identification. Our evaluation demonstrates the effectiveness and usability of our approach. Our work is open-sourced and available at https://github.com/Gk0Wk/XGraphRAG. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 09:14:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Ke",
""
],
[
"Pan",
"Bo",
""
],
[
"Feng",
"Yingchaojie",
""
],
[
"Wu",
"Yuwei",
""
],
[
"Chen",
"Jieyi",
""
],
[
"Zhu",
"Minfeng",
""
],
[
"Chen",
"Wei",
""
]
] |
2506.13783 | Soyeon Choi | Soyeon Choi, Kangwook Lee, Oliver Sng, Joshua M. Ackerman | Infected Smallville: How Disease Threat Shapes Sociality in LLM Agents | 8 pages | null | null | null | physics.soc-ph cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | How does the threat of infectious disease influence sociality among generative agents? We used generative agent-based modeling (GABM), powered by large language models, to experimentally test hypotheses about the behavioral immune system. Across three simulation runs, generative agents who read news about an infectious disease outbreak showed significantly reduced social engagement compared to agents who received no such news, including lower attendance at a social gathering, fewer visits to third places (e.g., cafe, store, park), and fewer conversations throughout the town. In interview responses, agents explicitly attributed their behavioral changes to disease-avoidance motivations. A validity check further indicated that they could distinguish between infectious and noninfectious diseases, selectively reducing social engagement only when there was a risk of infection. Our findings highlight the potential of GABM as an experimental tool for exploring complex human social dynamics at scale. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 18:59:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Choi",
"Soyeon",
""
],
[
"Lee",
"Kangwook",
""
],
[
"Sng",
"Oliver",
""
],
[
"Ackerman",
"Joshua M.",
""
]
] |
2506.13784 | Junting Zhou | Junting Zhou and Wang Li and Yiyan Liao and Nengyuan Zhang and Tingjia Miaoand Zhihui Qi and Yuhan Wu and Tong Yang | AcademicBrowse: Benchmarking Academic Browse Ability of LLMs | null | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs)' search capabilities have garnered significant attention. Existing benchmarks, such as OpenAI's BrowseComp, primarily focus on general search scenarios and fail to adequately address the specific demands of academic search. These demands include deeper literature tracing and organization, professional support for academic databases, the ability to navigate long-tail academic knowledge, and ensuring academic rigor. Here, we proposed AcademicBrowse, the first dataset specifically designed to evaluate the complex information retrieval capabilities of Large Language Models (LLMs) in academic research. AcademicBrowse possesses the following key characteristics: Academic Practicality, where question content closely mirrors real academic learning and research environments, avoiding deliberately misleading models; High Difficulty, with answers that are challenging for single models (e.g., Grok DeepSearch or Gemini Deep Research) to provide directly, often requiring at least three deep searches to derive; Concise Evaluation, where limiting conditions ensure answers are as unique as possible, accompanied by clear sources and brief solution explanations, greatly facilitating subsequent audit and verification, surpassing the current lack of analyzed search datasets both domestically and internationally; and Broad Coverage, as the dataset spans at least 15 different academic disciplines. Through AcademicBrowse, we expect to more precisely measure and promote the performance improvement of LLMs in complex academic information retrieval tasks. The data is available at: https://huggingface.co/datasets/PKU-DS-LAB/AcademicBrowse | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 02:05:23 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhou",
"Junting",
""
],
[
"Li",
"Wang",
""
],
[
"Liao",
"Yiyan",
""
],
[
"Zhang",
"Nengyuan",
""
],
[
"Qi",
"Tingjia Miaoand Zhihui",
""
],
[
"Wu",
"Yuhan",
""
],
[
"Yang",
"Tong",
""
]
] |
2506.13785 | Patrick Sutanto | Patrick Sutanto, Jonathan Kenrick, Max Lorenz, Joan Santoso | LLM-Driven Data Generation and a Novel Soft Metric for Evaluating Text-to-SQL in Aviation MRO | null | null | null | null | cs.DB cs.IR | http://creativecommons.org/licenses/by/4.0/ | The application of Large Language Models (LLMs) to text-to-SQL tasks promises to democratize data access, particularly in critical industries like aviation Maintenance, Repair, and Operation (MRO). However, progress is hindered by two key challenges: the rigidity of conventional evaluation metrics such as execution accuracy, which offer coarse, binary feedback, and the scarcity of domain-specific evaluation datasets. This paper addresses these gaps. To enable more nuanced assessment, we introduce a novel F1-score-based 'soft' metric that quantifies the informational overlap between generated and ground-truth SQL results. To address data scarcity, we propose an LLM-driven pipeline that synthesizes realistic question-SQL pairs from database schemas. We demonstrate our contributions through an empirical evaluation on an authentic MRO database. Our experiments show that the proposed soft metric provides more insightful performance analysis than strict accuracy, and our data generation technique is effective in creating a domain-specific benchmark. Together, these contributions offer a robust framework for evaluating and advancing text-to-SQL systems in specialized environments. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 04:04:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sutanto",
"Patrick",
""
],
[
"Kenrick",
"Jonathan",
""
],
[
"Lorenz",
"Max",
""
],
[
"Santoso",
"Joan",
""
]
] |
2506.13786 | Quang Vinh Tran | Vuong M. Ngo, Tran Quang Vinh, Patricia Kearney and Mark Roantree | Enhancing Bagging Ensemble Regression with Data Integration for Time Series-Based Diabetes Prediction | 17th International Conference on Computational Collective Intelligence, LNAI, Springer, 11 pages | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diabetes is a chronic metabolic disease characterized by elevated blood glucose levels, leading to complications like heart disease, kidney failure, and nerve damage. Accurate state-level predictions are vital for effective healthcare planning and targeted interventions, but in many cases, data for necessary analyses are incomplete. This study begins with a data engineering process to integrate diabetes-related datasets from 2011 to 2021 to create a comprehensive feature set. We then introduce an enhanced bagging ensemble regression model (EBMBag+) for time series forecasting to predict diabetes prevalence across U.S. cities. Several baseline models, including SVMReg, BDTree, LSBoost, NN, LSTM, and ERMBag, were evaluated for comparison with our EBMBag+ algorithm. The experimental results demonstrate that EBMBag+ achieved the best performance, with an MAE of 0.41, RMSE of 0.53, MAPE of 4.01, and an R2 of 0.9. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 04:21:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ngo",
"Vuong M.",
""
],
[
"Vinh",
"Tran Quang",
""
],
[
"Kearney",
"Patricia",
""
],
[
"Roantree",
"Mark",
""
]
] |
2506.13787 | Haoyang Feng | Yanjun Dai, Haoyang Feng, Yuan Gao | Analysis of Anonymous User Interaction Relationships and Prediction of Advertising Feedback Based on Graph Neural Network | null | null | null | null | cs.IR cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | While online advertising is highly dependent on implicit interaction networks of anonymous users for engagement inference, and for the selection and optimization of delivery strategies, existing graph models seldom can capture the multi-scale temporal, semantic and higher-order dependency features of these interaction networks, thus it's hard to describe the complicated patterns of the anonymous behavior. In this paper, we propose Decoupled Temporal-Hierarchical Graph Neural Network (DTH-GNN), which achieves three main contributions. Above all, we introduce temporal edge decomposition, which divides each interaction into three types of channels: short-term burst, diurnal cycle and long-range memory, and conducts feature extraction using the convolution kernel of parallel dilated residuals; Furthermore, our model builds a hierarchical heterogeneous aggregation, where user-user, user-advertisement, advertisement-advertisement subgraphs are combined through the meta-path conditional Transformer encoder, where the noise structure is dynamically tamped down via the synergy of cross-channel self-attention and gating relationship selector. Thirdly, the contrast regularity of feedback perception is formulated, the consistency of various time slices is maximized, the entropy of control exposure information with dual-view target is maximized, the global prototype of dual-momentum queue distillation is presented, and the strategy gradient layer with light weight is combined with delaying transformation signal to fine-tune the node representation for benefit-oriented. The AUC of DTH-GNN improved by 8.2% and the logarithmic loss improved by 5.7% in comparison with the best baseline model. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 04:40:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dai",
"Yanjun",
""
],
[
"Feng",
"Haoyang",
""
],
[
"Gao",
"Yuan",
""
]
] |
