id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 3
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-12-13 00:00:00
| authors_parsed
listlengths 1
3.44k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2512.10409
|
Shubhrangshu Biswas
|
Shubhrangshu Biswas, Jai Sukhatme, Bishakhdatta Gayen
|
Impact of Background Conditions on the Structure and Propagation of the Boreal Summer Quasi-Biweekly Oscillation
|
23 pages, 12 figures, 2 tables; Submitted to "Quarterly Journal of the Royal Meteorological Society"
| null | null | null |
physics.ao-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We examine the westward-propagating quasi-biweekly oscillation (QBWO) during boreal summer, with a focus on how background moisture and winds shape its structure and propagation. In dry regions, convection lags the circulation by nearly a quarter cycle, whereas in very moist regions it becomes nearly in-phase and extends across the QBWO gyre. As the background moistens, moisture anomalies increasingly align with the QBWO circulation. Despite differences in environmental moisture and wind conditions, several structural features remain robust: outgoing longwave radiation and moisture anomalies stay collocated, moisture and pressure-velocity anomalies remain vertically upright, and the filtered winds retain a first-baroclinic mode structure. A vorticity budget shows that, although the planetary vorticity-gradient term is important, both planetary stretching and horizontal advection are needed to explain the vorticity tendency- and their relative importance shifts with the moisture regime. In dry and moderately moist regions with easterly mean flow, mean winds primarily advect vorticity anomalies. In contrast, in very moist regions with westerly flow, anomalous winds instead advect the background vorticity. An analogous transition occurs in the moisture budget: in dry and moderately moist environments, zonal mean flow advection dominates, but in very moist regions, strong background moisture gradients allow eddy advection of the mean moisture field to become the leading term. In the moist regime, vertical advection, precipitation, and evaporation also contribute substantially to the moisture tendency. Overall, the QBWO behaves like a mean-flow-driven linear mode in dry and moderately moist regions with easterly background winds, but shifts toward a regime dominated by eddy advection of background vorticity and moisture in very moist regions characterized by westerly flow.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:19:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Biswas",
"Shubhrangshu",
""
],
[
"Sukhatme",
"Jai",
""
],
[
"Gayen",
"Bishakhdatta",
""
]
] |
2512.10410
|
Mikael Rordam
|
Magdalena Musat and Mikael R{\o}rdam
|
Entanglement in C$^*$-algebras: tensor products of state spaces
|
25 pages
| null | null | null |
math.OA math.FA quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
We analyze the Namioka-Phelps minimal and maximal tensor products of compact convex sets arising as state spaces of C$^*$-algebras, and, relatedly, study entanglement in (infinite dimensional) C$^*$-algebras. The minimal Namioka-Phelps tensor product of the state spaces of two C$^*$-algebras is shown to correspond to the set of separable (= un-entangled) states on the tensor product of the C$^*$-algebras. We show that these maximal and minimal tensor product of the state spaces agree precisely when one of the two C$^*$-algebras is commutative. This confirms an old conjecture by Barker in the case where the compact convex sets are state spaces of C$^*$-algebras.
The Namioka-Phelps tensor product of the trace simplexes of two C$^*$-algebras is shown always to be the trace simplex of the tensor product of the C$^*$-algebras. This can be used, for example, to show that the trace simplex of (any) tensor product of two C$^*$-algebras is the Poulsen simplex if and only if the trace simplex of each of the C$^*$-algebras is the Poulsen simplex.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:21:15 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Musat",
"Magdalena",
""
],
[
"Rørdam",
"Mikael",
""
]
] |
2512.10411
|
Yijiong Yu
|
Yijiong Yu, Jiale Liu, Qingyun Wu, Huazheng Wang, Ji Pei
|
Sliding Window Attention Adaptation
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by-sa/4.0/
|
The self-attention mechanism in Transformer-based Large Language Models (LLMs) scales quadratically with input length, making long-context inference expensive. Sliding window attention (SWA) reduces this cost to linear complexity, but naively enabling complete SWA at inference-time for models pretrained with full attention (FA) causes severe long-context performance degradation due to training-inference mismatch. This makes us wonder: Can FA-pretrained LLMs be well adapted to SWA without pretraining? We investigate this by proposing Sliding Window Attention Adaptation (SWAA), a set of practical recipes that combine five methods for better adaptation: (1) applying SWA only during prefilling; (2) preserving "sink" tokens; (3) interleaving FA/SWA layers; (4) chain-of-thought (CoT); and (5) fine-tuning. Our experiments show that SWA adaptation is feasible while non-trivial: no single method suffices, yet specific synergistic combinations effectively recover the original long-context performance. We further analyze the performance-efficiency trade-offs of different SWAA configurations and provide recommended recipes for diverse scenarios. Our code is available at https://github.com/yuyijiong/sliding-window-attention-adaptation
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:21:24 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yu",
"Yijiong",
""
],
[
"Liu",
"Jiale",
""
],
[
"Wu",
"Qingyun",
""
],
[
"Wang",
"Huazheng",
""
],
[
"Pei",
"Ji",
""
]
] |
2512.10412
|
Young-Jin Sim
|
Kyudong Choi and In-Jee Jeong and Young-Jin Sim
|
Vortex atmospheres of traveling vortices: rigorous definition, existence, and topological classification
|
38 pages, 11 figures
| null | null | null |
math.AP
|
http://creativecommons.org/licenses/by/4.0/
|
In incompressible and inviscid fluids, the vortex atmosphere refers to the collection of fluid particles outside the support of a traveling vortex that are nevertheless carried along with it. This phenomenon has been recognized since the nineteenth century, e.g., in the classical works of O. Reynolds [Nature, 1876] and O. Lodge [Lond. Edinb. Dubl. Phil. Mag., 1885], yet rigorous mathematical definitions and proofs have remained largely undeveloped, with most subsequent studies relying on thin-core approximations or asymptotic analyses. In this paper, we give a rigorous definition of a vortex atmosphere and establish its existence and uniqueness. We further compare the planar atmosphere surrounding a 2D vortex dipole with the axisymmetric atmosphere surrounding a 3D vortex ring. In particular, we emphasize and prove the topological distinctions observed by W. Hicks [Lond. Edinb. Dubl. Phil. Mag., 1919]: under natural assumptions, every 2D dipole with its atmosphere forms an oval-shaped region, whereas for 3D rings, both spheroidal and toroidal configurations may occur. Our proof is based on showing that each atmosphere can be characterized precisely as a specific superlevel set of its corresponding stream function.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:22:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Choi",
"Kyudong",
""
],
[
"Jeong",
"In-Jee",
""
],
[
"Sim",
"Young-Jin",
""
]
] |
2512.10413
|
J\k{e}drzej Hodor
|
J\k{e}drzej Hodor and Jakub Sordyl
|
Local dimension of a Boolean lattice
| null | null | null | null |
math.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
For every integer $n$ with $n \geq 4$, we prove that the local dimension of a poset consisting of all the subsets of $\{1,\dots,n\}$ equipped with the inclusion relation is strictly less than $n$, answering a question of Kim, Martin, Masa\v{r}\'ik, Shull, Smith, Uzzell, and Wang (Eur. J. Comb. 2020). We also study several related problems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:24:57 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Hodor",
"Jędrzej",
""
],
[
"Sordyl",
"Jakub",
""
]
] |
2512.10415
|
Devanshu Sahoo
|
Devanshu Sahoo, Vasudev Majhi, Arjun Neekhra, Yash Sinha, Murari Mandal, Dhruv Kumar
|
How to Trick Your AI TA: A Systematic Study of Academic Jailbreaking in LLM Code Evaluation
|
Under Review
| null | null | null |
cs.SE cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The use of Large Language Models (LLMs) as automatic judges for code evaluation is becoming increasingly prevalent in academic environments. But their reliability can be compromised by students who may employ adversarial prompting strategies in order to induce misgrading and secure undeserved academic advantages. In this paper, we present the first large-scale study of jailbreaking LLM-based automated code evaluators in academic context. Our contributions are: (i) We systematically adapt 20+ jailbreaking strategies for jailbreaking AI code evaluators in the academic context, defining a new class of attacks termed academic jailbreaking. (ii) We release a poisoned dataset of 25K adversarial student submissions, specifically designed for the academic code-evaluation setting, sourced from diverse real-world coursework and paired with rubrics and human-graded references, and (iii) In order to capture the multidimensional impact of academic jailbreaking, we systematically adapt and define three jailbreaking metrics (Jailbreak Success Rate, Score Inflation, and Harmfulness). (iv) We comprehensively evalulate the academic jailbreaking attacks using six LLMs. We find that these models exhibit significant vulnerability, particularly to persuasive and role-play-based attacks (up to 97% JSR). Our adversarial dataset and benchmark suite lay the groundwork for next-generation robust LLM-based evaluators in academic code assessment.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:28:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Sahoo",
"Devanshu",
""
],
[
"Majhi",
"Vasudev",
""
],
[
"Neekhra",
"Arjun",
""
],
[
"Sinha",
"Yash",
""
],
[
"Mandal",
"Murari",
""
],
[
"Kumar",
"Dhruv",
""
]
] |
2512.10416
|
Wenfei Guan
|
Wenfei Guan, Jilin Mei, Tong Shen, Xumin Wu, Shuo Wang, Cheng Min, Yu Hu
|
Beyond Endpoints: Path-Centric Reasoning for Vectorized Off-Road Network Extraction
| null | null | null | null |
cs.CV cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Deep learning has advanced vectorized road extraction in urban settings, yet off-road environments remain underexplored and challenging. A significant domain gap causes advanced models to fail in wild terrains due to two key issues: lack of large-scale vectorized datasets and structural weakness in prevailing methods. Models such as SAM-Road employ a node-centric paradigm that reasons at sparse endpoints, making them fragile to occlusions and ambiguous junctions in off-road scenes, leading to topological errors.This work addresses these limitations in two complementary ways. First, we release WildRoad, a gloabal off-road road network dataset constructed efficiently with a dedicated interactive annotation tool tailored for road-network labeling. Second, we introduce MaGRoad (Mask-aware Geodesic Road network extractor), a path-centric framework that aggregates multi-scale visual evidence along candidate paths to infer connectivity robustly.Extensive experiments show that MaGRoad achieves state-of-the-art performance on our challenging WildRoad benchmark while generalizing well to urban datasets. A streamlined pipeline also yields roughly 2.5x faster inference, improving practical applicability. Together, the dataset and path-centric paradigm provide a stronger foundation for mapping roads in the wild.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:29:27 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Guan",
"Wenfei",
""
],
[
"Mei",
"Jilin",
""
],
[
"Shen",
"Tong",
""
],
[
"Wu",
"Xumin",
""
],
[
"Wang",
"Shuo",
""
],
[
"Min",
"Cheng",
""
],
[
"Hu",
"Yu",
""
]
] |
2512.10417
|
Yudong Luo
|
Yudong Luo, Shuai Zha and Toshitaka Kajino
|
A Systematic Study of Magnetic Fields Impacts on Neutrino Transport in Core-Collapse Supernovae
|
20 pages, 7 figures, accepted by Phy. Rev. D
| null | null | null |
astro-ph.HE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We quantify the impact of strong magnetic fields (assuming $B=B_0\cdot r_0^3/r^3$ with $B_0\gtrsim 10^{16}$ G) on the neutrino transport in core-collapse supernovae (CCSNe). Magnetic fields quantize the momenta of electrons and positrons, resulting in an enhanced absorption cross section for low-energy neutrinos and suppressed chemical potentials for $e^\pm$. We include these changes in the M1 scheme for neutrino transport and perform 1-D CCSNe simulations with \texttt{GR1D}. The increased low-energy cross sections reduce the $\bar{\nu}_e$ mean energy $\langle E_{\bar\nu_e}\rangle$ while elevating the neutrino number luminosities $\mathcal{L_\nu}$ for both ${\nu}_e$ and $\bar{\nu}_e$ due to the lower energy weighted spectra. The reduction of chemical potential enhances the $\bar{\nu}_e$ emission while suppressing that of $\nu_e$, thereby driving an increase in the electron fraction behind the stalled shock at $\sim30$--$100$ km. This further amplifies $\langle E_{\nu_e}\rangle$ through an increased electron density. Consequently, magnetic fields amplify $L_{\nu_e}$ by increasing both $\mathcal{L}_{\nu_e}$ and $\langle E_{\nu_e}\rangle$ whereas for $\bar\nu_e$, the rise in $\mathcal{L}_{\bar\nu_e}$ is offset by a decreased $\langle E_{\bar\nu_e}\rangle$, leading to a minimal change in $L_{\bar\nu_e}$. A systematic parameter scan of dipole field configurations suggests that, for $r_0 > 30$ km, $\langle E_{\bar{\nu}_e} \rangle$ is significantly suppressed and $L_{\nu_e}$ is enhanced if $B_0 \geq {2.7} \times 10^{16}$ G. These magnetic effects become negligible for $B_0$ below $\sim {7.4} \times 10^{15}$ G.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:31:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Luo",
"Yudong",
""
],
[
"Zha",
"Shuai",
""
],
[
"Kajino",
"Toshitaka",
""
]
] |
2512.10418
|
Gustavo Berganti\~nos
|
Gustavo Berganti\~nos, Leticia Lorenzo
|
Applying an axiomatic approach to revenue allocation in airlines problems
|
22 pages
| null | null | null |
econ.TH
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The International Air Transport Association (IATA) states that the revenue from interline tickets must be shared among the different airlines according to a weighted system. We analyze this problem following an axiomatic approach, and our theoretical results support IATA's procedure. Our first result justifies the use of a weighted system, but it does not specify which weights should be applied. Assuming that the weights are fixed, we provide several results that further support the use of IATA's mechanism. Finally, we provide results for the case in which all flights can be considered equivalent and no weighting is required.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:33:10 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bergantiños",
"Gustavo",
""
],
[
"Lorenzo",
"Leticia",
""
]
] |
2512.10419
|
Phu Pham
|
Phu Pham, Damon Conover, Aniket Bera
|
TransLocNet: Cross-Modal Attention for Aerial-Ground Vehicle Localization with Contrastive Learning
|
8 pages, 4 figures, 4 tables
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Aerial-ground localization is difficult due to large viewpoint and modality gaps between ground-level LiDAR and overhead imagery. We propose TransLocNet, a cross-modal attention framework that fuses LiDAR geometry with aerial semantic context. LiDAR scans are projected into a bird's-eye-view representation and aligned with aerial features through bidirectional attention, followed by a likelihood map decoder that outputs spatial probability distributions over position and orientation. A contrastive learning module enforces a shared embedding space to improve cross-modal alignment. Experiments on CARLA and KITTI show that TransLocNet outperforms state-of-the-art baselines, reducing localization error by up to 63% and achieving sub-meter, sub-degree accuracy. These results demonstrate that TransLocNet provides robust and generalizable aerial-ground localization in both synthetic and real-world settings.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:34:26 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Pham",
"Phu",
""
],
[
"Conover",
"Damon",
""
],
[
"Bera",
"Aniket",
""
]
] |
2512.10420
|
Lorenzo Gavassino
|
Lorenzo Gavassino, Patrick Niekamp, S\"oren Schlichting, Gabriel S Denicol
|
Stationary Couette-type flows in relativistic fluids
|
10 pages, 6 figures, comments welcome!