2506.13788 | Bilel HAFSI | Alexandre Westrelin (IEMN), Pierre Debavelear (ICAM), Mathieu Lefevbre (ICAM), Nour Abdallah (IEMN), Kamal Lmimouni (IEMN), Olivier Stienne, Bilel Hafsi (IEMN) | Design and Fabrication Of Multiplexed One-Port SAW Resonators On A Single Chip | null | null | null | null | physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Shear horizontal surface acoustic wave (SH-SAW) sensors are considered as a viable option for label-free, sensitive, real-time, and cost-effective detection. In this paper, we present the design, the fabrication and the characterization of multiplexed SAW sensors based on a single chip SAW resonator. Operating near the industrial, scientific and medical (ISM) bands, resonators were microfabricated on 42.75{\textdegree} Y (ST) cut Quartz. This substrate was selected due to its traditional association with temperature stability in the design of SAW devices. Parameters like aperture, number of interdigitated electrode (IDT) pairs, IDT wavelengths, and number of reflectors were studied experimentally to enhance electrical responses. Resonance frequency, quality factor and electromechanical coupling were extracted through impedance measurements and fitted with a modified Butterworth-Van Dyke (MBVD) model. Based on these results, we propose a novel design for a multi-frequency device that can serve as a single electronic nose for complex gases detection. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 06:22:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Westrelin",
"Alexandre",
"",
"IEMN"
],
[
"Debavelear",
"Pierre",
"",
"ICAM"
],
[
"Lefevbre",
"Mathieu",
"",
"ICAM"
],
[
"Abdallah",
"Nour",
"",
"IEMN"
],
[
"Lmimouni",
"Kamal",
"",
"IEMN"
],
[
"Stienne",
"Olivier",
"",
"IEMN"
],
[
"Hafsi",
"Bilel",
"",
"IEMN"
]
] |
2506.13789 | Jaekwang Kim | Subin Choi, Chanmi Jung, Dae Hoon Lee, Jeongan Choi, Jaekwang Kim | Numerical Modeling of n-Hexane Pyrolysis with an Optimized Kinetic Mechanism in a Hydrogen Plasma Reactor | 31 pages, 13 figures | null | null | null | physics.plasm-ph physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | The physicochemical mechanisms underlying the pyrolysis of n-hexane in a high temperature Ar-H2 environment were investigated for plasma pyrolysis process. An optimal chemical kinetics model was developed using the Reaction Mechanism Generator (RMG), an automated tool for constructing reaction mechanisms. This model was validated through 0-D analyses, where simulation result were compared with existing kinetic models (LLNL,JetSurf) and experimental data from conventional n-hexane pyrolysis. Subsequently, 1-D analysis were conducted to identify the optimal operational flow rate in plasma pyrolysis reactor, the results of which informed detailed three-dimensional (2-D) computational fluid dynamics (CFD) modeling of the plasma reactor. The CFD simulations reveal that fluid mixing dynamics play a dominant role in determining the extent of conversion and product selectivity, highlighting the limitations of lower-dimensional models in capturing essential transport phenomena. Notably, the simulations indicate a higher C2 monomer selectivity of approximately 50 % under plasma-based n-hexane pyrolysis, in contrast to the roughly 30 % selectivity achieved via conventional fossil-fuel-based methods. These findings underscore the potential advantages of plasma-driven pyrolysis and represent a critical step toward a comprehensive understanding of the complex thermochemical behavior governing plasma-assisted processes. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 07:13:46 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Choi",
"Subin",
""
],
[
"Jung",
"Chanmi",
""
],
[
"Lee",
"Dae Hoon",
""
],
[
"Choi",
"Jeongan",
""
],
[
"Kim",
"Jaekwang",
""
]
] |
2506.13791 | Pedro Hack | Pedro Hack | Belief propagation for networks with loops: The neighborhoods-intersections-based method | null | null | null | null | physics.soc-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In order to diminish the damaging effect of loops on belief propagation (BP), the first explicit version of generalized BP for networks, the KCN-method, was recently introduced. Despite its success, the KCN-method spends computational resources inefficiently. Such inefficiencies can quickly turn the exact application of the method unfeasible, since its time complexity increases exponentially with them. This affects for instance tree networks, for which, despite not offering any accuracy advantage with respect to BP, the time complexity of the KCN-method grows exponentially with the nodes' degree. To avoid these issues, we introduce here a new generalized BP scheme, the NIB-method, which only spends computational resources provided they are needed in order to account for correlations in the network. In fact, we show that, given a network with only short loops, the NIB-method is exact and optimal, and we characterize its time complexity reduction with respect to the KCN-method. If long loops are also present, both methods become approximate. In this scenario, we discuss the relation between the methods and we show how to interpolate between them, obtaining a richer family of generalized BP algorithms that trade accuracy for complexity. Lastly, we find a good agreement between the (approximate) KCN and NIB methods when computing the partition function for two artificial networks. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 13:01:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hack",
"Pedro",
""
]
] |
2506.13792 | Gon\c{c}alo Carvalho Hora de | Gon\c{c}alo Hora de Carvalho, Lazar S. Popov, Sander Kaatee, Kristinn R. Th\'orisson, Tangrui Li, P\'etur H\'uni Bj\"ornsson, Jilles S. Dibangoye | ICE-ID: A Novel Historical Census Data Benchmark Comparing NARS against LLMs, \& a ML Ensemble on Longitudinal Identity Resolution | null | null | null | null | cs.AI cs.CL cs.LG stat.AP | http://creativecommons.org/licenses/by/4.0/ | We introduce ICE-ID, a novel benchmark dataset for historical identity resolution, comprising 220 years (1703-1920) of Icelandic census records. ICE-ID spans multiple generations of longitudinal data, capturing name variations, demographic changes, and rich genealogical links. To the best of our knowledge, this is the first large-scale, open tabular dataset specifically designed to study long-term person-entity matching in a real-world population. We define identity resolution tasks (within and across census waves) with clearly documented metrics and splits. We evaluate a range of methods: handcrafted rule-based matchers, a ML ensemble as well as LLMs for structured data (e.g. transformer-based tabular networks) against a novel approach to tabular data called NARS (Non-Axiomatic Reasoning System) - a general-purpose AI framework designed to reason with limited knowledge and resources. Its core is Non-Axiomatic Logic (NAL), a term-based logic. Our experiments show that NARS is suprisingly simple and competitive with other standard approaches, achieving SOTA at our task. By releasing ICE-ID and our code, we enable reproducible benchmarking of identity resolution approaches in longitudinal settings and hope that ICE-ID opens new avenues for cross-disciplinary research in data linkage and historical analytics. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 13:46:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"de Carvalho",
"Gonçalo Hora",
""
],
[
"Popov",
"Lazar S.",
""
],
[
"Kaatee",
"Sander",
""
],
[
"Thórisson",
"Kristinn R.",
""
],
[
"Li",
"Tangrui",
""
],
[
"Björnsson",
"Pétur Húni",
""
],
[
"Dibangoye",
"Jilles S.",
""
]
] |
2506.13793 | Zongxian Yang | Zongxian Yang, Jiayu Qian, Zegao Peng, Haoyu Zhang, Zhi-An Huang | Med-REFL: Medical Reasoning Enhancement via Self-Corrected Fine-grained Reflection | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large reasoning models have recently made significant strides in mathematical and code reasoning, yet their success has not transferred smoothly to the medical domain. While multiple factors contribute to this disparity, a critical issue is the inadequate focus on the quality of intermediate reflection steps, which is particularly crucial in high-stakes medical scenarios. To address this challenge, we propose Med-REFL, a \underline{\textbf{Med}}ical \underline{\textbf{R}}easoning \underline{\textbf{E}}nhancement via self-corrected \underline{\textbf{F}}ine-grained ref\underline{\textbf{L}}ection. Our method leverages a tree-of-thought approach to decompose medical questions into fine-grained reasoning paths, quantitatively evaluating each step and its subsequent reflections. These assessments enable automatic construction of direct preference optimization data, reducing reliance on expensive expert annotations while guiding models to identify and correct reasoning errors. Experimental results on the MedQA-USMLE benchmark demonstrate Med-REFL achieves consistent improvements, with average gains up to 4.11\%. Notably, it further boosts the state-of-the-art performance of 7B/8B models by an additional 4.13\%. Furthermore, Med-REFL exhibits strong generalization capabilities and robustness across several challenging medical question-answering datasets. Our work illustrates that prioritizing reflection quality leads to more accurate and trustworthy reasoning in medical AI applications. Checkpoints, code, and data can be found \href{https://github.com/TianYin123/Med-REFL}{here}. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 14:58:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yang",
"Zongxian",
""
],
[
"Qian",
"Jiayu",
""
],
[
"Peng",
"Zegao",
""
],
[
"Zhang",
"Haoyu",
""
],
[
"Huang",
"Zhi-An",
""
]
] |
2506.13794 | Jared James Grogan | Jared James Grogan | AgentFacts: Universal KYA Standard for Verified AI Agent Metadata & Deployment | 15 pages, 1 figure, technical specification with appendix. Apache 2.0 license. Prior art established 2022 | null | null | null | cs.MA | http://creativecommons.org/licenses/by/4.0/ | Enterprise AI deployment faces critical "Know Your Agent" (KYA) challenges where organizations must verify third-party agent capabilities and establish trust without standardized metadata or verification infrastructure. Current approaches rely on self-declared capabilities and custom integration processes that create trust gaps and coordination friction limiting confident enterprise adoption. This paper presents AgentFacts, a universal metadata standard that enables systematic agent verification through cryptographically-signed capability declarations, multi-authority validation, and dynamic permission management. The specification introduces domain-specialized verification where different trusted authorities validate specific metadata aspects based on their expertise, eliminating single points of trust failure while enabling graduated confidence assessment. AgentFacts transforms agent procurement from custom integration projects into standardized workforce management, providing the transparency and governance infrastructure necessary for enterprise AI coordination at scale. | [
{
"version": "v1",
"created": "Wed, 11 Jun 2025 17:28:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Grogan",
"Jared James",
""
]
] |
2506.13795 | Boshen Shi | Boshen Shi, Yongqing Wang, Fangda Guo, Jiangli Shao, Huawei Shen, Xueqi Cheng | BotTrans: A Multi-Source Graph Domain Adaptation Approach for Social Bot Detection | Accetpted to ECML-PKDD 2025 Research Track as oral; Code&data: https://github.com/Skyorca/BotTrans | null | null | null | cs.AI cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transferring extensive knowledge from relevant social networks has emerged as a promising solution to overcome label scarcity in detecting social bots and other anomalies with GNN-based models. However, effective transfer faces two critical challenges. Firstly, the network heterophily problem, which is caused by bots hiding malicious behaviors via indiscriminately interacting with human users, hinders the model's ability to learn sufficient and accurate bot-related knowledge from source domains. Secondly, single-source transfer might lead to inferior and unstable results, as the source network may embody weak relevance to the task and provide limited knowledge. To address these challenges, we explore multiple source domains and propose a multi-source graph domain adaptation model named \textit{BotTrans}. We initially leverage the labeling knowledge shared across multiple source networks to establish a cross-source-domain topology with increased network homophily. We then aggregate cross-domain neighbor information to enhance the discriminability of source node embeddings. Subsequently, we integrate the relevance between each source-target pair with model optimization, which facilitates knowledge transfer from source networks that are more relevant to the detection task. Additionally, we propose a refinement strategy to improve detection performance by utilizing semantic knowledge within the target domain. Extensive experiments on real-world datasets demonstrate that \textit{BotTrans} outperforms the existing state-of-the-art methods, revealing its efficacy in leveraging multi-source knowledge when the target detection task is unlabeled. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 02:10:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shi",
"Boshen",
""
],
[
"Wang",
"Yongqing",
""
],
[
"Guo",
"Fangda",
""
],
[
"Shao",
"Jiangli",
""
],
[
"Shen",
"Huawei",
""
],
[
"Cheng",
"Xueqi",
""
]
] |
2506.13796 | Zhou Chen | Zhou Chen, Xiao Wang, Yuanhong Liao, Ming Lin, Yuqi Bai | ClimateChat: Designing Data and Methods for Instruction Tuning LLMs to Answer Climate Change Queries | ICLR 2025 camera ready, 13 pages, 4 figures, 4 tables | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As the issue of global climate change becomes increasingly severe, the demand for research in climate science continues to grow. Natural language processing technologies, represented by Large Language Models (LLMs), have been widely applied to climate change-specific research, providing essential information support for decision-makers and the public. Some studies have improved model performance on relevant tasks by constructing climate change-related instruction data and instruction-tuning LLMs. However, current research remains inadequate in efficiently producing large volumes of high-precision instruction data for climate change, which limits further development of climate change LLMs. This study introduces an automated method for constructing instruction data. The method generates instructions using facts and background knowledge from documents and enhances the diversity of the instruction data through web scraping and the collection of seed instructions. Using this method, we constructed a climate change instruction dataset, named ClimateChat-Corpus, which was used to fine-tune open-source LLMs, resulting in an LLM named ClimateChat. Evaluation results show that ClimateChat significantly improves performance on climate change question-and-answer tasks. Additionally, we evaluated the impact of different base models and instruction data on LLM performance and demonstrated its capability to adapt to a wide range of climate change scientific discovery tasks, emphasizing the importance of selecting an appropriate base model for instruction tuning. This research provides valuable references and empirical support for constructing climate change instruction data and training climate change-specific LLMs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 08:43:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chen",
"Zhou",
""
],
[
"Wang",
"Xiao",
""
],
[
"Liao",
"Yuanhong",
""
],
[
"Lin",
"Ming",
""
],
[
"Bai",
"Yuqi",
""
]
] |
2506.13797 | Gaspare Galati | Gaspare Galati and Gabriele Pavan and Frederick Daum | Quantum Radar and Research Assessment | five Sections; three Tables; three Figures | null | null | null | physics.soc-ph eess.SP | http://creativecommons.org/licenses/by/4.0/ | Quantum Radar was studied in many Nations for about fifteen years with the production of some hundred publications. In the post 2020 literature, it is shown that, due to the exceedingly low transmitted power, Quantum Radar cannot produce neither significant results nor real world applications. Regrettably, most of the scientific community ignored this negative outcome: a fact worth of exam. A detailed study of such an assessment failure depicts the main shortcomings of the present situation, calling for a redesign of the research assessment at the international level, with proposals shown in the ending section of this paper. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 10:43:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Galati",
"Gaspare",
""
],
[
"Pavan",
"Gabriele",
""
],
[
"Daum",
"Frederick",
""
]
] |
2506.13798 | Roger Brent | Roger Brent and T. Greg McKelvey Jr | Contemporary AI foundation models increase biological weapons risk | 58 pages, 10 figures, 4 tables | null | null | WR-A3853-1 | cs.CY cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The rapid advancement of artificial intelligence has raised concerns about its potential to facilitate biological weapons development. We argue existing safety assessments of contemporary foundation AI models underestimate this risk, largely due to flawed assumptions and inadequate evaluation methods. First, assessments mistakenly assume biological weapons development requires tacit knowledge, or skills gained through hands-on experience that cannot be easily verbalized. Second, they rely on imperfect benchmarks that overlook how AI can uplift both nonexperts and already-skilled individuals. To challenge the tacit knowledge assumption, we examine cases where individuals without formal expertise, including a 2011 Norwegian ultranationalist who synthesized explosives, successfully carried out complex technical tasks. We also review efforts to document pathogen construction processes, highlighting how such tasks can be conveyed in text. We identify "elements of success" for biological weapons development that large language models can describe in words, including steps such as acquiring materials and performing technical procedures. Applying this framework, we find that advanced AI models Llama 3.1 405B, ChatGPT-4o, and Claude 3.5 Sonnet can accurately guide users through the recovery of live poliovirus from commercially obtained synthetic DNA, challenging recent claims that current models pose minimal biosecurity risk. We advocate for improved benchmarks, while acknowledging the window for meaningful implementation may have already closed. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 19:53:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Brent",
"Roger",
""
],
[
"McKelvey",
"T. Greg",
"Jr"
]
] |