| null | null | null |
nucl-th hep-th physics.flu-dyn
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We investigate a class of stationary, planar-symmetric solutions of relativistic hydrodynamics, in which a dissipative fluid is confined between two parallel plates that move relative to each other and/or are maintained at different temperatures. We find that neglecting the heat flux leads to qualitatively incorrect flow profiles, even in systems with temperature-independent viscosity. This arises from the fact that, in special relativity, the heat flux itself contributes to the momentum density (the so-called "inertia of heat"). This effect is most evident in the Landau frame, where the fluid removes the excess energy generated by viscous heating by streaming across the boundaries. The analysis is further extended to the limit of vanishing chemical potential.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:34:40 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Gavassino",
"Lorenzo",
""
],
[
"Niekamp",
"Patrick",
""
],
[
"Schlichting",
"Sören",
""
],
[
"Denicol",
"Gabriel S",
""
]
] |
2512.10421
|
Jiazhen Huang
|
Xiao Chen, Zhongjing Du, Jiazhen Huang, Xu Jiang, Li Lu, Jingyan Jiang, Zhi Wang
|
Neural Collapse in Test-Time Adaptation
|
10 pages, 8 figures
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Test-Time Adaptation (TTA) enhances model robustness to out-of-distribution (OOD) data by updating the model online during inference, yet existing methods lack theoretical insights into the fundamental causes of performance degradation under domain shifts. Recently, Neural Collapse (NC) has been proposed as an emergent geometric property of deep neural networks (DNNs), providing valuable insights for TTA. In this work, we extend NC to the sample-wise level and discover a novel phenomenon termed Sample-wise Alignment Collapse (NC3+), demonstrating that a sample's feature embedding, obtained by a trained model, aligns closely with the corresponding classifier weight. Building on NC3+, we identify that the performance degradation stems from sample-wise misalignment in adaptation which exacerbates under larger distribution shifts. This indicates the necessity of realigning the feature embeddings with their corresponding classifier weights. However, the misalignment makes pseudo-labels unreliable under domain shifts. To address this challenge, we propose NCTTA, a novel feature-classifier alignment method with hybrid targets to mitigate the impact of unreliable pseudo-labels, which blends geometric proximity with predictive confidence. Extensive experiments demonstrate the effectiveness of NCTTA in enhancing robustness to domain shifts. For example, NCTTA outperforms Tent by 14.52% on ImageNet-C.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:34:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chen",
"Xiao",
""
],
[
"Du",
"Zhongjing",
""
],
[
"Huang",
"Jiazhen",
""
],
[
"Jiang",
"Xu",
""
],
[
"Lu",
"Li",
""
],
[
"Jiang",
"Jingyan",
""
],
[
"Wang",
"Zhi",
""
]
] |
2512.10422
|
Youmin Ko
|
Youmin Ko, Sungjong Seo, Hyunjoon Kim
|
Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers
|
Accepted to NeurIPS 2025
| null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Since large language models (LLMs) have a tendency to generate factually inaccurate output, retrieval-augmented generation (RAG) has gained significant attention as a key means to mitigate this downside of harnessing only LLMs. However, existing RAG methods for simple and multi-hop question answering (QA) are still prone to incorrect retrievals and hallucinations. To address these limitations, we propose CoopRAG, a novel RAG framework for the question answering task in which a retriever and an LLM work cooperatively with each other by exchanging informative knowledge, and the earlier and later layers of the retriever model work cooperatively with each other to accurately rank the retrieved documents relevant to a given query. In this framework, we (i) unroll a question into sub-questions and a reasoning chain in which uncertain positions are masked, (ii) retrieve the documents relevant to the question augmented with the sub-questions and the reasoning chain, (iii) rerank the documents by contrasting layers of the retriever, and (iv) reconstruct the reasoning chain by filling the masked positions via the LLM. Our experiments demonstrate that CoopRAG consistently outperforms state-of-the-art QA methods on three multi-hop QA datasets as well as a simple QA dataset in terms of both the retrieval and QA performances. Our code is available.\footnote{https://github.com/meaningful96/CoopRAG}
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:35:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ko",
"Youmin",
""
],
[
"Seo",
"Sungjong",
""
],
[
"Kim",
"Hyunjoon",
""
]
] |
2512.10423
|
Ildefonso Castro-Infantes
|
Paula Carretero, Ildefonso Castro and Ildefonso Castro-Infantes
|
Rotational Weingarten surfaces in Lorentz-Minkowski space
|
26 pages, 14 figures
| null | null | null |
math.DG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose a new approach to the study of rotational surfaces in Lorentz-Minkowski space based on the notion of the geometric linear momentum of the generatrix curves with respect to the axes of revolution. This technique allows us to reduce any Weingarten condition on the surface to a first-order ordinary differential equation for the momentum as a function of the distance to the corresponding axis, providing a unified framework that encompasses the three causal types of rotation axes. As a direct application, we classify important families of rotational Weingarten surfaces in this setting, including some linear and quadratic cases. Furthermore, we introduce the non-degenerate quadric surfaces of revolution in Lorentz-Minkowski space and characterize them in terms of a specific cubic Weingarten relation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:35:59 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Carretero",
"Paula",
""
],
[
"Castro",
"Ildefonso",
""
],
[
"Castro-Infantes",
"Ildefonso",
""
]
] |
2512.10424
|
Hai-Long Qin
|
Hai-Long Qin, Sixian Wang, Guo Lu, Jincheng Dai
|
Neural Hamiltonian Deformation Fields for Dynamic Scene Rendering
|
Accepted by ACM SIGGRAPH Asia 2025, project page: https://qin-jingyun.github.io/NeHaD
| null | null | null |
cs.GR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Representing and rendering dynamic scenes with complex motions remains challenging in computer vision and graphics. Recent dynamic view synthesis methods achieve high-quality rendering but often produce physically implausible motions. We introduce NeHaD, a neural deformation field for dynamic Gaussian Splatting governed by Hamiltonian mechanics. Our key observation is that existing methods using MLPs to predict deformation fields introduce inevitable biases, resulting in unnatural dynamics. By incorporating physics priors, we achieve robust and realistic dynamic scene rendering. Hamiltonian mechanics provides an ideal framework for modeling Gaussian deformation fields due to their shared phase-space structure, where primitives evolve along energy-conserving trajectories. We employ Hamiltonian neural networks to implicitly learn underlying physical laws governing deformation. Meanwhile, we introduce Boltzmann equilibrium decomposition, an energy-aware mechanism that adaptively separates static and dynamic Gaussians based on their spatial-temporal energy states for flexible rendering. To handle real-world dissipation, we employ second-order symplectic integration and local rigidity regularization as physics-informed constraints for robust dynamics modeling. Additionally, we extend NeHaD to adaptive streaming through scale-aware mipmapping and progressive optimization. Extensive experiments demonstrate that NeHaD achieves physically plausible results with a rendering quality-efficiency trade-off. To our knowledge, this is the first exploration leveraging Hamiltonian mechanics for neural Gaussian deformation, enabling physically realistic dynamic scene rendering with streaming capabilities.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:36:30 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Qin",
"Hai-Long",
""
],
[
"Wang",
"Sixian",
""
],
[
"Lu",
"Guo",
""
],
[
"Dai",
"Jincheng",
""
]
] |
2512.10425
|
Fan Yu
|
Fan Yu, Guodong Li, Si Wu, Weijun Fang, and Sihuang Hu
|
Making Wide Stripes Practical: Cascaded Parity LRCs for Efficient Repair and High Reliability
| null | null | null | null |
cs.DC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Erasure coding with wide stripes is increasingly adopted to reduce storage overhead in large-scale storage systems. However, existing Locally Repairable Codes (LRCs) exhibit structural limitations in this setting: inflated local groups increase single-node repair cost, multi-node failures frequently trigger expensive global repair, and reliability degrades sharply. We identify a key root cause: local and global parity blocks are designed independently, preventing them from cooperating during repair. We present Cascaded Parity LRCs (CP-LRCs), a new family of wide stripe LRCs that embed structured dependency between parity blocks by decomposing a global parity block across all local parity blocks. This creates a cascaded parity group that preserves MDS-level fault tolerance while enabling low-bandwidth single-node and multi-node repairs. We provide a general coefficient-generation framework, develop repair algorithms exploiting cascading, and instantiate the design with CP-Azure and CP-Uniform. Evaluations on Alibaba Cloud show reductions in repair time of up to 41% for single-node failures and 26% for two-node failures.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:36:47 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yu",
"Fan",
""
],
[
"Li",
"Guodong",
""
],
[
"Wu",
"Si",
""
],
[
"Fang",
"Weijun",
""
],
[
"Hu",
"Sihuang",
""
]
] |
2512.10426
|
Murtaza Rangwala
|
N Mangala, Murtaza Rangwala, S Aishwarya, B Eswara Reddy, Rajkumar Buyya, KR Venugopal, SS Iyengar, LM Patnaik
|
Differential Privacy for Secure Machine Learning in Healthcare IoT-Cloud Systems
| null | null | null | null |
cs.CR cs.DC
|
http://creativecommons.org/licenses/by/4.0/
|
Healthcare has become exceptionally sophisticated, as wearables and connected medical devices are revolutionising remote patient monitoring, emergency response, medication management, diagnosis, and predictive and prescriptive analytics. Internet of Things and Cloud computing integrated systems (IoT-Cloud) facilitate sensing, automation, and processing for these healthcare applications. While real-time response is crucial for alleviating patient emergencies, protecting patient privacy is extremely important in data-driven healthcare. In this paper, we propose a multi-layer IoT, Edge and Cloud architecture to enhance the speed of response for emergency healthcare by distributing tasks based on response criticality and permanence of storage. Privacy of patient data is assured by proposing a Differential Privacy framework across several machine learning models such as K-means, Logistic Regression, Random Forest and Naive Bayes. We establish a comprehensive threat model identifying three adversary classes and evaluate Laplace, Gaussian, and hybrid noise mechanisms across varying privacy budgets, with supervised algorithms achieving up to 86% accuracy. The proposed hybrid Laplace-Gaussian noise mechanism with adaptive budget allocation provides a balanced approach, offering moderate tails and better privacy-utility trade-offs for both low and high dimension datasets. At the practical threshold of $\varepsilon = 5.0$, supervised algorithms achieve 82-84% accuracy while reducing attribute inference attacks by up to 18% and data reconstruction correlation by 70%. Blockchain security further ensures trusted communication through time-stamping, traceability, and immutability for analytics applications. Edge computing demonstrates 8$\times$ latency reduction for emergency scenarios, validating the hierarchical architecture for time-critical operations.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:37:37 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mangala",
"N",
""
],
[
"Rangwala",
"Murtaza",
""
],
[
"Aishwarya",
"S",
""
],
[
"Reddy",
"B Eswara",
""
],
[
"Buyya",
"Rajkumar",
""
],
[
"Venugopal",
"KR",
""
],
[
"Iyengar",
"SS",
""
],
[
"Patnaik",
"LM",
""
]
] |
2512.10428
|
Yu Haowen
|
Haowen Yu, Na Fan, Xing Liu, Ximin Lyu
|
Design and Implementation of a High-Precision Wind-Estimation UAV with Onboard Sensors
|
https://www.sciencedirect.com/science/article/abs/pii/S0263224125032415?via%3Dihub
| null |
10.1016/j.measurement.2025.119882
| null |
cs.ET cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Accurate real-time wind vector estimation is essential for enhancing the safety, navigation accuracy, and energy efficiency of unmanned aerial vehicles (UAVs). Traditional approaches rely on external sensors or simplify vehicle dynamics, which limits their applicability during agile flight or in resource-constrained platforms. This paper proposes a real-time wind estimation method based solely on onboard sensors. The approach first estimates external aerodynamic forces using a disturbance observer (DOB), and then maps these forces to wind vectors using a thin-plate spline (TPS) model. A custom-designed wind barrel mounted on the UAV enhances aerodynamic sensitivity, further improving estimation accuracy. The system is validated through comprehensive experiments in wind tunnels, indoor and outdoor flights. Experimental results demonstrate that the proposed method achieves consistently high-accuracy wind estimation across controlled and real-world conditions, with speed RMSEs as low as \SI{0.06}{m/s} in wind tunnel tests, \SI{0.22}{m/s} during outdoor hover, and below \SI{0.38}{m/s} in indoor and outdoor dynamic flights, and direction RMSEs under \ang{7.3} across all scenarios, outperforming existing baselines. Moreover, the method provides vertical wind estimates -- unavailable in baselines -- with RMSEs below \SI{0.17}{m/s} even during fast indoor translations.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:39:55 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yu",
"Haowen",
""
],
[
"Fan",
"Na",
""
],
[
"Liu",
"Xing",
""
],
[
"Lyu",
"Ximin",
""
]
] |
2512.10429
|
Ezequiel L\'opez-Rubio
|
Ezequiel Lopez-Rubio
|
Representation of the structure of graphs by sequences of instructions
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The representation of graphs is commonly based on the adjacency matrix concept. This formulation is the foundation of most algebraic and computational approaches to graph processing. The advent of deep learning language models offers a wide range of powerful computational models that are specialized in the processing of text. However, current procedures to represent graphs are not amenable to processing by these models. In this work, a new method to represent graphs is proposed. It represents the adjacency matrix of a graph by a string of simple instructions. The instructions build the adjacency matrix step by step. The transformation is reversible, i.e. given a graph the string can be produced and vice versa. The proposed representation is compact and it maintains the local structural patterns of the graph. Therefore, it is envisaged that it could be useful to boost the processing of graphs by deep learning models. A tentative computational experiment is reported, with favorable results.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:40:06 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lopez-Rubio",
"Ezequiel",
""
]
] |
2512.10430
|
Olga Tsymboi
|
Dmitrii Stoianov, Danil Taranets, Olga Tsymboi, Ramil Latypov, Almaz Dautov, Vladislav Kruglikov, Nikita Surkov, German Abramov, Pavel Gein, Dmitry Abulkhanov, Mikhail Gashkov, Viktor Zelenkovskiy, Artem Batalov, Aleksandr Medvedev, Anatolii Potapov
|
T-pro 2.0: An Efficient Russian Hybrid-Reasoning Model and Playground
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce T-pro 2.0, an open-weight Russian LLM for hybrid reasoning and efficient inference. The model supports direct answering and reasoning-trace generation, using a Cyrillic-dense tokenizer and an adapted EAGLE speculative-decoding pipeline to reduce latency. To enable reproducible and extensible research, we release the model weights, the T-Wix 500k instruction corpus, the T-Math reasoning benchmark, and the EAGLE weights on Hugging Face. These resources allow users to study Russian-language reasoning and to extend or adapt both the model and the inference pipeline. A public web demo exposes reasoning and non-reasoning modes and illustrates the speedups achieved by our inference stack across domains. T-pro 2.0 thus serves as an accessible open system for building and evaluating efficient, practical Russian LLM applications.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:40:10 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Stoianov",
"Dmitrii",
""
],
[
"Taranets",
"Danil",
""
],
[
"Tsymboi",
"Olga",
""
],
[
"Latypov",
"Ramil",
""
],
[
"Dautov",
"Almaz",
""
],
[
"Kruglikov",
"Vladislav",
""
],
[
"Surkov",
"Nikita",
""
],
[
"Abramov",
"German",
""
],
[
"Gein",
"Pavel",
""
],
[
"Abulkhanov",
"Dmitry",
""
],
[
"Gashkov",
"Mikhail",
""
],
[
"Zelenkovskiy",
"Viktor",
""
],
[
"Batalov",
"Artem",
""
],
[
"Medvedev",
"Aleksandr",
""
],
[
"Potapov",
"Anatolii",
""
]
] |
2512.10431
|
JiLei Hu
|
Jilei Hu, Fenglin He, Lianming Huang, Qianfeng Wang
|
A Sampling Strategy Benchmark for Machine-Learning-Based Seismic Liquefaction Prediction
| null | null | null | null |
physics.geo-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Sampling strategy including sampling methods and training set configurations (training set sample size, train-test split ratio, and class distribution) significantly affects machine-learning (ML) model performance in seismic liquefaction prediction. However, existing ML applications in seismic liquefaction prediction remain fragmented: sampling strategies vary widely across studies without a unified benchmark. Moreover, these studies generally optimize the sample set configuration independently, ignoring the interaction among training set configurations. To address these limitations, this study establishes a benchmark that systematically evaluates sampling methods, training set sample sizes, train-test split ratios, class distributions, and training set configurations coupling on seven mainstream ML models performance, and further improves the predictive accuracy of seismic liquefaction-using a database of 250 historical liquefaction events, evaluated by Acc and F1. The results show that ordered systematic sampling yields the best performance across all models. The optimal model can be trained when the training set sample size is 200, the train-test split ratio is 80:20, and the class distribution range is 1-1.5. Among them, the train-test split ratio most significantly influenced performance, followed by the class distribution, with the training set sample size having the least effect. Furthermore, the Random Forest model achieves the highest performance, while the K-Nearest Neighbor model performs the weakest. Importantly, this study systematically identifies and verifies for the first time that there will be an interaction effect among training set configurations, rather than a simple additive effect. This study provides a benchmark for scholars to select the optimal sampling method and training set configurations to obtain high accuracy in ML-based liquefaction prediction.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:40:16 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Hu",
"Jilei",
""
],
[
"He",
"Fenglin",
""
],
[
"Huang",
"Lianming",
""
],
[
"Wang",
"Qianfeng",
""
]
] |
2512.10432
|
Andon A. Rangelov
|
Peter Chernev and Andon A. Rangelov
|
Robust population transfer by a detuning sign jump: from two-state quantum system to SU(2)-symmetric three-state quantum system
| null | null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose and analyze a robust population-transfer protocol in a driven two-level system based on a sudden sign change of the detuning at the maximum of a smooth coupling pulse. Away from the jump the dynamics is adiabatic, while the sign flip produces a single nonadiabatic kick in the adiabatic basis. Within a simple stepwise adiabatic-sudden approximation we obtain a compact analytic expression for the final transition probability, identify the parameter regimes that yield high-fidelity inversion, and show that the result depends only on the change of the mixing angle across the detuning jump, i.e., solely on the ratio of the peak Rabi frequency to the detuning. Numerical simulations of the full time-dependent Schr\"odinger equation confirm the validity and robustness of this description over a broad parameter range.
We then use the Majorana decomposition to extend the scheme to an SU(2)-symmetric three-state chain driven by the same coupling and detuning functions. In this setting the three-state propagator is expressed in closed form through the two-level Cayley-Klein parameters, which allows us to derive explicit transition probabilities for all three initial states. In particular, we show that for strong coupling the protocol yields almost complete population transfer between the two outer states, with only small transient population of the middle state, while retaining the same intrinsic robustness as in the underlying two-level model.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:40:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chernev",
"Peter",
""
],
[
"Rangelov",
"Andon A.",
""
]
] |
2512.10433
|
Mijin Koo
|
Hojun Lee, Mijin Koo, Yeji Song, Nojun Kwak
|
Targeted Data Protection for Diffusion Model by Matching Training Trajectory
|
AAAI 2026
| null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Recent advancements in diffusion models have made fine-tuning text-to-image models for personalization increasingly accessible, but have also raised significant concerns regarding unauthorized data usage and privacy infringement. Current protection methods are limited to passively degrading image quality, failing to achieve stable control. While Targeted Data Protection (TDP) offers a promising paradigm for active redirection toward user-specified target concepts, existing TDP attempts suffer from poor controllability due to snapshot-matching approaches that fail to account for complete learning dynamics. We introduce TAFAP (Trajectory Alignment via Fine-tuning with Adversarial Perturbations), the first method to successfully achieve effective TDP by controlling the entire training trajectory. Unlike snapshot-based methods whose protective influence is easily diluted as training progresses, TAFAP employs trajectory-matching inspired by dataset distillation to enforce persistent, verifiable transformations throughout fine-tuning. We validate our method through extensive experiments, demonstrating the first successful targeted transformation in diffusion models with simultaneous control over both identity and visual patterns. TAFAP significantly outperforms existing TDP attempts, achieving robust redirection toward target concepts while maintaining high image quality. This work enables verifiable safeguards and provides a new framework for controlling and tracing alterations in diffusion model outputs.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:47:41 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lee",
"Hojun",
""
],
[
"Koo",
"Mijin",
""
],
[
"Song",
"Yeji",
""
],
[
"Kwak",
"Nojun",
""
]
] |
2512.10434
|
Sikander Azam Dr
|
Sikander Azam, Qaiser Rafiq, Rajwali Khan, Hamdy Khamees Thabet
|
Engineering Multifunctional Response in Monolayer Fe3O4 via Zr Adsorption: From Half-Metallicity to Enhanced Piezoelectricity
| null | null | null | null |
cond-mat.mtrl-sci physics.comp-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Two-dimensional (2D) magnetic oxides are increasingly studied for their multifunctional potential in fields like spintronics, optoelectronics, and energy conversion. In this research, we conduct a detailed first-principles study of pure monolayer Fe3O4 and its modification through Zr adsorption at two sites: on top of an Fe atom and at the bridge between Fe atoms. Using spin-polarized density functional theory with the GGA plus U method, we examine how adsorption affects structure, electronic, magnetic, optical, elastic, and piezoelectric properties. The original monolayer shows half-metallicity, strong spin polarization, and a moderate in-plane piezoelectric effect. Zr adsorption causes local lattice distortions and orbital hybridization, resulting in intermediate electronic states, a reduced bandgap, and increased optical absorption in both spin channels. Notably, Zr at the bridge site greatly enhances dielectric response, optical conductivity, and piezoelectric coefficients, tripling e11 compared to the pristine layer. Elastic constants indicate mechanical softening after functionalization, and energy loss spectra display shifts in plasmon resonance. These findings suggest Zr adsorption offers a controllable, non-destructive way to tune spin, charge, and lattice interactions in Fe4O4 monolayers, connecting magnetic, optical, and piezoelectric functionalities within a single 2D material platform.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:47:54 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Azam",
"Sikander",
""
],
[
"Rafiq",
"Qaiser",
""
],
[
"Khan",
"Rajwali",
""
],
[
"Thabet",
"Hamdy Khamees",
""
]
] |
2512.10435
|
Agniva Maiti
|
Agniva Maiti, Prajwal Panth, and Suresh Chandra Satapathy
|
Semantic Reconstruction of Adversarial Plagiarism: A Context-Aware Framework for Detecting and Restoring "Tortured Phrases" in Scientific Literature
|
10 pages, 5 figures; unpublished manuscript; submitted to arXiv for dissemination
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The integrity and reliability of scientific literature is facing a serious threat by adversarial text generation techniques, specifically from the use of automated paraphrasing tools to mask plagiarism. These tools generate "tortured phrases", statistically improbable synonyms (e.g. "counterfeit consciousness" for "artificial intelligence"), that preserve the local grammar while obscuring the original source. Most existing detection methods depend heavily on static blocklists or general-domain language models, which suffer from high false-negative rates for novel obfuscations and cannot determine the source of the plagiarized content. In this paper, we propose Semantic Reconstruction of Adversarial Plagiarism (SRAP), a framework designed not only to detect these anomalies but to mathematically recover the original terminology. We use a two-stage architecture: (1) statistical anomaly detection with a domain-specific masked language model (SciBERT) using token-level pseudo-perplexity, and (2) source-based semantic reconstruction using dense vector retrieval (FAISS) and sentence-level alignment (SBERT). Experiments on a parallel corpus of adversarial scientific text show that while zero-shot baselines fail completely (0.00 percent restoration accuracy), our retrieval-augmented approach achieves 23.67 percent restoration accuracy, significantly outperforming baseline methods. We also show that static decision boundaries are necessary for robust detection in jargon-heavy scientific text, since dynamic thresholding fails under high variance. SRAP enables forensic analysis by linking obfuscated expressions back to their most probable source documents.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:53:25 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Maiti",
"Agniva",
""
],
[
"Panth",
"Prajwal",
""
],
[
"Satapathy",
"Suresh Chandra",
""
]
] |
2512.10436
|
Flavio S. Nogueira
|
Anastasiia Chyzhykova, Jeroen van den Brink, Flavio S. Nogueira
|
The Feynman paradox in a spherical axion insulator
|
9 pages, 4 figures
| null | null | null |
cond-mat.mes-hall hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
We show that a small charged probe near a spherical topological insulator causes the latter to rotate around a symmetry axis defined by the center of the sphere and the position of the charge outside the latter. The rotation occurs when the distance from the charge to the center of the sphere is changed. This phenomenon occurs due to induced static fields and is a consequence of the axion electrodynamics underlying the electromagnetic response of a topological insulator. Assuming a regime where the charged probe can be regarded as a point charge $q=Ne$, where $N$ is a positive integer and $e$ is the elementary electric charge, we obtain that the rotation frequency is given by $\omega=(N\alpha)^2\Upsilon(\epsilon,d/a)/I$, where $I$ is the moment of inertia, $\alpha$ is the fine-structure constant, and the function $\Upsilon$ depends on the dielectric constant $\epsilon$ and the relative distance $d/a$ of the charge from the center of the sphere of radius $a$. Since the point charge also induces Hall currents on the surface, we compute also their associated angular momentum. This allows us to derive an exact expression for the electronic velocity on the surface as a function of $a/d$.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:54:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chyzhykova",
"Anastasiia",
""
],
[
"Brink",
"Jeroen van den",
""
],
[
"Nogueira",
"Flavio S.",
""
]
] |
2512.10437
|
Christos Orfanopoulos
|
Stylianos Kandylakis, Christos Orfanopoulos, Georgios Siolas, and Panayiotis Tsanakas
|
An M-Health Algorithmic Approach to Identify and Assess Physiotherapy Exercises in Real Time
|
11 pages, 5 figures
| null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
This work presents an efficient algorithmic framework for real-time identification, classification, and evaluation of human physiotherapy exercises using mobile devices. The proposed method interprets a kinetic movement as a sequence of static poses, which are estimated from camera input using a pose-estimation neural network. Extracted body keypoints are transformed into trigonometric angle-based features and classified with lightweight supervised models to generate frame-level pose predictions and accuracy scores. To recognize full exercise movements and detect deviations from prescribed patterns, we employ a dynamic-programming scheme based on a modified Levenshtein distance algorithm, enabling robust sequence matching and localization of inaccuracies. The system operates entirely on the client side, ensuring scalability and real-time performance. Experimental evaluation demonstrates the effectiveness of the methodology and highlights its applicability to remote physiotherapy supervision and m-health applications.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:56:03 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kandylakis",
"Stylianos",
""
],
[
"Orfanopoulos",
"Christos",
""
],
[
"Siolas",
"Georgios",
""
],
[
"Tsanakas",
"Panayiotis",
""
]
] |
2512.10438
|
Yuval Wigderson
|
Jacob Fox, Benny Sudakov, and Yuval Wigderson
|
Color-avoiding Ramsey for directed paths and vector sequences
|
22 pages
| null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
We study two closely related Ramsey-type questions. The first is about vectors: how long can a sequence of vectors in $[n]^q$ be if, for all pairs in the sequence, the later vector is strictly larger than the earlier in at least $q-1$ coordinates? The second is graph-theoretic: given a $q$-edge-coloring of a tournament, how long of a directed path can we guarantee whose edges avoid one of the colors? These problems arise naturally in many areas, such as convex geometry and extremal hypergraph theory, and have been extensively studied over the past 50 years.
We prove that if $\varepsilon>0$ is fixed and $q$ is sufficiently large, then every $N$-vertex tournament contains a color-avoiding directed path of length $N^{1-\varepsilon}$. Consequently, the maximum length of a $(q-1)$-increasing sequence of vectors in $[n]^q$ has length at most $n^{1+\varepsilon}$. Our results answer a question of Gowers and Long, strengthen several of their results, and extend earlier works of Tiskin and Loh.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:56:48 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Fox",
"Jacob",
""
],
[
"Sudakov",
"Benny",
""
],
[
"Wigderson",
"Yuval",
""
]
] |
2512.10439
|
Niccol\`o Grillo
|
Niccol\`o Grillo, James Rowbottom, Pietro Li\`o, Carola Bibiane Sch\"onlieb, Stefania Fresca
|
HypeR Adaptivity: Joint $hr$-Adaptive Meshing via Hypergraph Multi-Agent Deep Reinforcement Learning
| null | null | null | null |
cs.CE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Adaptive mesh refinement is central to the efficient solution of partial differential equations (PDEs) via the finite element method (FEM). Classical $r$-adaptivity optimizes vertex positions but requires solving expensive auxiliary PDEs such as the Monge-Amp\`ere equation, while classical $h$-adaptivity modifies topology through element subdivision but suffers from expensive error indicator computation and is constrained by isotropic refinement patterns that impose accuracy ceilings. Combined $hr$-adaptive techniques naturally outperform single-modality approaches, yet inherit both computational bottlenecks and the restricted cost-accuracy trade-off. Emerging machine learning methods for adaptive mesh refinement seek to overcome these limitations, but existing approaches address $h$-adaptivity or $r$-adaptivity in isolation. We present HypeR, a deep reinforcement learning framework that jointly optimizes mesh relocation and refinement. HypeR casts the joint adaptation problem using tools from hypergraph neural networks and multi-agent reinforcement learning. Refinement is formulated as a heterogeneous multi-agent Markov decision process (MDP) where element agents decide discrete refinement actions, while relocation follows an anisotropic diffusion-based policy on vertex agents with provable prevention of mesh tangling. The reward function combines local and global error reduction to promote general accuracy. Across benchmark PDEs, HypeR reduces approximation error by up to 6--10$\times$ versus state-of-art $h$-adaptive baselines at comparable element counts, breaking through the uniform refinement accuracy ceiling that constrains subdivision-only methods. The framework produces meshes with improved shape metrics and alignment to solution anisotropy, demonstrating that jointly learned $hr$-adaptivity strategies can substantially enhance the capabilities of automated mesh generation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:02:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Grillo",
"Niccolò",
""
],
[
"Rowbottom",
"James",
""
],
[
"Liò",
"Pietro",
""
],
[
"Schönlieb",
"Carola Bibiane",
""
],
[
"Fresca",
"Stefania",
""
]
] |
2512.10440
|
Nour El Houda Ben Chaabene
|
Nour El Houda Ben Chaabene, Hamza Hammami
|
Enhancing Next-Generation Language Models with Knowledge Graphs: Extending Claude, Mistral IA, and GPT-4 via KG-BERT
|
This paper was accepted and scheduled for inclusion in the ICALT 2025 proceedings but was ultimately not published due to absence from the conference presentation. It appears in the official program booklet. Conference: 2025 IEEE International Conference on Advanced Learning Technologies (ICALT)
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language models (LLMs) like Claude, Mistral IA, and GPT-4 excel in NLP but lack structured knowledge, leading to factual inconsistencies. We address this by integrating Knowledge Graphs (KGs) via KG-BERT to enhance grounding and reasoning. Experiments show significant gains in knowledge-intensive tasks such as question answering and entity linking. This approach improves factual reliability and enables more context-aware next-generation LLMs.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:02:45 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chaabene",
"Nour El Houda Ben",
""
],
[
"Hammami",
"Hamza",
""
]
] |
2512.10441
|
Nour El Houda Ben Chaabene
|
Nour El Houda Ben Chaabene, Hamza Hammami, Laid Kahloul
|
Decoding Student Minds: Leveraging Conversational Agents for Psychological and Learning Analysis
|
This manuscript is currently under peer review in Expert Systems with Applications
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a psychologically-aware conversational agent designed to enhance both learning performance and emotional well-being in educational settings. The system combines Large Language Models (LLMs), a knowledge graph-enhanced BERT (KG-BERT), and a bidirectional Long Short-Term Memory (LSTM) with attention to classify students' cognitive and affective states in real time. Unlike prior chatbots limited to either tutoring or affective support, our approach leverages multimodal data-including textual semantics, prosodic speech features, and temporal behavioral trends-to infer engagement, stress, and conceptual understanding. A pilot study with university students demonstrated improved motivation, reduced stress, and moderate academic gains compared to baseline methods. These results underline the promise of integrating semantic reasoning, multimodal fusion, and temporal modeling to support adaptive, student-centered educational interventions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:06:45 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chaabene",
"Nour El Houda Ben",
""
],
[
"Hammami",
"Hamza",
""
],
[
"Kahloul",
"Laid",
""
]
] |
2512.10442
|
Helena Del R\'io
|
Helena Del R\'io
|
Bollob\'as-type theorems for range strongly exposing operators
| null | null | null | null |
math.FA
|
http://creativecommons.org/licenses/by/4.0/
|
We study Bollob\'as-type theorems for range strongly exposing operators. When such a theorem holds for operators from a Banach space $X$ into another Banach space $Y$, we say that the pair $(X,Y)$ satisfies the Bishop-Phelps-Bollob\'as property for range strongly exposing operators (BPBp-RSE, for short). We provide new characterisations of uniform convexity and complex uniform convexity via the BPBp-RSE, including for pairs involving spaces such as $L_1(\mu), L_\infty(\mu)$ and $c_0$. In particular, we show that $(L_1(\mu), Y)$ satisfies the BPBp-RSE if and only if $Y$ is uniformly convex, and that $(L_\infty(\mu), Y)$ or $(c_0, Y)$ satisfy the BPBp-RSE if and only if $Y$ is $\mathbb{C}$-uniformly convex. We also highlight differences between the real and complex cases, showing that there exist pairs $(X, Y)$ for which the BPBp-RSE holds in the complex setting but fails for their respective underlying real spaces. Additionally, we consider various subspaces of operators, such as compact and finite-rank, and extend several results from the literature to this new setting. The paper concludes with a collection of open problems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:07:08 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Del Río",
"Helena",
""
]
] |
2512.10443
|
Sabtain Ahmad
|
Sabtain Ahmad, Meerzhan Kanatbekova, Ivona Brandic, Atakan Aral
|
Clustered Federated Learning with Hierarchical Knowledge Distillation
| null | null | null | null |
cs.DC cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Clustered Federated Learning (CFL) has emerged as a powerful approach for addressing data heterogeneity and ensuring privacy in large distributed IoT environments. By clustering clients and training cluster-specific models, CFL enables personalized models tailored to groups of heterogeneous clients. However, conventional CFL approaches suffer from fragmented learning for training independent global models for each cluster and fail to take advantage of collective cluster insights. This paper advocates a shift to hierarchical CFL, allowing bi-level aggregation to train cluster-specific models at the edge and a unified global model at the cloud. This shift improves training efficiency yet might introduce communication challenges. To this end, we propose CFLHKD, a novel personalization scheme for integrating hierarchical cluster knowledge into CFL. Built upon multi-teacher knowledge distillation, CFLHKD enables inter-cluster knowledge sharing while preserving cluster-specific personalization. CFLHKD adopts a bi-level aggregation to bridge the gap between local and global learning. Extensive evaluations of standard benchmark datasets demonstrate that CFLHKD outperforms representative baselines in cluster-specific and global model accuracy and achieves a performance improvement of 3.32-7.57\%.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:08:35 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ahmad",
"Sabtain",
""
],
[
"Kanatbekova",
"Meerzhan",
""
],
[
"Brandic",
"Ivona",
""
],
[
"Aral",
"Atakan",
""
]
] |
2512.10445
|
Francesco Freni
|
Francesco Freni, Anya Fries, Linus K\"uhne, Markus Reichstein, Jonas Peters
|
Maximum Risk Minimization with Random Forests
|
47 pages, 13 figures
| null | null | null |
stat.ML cs.AI cs.LG stat.ME
|
http://creativecommons.org/licenses/by/4.0/
|
We consider a regression setting where observations are collected in different environments modeled by different data distributions. The field of out-of-distribution (OOD) generalization aims to design methods that generalize better to test environments whose distributions differ from those observed during training. One line of such works has proposed to minimize the maximum risk across environments, a principle that we refer to as MaxRM (Maximum Risk Minimization). In this work, we introduce variants of random forests based on the principle of MaxRM. We provide computationally efficient algorithms and prove statistical consistency for our primary method. Our proposed method can be used with each of the following three risks: the mean squared error, the negative reward (which relates to the explained variance), and the regret (which quantifies the excess risk relative to the best predictor). For MaxRM with regret as the risk, we prove a novel out-of-sample guarantee over unseen test distributions. Finally, we evaluate the proposed methods on both simulated and real-world data.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:10:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Freni",
"Francesco",
""
],
[
"Fries",
"Anya",
""
],
[
"Kühne",
"Linus",
""
],
[
"Reichstein",
"Markus",
""
],
[
"Peters",
"Jonas",
""
]
] |
2512.10446
|
Chiara Boetti
|
Chiara Boetti, Matthew A. Nunes, Marina I. Knight
|
Long memory network time series
| null | null | null | null |
stat.ME
|
http://creativecommons.org/licenses/by/4.0/
|