2506.13799 | Soroush Vahidi | Soroush Vahidi | Feedforward Ordering in Neural Connectomes via Feedback Arc Minimization | This is a preliminary paper | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a suite of scalable algorithms for minimizing feedback arcs in large-scale weighted directed graphs, with the goal of revealing biologically meaningful feedforward structure in neural connectomes. Using the FlyWire Connectome Challenge dataset, we demonstrate the effectiveness of our ranking strategies in maximizing the total weight of forward-pointing edges. Our methods integrate greedy heuristics, gain-aware local refinements, and global structural analysis based on strongly connected components. Experiments show that our best solution improves the forward edge weight over previous top-performing methods. All algorithms are implemented efficiently in Python and validated using cloud-based execution on Google Colab Pro+. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 03:03:55 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Vahidi",
"Soroush",
""
]
] |
2506.13800 | Abul Ehtesham | Abul Ehtesham, Aditi Singh, Saket Kumar | Enhancing Clinical Decision Support and EHR Insights through LLMs and the Model Context Protocol: An Open-Source MCP-FHIR Framework | null | null | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Enhancing clinical decision support (CDS), reducing documentation burdens, and improving patient health literacy remain persistent challenges in digital health. This paper presents an open-source, agent-based framework that integrates Large Language Models (LLMs) with HL7 FHIR data via the Model Context Protocol (MCP) for dynamic extraction and reasoning over electronic health records (EHRs). Built on the established MCP-FHIR implementation, the framework enables declarative access to diverse FHIR resources through JSON-based configurations, supporting real-time summarization, interpretation, and personalized communication across multiple user personas, including clinicians, caregivers, and patients. To ensure privacy and reproducibility, the framework is evaluated using synthetic EHR data from the SMART Health IT sandbox (https://r4.smarthealthit.org/), which conforms to the FHIR R4 standard. Unlike traditional approaches that rely on hardcoded retrieval and static workflows, the proposed method delivers scalable, explainable, and interoperable AI-powered EHR applications. The agentic architecture further supports multiple FHIR formats, laying a robust foundation for advancing personalized digital health solutions. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 04:07:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Ehtesham",
"Abul",
""
],
[
"Singh",
"Aditi",
""
],
[
"Kumar",
"Saket",
""
]
] |
2506.13802 | Soumyaranjan Panda | S. Panda (a and b), P. K. Netrakanti (a), S. P. Behera (a and b), R. R. Sahu (a), K. Kumar (a and b), R. Sehgal (a), D. K. Mishra (a and b), V. Jha (a and b) ((a) Nuclear Physics Division, Bhabha Atomic Research Centre, Trombay, Mumbai, India, (b) Homi Bhabha National Institute, Anushakti Nagar, Mumbai, India) | Discrimination of neutron-$\gamma$ in the low energy regime using machine learning for an EJ-276D plastic scintillator | null | null | null | null | physics.ins-det hep-ex nucl-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present results for discrimination of neutron and $\gamma$ events using a plastic scintillator detector with pulse shape discrimination capabilities. Machine learning (ML) algorithms are used to improve the discriminatory power between neutron and $\gamma$ events at lower energy ranges which otherwise are not addressed by the conventional pulse shape discrimination techniques. The use of a multilayer perceptron with Bayesian inference (MLPBNN) and support vector machine (SVM) algorithms are studied using the recorded waveforms from the detector. Three variables are constructed for the input of ML algorithms, which captures the essence of the differences in the head and tail part of the neutron and $\gamma$ waveforms. The training and the testing of the ML algorithms are done using an AmBe neutron source. In the lower energy region, the results obtained from the ML predictions are compared with the results obtained from a time of flight (ToF) technique to benchmark the overall performance of the ML algorithms. A reasonable agreement is observed between the results obtained from ML algorithm and the ToF experiment in the studied energy range. The MLPBNN gives better discriminatory power for the neutron and $\gamma$ events than the SVM algorithm. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 12:47:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Panda",
"S.",
"",
"a and b"
],
[
"Netrakanti",
"P. K.",
"",
"a"
],
[
"Behera",
"S. P.",
"",
"a and b"
],
[
"Sahu",
"R. R.",
"",
"a"
],
[
"Kumar",
"K.",
"",
"a and b"
],
[
"Sehgal",
"R.",
"",
"a"
],
[
"Mishra",
"D. K.",
"",
"a and b"
],
[
"Jha",
"V.",
"",
"a and b"
]
] |
2506.13803 | Richard Lange | Richard D. Lange, Konrad P. Kording | Causality in the human niche: lessons for machine learning | 23 pages, 2 figures | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Humans interpret the world around them in terms of cause and effect and communicate their understanding of the world to each other in causal terms. These causal aspects of human cognition are thought to underlie humans' ability to generalize and learn efficiently in new domains, an area where current machine learning systems are weak. Building human-like causal competency into machine learning systems may facilitate the construction of effective and interpretable AI. Indeed, the machine learning community has been importing ideas on causality formalized by the Structural Causal Model (SCM) framework, which provides a rigorous formal language for many aspects of causality and has led to significant advances. However, the SCM framework fails to capture some salient aspects of human causal cognition and has likewise not yet led to advances in machine learning in certain critical areas where humans excel. We contend that the problem of causality in the ``human niche'' -- for a social, autonomous, and goal-driven agent sensing and acting in the world in which humans live -- is quite different from the kind of causality captured by SCMs. For example, everyday objects come in similar types that have similar causal properties, and so humans readily generalize knowledge of one type of object (cups) to another related type (bowls) by drawing causal analogies between objects with similar properties, but such analogies are at best awkward to express in SCMs. We explore how such causal capabilities are adaptive in, and motivated by, the human niche. By better appreciating properties of human causal cognition and, crucially, how those properties are adaptive in the niche in which humans live, we hope that future work at the intersection of machine learning and causality will leverage more human-like inductive biases to create more capable, controllable, and interpretable systems. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 14:13:52 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lange",
"Richard D.",
""
],
[
"Kording",
"Konrad P.",
""
]
] |
2506.13804 | Sarah McDaid PhD | Edward McDaid, Sarah McDaid | Instruction and Solution Probabilities as Heuristics for Inductive Programming | 10 pages, 10 figures | null | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Instruction subsets (ISs) are heuristics that can shrink the size of the inductive programming (IP) search space by tens of orders of magnitude. Here, we extend the IS approach by introducing instruction and solution probabilities as additional heuristics. Instruction probability reflects the expectation of an instruction occurring in a solution, based on the frequency of instruction occurrence in a large code sample. The solution probability for a partial or complete program is simply the product of all constituent instruction probabilities, including duplicates. We treat the minimum solution probabilities observed in code sample program units of different sizes as solution probability thresholds. These thresholds are used to prune the search space as partial solutions are constructed, thereby eliminating any branches containing unlikely combinations of instructions. The new approach has been evaluated using a large sample of human code. We tested two formulations of instruction probability: one based on instruction occurrence across the entire code sample and another that measured the distribution separately for each IS. Our results show that both variants produce substantial further reductions in the IP search space size of up to tens of orders of magnitude, depending on solution size. In combination with IS, reductions of over 100 orders of magnitude can be achieved. We also carried out cross-validation testing to show that the heuristics should work effectively with unseen code. The approach is described and the results and some ideas for future work are discussed. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 15:24:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"McDaid",
"Edward",
""
],
[
"McDaid",
"Sarah",
""
]
] |