Many scientific areas, from computer science to the environmental sciences and finance, give rise to multivariate time series which exhibit long memory, or loosely put, a slow decay in their autocorrelation structure. Efficient modelling and estimation in such settings is key for a number of analysis tasks, such as accurate prediction. However, traditional approaches for modelling such data, for example long memory vector autoregressive processes, are challenging even in modest dimensions, as the number of parameters grows quadratically with the number of modelled variables. Additionally, in many practical data settings, the observed series is accompanied by a (possibly inferred) network that provides information about the presence or absence of between-component associations via the graph edge topology. This article proposes two new models for capturing the dynamics of long memory time series where a network is accounted for. Our approach not only facilitates the analysis of graph-structured long memory time series, but also improves computational efficiency over traditional multivariate long memory models by leveraging the inherent low-dimensional parameter space by adapting likelihood-based estimation algorithms to the network setting. Simulation studies show that our proposed estimation is more stable than traditional models, and is able to tackle data scenarios where current models fail due to computational challenges. While widely applicable, here we demonstrate the efficacy of our proposed models on datasets arising in environmental science and finance.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:12:37 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Boetti",
"Chiara",
""
],
[
"Nunes",
"Matthew A.",
""
],
[
"Knight",
"Marina I.",
""
]
] |
2512.10447
|
Kaisei Higeta
|
Kaisei Higeta (Sophia University), Masakatsu Ogawa (Sophia University), Tomoki Murakami (Access Network Service Laboratories, NTT, Inc), Kazuya Ohara (Access Network Service Laboratories, NTT, Inc), Shinya Otsuki (Access Network Service Laboratories, NTT, Inc)
|
Outdoor Crowd Flow Estimation Using RSRP from Commercial LTE Base Station: A Field Study
|
6 pages, 15 figures, Accepted for presentation at the International Conference on Information Networking (ICOIN) 2026
| null | null | null |
eess.SP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
With the advent of the 6G era, Integrated Sensing and Communications (ISAC) has attracted increasing attention. One representative of use cases is crowd flow estimation on outdoor streets. However, most existing studies have focused on indoor environments or vehicles, and demonstrations of outdoor crowd flow estimation using commercial LTE base station remain limited. This study addresses this use case and proposes an analysis of a crowd flow estimation method using Reference Signal Received Power (RSRP) obtained from a commercial LTE base station. Specifically, pedestrian counts derived from a camera-based object recognition algorithm were associated with the variance of RSRP. The features obtained from the variance were quantitatively evaluated by combining a CatBoost regression model with SHapley Additive exPlanations (SHAP) analysis. Through this investigation, we clarified that an optimal variance window size for RSRP is 0.1 to 0.2 seconds and that enlarging the counting area increased the features obtained from the variance of RSRP, for machine learning. Consequently, this study is the first to quantitatively demonstrate the effectiveness of outdoor crowd flow estimation using commercial LTE, while also revealing the characteristic behavior of variance window size and counting area size in feature design.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:13:09 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Higeta",
"Kaisei",
"",
"Sophia University"
],
[
"Ogawa",
"Masakatsu",
"",
"Sophia University"
],
[
"Murakami",
"Tomoki",
"",
"Access Network Service Laboratories, NTT, Inc"
],
[
"Ohara",
"Kazuya",
"",
"Access Network Service Laboratories, NTT, Inc"
],
[
"Otsuki",
"Shinya",
"",
"Access Network Service Laboratories, NTT, Inc"
]
] |
2512.10448
|
Sanket Bajad
|
Sanket Narayan Bajad and Pawan Bharadwaj
|
Coherent Source Subsampling: A Data-Driven Strategy for Restoring Causal-Acausal Symmetry in Ambient Seismic Wavefield Correlations
| null | null | null | null |
physics.geo-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Ambient noise tomography relies on the assumption that the seismic wavefield is equipartitioned, meaning that energy is uniformly distributed among all directions. However, in practice, ambient noise sources are highly non-uniform in both spatial and temporal dimensions, resulting in biased estimation of the Green's function between stations. We introduce a data-driven method, Coherent Source Subsampling (CSS), which selects and averages only those cross-correlation time windows that are associated with the excitation of sources in the stationary-zone. By confining the ensemble average to these coherent subsets, CSS effectively mitigates the influence of anisotropic or intermittent sources and restores causal-acausal symmetry in the retrieved Green's functions. Applications to regional-scale ambient noise datasets demonstrate that CSS boosts inter-station coherence and enhances the reliability of surface-wave dispersion measurements, providing a physically interpretable bridge between source statistics and noise correlation theory.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:13:28 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bajad",
"Sanket Narayan",
""
],
[
"Bharadwaj",
"Pawan",
""
]
] |
2512.10449
|
Devanshu Sahoo
|
Devanshu Sahoo, Manish Prasad, Vasudev Majhi, Jahnvi Singh, Vinay Chamola, Yash Sinha, Murari Mandal, Dhruv Kumar
|
When Reject Turns into Accept: Quantifying the Vulnerability of LLM-Based Scientific Reviewers to Indirect Prompt Injection
| null | null | null | null |
cs.AI cs.CL cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The landscape of scientific peer review is rapidly evolving with the integration of Large Language Models (LLMs). This shift is driven by two parallel trends: the widespread individual adoption of LLMs by reviewers to manage workload (the "Lazy Reviewer" hypothesis) and the formal institutional deployment of AI-powered assessment systems by conferences like AAAI and Stanford's Agents4Science. This study investigates the robustness of these "LLM-as-a-Judge" systems (both illicit and sanctioned) to adversarial PDF manipulation. Unlike general jailbreaks, we focus on a distinct incentive: flipping "Reject" decisions to "Accept," for which we develop a novel evaluation metric which we term as WAVS (Weighted Adversarial Vulnerability Score). We curated a dataset of 200 scientific papers and adapted 15 domain-specific attack strategies to this task, evaluating them across 13 Language Models, including GPT-5, Claude Haiku, and DeepSeek. Our results demonstrate that obfuscation strategies like "Maximum Mark Magyk" successfully manipulate scores, achieving alarming decision flip rates even in large-scale models. We will release our complete dataset and injection framework to facilitate more research on this topic.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:13:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Sahoo",
"Devanshu",
""
],
[
"Prasad",
"Manish",
""
],
[
"Majhi",
"Vasudev",
""
],
[
"Singh",
"Jahnvi",
""
],
[
"Chamola",
"Vinay",
""
],
[
"Sinha",
"Yash",
""
],
[
"Mandal",
"Murari",
""
],
[
"Kumar",
"Dhruv",
""
]
] |
2512.10450
|
Han Li
|
Han Li, Shaohui Li, Wenrui Dai, Chenglin Li, Xinlong Pan, Haipeng Wang, Junni Zou, Hongkai Xiong
|
Error-Propagation-Free Learned Video Compression With Dual-Domain Progressive Temporal Alignment
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Existing frameworks for learned video compression suffer from a dilemma between inaccurate temporal alignment and error propagation for motion estimation and compensation (ME/MC). The separate-transform framework employs distinct transforms for intra-frame and inter-frame compression to yield impressive rate-distortion (R-D) performance but causes evident error propagation, while the unified-transform framework eliminates error propagation via shared transforms but is inferior in ME/MC in shared latent domains. To address this limitation, in this paper, we propose a novel unifiedtransform framework with dual-domain progressive temporal alignment and quality-conditioned mixture-of-expert (QCMoE) to enable quality-consistent and error-propagation-free streaming for learned video compression. Specifically, we propose dualdomain progressive temporal alignment for ME/MC that leverages coarse pixel-domain alignment and refined latent-domain alignment to significantly enhance temporal context modeling in a coarse-to-fine fashion. The coarse pixel-domain alignment efficiently handles simple motion patterns with optical flow estimated from a single reference frame, while the refined latent-domain alignment develops a Flow-Guided Deformable Transformer (FGDT) over latents from multiple reference frames to achieve long-term motion refinement (LTMR) for complex motion patterns. Furthermore, we design a QCMoE module for continuous bit-rate adaptation that dynamically assigns different experts to adjust quantization steps per pixel based on target quality and content rather than relies on a single quantization step. QCMoE allows continuous and consistent rate control with appealing R-D performance. Experimental results show that the proposed method achieves competitive R-D performance compared with the state-of-the-arts, while successfully eliminating error propagation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:14:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Han",
""
],
[
"Li",
"Shaohui",
""
],
[
"Dai",
"Wenrui",
""
],
[
"Li",
"Chenglin",
""
],
[
"Pan",
"Xinlong",
""
],
[
"Wang",
"Haipeng",
""
],
[
"Zou",
"Junni",
""
],
[
"Xiong",
"Hongkai",
""
]
] |
2512.10451
|
Le Tuan Minh Trinh
|
Le Tuan Minh Trinh, Le Minh Vu Pham, Thi Minh Anh Pham, An Duc Nguyen
|
Metacognitive Sensitivity for Test-Time Dynamic Model Selection
|
Accepted at the NeurIPS 2025 CogInterp Workshop
| null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
A key aspect of human cognition is metacognition - the ability to assess one's own knowledge and judgment reliability. While deep learning models can express confidence in their predictions, they often suffer from poor calibration, a cognitive bias where expressed confidence does not reflect true competence. Do models truly know what they know? Drawing from human cognitive science, we propose a new framework for evaluating and leveraging AI metacognition. We introduce meta-d', a psychologically-grounded measure of metacognitive sensitivity, to characterise how reliably a model's confidence predicts its own accuracy. We then use this dynamic sensitivity score as context for a bandit-based arbiter that performs test-time model selection, learning which of several expert models to trust for a given task. Our experiments across multiple datasets and deep learning model combinations (including CNNs and VLMs) demonstrate that this metacognitive approach improves joint-inference accuracy over constituent models. This work provides a novel behavioural account of AI models, recasting ensemble selection as a problem of evaluating both short-term signals (confidence prediction scores) and medium-term traits (metacognitive sensitivity).
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:15:05 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Trinh",
"Le Tuan Minh",
""
],
[
"Pham",
"Le Minh Vu",
""
],
[
"Pham",
"Thi Minh Anh",
""
],
[
"Nguyen",
"An Duc",
""
]
] |
2512.10452
|
Yang Yang
|
Yang Yang, Li Kuang, Jiakun Liu, Zhongxin Liu, Yingjie Xia, David Lo
|
UniCoR: Modality Collaboration for Robust Cross-Language Hybrid Code Retrieval
|
Accepted by the 48th IEEE/ACM International Conference on Software Engineering (ICSE 2026)
| null | null | null |
cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Effective code retrieval is indispensable and it has become an important paradigm to search code in hybrid mode using both natural language and code snippets. Nevertheless, it remains unclear whether existing approaches can effectively leverage such hybrid queries, particularly in cross-language contexts. We conduct a comprehensive empirical study of representative code models and reveal three challenges: (1) insufficient semantic understanding; (2) inefficient fusion in hybrid code retrieval; and (3) weak generalization in cross-language scenarios. To address these challenges, we propose UniCoR, a novel self-supervised framework that learns Unified Code Representations framework designed to learn unified and robust code representations. Firstly, we design a multi-perspective supervised contrastive learning module to enhance semantic understanding and modality fusion. It aligns representations from multiple perspectives, including code-to-code, natural language-to-code, and natural language-to-natural language, enforcing the model to capture a semantic essence among modalities. Secondly, we introduce a representation distribution consistency learning module to improve cross-language generalization, which explicitly aligns the feature distributions of different programming languages, enabling language-agnostic representation learning. Extensive experiments on both empirical benchmark and large-scale benchmark show that UniCoR outperforms all baseline models, achieving an average improvement of 8.64% in MRR and 11.54% in MAP over the best-performing baseline. Furthermore, UniCoR exhibits stability in hybrid code retrieval and generalization capability in cross-language scenarios.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:15:38 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yang",
"Yang",
""
],
[
"Kuang",
"Li",
""
],
[
"Liu",
"Jiakun",
""
],
[
"Liu",
"Zhongxin",
""
],
[
"Xia",
"Yingjie",
""
],
[
"Lo",
"David",
""
]
] |
2512.10453
|
Lars G Bag{\o}ien Johnsen
|
Lars G.B. Johnsen
|
Grammaticality Judgments in Humans and Language Models: Revisiting Generative Grammar with LLMs
|
2 figures
| null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
What counts as evidence for syntactic structure? In traditional generative grammar, systematic contrasts in grammaticality such as subject-auxiliary inversion and the licensing of parasitic gaps are taken as evidence for an internal, hierarchical grammar. In this paper, we test whether large language models (LLMs), trained only on surface forms, reproduce these contrasts in ways that imply an underlying structural representation.
We focus on two classic constructions: subject-auxiliary inversion (testing recognition of the subject boundary) and parasitic gap licensing (testing abstract dependency structure). We evaluate models including GPT-4 and LLaMA-3 using prompts eliciting acceptability ratings. Results show that LLMs reliably distinguish between grammatical and ungrammatical variants in both constructions, and as such support that they are sensitive to structure and not just linear order. Structural generalizations, distinct from cognitive knowledge, emerge from predictive training on surface forms, suggesting functional sensitivity to syntax without explicit encoding.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:17:35 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Johnsen",
"Lars G. B.",
""
]
] |
2512.10454
|
Ian Mochida
|
Ian F. Mochida, Tetsuyuki Takayama, Shoichi Yamaguchi, and Tetsuya Hama
|
An Extended Mixed Quantum/Classical Approach for Quantitative Calculation of Complex Refractive Index
| null |
J. Phys. Chem. Lett. 2025, 16, XXX, 12741-12751
|
10.1021/acs.jpclett.5c02887
| null |
physics.optics physics.chem-ph
|
http://creativecommons.org/licenses/by/4.0/
|
The mixed quantum/classical approach of Skinner and co-workers has been widely used to calculate the line shapes of the infrared spectra of water (H2O), but less attention has been paid to the use of this approach in quantitatively calculating spectral intensity, thereby limiting direct comparisons of calculated and experimental spectra. Here, we extend this theoretical framework to facilitate direct computation of the full complex refractive index of water, replacing the normalized ordinate used in previous studies. Our results for the OH stretching region of H2O capture both the shapes and intensities of the experimental spectra. They reveal that inclusion of the local field effect is crucial to the accurate reproduction of spectral intensity. This extended approach enables new areas of analysis of the bulk, thin-film, and cluster spectra of water.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:19:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mochida",
"Ian F.",
""
],
[
"Takayama",
"Tetsuyuki",
""
],
[
"Yamaguchi",
"Shoichi",
""
],
[
"Hama",
"Tetsuya",
""
]
] |
2512.10455
|
Marc Abboud
|
Marc Abboud
|
A dynamical characterisation of smooth cubic affine surfaces of Markov type
| null | null | null | null |
math.AG math.CV math.DS
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Using valuative techniques, we show that a smooth affine surface with a non-elementary automorphism group and completable by a cycle of rational curves is either the algebraic torus or a smooth cubic affine surface of Markov type. Furthermore we show that smooth cubic affine surfaces of Markov type do no admit dominant endomorphisms which are not automorphisms.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:25:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Abboud",
"Marc",
""
]
] |
2512.10456
|
Xizhuang Xie
|
Lei Niu, Yi Wang, and Xizhuang Xie
|
Global dynamics of three-dimensional Lotka-Volterra competition models with seasonal succession: II. Uniqueness of positive fixed points
| null | null | null | null |
math.DS
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In this second part of the series, we investigate the uniqueness of positive fixed points of the Poincare map associated with the 3-dimensional Lotka-Volterra competition model with seasonal succession. Building on our first part of the series on the classification of 33 dynamical equivalence classes (regardless of the uniqueness of positive fixed points), we demonstrate in this paper that classes 26 and 27 may indeed exhibit multiple positive fixed points. This reveals a fundamental distinction from both its 2-dimensional analogue and the classical 3-dimensional competitive Lotka-Volterra model. More concretely, by focusing on the model with identical growth and death rates, we establish an equivalent characterization for the (non)uniqueness of positive fixed points. Based on this characterization, we further show that classes 19-25 and 28-33 admit a unique positive fixed point and exhibit trivial dynamics: all trajectories converge to some fixed point (corresponding to harmonic solutions). In contrast, classes 26 and 27 possess richer dynamical scenarios: there can contain a continuum of invariant closed curves, on which orbits may be periodic (corresponding to subharmonic solutions), dense (corresponding to quasi-periodic solutions), or may even consist entirely of positive fixed points (which exhibits the nonuniqueness of positive fixed points).
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:26:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Niu",
"Lei",
""
],
[
"Wang",
"Yi",
""
],
[
"Xie",
"Xizhuang",
""
]
] |
2512.10457
|
Shiv Ratn
|
Shiv Ratn, Shivang Rampriyan, Bahni Ray
|
Hybrid Physics-ML Model for Forward Osmosis Flux with Complete Uncertainty Quantification
|
12 pages, 6 figures
| null | null | null |
cs.LG stat.ML
|
http://creativecommons.org/licenses/by/4.0/
|
Forward Osmosis (FO) is a promising low-energy membrane separation technology, but challenges in accurately modelling its water flux (Jw) persist due to complex internal mass transfer phenomena. Traditional mechanistic models struggle with empirical parameter variability, while purely data-driven models lack physical consistency and rigorous uncertainty quantification (UQ). This study introduces a novel Robust Hybrid Physics-ML framework employing Gaussian Process Regression (GPR) for highly accurate, uncertainty-aware Jw prediction. The core innovation lies in training the GPR on the residual error between the detailed, non-linear FO physical model prediction (Jw_physical) and the experimental water flux (Jw_actual). Crucially, we implement a full UQ methodology by decomposing the total predictive variance (sigma2_total) into model uncertainty (epistemic, from GPR's posterior variance) and input uncertainty (aleatoric, analytically propagated via the Delta method for multi-variate correlated inputs). Leveraging the inherent strength of GPR in low-data regimes, the model, trained on a meagre 120 data points, achieved a state-of-the-art Mean Absolute Percentage Error (MAPE) of 0.26% and an R2 of 0.999 on the independent test data, validating a truly robust and reliable surrogate model for advanced FO process optimization and digital twin development.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:27:44 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ratn",
"Shiv",
""
],
[
"Rampriyan",
"Shivang",
""
],
[
"Ray",
"Bahni",
""
]
] |
2512.10458
|
Zhixiu Han
|
Zhixiu Han, Fanxu Meng, Weidong Li, Xutao Yu, Zaichen Zhang
|
Shot and Architecture Adaptive Subspace Variational Quantum Eigensolver for Microwave Simulation
| null | null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Quantum computing offers a promising paradigm for electromagnetic eigenmode analysis, enabling compact representations of complex field interactions and potential exponential speedup over classical numerical solvers. Recent efforts have applied variational quantum eigensolver (VQE) based methods to compute waveguide modes, demonstrating the feasibility of simulating TE and TM field distributions on noisy intermediate-scale quantum (NISQ) hardware. However, these studies typically employ manually designed, fixed-depth parameterized quantum circuits and uniform measurement-shot strategies, resulting in excessive quantum resource consumption, limited circuit expressivity, and reduced robustness under realistic noise conditions. To address these limitations, we propose an architecture and shot adaptive subspace variational quantum eigensolver for efficient microwave waveguide eigenmode simulation on NISQ devices. The proposed framework integrates a reinforcement learning (RL) based circuit design strategy and an adaptive shot allocation mechanism to jointly reduce quantum resource overhead. Specifically, the RL agent autonomously explores the quantum circuit space to generate hardware-efficient parameterized quantum circuits, while the adaptive measurement scheme allocates sampling resources according to Hamiltonian term weights. Numerical experiments on three- and five-qubit systems demonstrate that the proposed framework achieves accurate estimation of TE and TM mode eigenvalues, with a minimum absolute error down to $10^{-8}$ and reconstructed field distributions under noiseless conditions in excellent agreement with classical electromagnetic solutions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:34:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Han",
"Zhixiu",
""
],
[
"Meng",
"Fanxu",
""
],
[
"Li",
"Weidong",
""
],
[
"Yu",
"Xutao",
""
],
[
"Zhang",
"Zaichen",
""
]
] |
2512.10459
|
Zhi-Wei Liu
|
Zhi-Wei Liu, Jia-Ming Xie, Jun-Xu Lu, Li-Sheng Geng
|
Probing the di-$J/\Psi$ interaction and the nature of $X(6200)$ with femtoscopic correlation functions
|
7 pages, 2 figures; Comments and suggestions are welcome
| null | null | null |
hep-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent re-analyses of the di-$J/\Psi$ invariant mass spectra reveal a state near the di-$J/\Psi$ threshold, referred to as the $X(6200)$. Yet the nature of this near-threshold pole--whether it is a resonant, bound, or virtual state--remains unresolved due to our limited understanding of the di-$J/\Psi$ interaction. To address this question, we predict the di-$J/\Psi$ and $J/\Psi\Psi(2S)$ femtoscopic correlation functions based on the Koonin-Pratt formula with a Gaussian source and the coupled-channel dynamics. Our results show that the di-$J/\Psi$ correlation function exhibits distinctly different behaviors in each scenario, especially for small source sizes ($R\sim1$ fm), providing a clear experimental observable to distinguish the nature of $X(6200)$. These distinguishing features persist even when quantum statistical effects and coupled-channel dynamics are included and show negligible sensitivity to off-shell ambiguities. Given the high $J/\Psi$ production rates and clean detection channels at the LHC, we hope that these discoveries will stimulate further experimental studies and help clarify the nature of double-vector-charmonium interactions and the nonperturbative dynamics of fully-heavy tetraquark systems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:34:30 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Zhi-Wei",
""
],
[
"Xie",
"Jia-Ming",
""
],
[
"Lu",
"Jun-Xu",
""
],
[
"Geng",
"Li-Sheng",
""
]
] |
2512.10460
|
Nils Berglund
|
Baptiste Bergeot, Nils Berglund and Israa Zogheib
|
The dynamic saddle-node bifurcation with noise on the slow variable
|
37 pages, 8 figures
| null | null | null |
math.PR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this work, we analyse the effect of adding Gaussian white noise to the slow variable of a slow--fast system passing through a saddle--node (or fold) bifurcation. This problem is mainly motivated by applications to non-equilibrium energy sinks. While the effect of adding noise to the fast variable, which is important for noise-induced tipping, has been previously analysed in detail, the case where the slow variable is perturbed by noise has not been considered before. Our main result is that the noise increases the slow variable on average. We compute the effect of the noise, to lowest order, on the expectation and variance of the slow variable after the bifurcation. The contribution of the noise can be explicitly expressed in terms of Airy functions. We also provide numerical simulations, which show that the expansion to lowest order matches the observations for fairly large values of the noise intensity.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:34:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bergeot",
"Baptiste",
""
],
[
"Berglund",
"Nils",
""
],
[
"Zogheib",
"Israa",
""
]
] |
2512.10461
|
Haoyu Zhu
|
Haoyu Zhu, Yao Zhang, Jiashen Ren, Qingchun Hou
|
T-SKM-Net: Trainable Neural Network Framework for Linear Constraint Satisfaction via Sampling Kaczmarz-Motzkin Method
| null | null | null | null |
cs.LG cs.AI math.OC
|
http://creativecommons.org/licenses/by/4.0/
|
Neural network constraint satisfaction is crucial for safety-critical applications such as power system optimization, robotic path planning, and autonomous driving. However, existing constraint satisfaction methods face efficiency-applicability trade-offs, with hard constraint methods suffering from either high computational complexity or restrictive assumptions on constraint structures. The Sampling Kaczmarz-Motzkin (SKM) method is a randomized iterative algorithm for solving large-scale linear inequality systems with favorable convergence properties, but its argmax operations introduce non-differentiability, posing challenges for neural network applications. This work proposes the Trainable Sampling Kaczmarz-Motzkin Network (T-SKM-Net) framework and, for the first time, systematically integrates SKM-type methods into neural network constraint satisfaction. The framework transforms mixed constraint problems into pure inequality problems through null space transformation, employs SKM for iterative solving, and maps solutions back to the original constraint space, efficiently handling both equality and inequality constraints. We provide theoretical proof of post-processing effectiveness in expectation and end-to-end trainability guarantees based on unbiased gradient estimators, demonstrating that despite non-differentiable operations, the framework supports standard backpropagation. On the DCOPF case118 benchmark, our method achieves 4.27ms/item GPU serial forward inference with 0.0025% max optimality gap with post-processing mode and 5.25ms/item with 0.0008% max optimality gap with joint training mode, delivering over 25$\times$ speedup compared to the pandapower solver while maintaining zero constraint violations under given tolerance.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:35:13 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhu",
"Haoyu",
""
],
[
"Zhang",
"Yao",
""
],
[
"Ren",
"Jiashen",
""
],
[
"Hou",
"Qingchun",
""
]
] |
2512.10462
|
Zejie Yu
|
Haoran Li, Fei Huang, Jingyan Guo, He Gao, Hanwen Li, Zhile Wu, Xinmin Yao, Zhengyuan Bao, Huan Li, Yaocheng Shi, Zejie Yu, Daoxin Dai
|
Pockels effect induced strong Kerr nonlinearity in a lithium niobate waveguide
|
18 pages, 6 figures
| null | null | null |
physics.optics
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The utilization of Kerr nonlinearity in lithium niobate has been extensively investigated over the years. Nevertheless, the practical implementation of Kerr nonlinearity in waveguides has been constrained by the material's inherently low third-order nonlinear coefficients. Here, we present a significant advancement by demonstrating Pockels effect-induced strong Kerr nonlinearity in a periodically poled thin-film lithium niobate waveguide. Both effective four-wave mixing (FWM) and cascaded effective FWM processes are experimentally observed. The induced FWM process achieves a remarkable maximum output power of -8.5 dBm, spanning a wavelength spectrum of over 116.8 nm. Analysis reveals that the induced effective Kerr nonlinearity exhibits a substantial effective nonlinear refractive index as $2.9\times 10^{-15} m^{2}W^{-1}$, corresponding to an effective nonlinear refractive index enhancement factor of $1.6\times 10^{4}$ relative to the intrinsic value. Moreover, a wavelength-converting experiment demonstrates a flat optic-to-optic response over a broadband radiofrequency spectrum, confirming that signal integrity is well preserved after on-chip effective FWM conversion. Therefore, the demonstrated efficient and broadband Pockels effect induced effective Kerr nonlinearity paves the way for novel applications in diverse fields, including spectroscopy, parametric amplification, quantum correlation studies, and wavelength conversion technologies.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:35:45 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Haoran",
""
],
[
"Huang",
"Fei",
""
],
[
"Guo",
"Jingyan",
""
],
[
"Gao",
"He",
""
],
[
"Li",
"Hanwen",
""
],
[
"Wu",
"Zhile",
""
],
[
"Yao",
"Xinmin",
""
],
[
"Bao",
"Zhengyuan",
""
],
[
"Li",
"Huan",
""
],
[
"Shi",
"Yaocheng",
""
],
[
"Yu",
"Zejie",
""
],
[
"Dai",
"Daoxin",
""
]
] |
2512.10463
|
Vishnu Rajagopal
|
Suman Kumar Panja and Vishnu Rajagopal
|
Thermal Casimir effect in $\kappa$-Minkowski space-time
| null | null | null | null |
hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
We study the finite temperature Casimir effect for parallel plates in the $\kappa$-Minkowski space-time. Using the Matsubara formalism and imposing the Dirichlet boundary conditions on a massless $\kappa$-scalar field, we compute the $\kappa$-deformed corrections to thermal Casimir free energy, pressure, entropy, and internal energy. Our results demonstrate that space-time non-commutativity enhances the attractive nature of the thermal Casimir force while preserving thermodynamic consistency; the system satisfies the Nernst theorem and laws of thermodynamics remain intact in $\kappa$-deformed space-time. Our analysis yields an upper bound on the deformation parameter as $a\leq10^{-18}m$. Furthermore, our results indicate that non-commutative effects become experimentally observable in Casimir effect studies when the ratio of the non-commutative scale to plate separation satisfies $a/L\leq 10^{-12}$. We also obtain the expression for Stefan-Boltzmann's law in $\kappa$-Minkowski space-time.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:36:24 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Panja",
"Suman Kumar",
""
],
[
"Rajagopal",
"Vishnu",
""
]
] |
2512.10464
|
Satyandra Mohan Sharma
|
Satyandra M. Sharma, Varun Sheel, Martin P\"atzold
|
Identification and Characterization of the Topside Bulge of the Venusian Ionosphere
|
ICARUS
| null | null | null |
physics.space-ph astro-ph.EP
|
http://creativecommons.org/licenses/by/4.0/
|
Venus, in the absence of an intrinsic magnetic field, undergoes a direct interaction between its ionosphere and the solar wind. Previous missions, including Mariner, Venera, and the Pioneer Venus Orbiter (PVO), reported a recurring localized increase in electron density, often termed a "bulge," at altitudes between 160 and 200 km. This study investigates this topside bulge using over 200 dayside electron density profiles derived from the Venus Radio Science experiment (VeRa) onboard the Venus Express (VEX). We employ an automated, gradient-based classification algorithm to provide a quantitative and reproducible method for identifying and categorizing the bulge morphology into three types. Type 1 profiles exhibit a distinct secondary peak above the main V2 layer. Type 2 profiles display a shoulder-like feature near the bulge altitude. Type 3 bulges are not visually apparent but can be identified through residuals obtained after subtracting a Chapman layer fit to the V2 peak. The bulge is detected in over 80\% of the analyzed profiles, with a higher occurrence during periods of low solar activity and at lower solar zenith angles (SZA). Type 1 morphologies are only observed at low latitudes (within $\pm 40^\circ$). The peak altitude of the bulge negatively correlates with SZA, suggesting that thermospheric cooling toward the terminator significantly influences the bulge altitude. The occurrence patterns and morphological characteristics indicate that the bulge is likely influenced by external drivers, such as solar wind interaction, rather than being solely a result of local photochemical processes.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:39:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Sharma",
"Satyandra M.",
""
],
[
"Sheel",
"Varun",
""
],
[
"Pätzold",
"Martin",
""
]
] |
2512.10465
|
Congcong Zheng
|
Congcong Zheng, Kun Wang, Xutao Yu, Ping Xu, Zaichen Zhang
|
Optimal Distributed Similarity Estimation for Unitary Channels
| null | null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study distributed similarity estimation for unitary channels (DSEU), the task of estimating the similarity between unitary channels implemented on different quantum devices. We completely address DSEU by showing that, for $n$-qubit unitary channels, the query complexity of DSEU is $\Theta(\sqrt{d})$, where $d=2^n$, for both incoherent and coherent accesses. First, we propose two estimation algorithms for DSEU with these accesses utilizing the randomized measurement toolbox. The query complexities of these algorithms are both $O(\sqrt{d})$. Although incoherent access is generally weaker than coherent access, our incoherent algorithm matches this complexity by leveraging additional shared randomness between devices, highlighting the power of shared randomness in distributed quantum learning. We further establish matching lower bounds, proving that $\Theta(\sqrt{d})$ queries are both necessary and sufficient for DSEU. Finally, we compare our algorithms with independent classical shadow and show that ours have a square-root advantage. Our results provide practical and theoretically optimal tools for quantum devices benchmarking and for distributed quantum learning.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:40:31 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zheng",
"Congcong",
""
],
[
"Wang",
"Kun",
""
],
[
"Yu",
"Xutao",
""
],
[
"Xu",
"Ping",
""
],
[
"Zhang",
"Zaichen",
""
]
] |
2512.10466
|
Siarhei Finski
|
Siarhei Finski
|
Geometric quantization on big line bundles
|
57 pages + references
| null | null | null |
math.CV math.AG math.DG math.FA
|
http://creativecommons.org/licenses/by/4.0/
|
We extend several geometric quantization results to the setting of big line bundles. More precisely, we prove the asymptotic isometry property for the map that associates to a metric on a big line bundle the corresponding sup-norms on the spaces of holomorphic sections of its tensor powers. Building on this, we show that submultiplicative norms on section rings of big line bundles are asymptotically equivalent to sup-norms. As an application, we show that any bounded submultiplicative filtration on the section ring of a big line bundle naturally gives rise to a Mabuchi geodesic ray, and the speed of this ray encodes the statistical invariants of the filtration.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:42:39 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Finski",
"Siarhei",
""
]
] |
2512.10467
|
Lujia Bai
|
Bufan Li and Lujia Bai and Weichi Wu
|
Learning Time-Varying Correlation Networks with FDR Control via Time-Varying P-values
| null | null | null | null |
stat.ME econ.EM math.ST stat.TH
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a systematic framework for controlling false discovery rate in learning time-varying correlation networks from high-dimensional, non-linear, non-Gaussian and non-stationary time series with an increasing number of potential abrupt change points in means. We propose a bootstrap-assisted approach to derive dependent and time-varying P-values from a robust estimate of time-varying correlation functions, which are not sensitive to change points. Our procedure is based on a new high-dimensional Gaussian approximation result for the uniform approximation of P-values across time and different coordinates. Moreover, we establish theoretically guaranteed Benjamini--Hochberg and Benjamini--Yekutieli procedures for the dependent and time-varying P-values, which can achieve uniform false discovery rate control. The proposed methods are supported by rigorous mathematical proofs and simulation studies. We also illustrate the real-world application of our framework using both brain electroencephalogram and financial time series data.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:43:05 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Bufan",
""
],
[
"Bai",
"Lujia",
""
],
[
"Wu",
"Weichi",
""
]
] |
2512.10468
|
Marco Bertola
|
Marco Bertola
|
Algebraic approach to the inverse spectral problem for rational matrices
|
22 pages, 1 figure
| null | null | null |
math-ph math.DS math.MP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We consider the problem of reconstruction of an $n\times n$ matrix with coefficients depending rationally on $x\in \mathbb P^1$ from the data of: (a) its characteristic polynomial and (b) a line bundle of degree $g+n-1$, with $g$ the geometric genus of the spectral curve, represented by a choice of $g+n+1$ points forming a (non-positive) divisor of the given degree.
We thus provide a reconstruction formula that does not involve transcendental functions; this includes formulas for the spectral projectors and for the change of line bundle, thus integrating the isospectral flows.
The formula is a single residue formula which depends rationally on the coordinates of the points involved, the coefficients of the spectral curve, and the position of the finite poles of $L$. We also discuss the canonical bi-differential associated with the Lax matrix and its relationship with other bi-differentials that appear in Topological Recursion and integrable systems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:44:11 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bertola",
"Marco",
""
]
] |
2512.10469
|
Wen Zhou
|
Hanyi Zhang, Xueqi Xing, Jiang-Jing Wang, Chao Nie, Yuxin Du, Junying Zhang, Xueyang Shen, Wen Zhou, Matthias Wuttig, Riccardo Mazzarello, Wei Zhang
|
Atomistic understanding of two-dimensional monatomic phase-change material for non-volatile optical applications
| null | null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Elemental antimony (Sb) is a promising material for phase-change memory, neuromorphic computing and nanophotonic applications, because its compositional simplicity can prevent phase segregation upon extensive programming. Scaling down the film thickness is a necessary step to prolong the lifetime of amorphous Sb, but the optical properties of Sb are also significantly altered as the thickness is reduced to a few nanometers, adding complexity to device optimization. In this work, we aim to provide atomistic understanding of the thickness-dependent optical responses in Sb thin films. As thickness decreases, both the extinction coefficient and optical contrast reduce in the near-infrared spectrum, consistent with previous optical measurements. Such thickness dependence gives rise to a bottom thickness limit of 2 nm in photonic applications, as predicted by coarse-grained device simulations. Further bonding analysis reveals a fundamentally different behavior for amorphous and crystalline Sb upon downscaling, resulting in the reduction of optical contrast. Thin film experiments are also carried out to validate our predictions. The thickness-dependent optical trend is fully demonstrated by our ellipsometric spectroscopy experiments, and the bottom thickness limit of 2 nm is confirmed by structural characterization experiments. Finally, we show that the greatly improved amorphous-phase stability of the 2 nm Sb thin film enables robust and reversible optical switching in a silicon-based waveguide device.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:45:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Hanyi",
""
],
[
"Xing",
"Xueqi",
""
],
[
"Wang",
"Jiang-Jing",
""
],
[
"Nie",
"Chao",
""
],
[
"Du",
"Yuxin",
""
],
[
"Zhang",
"Junying",
""
],
[
"Shen",
"Xueyang",
""
],
[
"Zhou",
"Wen",
""
],
[
"Wuttig",
"Matthias",
""
],
[
"Mazzarello",
"Riccardo",
""
],
[
"Zhang",
"Wei",
""
]
] |
2512.10470
|
Kaleb Bacztub
|
Kaleb Bacztub, Braden Vester, Matteo Hodge, Liulseged Abate
|
Stealth and Evasion in Rogue AP Attacks: An Analysis of Modern Detection and Bypass Techniques
|
5 pages, 3 figures, experimental paper
| null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by/4.0/
|
Wireless networks act as the backbone of modern digital connectivity, making them a primary target for cyber adversaries. Rogue Access Point attacks, specifically the Evil Twin variant, enable attackers to clone legitimate wireless network identifiers to deceive users into connecting. Once a connection is established, the adversary can intercept traffic and harvest sensitive credentials. While modern defensive architectures often employ Network Intrusion Detection Systems (NIDS) to identify malicious activity, the effectiveness of these systems against Layer 2 wireless threats remains a subject of critical inquiry. This project aimed to design a stealth-capable Rogue AP and evaluate its detectability against Suricata, an open-source NIDS/IPS. The methodology initially focused on a hardware-based deployment using Raspberry Pi platforms but transitioned to a virtualized environment due to severe system compatibility issues. Using Wifipumpkin3, the research team successfully deployed a captive portal that harvested user credentials from connected devices. However, the Suricata NIDS failed to flag the attack, highlighting a significant blind spot in traditional intrusion detection regarding wireless management frame attacks. This paper details the construction of the attack, the evasion techniques employed, and the limitations of current NIDS solutions in detecting localized wireless threats
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:45:48 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bacztub",
"Kaleb",
""
],
[
"Vester",
"Braden",
""
],
[
"Hodge",
"Matteo",
""
],
[
"Abate",
"Liulseged",
""
]
] |
2512.10472
|
Matthias Lanzinger
|
Markus Hecher, Matthias Lanzinger
|
From Alternation to FPRAS: Toward a Complexity Classification of Approximate Counting
| null | null | null | null |