2506.13805 | Bonam Mingole | Bonam Mingole, Aditya Majumdar, Firdaus Ahmed Choudhury, Jennifer L. Kraschnewski, Shyam S. Sundar, Amulya Yadav | Dr. GPT Will See You Now, but Should It? Exploring the Benefits and Harms of Large Language Models in Medical Diagnosis using Crowdsourced Clinical Cases | null | null | null | null | cs.CY cs.AI | http://creativecommons.org/licenses/by/4.0/ | The proliferation of Large Language Models (LLMs) in high-stakes applications such as medical (self-)diagnosis and preliminary triage raises significant ethical and practical concerns about the effectiveness, appropriateness, and possible harmfulness of the use of these technologies for health-related concerns and queries. Some prior work has considered the effectiveness of LLMs in answering expert-written health queries/prompts, questions from medical examination banks, or queries based on pre-existing clinical cases. Unfortunately, these existing studies completely ignore an in-the-wild evaluation of the effectiveness of LLMs in answering everyday health concerns and queries typically asked by general users, which corresponds to the more prevalent use case for LLMs. To address this research gap, this paper presents the findings from a university-level competition that leveraged a novel, crowdsourced approach for evaluating the effectiveness of LLMs in answering everyday health queries. Over the course of a week, a total of 34 participants prompted four publicly accessible LLMs with 212 real (or imagined) health concerns, and the LLM generated responses were evaluated by a team of nine board-certified physicians. At a high level, our findings indicate that on average, 76% of the 212 LLM responses were deemed to be accurate by physicians. Further, with the help of medical professionals, we investigated whether RAG versions of these LLMs (powered with a comprehensive medical knowledge base) can improve the quality of responses generated by LLMs. Finally, we also derive qualitative insights to explain our quantitative findings by conducting interviews with seven medical professionals who were shown all the prompts in our competition. This paper aims to provide a more grounded understanding of how LLMs perform in real-world everyday health communication. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 17:12:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mingole",
"Bonam",
""
],
[
"Majumdar",
"Aditya",
""
],
[
"Choudhury",
"Firdaus Ahmed",
""
],
[
"Kraschnewski",
"Jennifer L.",
""
],
[
"Sundar",
"Shyam S.",
""
],
[
"Yadav",
"Amulya",
""
]
] |
2506.13806 | Nick P. Proukakis | Nick P. Proukakis | A Century of Bose-Einstein Condensation | Brief Comment to appear in Communications Physics (Nature) | null | null | null | cond-mat.quant-gas cond-mat.stat-mech quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bose-Einstein Condensation is a phenomenon at the heart of many of the past century's most intriguing and fundamental manifestations, such as superfluidity and superconductivity. It was discovered theoretically some 100 years ago, and unequivocally experimentally demonstrated in the context of weakly interacting gases 30 years ago. Since then, it has spawned a revolution in our understanding of fundamental phases of matter and collective quantum dynamics extending across all physical scales and energies, with unforeseen implications and the potential for envisaged quantum technological applications. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 19:20:26 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Proukakis",
"Nick P.",
""
]
] |
2506.13807 | Florian Kofler | Florian Kofler, Marcel Rosier, Mehdi Astaraki, Ujjwal Baid, Hendrik M\"oller, Josef A. Buchner, Felix Steinbauer, Eva Oswald, Ezequiel de la Rosa, Ivan Ezhov, Constantin von See, Jan Kirschke, Anton Schmick, Sarthak Pati, Akis Linardos, Carla Pitarch, Sanyukta Adap, Jeffrey Rudie, Maria Correia de Verdier, Rachit Saluja, Evan Calabrese, Dominic LaBella, Mariam Aboian, Ahmed W. Moawad, Nazanin Maleki, Udunna Anazodo, Maruf Adewole, Marius George Linguraru, Anahita Fathi Kazerooni, Zhifan Jiang, Gian Marco Conte, Hongwei Li, Juan Eugenio Iglesias, Spyridon Bakas, Benedikt Wiestler, Marie Piraud and Bjoern Menze | BraTS orchestrator : Democratizing and Disseminating state-of-the-art brain tumor image analysis | 27p, 2figs, 3tabs | null | null | null | eess.IV cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Brain Tumor Segmentation (BraTS) cluster of challenges has significantly advanced brain tumor image analysis by providing large, curated datasets and addressing clinically relevant tasks. However, despite its success and popularity, algorithms and models developed through BraTS have seen limited adoption in both scientific and clinical communities. To accelerate their dissemination, we introduce BraTS orchestrator, an open-source Python package that provides seamless access to state-of-the-art segmentation and synthesis algorithms for diverse brain tumors from the BraTS challenge ecosystem. Available on GitHub (https://github.com/BrainLesion/BraTS), the package features intuitive tutorials designed for users with minimal programming experience, enabling both researchers and clinicians to easily deploy winning BraTS algorithms for inference. By abstracting the complexities of modern deep learning, BraTS orchestrator democratizes access to the specialized knowledge developed within the BraTS community, making these advances readily available to broader neuro-radiology and neuro-oncology audiences. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 21:28:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kofler",
"Florian",
""
],
[
"Rosier",
"Marcel",
""
],
[
"Astaraki",
"Mehdi",
""
],
[
"Baid",
"Ujjwal",
""
],
[
"Möller",
"Hendrik",
""
],
[
"Buchner",
"Josef A.",
""
],
[
"Steinbauer",
"Felix",
""
],
[
"Oswald",
"Eva",
""
],
[
"de la Rosa",
"Ezequiel",
""
],
[
"Ezhov",
"Ivan",
""
],
[
"von See",
"Constantin",
""
],
[
"Kirschke",
"Jan",
""
],
[
"Schmick",
"Anton",
""
],
[
"Pati",
"Sarthak",
""
],
[
"Linardos",
"Akis",
""
],
[
"Pitarch",
"Carla",
""
],
[
"Adap",
"Sanyukta",
""
],
[
"Rudie",
"Jeffrey",
""
],
[
"de Verdier",
"Maria Correia",
""
],
[
"Saluja",
"Rachit",
""
],
[
"Calabrese",
"Evan",
""
],
[
"LaBella",
"Dominic",
""
],
[
"Aboian",
"Mariam",
""
],
[
"Moawad",
"Ahmed W.",
""
],
[
"Maleki",
"Nazanin",
""
],
[
"Anazodo",
"Udunna",
""
],
[
"Adewole",
"Maruf",
""
],
[
"Linguraru",
"Marius George",
""
],
[
"Kazerooni",
"Anahita Fathi",
""
],
[
"Jiang",
"Zhifan",
""
],
[
"Conte",
"Gian Marco",
""
],
[
"Li",
"Hongwei",
""
],
[
"Iglesias",
"Juan Eugenio",
""
],
[
"Bakas",
"Spyridon",
""
],
[
"Wiestler",
"Benedikt",
""
],
[
"Piraud",
"Marie",
""
],
[
"Menze",
"Bjoern",
""
]
] |
2506.13808 | Marcus Hans | Sebastian Lellig, Subisha Balakumar, Peter Schweizer, Eva P. Mayer, Simon Evertz, Marcus Hans, Damian M. Holzapfel, Yin Du, Qing Zhou, Martin Dienwiebel, Johann Michler, Jochen M. Schneider | Designing artificial zinc phosphate tribofilms with tailored mechanical properties by altering the chain length | Manuscript: 19 pages, 6 figures. Supplementary information: 2 pages, 3 figures | null | null | null | cond-mat.mtrl-sci physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Zinc dialkyldithiophosphate (ZDDP), as the most prominent lubrication additive, forms tribofilms consisting primarily of zinc phosphate glasses containing sulfides. As sulfur is linked to environmental concerns, sulfur-free zinc phosphate coatings have been sputtered from a Zn3(PO4)2 target and investigated here. Based on the bridging to non-bridging oxygen ratio, determined by X-ray photoelectron spectroscopy (XPS), the as deposited coatings are classified as metaphosphates. As the annealing temperature is increased, the chain lengths are reduced, as witnessed by XPS data indicated by a loss of phosphorus and oxygen of the coating surface, likely due to hydrolysis with water from the atmosphere. Transmission electron microscopy energy-dispersive X-ray spectroscopy line scans show that the XPS-revealed composition change of the coating surface upon annealing occurs over the whole thickness of the coating. This alteration in composition and chain length reductions causes a rise in hardness, reduced Young's modulus, and wear resistance. Therefore, the properties of the artificial zinc phosphate tribofilms can be tailored via a thermally stimulated composition change, causing an alternation in chain length from meta- to orthophosphate and thereby enabling the design of coatings with desired mechanical properties. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 21:46:34 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lellig",
"Sebastian",
""
],
[
"Balakumar",
"Subisha",
""
],
[
"Schweizer",
"Peter",
""
],
[
"Mayer",
"Eva P.",
""
],
[
"Evertz",
"Simon",
""
],
[
"Hans",
"Marcus",
""
],
[
"Holzapfel",
"Damian M.",
""
],
[
"Du",
"Yin",
""
],
[
"Zhou",
"Qing",
""
],
[
"Dienwiebel",
"Martin",
""
],
[
"Michler",
"Johann",
""
],
[
"Schneider",
"Jochen M.",
""
]
] |
2506.13809 | Roman Belavkin | Roman V. Belavkin | Analysis and Optimization of Probabilities of Beneficial Mutation and Crossover Recombination in a Hamming Space | 42 pages | null | null | null | q-bio.PE cs.AI math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inspired by Fisher's geometric approach to study beneficial mutations, we analyse probabilities of beneficial mutation and crossover recombination of strings in a general Hamming space with arbitrary finite alphabet. Mutations and recombinations that reduce the distance to an optimum are considered as beneficial. Geometric and combinatorial analysis is used to derive closed-form expressions for transition probabilities between spheres around an optimum giving a complete description of Markov evolution of distances from an optimum over multiple generations. This paves the way for optimization of parameters of mutation and recombination operators. Here we derive optimality conditions for mutation and recombination radii maximizing the probabilities of mutation and crossover into the optimum. The analysis highlights important differences between these evolutionary operators. While mutation can potentially reach any part of the search space, the probability of beneficial mutation decreases with distance to an optimum, and the optimal mutation radius or rate should also decrease resulting in a slow-down of evolution near the optimum. Crossover recombination, on the other hand, acts in a subspace of the search space defined by the current population of strings. However, probabilities of beneficial and deleterious crossover are balanced, and their characteristics, such as variance, are translation invariant in a Hamming space, suggesting that recombination may complement mutation and boost the rate of evolution near the optimum. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 22:24:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Belavkin",