cs.CC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Counting problems are fundamental across mathematics and computer science. Among the most subtle are those whose associated decision problem is solvable in polynomial time, yet whose exact counting version appears intractable. For some such problems, however, one can still obtain efficient randomized approximation in the form of a fully polynomial randomized approximation scheme (FPRAS). Existing proofs of FPRAS existence are often highly technical and problem-specific, offering limited insight into a more systematic complexity-theoretic account of approximability. In this work, we propose a machine-based framework for establishing the existence of an FPRAS beyond previous uniform criteria. Our starting point is alternating computation: we introduce a counting model obtained by equipping alternating Turing machines with a transducer-style output mechanism, and we use it to define a corresponding counting class spanALP. We show that every problem in spanALP admits an FPRAS, yielding a reusable sufficient condition that can be applied via reductions to alternating logspace, polynomial-time computation with output. We situate spanALP in the counting complexity landscape as strictly between #L and TotP (assuming RP $\neq$ NP) and observe interesting conceptual and technical gaps in the current machinery counting complexity. Moreover, as an illustrative application, we obtain an FPRAS for counting answers to counting the answers Dyck-constrained path queries in edge-labeled graphs, i.e., counting the number of distinct labelings realized by s-t walks whose label sequence is well-formed with respect to a Dyck-like language. To our knowledge, no FPRAS was previously known for this setting. We expect the alternating-transducer characterization to provide a broadly applicable tool for establishing FPRAS existence for further counting problems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:47:54 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Hecher",
"Markus",
""
],
[
"Lanzinger",
"Matthias",
""
]
] |
2512.10473
|
Carlos N\'u\~nez Fern\'andez
|
Mejdi Aza\"iez, Yayu Guo, Carlos N\'u\~nez Fern\'andez, Samuele Rubino and Chuanju Xu
|
Second order reduced model via incremental projection for Navier Stokes
| null | null | null | null |
math.NA cs.NA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The numerical simulation of incompressible flows is challenging due to the tight coupling of velocity and pressure. Projection methods offer an effective solution by decoupling these variables, making them suitable for large-scale computations. This work focuses on reduced-order modeling using incremental projection schemes for the Stokes equations. We present both semi-discrete and fully discrete formulations, employing BDF2 in time and finite elements in space. A proper orthogonal decomposition (POD) approach is adopted to construct a reduced-order model for the Stokes problem. The method enables explicit computation of reduced velocity and pressure while preserving accuracy. We provide a detailed stability analysis and derive error estimates, showing second-order convergence in time. Numerical experiments are conducted to validate the theoretical results and demonstrate computational efficiency.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:51:49 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Azaïez",
"Mejdi",
""
],
[
"Guo",
"Yayu",
""
],
[
"Fernández",
"Carlos Núñez",
""
],
[
"Rubino",
"Samuele",
""
],
[
"Xu",
"Chuanju",
""
]
] |
2512.10474
|
Lucy Panier
|
L. Panier, G. Buldgen, M. Matteuzzi, R. Scuflaire, M. A. Dupret, A. Noels and A. Miglio
|
Detailed theoretical analysis of core Helium-burning stars: Mixed mode patterns I. Impact of the He-flash discontinuity and of induced semi-convection
|
12 pages, 15 figures. Version of the manuscript before the language editor
|
A&A Volume 704, December 2025, article number A22
|
10.1051/0004-6361/202555524
| null |
astro-ph.SR
|
http://creativecommons.org/licenses/by/4.0/
|
Space missions like CoRoT, Kepler, and TESS have made asteroseismology a powerful probe of stellar interiors. Red giants are key targets thanks to their rich mixed-mode oscillation spectra, which reveal properties of both core and envelope. However, current models of core helium-burning red giants still fail to fully reproduce observed oscillation patterns, largely due to uncertainties in mixing processes such as overshooting and semi-convection. This motivates the need for better seismic constraints to refine stellar models. We investigate how internal structural features shape asteroseismic signatures in core helium-burning stars, focusing on the links between seismic properties and internal chemical profiles. Using an updated version of the Liege stellar evolution code and its adiabatic oscillation code, we compute and analyse mixed-mode patterns for a range of stellar models. Our results show that sharp chemical gradients and central overshooting strongly influence the mixed-mode spectra. Changes in overshooting modify the extent of the semi-convective region, altering the local Brunt-Vaisala frequency and thus the observed period spacing. Variations in overshooting are compensated by shifts in semi-convective layers, keeping the total mixed-core size nearly constant across models. As a result, stellar evolution is only mildly affected, while the seismic signatures, especially the Brunt-Vaisala frequency profile, are highly sensitive to these internal adjustments.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:53:03 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Panier",
"L.",
""
],
[
"Buldgen",
"G.",
""
],
[
"Matteuzzi",
"M.",
""
],
[
"Scuflaire",
"R.",
""
],
[
"Dupret",
"M. A.",
""
],
[
"Noels",
"A.",
""
],
[
"Miglio",
"A.",
""
]
] |
2512.10475
|
Yicheng Zhong
|
Yicheng Zhong, Fu-Chun Zhang, Kun Jiang
|
Yamaji effect and quantum oscillation in Yang-Rice-Zhang model of underdoped cuprates
|
5 pages, 4 figures with appendix
| null | null | null |
cond-mat.supr-con cond-mat.str-el
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Recent experiments have revealed signatures of small Fermi pockets in the pseudogap phase of cuprate superconductors, most notably the Yamaji effect observed in $\mathrm{HgBa}_2\mathrm{CuO}_{4+\delta}$. The Yang-Rice-Zhang (YRZ) model provides a successful phenomenological description of the pseudogap state and naturally predicts such small pockets. In this work, we use a microscopic framework to calculate angle-dependent magnetoresistance and quantum oscillation within the YRZ model. Our calculations simultaneously reproduce the experimentally observed Yamaji oscillations and the Shubnikov-de Haas oscillation corresponding to a pocket area of about $p/8$, with $p$ the hole density. By further testing the effect of Green's-function zeros, we confirm that isolated zeros leave the oscillation period unchanged, whereas an extended zero segment suppresses and modifies the oscillation. Our findings demonstrate that the YRZ model captures essential features of the pseudogap regime and provides a general quantum approach that can be applied to more complex electronic structures.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:54:47 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhong",
"Yicheng",
""
],
[
"Zhang",
"Fu-Chun",
""
],
[
"Jiang",
"Kun",
""
]
] |
2512.10476
|
Ricardo Javier Pe\v{n}a Rom\'an
|
Ricardo Javier Pe\~na Rom\'an, Sandip Maity, Fabian Samad, Dinesh Pinto, Simon Josephy, Andrea Morales, Attila K\'akay, Klaus Kern, Olav Hellwig, and Aparajita Singha
|
Nanoscale magnetometry of a synthetic three-dimensional spin texture
| null | null | null | null |
cond-mat.mes-hall
|
http://creativecommons.org/licenses/by/4.0/
|
Multilayered synthetic antiferromagnets (SAFs) are artificial three-dimensional (3D) architectures engineered to create novel, complex, and stable spin textures. Non-invasive and quantitative nanoscale magnetic imaging of the two-dimensional stray field profile at the sample surface is essential for understanding the fundamental properties of the spin-structure and being able to tailor them to achieve new functionalities. However, the deterministic detection of spin textures and their quantitative characterization on the nanoscale remain challenging. Here, we use nitrogen-vacancy scanning probe microscopy (NV-SPM) under ambient conditions to perform the first quantitative vector-field magnetometry measurements in the multilayered SAF [(Co/Pt)$_5$/Co/Ru]$_3$/(Co/Pt)$_6$. We investigate nanoscale static and dynamic properties of antiferromagnetic domains with boundaries hosting ``one-dimensional'' ferromagnetic stripes with ~ 100 nm of width and periodic modulation of the magnetization. By employing NV-SPM measurements in different imaging modes and involving NV-probes with various crystallographic orientations, we demonstrated distinct fingerprints emerging from GHz-range spin noise and constant stray fields on the order of several mT. This provides quantitative insights into the structure of domains and domain walls, as well as, into magnetic noise associated with thermal spin-waves. Our work opens up new opportunities for quantitative vector-field magnetometry of modern magnetic materials with tailored 3D spin textures and stray field profiles, and potentially novel spin-wave dispersions--in a quantitative and non-invasive manner, with exceptional magnetic sensitivity and nanometer scale spatial resolution.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:55:12 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Román",
"Ricardo Javier Peña",
""
],
[
"Maity",
"Sandip",
""
],
[
"Samad",
"Fabian",
""
],
[
"Pinto",
"Dinesh",
""
],
[
"Josephy",
"Simon",
""
],
[
"Morales",
"Andrea",
""
],
[
"Kákay",
"Attila",
""
],
[
"Kern",
"Klaus",
""
],
[
"Hellwig",
"Olav",
""
],
[
"Singha",
"Aparajita",
""
]
] |
2512.10477
|
Timur Ishuov
|
Timur Ishuov, Michele Folgheraiter, Madi Nurmanov, Goncalo Gordo, Rich\'ard Farkas, J\'ozsef Dombi
|
Symphony: A Heuristic Normalized Calibrated Advantage Actor and Critic Algorithm in application for Humanoid Robots
| null | null | null | null |
cs.RO cs.NE
|
http://creativecommons.org/licenses/by/4.0/
|
In our work we not explicitly hint that it is a misconception to think that humans learn fast. Learning process takes time. Babies start learning to move in the restricted liquid area called placenta. Children often are limited by underdeveloped body. Even adults are not allowed to participate in complex competitions right away. However, with robots, when learning from scratch, we often don't have the privilege of waiting for dozen millions of steps. "Swaddling" regularization is responsible for restraining an agent in rapid but unstable development penalizing action strength in a specific way not affecting actions directly. The Symphony, Transitional-policy Deterministic Actor and Critic algorithm, is a concise combination of different ideas for possibility of training humanoid robots from scratch with Sample Efficiency, Sample Proximity and Safety of Actions in mind. It is no secret that continuous increase in Gaussian noise without appropriate smoothing is harmful for motors and gearboxes. Compared to Stochastic algorithms, we set a limited parametric noise and promote a reduced strength of actions, safely increasing entropy, since the actions are kind of immersed in weaker noise. When actions require more extreme values, actions rise above the weak noise. Training becomes empirically much safer for both the environment around and the robot's mechanisms. We use Fading Replay Buffer: using a fixed formula containing the hyperbolic tangent, we adjust the batch sampling probability: the memory contains a recent memory and a long-term memory trail. Fading Replay Buffer allows us to use Temporal Advantage when we improve the current Critic Network prediction compared to the exponential moving average. Temporal Advantage allows us to update Actor and Critic in one pass, as well as combine Actor and Critic in one Object and implement their Losses in one line.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:55:49 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ishuov",
"Timur",
""
],
[
"Folgheraiter",
"Michele",
""
],
[
"Nurmanov",
"Madi",
""
],
[
"Gordo",
"Goncalo",
""
],
[
"Farkas",
"Richárd",
""
],
[
"Dombi",
"József",
""
]
] |
2512.10478
|
Yumeng Zhang
|
Yumeng Zhang, Huayan Guo and Vincent Lau
|
A Novel Pilot Scheme for Uplink Channel Estimation for Sub-array Structured ELAA in XL-MIMO systems
| null | null | null | null |
eess.SP
|
http://creativecommons.org/licenses/by/4.0/
|
This paper proposes a novel pilot scheme for multi-user uplink channel estimation in extra-large-scale massive MIMO (XL-MIMO) systems with extremely large aperture arrays (ELAA). The large aperture of ELAA introduces spatial non-stationarity, where far-apart users have significantly distinct visibility at the antennas, thereby reducing inter-user interference. This insight motivates our novel pilot scheme to group users with distinct visibility regions to share the same frequency subcarriers for channel estimation, so that more users can be served with reduced pilot overhead. Specifically, the proposed pilot scheme employs frequency-division multiplexing for inter-group channel estimation, while intra-group users -- benefiting from strong spatial orthogonality -- are distinguished by shifted cyclic codes, similar to code-division multiplexing. Additionally, we introduce a sub-array structured ELAA, where each sub-array is a traditional MIMO array and treated as spatial stationary, while the distances between sub-arrays can be significantly larger to achieve an expanded aperture. The channel support for sub-arrays features clustered sparsity in the antenna-delay domain and is modeled by a 2-dimensional (2-D) Markov random field (MRF). Based on this, we propose a low-complexity channel estimation algorithm within a turbo Bayesian inference framework that incorporates the 2-D MRF prior model. Simulations show that the proposed scheme and algorithm allow the XL-MIMO system to support more users, and deliver superior channel estimation performance.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:55:59 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Yumeng",
""
],
[
"Guo",
"Huayan",
""
],
[
"Lau",
"Vincent",
""
]
] |
2512.10479
|
Jerome Petri
|
J\'er\^ome P\'etri
|
Multi-wavelength emission in resistive pulsar magnetospheres
|
Accepted for publication in A&A
| null | null | null |
astro-ph.HE
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In this paper, we compute a full set of neutron star magnetosphere structures from the basic vacuum regime to the dissipation-less force-free regime by implementing a resistive prescription for the plasma. A comparison to the radiation reaction limit is also discussed. We investigated the impact of these resistive magnetospheres onto the multi-wavelength emission properties based on the polar cap model for radio wavelengths, on the slot gap model for X-rays and on the striped wind model for $\gamma$-rays.} % methods heading (mandatory) {We performed time-dependent pseudo-spectral simulations of the full Maxwell equations including a resistive Ohm's law. We deduced the polar cap shape and size, the Poynting flux, the magnetic field structure and the current sheet surface, depending on the magnetic obliquity~$\chi$ and on the conductivity~$\sigma$. We found that the geometry of the magnetosphere close to the stellar surface is not impacted by the amount of resistivity. Polar cap rims remain very similar in shape and size. However the Poynting flux varies significantly as well as the magnetic field sweep-back in the vicinity of the light-cylinder. This bending of field lines reflects into the $\gamma$-ray pulse profiles, changing the $\gamma$-ray peak separation~$\Delta$ as well as the time lag~$\delta$ between the radio pulse and $\gamma$-ray peaks. X-ray pulse profiles are also drastically affected by the resistivity. A full set of multi-wavelength light-curves can be compiled for future comparison with the third $\gamma$-ray pulsar catalogue. This systematic study will help to constrain the amount of magnetic energy flowing into particle kinetic energy and shared by radiation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:57:48 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Pétri",
"Jérôme",
""
]
] |
2512.10480
|
Jiaqiang Zhang
|
Jiaqiang Zhang, Xianjia Yu, Sier Ha, Paola Torrico Moron, Sahar Salimpour, Farhad Kerama, Haizhou Zhang, Tomi Westerlund
|
Seamless Outdoor-Indoor Pedestrian Positioning System with GNSS/UWB/IMU Fusion: A Comparison of EKF, FGO, and PF
|
8 pages, 4 figures, submitted to The 17th International Conference on Ambient Systems, Networks and Technologies
| null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Accurate and continuous pedestrian positioning across outdoor-indoor environments remains challenging because GNSS, UWB, and inertial PDR are complementary yet individually fragile under signal blockage, multipath, and drift. This paper presents a unified GNSS/UWB/IMU fusion framework for seamless pedestrian localization and provides a controlled comparison of three probabilistic back-ends: an error-state extended Kalman filter, sliding-window factor graph optimization, and a particle filter. The system uses chest-mounted IMU-based PDR as the motion backbone and integrates absolute updates from GNSS outdoors and UWB indoors. To enhance transition robustness and mitigate urban GNSS degradation, we introduce a lightweight map-based feasibility constraint derived from OpenStreetMap building footprints, treating most building interiors as non-navigable while allowing motion inside a designated UWB-instrumented building. The framework is implemented in ROS 2 and runs in real time on a wearable platform, with visualization in Foxglove. We evaluate three scenarios: indoor (UWB+PDR), outdoor (GNSS+PDR), and seamless outdoor-indoor (GNSS+UWB+PDR). Results show that the ESKF provides the most consistent overall performance in our implementation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:59:03 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Jiaqiang",
""
],
[
"Yu",
"Xianjia",
""
],
[
"Ha",
"Sier",
""
],
[
"Moron",
"Paola Torrico",
""
],
[
"Salimpour",
"Sahar",
""
],
[
"Kerama",
"Farhad",
""
],
[
"Zhang",
"Haizhou",
""
],
[
"Westerlund",
"Tomi",
""
]
] |
2512.10481
|
Gaozhao Wang
|
Gaozhao Wang, Xing Liu, Zhenduo Ye, Zhengxiong Liu and Panfeng Huang
|
Contact SLAM: An Active Tactile Exploration Policy Based on Physical Reasoning Utilized in Robotic Fine Blind Manipulation Tasks
|
8 pages, 8 figures
| null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Contact-rich manipulation is difficult for robots to execute and requires accurate perception of the environment. In some scenarios, vision is occluded. The robot can then no longer obtain real-time scene state information through visual feedback. This is called ``blind manipulation". In this manuscript, a novel physically-driven contact cognition method, called ``Contact SLAM", is proposed. It estimates the state of the environment and achieves manipulation using only tactile sensing and prior knowledge of the scene. To maximize exploration efficiency, this manuscript also designs an active exploration policy. The policy gradually reduces uncertainties in the manipulation scene. The experimental results demonstrated the effectiveness and accuracy of the proposed method in several contact-rich tasks, including the difficult and delicate socket assembly task and block-pushing task.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:59:08 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wang",
"Gaozhao",
""
],
[
"Liu",
"Xing",
""
],
[
"Ye",
"Zhenduo",
""
],
[
"Liu",
"Zhengxiong",
""
],
[
"Huang",
"Panfeng",
""
]
] |
2512.10482
|
Liana David
|
Vicente Cort\'es and Liana David
|
Components of generalised complex structures on transitive Courant algebroids
|
32 pages
| null | null | null |
math.DG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Generalised almost complex structures $\mathcal J$ on transitive Courant algebroids $E$ are studied in terms of their components with respect to a splitting $E\cong TM \oplus T^*M \oplus \mathcal G$, where $M$ denotes the base of $E$ and $\mathcal G$ its bundle of quadratic Lie algebras. Necessary and sufficient integrability equations for $\mathcal J$ are established in this formalism. As an application, it is shown that the integrability of $\mathcal J$ implies that one of the components defines a Poisson structure on $M$. Then the structure (normal form) of generalised complex structures for which the Poisson structure is non-degenerate is determined. It is shown that it is fully encoded in a pair $(\omega , \rho )$ consisting of a symplectic structure $\omega$ on $M$ and a representation $\rho : \pi_1(M) \to \mathrm{Aut}(\mathfrak g, \langle \cdot ,\cdot \rangle_{\mathfrak{g}}, J_{\mathfrak{g}})$ by automorphism of a quadratic Lie algebra $(\mathfrak g, \langle \cdot ,\cdot \rangle_{\mathfrak{g}})$ commuting with an integrable (in the sense of Lie algebras) skew-symmetric complex structure $J_{\mathfrak{g}}$. Examples of such representations and obstructions for the existence of non-degenerate generalised complex structures are discussed. Finally, a construction of generalised complex structures on transitive Courant algebroids over complex manifolds for which the Poisson structure degenerates along a complex analytic hypersurface is presented.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 09:59:22 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Cortés",
"Vicente",
""
],
[
"David",
"Liana",
""
]
] |
2512.10483
|
Mladen Pavicic
|
Mladen Pavicic
|
On Simplest Kochen-Specker Sets
|
4 pages, 4 figures, submitted to Phys. Rev. Lett
| null | null | null |
quant-ph cs.IT math-ph math.IT math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