"Roman V.",
""
]
] |
2506.13810 | Olivier Saidi | Olivier Saidi | Bridging Pattern-Aware Complexity with NP-Hard Optimization: A Unifying Framework and Empirical Study | null | null | 10.5281/zenodo.15006676 | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | NP hard optimization problems like the Traveling Salesman Problem (TSP) defy efficient solutions in the worst case, yet real-world instances often exhibit exploitable patterns. We propose a novel patternaware complexity framework that quantifies and leverages structural regularities e.g., clustering, symmetry to reduce effective computational complexity across domains, including financial forecasting and LLM optimization. With rigorous definitions, theorems, and a meta learning driven solver pipeline, we introduce metrics like Pattern Utilization Efficiency (PUE) and achieve up to 79 percent solution quality gains in TSP benchmarks (22 to 2392 cities). Distinct from theoretical NP hardness, our approach offers a unified, practical lens for pattern-driven efficiency. | [
{
"version": "v1",
"created": "Wed, 12 Mar 2025 11:05:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Saidi",
"Olivier",
""
]
] |
2506.13811 | Sompote Youwai Dr. | Sompote Youwai, David Phim, Vianne Gayl Murcia, Rianne Clair Onas | Investigating the Potential of Large Language Model-Based Router Multi-Agent Architectures for Foundation Design Automation: A Task Classification and Expert Selection Study | null | null | null | null | cs.MA cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | This study investigates router-based multi-agent systems for automating foundation design calculations through intelligent task classification and expert selection. Three approaches were evaluated: single-agent processing, multi-agent designer-checker architecture, and router-based expert selection. Performance assessment utilized baseline models including DeepSeek R1, ChatGPT 4 Turbo, Grok 3, and Gemini 2.5 Pro across shallow foundation and pile design scenarios. The router-based configuration achieved performance scores of 95.00% for shallow foundations and 90.63% for pile design, representing improvements of 8.75 and 3.13 percentage points over standalone Grok 3 performance respectively. The system outperformed conventional agentic workflows by 10.0 to 43.75 percentage points. Grok 3 demonstrated superior standalone performance without external computational tools, indicating advances in direct LLM mathematical reasoning for engineering applications. The dual-tier classification framework successfully distinguished foundation types, enabling appropriate analytical approaches. Results establish router-based multi-agent systems as optimal for foundation design automation while maintaining professional documentation standards. Given safety-critical requirements in civil engineering, continued human oversight remains essential, positioning these systems as advanced computational assistance tools rather than autonomous design replacements in professional practice. | [
{
"version": "v1",
"created": "Fri, 13 Jun 2025 23:45:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Youwai",
"Sompote",
""
],
[
"Phim",
"David",
""
],
[
"Murcia",
"Vianne Gayl",
""
],
[
"Onas",
"Rianne Clair",
""
]
] |
2506.13812 | Humberto Mart\'inez-Huerta | Jonathan Rinc\'on, Humberto Mart\'inez-Huerta, Adolfo Huet, Alberto Hern\'andez-Almada, and Miguel A. Garc\'ia-Aspeitia | Late-Time Cosmic Acceleration from QCD Confinement Dynamics | 12 pages, 6 figures | null | null | null | physics.gen-ph gr-qc | http://creativecommons.org/licenses/by/4.0/ | We explore a phenomenological extension of the Polyakov-Nambu-Jona-Lasinio (PNJL) model by introducing a coupling between the Polyakov loop potential and the Hubble parameter $H(t)$, inspired by the hypothesis that the QCD vacuum, particularly in the confined phase, may be sensitive to the large-scale dynamics of the universe. The coupling term, proportional to $(H(t)/H_0)^d f(\Phi)$, vanishes in the deconfined regime and acts as a dynamical dark energy contribution at late times, without invoking a fundamental cosmological constant. This framework maintains the thermodynamic consistency of the PNJL model and introduces a minimal modification to the effective potential, controlled by a single exponent $d$. We confront the model with low-redshift cosmological data, including cosmic chronometers, Type Ia supernovae, HII galaxies, and quasars. Using a Bayesian Monte Carlo analysis, we derive constraints on the model parameters and compare its performance with LambdaCDM. Our results show that the modified PNJL cosmology provides a statistically competitive fit to current data within 68% C.L.; while offering a theoretically motivated alternative to standard dark energy scenarios. We discuss the implications for QCD in curved spacetime and suggest avenues for further theoretical and observational investigation. | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 12:21:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Rincón",
"Jonathan",
""
],
[
"Martínez-Huerta",
"Humberto",
""
],
[
"Huet",
"Adolfo",
""
],
[
"Hernández-Almada",
"Alberto",
""
],
[
"García-Aspeitia",
"Miguel A.",
""
]
] |
2506.13813 | Rubens Machado | Rubens E. G. Machado, Giovanni C. Tauil, Nicholas Schweder-Souza | Accuracy of analytic potentials for orbits of satellites around a Milky Way-like galaxy: comparison with $N$-body simulations | 18 pages, 11 figures, accepted for publication in Universe | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | To study the orbits of satellites, a galaxy could be modelled either by means of a static gravitational potential, or by live $N$-body particles. Analytic potentials allow for fast calculations, but are idealized and non-responsive. On the other hand, $N$-body simulations are more realistic, but demand higher computational cost. Our goal is to characterize the regimes in which analytic potentials provide a sufficient approximation, and those where $N$-bodies are necessary. We perform two sets of simulations using both Gala and Gadget, in order to closely compare the orbital evolution of satellites around a Milky Way-like galaxy. Focusing on the periods when the satellite has not yet been severely disrupted by tidal forces, we find that the orbits of satellites up to $10^{8} {\rm M_{\odot}}$ can be reliably computed with analytic potentials to within 5% error, if they are circular or moderately eccentric. If the satellite is as massive as $10^{9} {\rm M_{\odot}}$, errors of 9% are to be expected. However, if the orbital radius is smaller than 30 kpc, the results may not be relied upon with the same accuracy beyond 1--2 Gyr. | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 18:04:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Machado",
"Rubens E. G.",
""
],
[
"Tauil",
"Giovanni C.",
""
],
[
"Schweder-Souza",
"Nicholas",
""
]
] |
2506.13814 | Lufei Liu | Lufei Liu, Tor M. Aamodt | ReFrame: Layer Caching for Accelerated Inference in Real-Time Rendering | Published at ICML 2025 | null | null | null | cs.GR cs.LG eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graphics rendering applications increasingly leverage neural networks in tasks such as denoising, supersampling, and frame extrapolation to improve image quality while maintaining frame rates. The temporal coherence inherent in these tasks presents an opportunity to reuse intermediate results from previous frames and avoid redundant computations. Recent work has shown that caching intermediate features to be reused in subsequent inferences is an effective method to reduce latency in diffusion models. We extend this idea to real-time rendering and present ReFrame, which explores different caching policies to optimize trade-offs between quality and performance in rendering workloads. ReFrame can be applied to a variety of encoder-decoder style networks commonly found in rendering pipelines. Experimental results show that we achieve 1.4x speedup on average with negligible quality loss in three real-time rendering tasks. Code available: https://ubc-aamodt-group.github.io/reframe-layer-caching/ | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 20:17:43 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Lufei",
""
],
[
"Aamodt",
"Tor M.",
""
]
] |
2506.13815 | Shrinivass Arunachalam Balasubramanian | Shrinivass Arunachalam Balasubramanian | Signal-First Architectures: Rethinking Front-End Reactivity | 18 pages, 4 figures | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Modern front-end frameworks face escalating reactivity management challenges, including performance degradation from complex observable chains and unpredictable re-renders. This paper introduces Signal-First Architecture--a novel paradigm where granular, dependency-tracked signals are the atomic unit of reactivity. Unlike traditional RxJS or NgRx patterns, Signal-First enforces reactive flows from explicit signal declarations, with derived values via computed() and side effects scoped to effect(). This model ensures deterministic behavior by eliminating implicit subscriptions and optimizing reactive graph evaluation.
We present a comparative analysis of three Angular reactivity models: RxJS service-based, NgRx global stores, and pure Signal-First implementations. Through controlled benchmarking, including Chrome DevTools performance tracing, memory heap snapshots, and Lighthouse audits, this study quantifies Signal-First advantages. | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 20:34:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Balasubramanian",
"Shrinivass Arunachalam",
""
]
] |
2506.13816 | Florian Neukart | Florian Neukart, Eike Marx, Valerii Vinokur | Information Wells and the Emergence of Primordial Black Holes in a Cyclic Quantum Universe | 18 pages, 12 figures | null | null | null | physics.gen-ph | http://creativecommons.org/licenses/by/4.0/ | Primordial black holes (PBHs) remain one of the most intriguing candidates for dark matter and a unique probe of physics at extreme curvatures. Here, we examine their formation in a bounce cosmology when the post-crunch universe inherits a highly inhomogeneous distribution of imprint entropy from the Quantum Memory Matrix (QMM). Within QMM, every Planck-scale cell stores quantum information about infalling matter; the surviving entropy field S(x) contributes an effective dust component T^QMM_{\mu\nu} = lambda * [ (nabla_mu S)(nabla_nu S) - (1/2) * g_{\mu\nu} * (nabla S)^2 + ... ] that deepens curvature wherever S is large. We show that (i) reasonable bounce temperatures and a QMM coupling lambda ~ O(1) naturally amplify these "information wells" until the density contrast exceeds the critical value delta_c ~ 0.3; (ii) the resulting PBH mass spectrum spans 10^{-16} to 10^3 solar masses, matching current microlensing and PTA windows; and (iii) the same mechanism links PBH abundance to earlier QMM explanations of dark matter and the cosmic matter-antimatter imbalance. Observable signatures include a mild blue tilt in small-scale power, characteristic mu-distortions, and an enhanced integrated Sachs-Wolfe signal - all of which will be tested by upcoming CMB, PTA, and lensing surveys. | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 22:33:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Neukart",
"Florian",
""
],
[
"Marx",
"Eike",
""
],
[
"Vinokur",
"Valerii",
""
]
] |
2506.13817 | Saleem Al Dajani | Saleem A. Al Dajani, Abel Sanchez, John R. Williams | DeepSeq: High-Throughput Single-Cell RNA Sequencing Data Labeling via Web Search-Augmented Agentic Generative AI Foundation Models | 4 pages, 5 figures, Accepted by ICML 2025 FM4LS https://openreview.net/forum?id=zNjXOZxEYB . Workshop on Multi-modal Foundation Models and Large Language Models for Life Sciences (FM4LS)}, July 2025 | International Conference on Machine Learning (ICML). Workshop on Multi-modal Foundation Models and Large Language Models for Life Sciences (FM4LS), July 2025 | null | null | q-bio.GN cs.AI cs.LG cs.SE q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative AI foundation models offer transformative potential for processing structured biological data, particularly in single-cell RNA sequencing, where datasets are rapidly scaling toward billions of cells. We propose the use of agentic foundation models with real-time web search to automate the labeling of experimental data, achieving up to 82.5% accuracy. This addresses a key bottleneck in supervised learning for structured omics data by increasing annotation throughput without manual curation and human error. Our approach enables the development of virtual cell foundation models capable of downstream tasks such as cell-typing and perturbation prediction. As data volume grows, these models may surpass human performance in labeling, paving the way for reliable inference in large-scale perturbation screens. This application demonstrates domain-specific innovation in health monitoring and diagnostics, aligned with efforts like the Human Cell Atlas and Human Tumor Atlas Network. | [
{
"version": "v1",
"created": "Sat, 14 Jun 2025 23:30:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dajani",
"Saleem A. Al",
""
],
[
"Sanchez",
"Abel",
""
],
[
"Williams",
"John R.",
""
]
] |
2506.13818 | Marcelle Momha | Marcelle Momha | The Synthetic Mirror -- Synthetic Data at the Age of Agentic AI | null | null | null | null | cs.CY cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Synthetic data, which is artificially generated and intelligently mimicking or supplementing the real-world data, is increasingly used. The proliferation of AI agents and the adoption of synthetic data create a synthetic mirror that conceptualizes a representation and potential distortion of reality, thus generating trust and accountability deficits. This paper explores the implications for privacy and policymaking stemming from synthetic data generation, and the urgent need for new policy instruments and legal framework adaptation to ensure appropriate levels of trust and accountability for AI agents relying on synthetic data. Rather than creating entirely new policy or legal regimes, the most practical approach involves targeted amendments to existing frameworks, recognizing synthetic data as a distinct regulatory category with unique characteristics. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 02:10:02 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Momha",
"Marcelle",
""
]
] |
2506.13819 | El Arbi Belfarsi | El Arbi Belfarsi, Henry Flores, and Maria Valero | Reliable Noninvasive Glucose Sensing via CNN-Based Spectroscopy | Submitted to the IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI 2025) | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In this study, we present a dual-modal AI framework based on short-wave infrared (SWIR) spectroscopy. The first modality employs a multi-wavelength SWIR imaging system coupled with convolutional neural networks (CNNs) to capture spatial features linked to glucose absorption. The second modality uses a compact photodiode voltage sensor and machine learning regressors (e.g., random forest) on normalized optical signals. Both approaches were evaluated on synthetic blood phantoms and skin-mimicking materials across physiological glucose levels (70 to 200 mg/dL). The CNN achieved a mean absolute percentage error (MAPE) of 4.82% at 650 nm with 100% Zone A coverage in the Clarke Error Grid, while the photodiode system reached 86.4% Zone A accuracy. This framework constitutes a state-of-the-art solution that balances clinical accuracy, cost efficiency, and wearable integration, paving the way for reliable continuous non-invasive glucose monitoring. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 03:01:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Belfarsi",
"El Arbi",
""
],
[
"Flores",
"Henry",
""
],
[
"Valero",
"Maria",
""
]
] |
2506.13820 | Shraddha Surana | Shraddha Surana, Ashwin Srinivasan and Michael Bain | Structured Program Synthesis using LLMs: Results and Insights from the IPARC Challenge | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The IPARC Challenge, inspired by ARC, provides controlled program synthesis tasks over synthetic images to evaluate automatic program construction, focusing on sequence, selection, and iteration. This set of 600 tasks has resisted automated solutions. This paper presents a structured inductive programming approach with LLMs that successfully solves tasks across all IPARC categories. The controlled nature of IPARC reveals insights into LLM-based code generation, including the importance of prior structuring, LLMs' ability to aid structuring (requiring human refinement), the need to freeze correct code, the efficiency of code reuse, and how LLM-generated code can spark human creativity. These findings suggest valuable mechanisms for human-LLM collaboration in tackling complex program synthesis. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 04:33:00 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Surana",
"Shraddha",
""
],
[
"Srinivasan",
"Ashwin",
""
],
[
"Bain",
"Michael",
""
]
] |
2506.13821 | Giovanni Bernardi | Giovanni Bernardi, Adrian Francalanza, Marco Peressotti, Mohammad Reza Mousavi | Role, cost, and complexity of software in the real-world: a case for formal methods | null | null | null | null | cs.SE cs.CY | http://creativecommons.org/licenses/by/4.0/ | In this chapter we outline the role that software has in modern society, along with the staggering costs of poor software quality. To lay this bare, we recall the costs of some of the major software failures that happened during the last~$40$ years. We argue that these costs justify researching, studying and applying formal software verification and in particular program analysis. This position is supported by successful industrial experiences. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 07:40:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bernardi",
"Giovanni",
""
],
[
"Francalanza",
"Adrian",
""
],
[
"Peressotti",
"Marco",
""
],
[
"Mousavi",
"Mohammad Reza",
""
]
] |
2506.13822 | Yang Liu | Yang Liu | The Trip to ZigBee Backscatter across a Decade, a Systematic Review | null | null | null | null | cs.NI eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The field of backscatter communication has undergone a profound transformation, evolving from a niche technology for radio-frequency identification (RFID) into a sophisticated paradigm poised to enable a truly battery-free Internet of Things (IoT). This evolution is built upon a deepening understanding of the fundamental principles governing these ultra-low-power links. Modern backscatter systems are no longer simple reflectors of continuous waves but are increasingly designed to interact with complex, data-carrying ambient signals from ubiquitous sources like WiFi, ZigBee, and cellular networks. This review systematically charts the journey of ambient backscatter, particularly focusing on its interaction with ZigBee and other commodity wireless protocols over the last decade. We analyze the progression from foundational proof-of-concept systems that established productive backscatter to modern high-throughput, concurrent, and cross-technology communication architectures. Key advancements in fine-grained modulation, robust synchronization, cross-technology physical layer emulation, and multi-tag coordination are detailed. A comparative analysis of state-of-the-art systems highlights the core trade-offs between performance metrics like data rate and range, power consumption, and compatibility with commodity hardware. Finally, we synthesize the primary challenges, including networking scalability, security vulnerabilities, the near-far problem, and practical deployment hurdles, and outline future research directions, such as integration with Reconfigurable Intelligent Surfaces (RIS) and 6G networks, that promise to further expand the capabilities of this transformative technology. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 07:48:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Yang",