In Phys. Rev. Lett. 135, 190203 (2025) a discovery of the simplest 3D contextual set with 33 vertices, 50 bases, and 14 complete bases is claimed. In this paper, we show that it was previously generated in Quantum 7, 953 (2023) and analyze the meaning, origin, and significance of the simplest contextual sets in any dimension. In particular, we prove that there is no ground to consider the aforementioned set as fundamental since there are many 3D contextual sets with a smaller number of complete bases. We also show that automatic generation of contextual sets from basic vector components automatically yields all known minimal contextual sets of any kind in any dimension and therefore also the aforementioned set in no CPU-time. In the end, we discuss varieties of contextual sets, in particular Kochen-Specker (KS), extended KS, and non-KS sets as well as ambiguities in their definitions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:01:34 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Pavicic",
"Mladen",
""
]
] |
2512.10484
|
Anastasiia Tiutiakina
|
Anastasiia Tiutiakina
|
Chaos, Entanglement and Measurement: Field-Theoretic Perspectives on Quantum Information Dynamics
|
233 pages, 16 figures
| null | null | null |
quant-ph cond-mat.stat-mech
|
http://creativecommons.org/licenses/by/4.0/
|
This work develops tools to understand how quantum information spreads, scrambles, and is reshaped by measurements in many-body systems. First, I study scrambling and pseudorandomness in the Brownian Sachdev-Ye-Kitaev (SYK) model, quantifying pseudorandomness using unitary k-designs and frame potentials. Using Keldysh path integrals with replicas and disorder averaging, I obtain analytic control of the approach to randomness, identify collective modes that delay convergence to Haar-like behavior, and estimate design times as functions of model parameters, clarifying links between scrambling, complexity growth, and random-circuit phenomenology. Second, I construct a field theory for weakly measured SYK clusters. Starting from a system-ancilla description and a continuum monitoring limit, and using fermionic coherent states with replicas and disorder averaging, I derive a nonlinear sigma model that captures measurement back-action and the competition between interaction-induced scrambling and information extraction, predicting characteristic crossover scales and response signatures that distinguish weak monitoring from fully unitary evolution. Third, I develop a strong-disorder renormalization group for measurement-only SYK clusters, based on the SO(2n) replica algebra and Dasgupta-Ma decimation rules. The flow shows features reminiscent of infinite-randomness behavior, but an order-of-limits subtlety renders the leading recursions non-robust, so the analytic evidence for an infinite-randomness fixed point is inconclusive, even though the average second Renyi entropy displays logarithmic scaling. Together, these results provide a unified language to diagnose when many-body dynamics generate operational randomness and how measurements redirect that flow.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:04:30 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Tiutiakina",
"Anastasiia",
""
]
] |
2512.10485
|
Chaomeng Lu
|
Chaomeng Lu and Bert Lagaisse
|
From Lab to Reality: A Practical Evaluation of Deep Learning Models and LLMs for Vulnerability Detection
| null | null | null | null |
cs.CR cs.LG cs.SE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vulnerability detection methods based on deep learning (DL) have shown strong performance on benchmark datasets, yet their real-world effectiveness remains underexplored. Recent work suggests that both graph neural network (GNN)-based and transformer-based models, including large language models (LLMs), yield promising results when evaluated on curated benchmark datasets. These datasets are typically characterized by consistent data distributions and heuristic or partially noisy labels. In this study, we systematically evaluate two representative DL models-ReVeal and LineVul-across four representative datasets: Juliet, Devign, BigVul, and ICVul. Each model is trained independently on each respective dataset, and their code representations are analyzed using t-SNE to uncover vulnerability related patterns. To assess realistic applicability, we deploy these models along with four pretrained LLMs, Claude 3.5 Sonnet, GPT-o3-mini, GPT-4o, and GPT-5 on a curated dataset, VentiVul, comprising 20 recently (May 2025) fixed vulnerabilities from the Linux kernel. Our experiments reveal that current models struggle to distinguish vulnerable from non-vulnerable code in representation space and generalize poorly across datasets with differing distributions. When evaluated on VentiVul, our newly constructed time-wise out-of-distribution dataset, performance drops sharply, with most models failing to detect vulnerabilities reliably. These results expose a persistent gap between academic benchmarks and real-world deployment, emphasizing the value of our deployment-oriented evaluation framework and the need for more robust code representations and higher-quality datasets.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:04:54 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lu",
"Chaomeng",
""
],
[
"Lagaisse",
"Bert",
""
]
] |
2512.10486
|
Nicolas Cuello
|
Nicol\'as Cuello, Eleonora Bianchi, Fran\c{c}ois M\'enard, Laurent Loinard, Ricardo Hern\'andez Garnica, Aurora Dur\'an, Cecilia Ceccarelli, Mar\'ia Jos\'e Maureira, Claire J. Chandler, Claudio Codella, Nami Sakai, Linda Podio, Giovanni Sabatini, Layal Chahine, Marta de Simone, Davide Fedele, Doug Johnstone, Tomoyuki Hanawa, Izaskun Jim\'enez-Serra, Satoshi Yamamoto
|
FAUST XXX: Dust enhancement in the young binary L1551 IRS 5
|
7 pages, 5 figures; Letter accepted for publication in A&A
| null | null | null |
astro-ph.SR astro-ph.EP astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
Young binary stars with discs provide unique laboratories to study the earliest stages of planet formation in star-forming environments. The detection of substructure in discs around Class I protostars challenges current models of disc evolution, suggesting that planets may form earlier than previously expected ($<1$ Myr). In the context of the ALMA Large Program FAUST, we present observations of the circumbinary disc (CBD) around the young binary system L1551 IRS 5. The CBD exhibits two prominent over-densities in the continuum emission at the edge of the cavity, with the Northern over-density being about 20% brighter than the Southern one. By analysing the disc morphology and kinematics of L1551 IRS 5, we delineate dynamical constraints on the binary's orbital parameters. Additionally, we present 3D hydrodynamical models of the CBD to predict both the dust and the gas surface densities. Then, we compare the resulting synthetic observations with ALMA observations of the continuum emission at 1.3 mm and the C$^{18}$O line emission. Our analysis suggests that the density enhancements observed with ALMA in L1551 IRS 5 can be caused by interactions between the binary stars and the CBD, leading to dust concentration within the disc. We conclude that the observed over-density corresponds to a location where could potentially grow under favourable conditions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:05:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Cuello",
"Nicolás",
""
],
[
"Bianchi",
"Eleonora",
""
],
[
"Ménard",
"François",
""
],
[
"Loinard",
"Laurent",
""
],
[
"Garnica",
"Ricardo Hernández",
""
],
[
"Durán",
"Aurora",
""
],
[
"Ceccarelli",
"Cecilia",
""
],
[
"Maureira",
"María José",
""
],
[
"Chandler",
"Claire J.",
""
],
[
"Codella",
"Claudio",
""
],
[
"Sakai",
"Nami",
""
],
[
"Podio",
"Linda",
""
],
[
"Sabatini",
"Giovanni",
""
],
[
"Chahine",
"Layal",
""
],
[
"de Simone",
"Marta",
""
],
[
"Fedele",
"Davide",
""
],
[
"Johnstone",
"Doug",
""
],
[
"Hanawa",
"Tomoyuki",
""
],
[
"Jiménez-Serra",
"Izaskun",
""
],
[
"Yamamoto",
"Satoshi",
""
]
] |
2512.10487
|
Vyron Kampourakis
|
Vyron Kampourakis, Georgios Kavallieratos, Georgios Spathoulas, Vasileios Gkioulos, Sokratis Katsikas
|
LLM-Assisted AHP for Explainable Cyber Range Evaluation
| null | null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by/4.0/
|
Cyber Ranges (CRs) have emerged as prominent platforms for cybersecurity training and education, especially for Critical Infrastructure (CI) sectors that face rising cyber threats. One way to address these threats is through hands-on exercises that bridge IT and OT domains to improve defensive readiness. However, consistently evaluating whether a CR platform is suitable and effective remains a challenge. This paper proposes an evaluation framework for CRs, emphasizing mission-critical settings by using a multi-criteria decision-making approach. We define a set of evaluation criteria that capture technical fidelity, training and assessment capabilities, scalability, usability, and other relevant factors. To weight and aggregate these criteria, we employ the Analytic Hierarchy Process (AHP), supported by a simulated panel of multidisciplinary experts implemented through a Large Language Model (LLM). This LLM-assisted expert reasoning enables consistent and reproducible pairwise comparisons across criteria without requiring direct expert convening. The framework's output equals quantitative scores that facilitate objective comparison of CR platforms and highlight areas for improvement. Overall, this work lays the foundation for a standardized and explainable evaluation methodology to guide both providers and end-users of CRs.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:07:15 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kampourakis",
"Vyron",
""
],
[
"Kavallieratos",
"Georgios",
""
],
[
"Spathoulas",
"Georgios",
""
],
[
"Gkioulos",
"Vasileios",
""
],
[
"Katsikas",
"Sokratis",
""
]
] |
2512.10488
|
Marie Turcicova
|
Natalia Stepanova, Marie Turcicova and Xiang Zhao
|
Adaptive almost full recovery in sparse nonparametric models
|
34 pages, 1 figure
| null | null | null |
math.ST stat.TH
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
We observe an unknown function of $d$ variables $f(\boldsymbol{t})$, $\boldsymbol{t} \in[0,1]^d$, in the Gaussian white noise model of intensity $\varepsilon>0$. We assume that the function $f$ is regular and that it is a sum of $k$-variate functions, where $k$ varies from $1$ to $s$ ($1\leq s\leq d$). These functions are unknown to us and only a few of them are nonzero. In this article, we address the problem of identifying the nonzero function components of $f$ almost fully in the case when $d=d_\varepsilon\to \infty$ as $\varepsilon\to 0$ and $s$ is either fixed or $s=s_\varepsilon\to \infty$, $s=o(d)$ as $\varepsilon\to 0$. This may be viewed as a variable selection problem. We derive the conditions when almost full variable selection in the model at hand is possible and provide a selection procedure that achieves this type of selection. The procedure is adaptive to the level of sparsity described by the sparsity index $\beta\in(0,1)$. We also derive conditions that make almost full variable selection in the model of our interest impossible. In view of these conditions, the proposed selector is seen to perform asymptotically optimal. The theoretical findings are illustrated numerically.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:09:49 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Stepanova",
"Natalia",
""
],
[
"Turcicova",
"Marie",
""
],
[
"Zhao",
"Xiang",
""
]
] |
2512.10489
|
Jacob Askew
|
J. Askew, D. J. Reardon, R. M. Shannon, M. Bailes, F. Camilo, A. Corongiu, M. Kramer, M. E. Lower, A. Parthasarathy, A. Possenti, V. Venkatraman Krishnan
|
Scintillating insights into PSR~J0737$-$3039A and the interstellar plasma of the Gum Nebula from MeerKAT
|
18 pages, 12 figures, Accepted for publication in MNRAS
| null | null | null |
astro-ph.HE
|
http://creativecommons.org/publicdomain/zero/1.0/
|
The double pulsar system PSR~J0737$-$3039A/B has enabled some of the most precise tests of strong-field gravity to date. Here, we present a scintillation analysis of the system based on an 18-month observation campaign with the MeerKAT radio telescope. We characterise this interference pattern to infer properties of scattering plasma and the orbital geometry of the system. Our preferred model supports a scattering screen located at a distance of $D_s = 360^{+30}_{-40}$ pc. This moderately anisotropic screen of ionized gas (axial ratio $A_R = 2.4 \pm 0.2$) lies near the edge of the Gum Nebula, which is believed to be a supernova remnant (SNR) or an \HII\, region. We estimate the expansion velocity of the nebula to be $V_{\textrm{s}} = 35 \pm 5$ km s$^{-1}$, implying a SNR age of $t \approx 1$ Myr. We also constrain the orbital orientation and inclination sense of the double pulsar to be $\Omega = 40^{\circ} \pm 3^{\circ}$ and $i > 90^{\circ}$, respectively. Assuming standard scattering geometry, our model yields a distance estimate consistent with the parallax-derived value of $D = 770 \pm 70$ pc from very long baseline interferometry. We conclude by discussing how future models of pulsar scintillation can enhance our understanding of the IISM and the properties of pulsars embedded within or lying behind such intervening structures.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:11:30 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Askew",
"J.",
""
],
[
"Reardon",
"D. J.",
""
],
[
"Shannon",
"R. M.",
""
],
[
"Bailes",
"M.",
""
],
[
"Camilo",
"F.",
""
],
[
"Corongiu",
"A.",
""
],
[
"Kramer",
"M.",
""
],
[
"Lower",
"M. E.",
""
],
[
"Parthasarathy",
"A.",
""
],
[
"Possenti",
"A.",
""
],
[
"Krishnan",
"V. Venkatraman",
""
]
] |
2512.10490
|
Volker Kraus
|
Volker Kraus, Wolfgang Hamm, Miklos Zrinyi
|
Relaxation in Polymer Networks under Uniaxial Extension and Biaxial Compression
|
5 pages, 3 figures
| null | null | null |
cond-mat.soft
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Predicting the time and temperature dependent behavior of polymer networks under complex loading is essential for the design of advanced elastomeric materials. Many practical applications involve combinations of deformation modes, such as uniaxial extension and biaxial compression, yet a unified description of their mechanical response remains challenging. In this study, we apply a consistent theoretical framework to describe both uniaxial and biaxial deformation modes, using the same constitutive formalism based on van der Waals network theory. The time dependence of the material response in both cases is governed by a substance specific relaxation spectrum, introduced through irreversible thermodynamics as a linear coupling to the quasi static reference state of the permanent network. The temperature dependence of the relaxation times is well described by the Williams Landel Ferry (WLF) equation in the high temperature or low strain rate regime, demonstrating that the same physical mechanisms underlie time dependent behavior across different loading geometries. Experimental results are presented for cross linked poly(methyl methacrylate) (PMMA) and polyvinyl acetate (PVAc), validating the theoretical model across both materials and deformation modes.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:12:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kraus",
"Volker",
""
],
[
"Hamm",
"Wolfgang",
""
],
[
"Zrinyi",
"Miklos",
""
]
] |
2512.10491
|
Myeonggi Kwon
|
Myeonggi Kwon, Takahiro Oba
|
Transfer maps in symplectic cohomology for convex symplectic domains
|
15 pages, 2 figures
| null | null | null |
math.SG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We construct transfer maps in symplectic cohomology for convex symplectic domains under the assumption that the complement of a subdomain is exact. We manipulate the action filtration by Reeb periods introduced by McLean--Ritter for the construction.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:13:39 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kwon",
"Myeonggi",
""
],
[
"Oba",
"Takahiro",
""
]
] |
2512.10492
|
Tiantian Zhang
|
Jiaxi Wu, Tiantian Zhang, Yuxing Wang, Yongzhe Chang, Xueqian Wang
|
UACER: An Uncertainty-Aware Critic Ensemble Framework for Robust Adversarial Reinforcement Learning
| null | null | null | null |
cs.LG cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Robust adversarial reinforcement learning has emerged as an effective paradigm for training agents to handle uncertain disturbance in real environments, with critical applications in sequential decision-making domains such as autonomous driving and robotic control. Within this paradigm, agent training is typically formulated as a zero-sum Markov game between a protagonist and an adversary to enhance policy robustness. However, the trainable nature of the adversary inevitably induces non-stationarity in the learning dynamics, leading to exacerbated training instability and convergence difficulties, particularly in high-dimensional complex environments. In this paper, we propose a novel approach, Uncertainty-Aware Critic Ensemble for robust adversarial Reinforcement learning (UACER), which consists of two strategies: 1) Diversified critic ensemble: a diverse set of K critic networks is exploited in parallel to stabilize Q-value estimation rather than conventional single-critic architectures for both variance reduction and robustness enhancement. 2) Time-varying Decay Uncertainty (TDU) mechanism: advancing beyond simple linear combinations, we develop a variance-derived Q-value aggregation strategy that explicitly incorporates epistemic uncertainty to dynamically regulate the exploration-exploitation trade-off while simultaneously stabilizing the training process. Comprehensive experiments across several MuJoCo control problems validate the superior effectiveness of UACER, outperforming state-of-the-art methods in terms of overall performance, stability, and efficiency.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:14:13 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wu",
"Jiaxi",
""
],
[
"Zhang",
"Tiantian",
""
],
[
"Wang",
"Yuxing",
""
],
[
"Chang",
"Yongzhe",
""
],
[
"Wang",
"Xueqian",
""
]
] |
2512.10493
|
Binquan Zhan
|
Binquan Zhang, Li Zhang, Haoyuan Zhang, Fang Liu, Song Wang, Bo Shen, An Fu, Lin Shi
|
Decoding Human-LLM Collaboration in Coding: An Empirical Study of Multi-Turn Conversations in the Wild
| null | null | null | null |
cs.SE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language models (LLMs) are increasingly acting as dynamic conversational interfaces, supporting multi-turn interactions that mimic human-like conversation and facilitate complex tasks like coding. While datasets such as LMSYS-Chat-1M and WildChat capture real-world user-LLM conversations, few studies systematically explore the mechanisms of human-LLM collaboration in coding scenarios. What tortuous paths do users experience during the interaction process? How well do the LLMs follow instructions? Are users satisfied? In this paper, we conduct an empirical analysis on human-LLM coding collaboration using LMSYS-Chat-1M and WildChat datasets to explore the human-LLM collaboration mechanism, LLMs' instruction following ability, and human satisfaction. This study yields interesting findings: 1) Task types shape interaction patterns(linear, star and tree), with code quality optimization favoring linear patterns, design-driven tasks leaning toward tree structures, and queries preferring star patterns; 2) Bug fixing and code refactoring pose greater challenges to LLMs' instruction following, with non-compliance rates notably higher than in information querying; 3) Code quality optimization and requirements-driven development tasks show lower user satisfaction, whereas structured knowledge queries and algorithm designs yield higher levels. These insights offer recommendations for improving LLM interfaces and user satisfaction in coding collaborations, while highlighting avenues for future research on adaptive dialogue systems. We believe this work broadens understanding of human-LLM synergies and supports more effective AI-assisted development.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:14:42 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Binquan",
""
],
[
"Zhang",
"Li",
""
],
[
"Zhang",
"Haoyuan",
""
],
[
"Liu",
"Fang",
""
],
[
"Wang",
"Song",
""
],
[
"Shen",
"Bo",
""
],
[
"Fu",
"An",
""
],
[
"Shi",
"Lin",
""
]
] |
2512.10494
|
Samuel Darwisman
|
Samuel Darwisman
|
Reconstructing Transportation Cost Planning Theory: A Multi-Layered Framework Integrating Stepwise Functions, AI-Driven Dynamic Pricing, and Sustainable Autonomy
|
20 pages, 3 tables, 2 figures
| null | null | null |
econ.TH
|
http://creativecommons.org/licenses/by/4.0/
|
The theoretical landscape of transportation cost planning is shifting from deterministic linear models to dynamic, data-driven optimization. As supply chains face volatility, static 20th-century cost assumptions prove increasingly inadequate. Despite rapid technological advancements, a unified framework linking economic production theory with the operational realities of autonomous, sustainable logistics remains absent. Existing models fail to address non-linear stepwise costs and real-time stochastic variables introduced by market dynamics. This study reconstructs transportation cost planning theory by synthesizing Grand, Middle-Range, and Applied theories. It aims to integrate stepwise cost functions, AI-driven decision-making, and environmental externalities into a cohesive planning model. A systematic theoretical synthesis was conducted using 28 high-impact papers published primarily between 2018 and 2025, employing multi-layered analysis to reconstruct cost drivers. The study identifies three critical shifts: the transition from linear to stepwise fixed costs, the necessity of AI-driven dynamic pricing for revenue optimization, and the role of Autonomous Electric Vehicles (AEVs) in minimizing long-term marginal costs. A "Dynamic-Sustainable Cost Planning Theory" is proposed, arguing that cost efficiency now depends on algorithmic prediction and autonomous fleet utilization rather than simple distance minimization.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:15:00 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Darwisman",
"Samuel",
""
]
] |
2512.10495
|
Antonio Amariti
|
Antonio Amariti, Pietro Glorioso, Chiara Mascherpa and Andrea Zanetti
|
Six Easy Pieces: interplays among dualities in 4d, 3d and 2d
|
44 pages, 22 figures
| null | null | null |