""
]
] |
2506.13823 | H.C. Manjunatha Dr | H.C. Manjunatha, P.S. Damodara Gupta, N. Sowmya, K.N. Sridhar | KHIFC-user friendly program for studying Heavy ion fusion barrier characteristics | null | null | null | null | nucl-th nucl-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We have developed an application for studying heavy ion fusion barrier characteristics such as such as fusion barrier heights ($V_B$), positions ($R_B$) and curvature of the inverted parabola ($\hbar\omega$). We call this application as KHIFC (Kolar Heavy Ion fusion barrier Characteristics). This software application hosted in the domain "https://systematics-of-heavy-ion-fusion.vercel.app/". This KHIFC produces fusion cross-sections with the simple input of projectile, target and center of mass energy. The values produced by the KHIFC is validated with experiments. Efficient tools like KHIFC are essential for researchers in nuclear physics, particularly when dealing with complex systems such as actinide and superheavy nuclei. By providing quick calculations and insights, it can significantly aid in experiment planning and theoretical investigations. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 10:52:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Manjunatha",
"H. C.",
""
],
[
"Gupta",
"P. S. Damodara",
""
],
[
"Sowmya",
"N.",
""
],
[
"Sridhar",
"K. N.",
""
]
] |
2506.13824 | Jinyang Huang | Jinyang Huang, Xiachong Feng, Qiguang Chen, Hanjie Zhao, Zihui Cheng, Jiesong Bai, Jingxuan Zhou, Min Li, Libo Qin | MLDebugging: Towards Benchmarking Code Debugging Across Multi-Library Scenarios | ACL 2025 Findings | null | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Code debugging is a crucial task in software engineering, which attracts increasing attention. While remarkable success has been made in the era of large language models (LLMs), current research still focuses on the simple no-library or single-library setting, ignoring the complex multi-library scenario in real-world applications. To address this limitation, we make the first attempt to introduce MLDebugging (Multi-Library Debugging), a comprehensive benchmark designed to assess debugging challenges within multi-library Python code. Specifically, MLDebugging encompasses 126 distinct Python libraries, covering a wide range of multi-library code issues, categorized into seven distinct types. Furthermore, we conduct a thorough evaluation of MLDebugging using both mainstream open-source and closed-source LLMs and highlight that current LLMs still struggle to correctly perform code debugging across multi-library scenarios. We hope this work can uncover the potential of LLMs in multi-library debugging scenario and offer insights for future research. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 13:02:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Huang",
"Jinyang",
""
],
[
"Feng",
"Xiachong",
""
],
[
"Chen",
"Qiguang",
""
],
[
"Zhao",
"Hanjie",
""
],
[
"Cheng",
"Zihui",
""
],
[
"Bai",
"Jiesong",
""
],
[
"Zhou",
"Jingxuan",
""
],
[
"Li",
"Min",
""
],
[
"Qin",
"Libo",
""
]
] |
2506.13825 | Gnankan Landry Regis N'guessan | Gnankan Landry Regis N'guessan and Issa Karambal | The Reflexive Integrated Information Unit: A Differentiable Primitive for Artificial Consciousness | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Research on artificial consciousness lacks the equivalent of the perceptron: a small, trainable module that can be copied, benchmarked, and iteratively improved. We introduce the Reflexive Integrated Information Unit (RIIU), a recurrent cell that augments its hidden state $h$ with two additional vectors: (i) a meta-state $\mu$ that records the cell's own causal footprint, and (ii) a broadcast buffer $B$ that exposes that footprint to the rest of the network. A sliding-window covariance and a differentiable Auto-$\Phi$ surrogate let each RIIU maximize local information integration online. We prove that RIIUs (1) are end-to-end differentiable, (2) compose additively, and (3) perform $\Phi$-monotone plasticity under gradient ascent. In an eight-way Grid-world, a four-layer RIIU agent restores $>90\%$ reward within 13 steps after actuator failure, twice as fast as a parameter-matched GRU, while maintaining a non-zero Auto-$\Phi$ signal. By shrinking "consciousness-like" computation down to unit scale, RIIUs turn a philosophical debate into an empirical mathematical problem. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 14:07:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"N'guessan",
"Gnankan Landry Regis",
""
],
[
"Karambal",
"Issa",
""
]
] |
2506.13826 | Roni Dey | R. Dey, P.K. Netrakanti, D.K. Mishra, S.P. Behera, R. Sehgal, V. Jha, and L.M. Pant | Study of neutron response using time of flight technique in ISMRAN detector | null | Springer Proc.Phys. 282 (2023) 103-108 | 10.1007/978-3-031-19268-5_14 | null | physics.ins-det nucl-ex | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We report the measurements of the fast neutron energy response in Indian Scintillator Matrix for Reactor Anti-Neutrinos (ISMRAN) detector consisting of an array of 9$\times$10 Plastic Scintillator Bars (PSBs) at BARC, Mumbai. ISMRAN is an above ground detector setup at $\sim$13 m from the Dhruva reactor core for the detection of reactor anti-neutrinos (${\overline{\ensuremath{\nu}}}_{e}$) via the inverse beta decay (IBD) process. The dominant sources of reactor-related background in the vicinity of the detector are high energy $\gamma$-rays and fast neutrons. Therefore, a good understanding of fast neutron response in PSB is an essential pre-requisite for suppression and discrimination of the fast neutron background from IBD events. Kinetic energies of the fast neutron were determined using the Time-of-Flight (TOF) technique, which is used to get the scintillation light yield due to recoiling proton in PSB. We also measured the fast neutron capture time distribution in ISMRAN array using a novel technique involving TOF of the measured fast neutrons. The observed characteristic neutron capture time ( $\tau$ ) of 68.29 $\pm$ 9.48 $\mu$s is in good agreement with GEANT4 based MC simulation. These experimentally measured results will be useful for discriminating correlated and uncorrelated (accidental) background events from the true IBD events in reactor ON and OFF conditions inside the reactor hall. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 14:43:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dey",
"R.",
""
],
[
"Netrakanti",
"P. K.",
""
],
[
"Mishra",
"D. K.",
""
],
[
"Behera",
"S. P.",
""
],
[
"Sehgal",
"R.",
""
],
[
"Jha",
"V.",
""
],
[
"Pant",
"L. M.",
""
]
] |
2506.13827 | Zhuoying Li | Zhuoying Li, Zhu Xu, Yuxin Peng, Yang Liu | Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing | null | null | null | null | cs.GR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Instruction-based image editing, which aims to modify the image faithfully according to the instruction while preserving irrelevant content unchanged, has made significant progress. However, there still lacks a comprehensive metric for assessing the editing quality. Existing metrics either require high human evaluation costs, which hinder large-scale evaluation, or are adapted from other tasks and lose task-specific concerns, failing to comprehensively evaluate both instruction-based modification and preservation of irrelevant regions, resulting in biased evaluation. To tackle this, we introduce a new metric called Balancing Preservation and Modification (BPM), tailored for instruction-based image editing by explicitly disentangling the image into editing-relevant and irrelevant regions for specific consideration. We first identify and locate editing-relevant regions, followed by a two-tier process to assess editing quality: Region-Aware Judge evaluates whether the position and size of the edited region align with the instruction, and Semantic-Aware Judge further assesses the instruction content compliance within editing-relevant regions as well as content preservation within irrelevant regions, yielding comprehensive and interpretable quality assessment. Moreover, the editing-relevant region localization in BPM can be integrated into image editing approaches to improve editing quality, demonstrating its broad applicability. We verify the effectiveness of the BPM metric on comprehensive instruction-editing data, and the results show the highest alignment with human evaluation compared to existing metrics, indicating its efficacy. Code is available at: https://joyli-x.github.io/BPM/ | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 17:12:57 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Zhuoying",
""
],
[
"Xu",
"Zhu",
""
],
[
"Peng",
"Yuxin",
""
],
[
"Liu",
"Yang",
""
]
] |
2506.13828 | Abdullah Burkan Bereketo\u{g}lu | Abdullah Burkan Bereketoglu | Hybrid Meta-Learning Framework for Anomaly Forecasting in Nonlinear Dynamical Systems via Physics-Inspired Simulation and Deep Ensembles | 6 pages, 5 figures, 5 algorithms | null | null | null | cs.LG cs.SY eess.SY physics.acc-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We propose a hybrid meta-learning framework for forecasting and anomaly detection in nonlinear dynamical systems characterized by nonstationary and stochastic behavior. The approach integrates a physics-inspired simulator that captures nonlinear growth-relaxation dynamics with random perturbations, representative of many complex physical, industrial, and cyber-physical systems. We use CNN-LSTM architectures for spatio-temporal feature extraction, Variational Autoencoders (VAE) for unsupervised anomaly scoring, and Isolation Forests for residual-based outlier detection in addition to a Dual-Stage Attention Recurrent Neural Network (DA-RNN) for one-step forecasting on top of the generated simulation data. To create composite anomaly forecasts, these models are combined using a meta-learner that combines forecasting outputs, reconstruction errors, and residual scores. The hybrid ensemble performs better than standalone models in anomaly localization, generalization, and robustness to nonlinear deviations, according to simulation-based experiments. The framework provides a broad, data-driven approach to early defect identification and predictive monitoring in nonlinear systems, which may be applied to a variety of scenarios where complete physical models might not be accessible. | [
{
"version": "v1",
"created": "Sun, 15 Jun 2025 21:17:34 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bereketoglu",
"Abdullah Burkan",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.