hep-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper we consider 4d $\mathrm{SU}(N)$ gauge theories with $N+1$ fundamentals, five antifundamentals and a conjugate two index antisymmetric tensor. The model has been shown to be in a mixed phase in the IR, splitting in an interacting non-Abelian Coulomb phase and a free magnetic phase. Through tensor deconfinement, we show that baryonic deformations lead to a non-Abelian free magnetic phase. Along the analysis we obtain a duality with symplectic SQCD that can be further reduced to 3d and 2d. In the 3d case the analysis of the three sphere partition function allows one to obtain dualities between $\mathrm{SU}(N)$ with a two index symmetric tensor and $\mathrm{SO}(N)$ theories. On the other hand, in 2d we recover dualities already known in the literature and propose new ones between special unitary and symplectic gauge theories.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:15:49 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Amariti",
"Antonio",
""
],
[
"Glorioso",
"Pietro",
""
],
[
"Mascherpa",
"Chiara",
""
],
[
"Zanetti",
"Andrea",
""
]
] |
2512.10496
|
Shilian Zheng
|
Shilian Zheng, Xiaoxiang Wu, Luxin Zhang, Keqiang Yue, Peihan Qi, Zhijin Zhao
|
T-ADD: Enhancing DOA Estimation Robustness Against Adversarial Attacks
| null | null | null | null |
eess.SP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Deep learning has achieved remarkable success in direction-of-arrival (DOA) estimation. However, recent studies have shown that adversarial perturbations can severely compromise the performance of such models. To address this vulnerability, we propose Transformer-based Adversarial Defense for DOA estimation (T-ADD), a transformer-based defense method designed to counter adversarial attacks. To achieve a balance between robustness and estimation accuracy, we formulate the adversarial defense as a joint reconstruction task and introduce a tailored joint loss function. Experimental results demonstrate that, compared with three state-of-the-art adversarial defense methods, the proposed T-ADD significantly mitigates the adverse effects of widely used adversarial attacks, leading to notable improvements in the adversarial robustness of the DOA model.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:17:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zheng",
"Shilian",
""
],
[
"Wu",
"Xiaoxiang",
""
],
[
"Zhang",
"Luxin",
""
],
[
"Yue",
"Keqiang",
""
],
[
"Qi",
"Peihan",
""
],
[
"Zhao",
"Zhijin",
""
]
] |
2512.10497
|
Karol Sajnok
|
Karol Sajnok, Kacper D\k{e}bski, Andrzej Dragan
|
The relativistic reason for quantum probability amplitudes
|
6 pages, 2 figures, 1 appendix
| null | null | null |
quant-ph gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We show that the quantum-mechanical probability distribution involving complex probability amplitudes can be derived from three natural conditions imposed on a relativistically invariant probability function describing the motion of a particle that can take multiple paths simultaneously. The conditions are: (i) pairwise Kolmogorov additivity, (ii) time symmetry, and (iii) Bayes' rule. The resulting solution, parameterized by a single constant, is the squared modulus of a sum of complex exponentials of the relativistic action, thereby recovering the Feynman path-integral formulation of quantum mechanics.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:18:27 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Sajnok",
"Karol",
""
],
[
"Dębski",
"Kacper",
""
],
[
"Dragan",
"Andrzej",
""
]
] |
2512.10498
|
Khurram Ashfaq
|
Khurram Ashfaq and Muhammad Tariq Mahmood
|
Robust Shape from Focus via Multiscale Directional Dilated Laplacian and Recurrent Network
|
Accepted to IJCV
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Shape-from-Focus (SFF) is a passive depth estimation technique that infers scene depth by analyzing focus variations in a focal stack. Most recent deep learning-based SFF methods typically operate in two stages: first, they extract focus volumes (a per pixel representation of focus likelihood across the focal stack) using heavy feature encoders; then, they estimate depth via a simple one-step aggregation technique that often introduces artifacts and amplifies noise in the depth map. To address these issues, we propose a hybrid framework. Our method computes multi-scale focus volumes traditionally using handcrafted Directional Dilated Laplacian (DDL) kernels, which capture long-range and directional focus variations to form robust focus volumes. These focus volumes are then fed into a lightweight, multi-scale GRU-based depth extraction module that iteratively refines an initial depth estimate at a lower resolution for computational efficiency. Finally, a learned convex upsampling module within our recurrent network reconstructs high-resolution depth maps while preserving fine scene details and sharp boundaries. Extensive experiments on both synthetic and real-world datasets demonstrate that our approach outperforms state-of-the-art deep learning and traditional methods, achieving superior accuracy and generalization across diverse focal conditions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:19:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ashfaq",
"Khurram",
""
],
[
"Mahmood",
"Muhammad Tariq",
""
]
] |
2512.10499
|
Ville Salo
|
Ville Salo and Ilkka T\"orm\"a
|
A Three-Dimensional SFT with Sparse Columns
|
29 pages, 5 figures
| null | null | null |
math.DS cs.DM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We construct a nontrivial three-dimensional subshift of finite type whose projective $\Z$-subdynamics, or $\Z$-trace, is 2-sparse, meaning that there are at most two nonzero symbols in any vertical column. The subshift is deterministic in the direction of the subdynamics, so it is topologically conjugate to the set of spacetime diagrams of a partial cellular automaton. We also present a variant of the subshift that is defined by Wang cubes, and one whose alphabet is binary.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:21:06 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Salo",
"Ville",
""
],
[
"Törmä",
"Ilkka",
""
]
] |
2512.10500
|
Yusuke Yanagisawa
|
Yusuke Yanagisawa, Shin-ichi Sasa
|
Discreteness-induced spatial chaos versus fluctuation-induced spatial order in stochastic Turing pattern formation
|
8 pages, 11 figures
| null | null | null |
cond-mat.stat-mech
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We investigate Turing pattern formation in a stochastic reaction-diffusion model defined on $N$ lattice sites, where each lattice site is associated with a reaction vessel of volume $\Omega$. We focus on a regime where spatial discreteness plays a crucial role, namely when the characteristic length of patterns is comparable to the lattice spacing. In this setting, we compare two different limiting procedures and show that they lead to qualitatively different outcomes. If we first take the deterministic limit $\Omega \to \infty$ and then the long-time limit $t \to \infty$, the stationary solutions of the corresponding spatially discrete deterministic equations become spatially chaotic in the limit $N\to\infty$. In contrast, if we first take the limit $t \to \infty$ and then take an appropriate limit of $\Omega \to \infty$ and $N\to\infty$, the resulting patterns are spatially periodic.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:21:08 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yanagisawa",
"Yusuke",
""
],
[
"Sasa",
"Shin-ichi",
""
]
] |
2512.10501
|
Ming Yan
|
Lim Chien Her, Ming Yan, Yunshu Bai, Ruihao Li, Hao Zhang
|
Zero-shot 3D Map Generation with LLM Agents: A Dual-Agent Architecture for Procedural Content Generation
| null | null | null | null |
cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Procedural Content Generation (PCG) offers scalable methods for algorithmically creating complex, customizable worlds. However, controlling these pipelines requires the precise configuration of opaque technical parameters. We propose a training-free architecture that utilizes LLM agents for zero-shot PCG parameter configuration. While Large Language Models (LLMs) promise a natural language interface for PCG tools, off-the-shelf models often fail to bridge the semantic gap between abstract user instructions and strict parameter specifications. Our system pairs an Actor agent with a Critic agent, enabling an iterative workflow where the system autonomously reasons over tool parameters and refines configurations to progressively align with human design preferences. We validate this approach on the generation of various 3D maps, establishing a new benchmark for instruction-following in PCG. Experiments demonstrate that our approach outperforms single-agent baselines, producing diverse and structurally valid environments from natural language descriptions. These results demonstrate that off-the-shelf LLMs can be effectively repurposed as generalized agents for arbitrary PCG tools. By shifting the burden from model training to architectural reasoning, our method offers a scalable framework for mastering complex software without task-specific fine-tuning.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:22:02 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Her",
"Lim Chien",
""
],
[
"Yan",
"Ming",
""
],
[
"Bai",
"Yunshu",
""
],
[
"Li",
"Ruihao",
""
],
[
"Zhang",
"Hao",
""
]
] |
2512.10502
|
Faranak Goodarzi
|
Faranak Goodarzi, Somayeh Ghafouri
|
Measures of inaccuracy based on Varextropy
| null | null | null | null |
math.ST stat.TH
|
http://creativecommons.org/licenses/by/4.0/
|
Recently, varextropy has been introduced as a new dispersion index and a measure of information. In this article, we derive the generating function of extropy and present its infinite series representation. Furthermore, we propose new variability measures: the inaccuracy and weighted inaccuracy measures between two random variables based on varextropy and we investigate their properties. We also obtain lower bounds for the inaccuracy measure and compare them with each other. In addition, we introduce a discrimination measure based on varextropy and employ it both for comparing probability distributions and for assessing the goodness of fit of distributions to data and we compare this measure with the dispersion index derived from the Kullback-Leibler divergence given in Balakrishnan et al. (2022).
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:22:02 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Goodarzi",
"Faranak",
""
],
[
"Ghafouri",
"Somayeh",
""
]
] |
2512.10503
|
Zhijing Wendy Wang
|
Zhijing Wendy Wang
|
Complexity of Hofer's geometry in higher dimensional manifolds
|
28 pages, 1 figure
| null | null | null |
math.SG math.DS
|
http://creativecommons.org/licenses/by/4.0/
|
This paper establishes robust obstructions to representing Hamiltonian diffeomorphisms as $k$-th powers ($k \geq 2$) or embedding them in flows for certain higher-dimensional symplectic manifolds $(M,\omega)$, including surface bundles. We prove that in the Hamiltonian group $(\mathrm{Ham}(M,\omega), d_H)$ equipped with the Hofer metric, there exist arbitrarily large balls that are disjoint from the set of $k$-th powers. Furthermore, we demonstrate that the free group on two generators embeds into every asymptotic cone of $(\mathrm{Ham}(M,\omega), d_H)$, revealing the large-scale geometric complexity of the Hamiltonian group.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:23:20 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wang",
"Zhijing Wendy",
""
]
] |
2512.10504
|
XinFang Zhang
|
Tianyan Quantum Group
|
Tianyan: Cloud services with quantum advantage
| null | null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Tianyan Quantum Cloud Platform offers cloud services demonstrating quantum advantage capabilities with a Zuchongzhi 3.0-like superconducting quantum processor. This cloud-accessible superconducting quantum prototype, named Tianyan-287, features 105 qubits and achieves high operational fidelities, with single-qubit gates, two-qubit gates, and readout fidelity at 99.90%, 99.56%, 98.7%, respectively. For a specific benchmark task involving random circuit sampling on a 74-qubit system over 24 cycles, the platform completes one million samples in just 18.4 minutes. In contrast, state-of-the-art classical supercomputers would require approximately 16,000 years to complete the equivalent calculation. To facilitate this, the platform provides access via Cqlib, an open-source SDK designed for working with quantum systems at the level of extended quantum circuits, operators, and primitives. The cloud service aims to democratize access to high-performance quantum hardware, enabling the community to validate and explore practical quantum advantages.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:23:39 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Tianyan Quantum Group",
"",
""
]
] |
2512.10505
|
Martin Rivas
|
Martin Rivas
|
Classical Dirac particle: Mass and Spin invariance and radiation reaction
|
13 pages
| null | null | null |
physics.class-ph
|
http://creativecommons.org/licenses/by/4.0/
|
According to the atomic principle an elementary particle has no excited states and under any interaction, if it is not annihilated, its internal structure cannot be modified. The intrinsic properties are the mass $m$ and the absolute value of the spin in the center of mass frame $S=\hbar/2$. We analyze the closed system made of a single Dirac particle and an external electromagnetic field. The Poincar\'e invariance of the dynamics implies that the energy, linear momentum and angular momentum of the whole system must be conserved. The Dirac particle has two distinguished points, the center of charge ${\bf r}$ and the center of mass ${\bf q}$. When interacting, the energy expended by the field is the work done by the external Lorentz force along the center of charge trajectory. The variation of the mechanical energy of the particle is the work done by the external Lorentz force along the center of mass trajectory. If these two works are different the excess of energy must be transformed into radiation returning that energy to the field. The accelerated Dirac particle radiates. We analyze the spin dynamics of the Dirac particle under an external electromagnetic field. The requirement that the absolute value of the spin for the center of mass observer cannot be modified by the interaction implies a modification of the dynamical equation which includes a new braking term along the center of mass velocity, that can be interpreted as the radiation reaction force.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:25:57 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Rivas",
"Martin",
""
]
] |
2512.10506
|
Tomohiko Mizutani
|
Tomohiko Mizutani
|
Hyperspectral Image Data Reduction for Endmember Extraction
| null | null | null | null |
eess.IV cs.LG eess.SP
|
http://creativecommons.org/licenses/by/4.0/
|
Endmember extraction from hyperspectral images aims to identify the spectral signatures of materials present in a scene. Recent studies have shown that self-dictionary methods can achieve high extraction accuracy; however, their high computational cost limits their applicability to large-scale hyperspectral images. Although several approaches have been proposed to mitigate this issue, it remains a major challenge. Motivated by this situation, this paper pursues a data reduction approach. Assuming that the hyperspectral image follows the linear mixing model with the pure-pixel assumption, we develop a data reduction technique that removes pixels that do not contain endmembers. We analyze the theoretical properties of this reduction step and show that it preserves pixels that lie close to the endmembers. Building on this result, we propose a data-reduced self-dictionary method that integrates the data reduction with a self-dictionary method based on a linear programming formulation. Numerical experiments demonstrate that the proposed method can substantially reduce the computational time of the original self-dictionary method without sacrificing endmember extraction accuracy.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:27:40 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mizutani",
"Tomohiko",
""
]
] |
2512.10507
|
Max Zimmer
|
Dominik Kuzinowicz, Pawe{\l} Lichocki, Gioni Mexi, Marc E. Pfetsch, Sebastian Pokutta, Max Zimmer
|
Objective Coefficient Rounding and Almost Symmetries in Binary Programs
|
8 pages, 2 pages references, 3 tables
| null | null | null |
math.OC
|
http://creativecommons.org/licenses/by/4.0/
|
This article investigates the interplay of rounding objective coefficients in binary programs and almost symmetries. Empirically, reducing the number of significant bits through rounding often leads to instances that are easier to solve. One reason can be that the amount of symmetries increases, which enables solvers to be more effective when they are exploited. This can signify that the original instance contains 'almost symmetries'. Furthermore, solving the rounded problems provides approximations to the original objective values. We empirically investigate these relations on instances of the capacitated facility location problem, the knapsack problem and a diverse collection of additional instances, using the solvers SCIP and CP-SAT. For all investigated problem classes, we show empirically that this yields faster algorithms with guaranteed solution quality. The influence of symmetry depends on the instance type and solver.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:28:09 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kuzinowicz",
"Dominik",
""
],
[
"Lichocki",
"Paweł",
""
],
[
"Mexi",
"Gioni",
""
],
[
"Pfetsch",
"Marc E.",
""
],
[
"Pokutta",
"Sebastian",
""
],
[
"Zimmer",
"Max",
""
]
] |
2512.10508
|
Vindya Vashishth
|
Vindya Vashishth and Bidya Binay Karak
|
Does the Babcock-Leighton dynamo operate in rapidly rotating solar-type stars? Exploration using a 3D dynamo model at different rotation rates
|
9 pages, 10 figures, Accepted in Monthly Notices of the Royal Astronomical Society
| null | null | null |
astro-ph.SR astro-ph.IM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The Babcock-Leighton dynamo, which relies on the generation of a poloidal field through the decay and dispersal of tilted bipolar magnetic regions (BMRs), is a promising paradigm for explaining the features of the solar magnetic cycle. In rapidly rotating stars, BMRs are expected to emerge at high latitudes, which are less efficient in generating the poloidal field due to poor cross-equatorial cancellation. The operation of the Babcock-Leighton dynamo in rapidly rotating stars is therefore questionable. We, for the first time, using a 3D kinematic dynamo model, STABLE, explore this question. By taking large-scale flows from mean-field hydrodynamics models for stars rotating at different speeds, We conduct a series of dynamo simulations in rapidly rotating stars, exploring the following four cases of spot deposition, each based on a different assumption about toroidal flux tube rise: (i) radial rise, (ii) parallel rise to the rotation axis, (iii) parallel rise combined with an increase in Joy's law slope with the stellar rotation rate, and (iv) increasing time delay and spot size. We find cyclic magnetic fields in all cases except case IV of the 1-day rotating star, for which the magnetic field is irregular. For the parallel-rise cases, the magnetic field becomes quadrupolar, and for all other cases, it is dipolar. Our work demonstrates that the Babcock-Leighton dynamo may operate even in rapidly rotating stars with starspots appearing at higher latitudes.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:28:25 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Vashishth",
"Vindya",
""
],
[
"Karak",
"Bidya Binay",
""
]
] |
2512.10509
|
George Willis
|
George A. Willis
|
Flat groups of automorphisms of totally disconnected, locally compact groups
| null | null | null | null |
math.GR
|
http://creativecommons.org/licenses/by/4.0/
|
A group, $\fl{H}$, of automorphisms of a totally disconnected locally compact group, $G$, is flat if there is a compact open $U\leq G$ such that the index $[\alpha(U):U\cap \alpha(U)]$ is mininimized for every $\alpha\in\fl{H}$. The stabilizer of $U$ in $\fl{H}$ is a normal subgroup, $\fl{H}_u$; the quotient $\fl{H}/\fl{H}_u$ is a free abelian group; and the rank of $\fl{H}$ is the rank of this free abelian group. Each singly generated group $\langle\alpha\rangle$ is flat and has rank either $0$ or $1$. Higher rank groups may be seen in Lie groups over local fields and automorphism groups of buildings.
Flat groups of automorphisms exhibit many of the features of these special examples, including analogues of roots and a factoring of $U$ into analogues of root subgroups. New proofs of improved versions of these results are presented here.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:28:59 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Willis",
"George A.",
""
]
] |
2512.10510
|
Jaewoo Lee
|
Chihyeon Song, Jaewoo Lee, Jinkyoo Park
|
Adaptive Replay Buffer for Offline-to-Online Reinforcement Learning
|
15 pages, 3 figures, 7 tables
| null | null | null |
cs.LG cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Offline-to-Online Reinforcement Learning (O2O RL) faces a critical dilemma in balancing the use of a fixed offline dataset with newly collected online experiences. Standard methods, often relying on a fixed data-mixing ratio, struggle to manage the trade-off between early learning stability and asymptotic performance. To overcome this, we introduce the Adaptive Replay Buffer (ARB), a novel approach that dynamically prioritizes data sampling based on a lightweight metric we call 'on-policyness'. Unlike prior methods that rely on complex learning procedures or fixed ratios, ARB is designed to be learning-free and simple to implement, seamlessly integrating into existing O2O RL algorithms. It assesses how closely collected trajectories align with the current policy's behavior and assigns a proportional sampling weight to each transition within that trajectory. This strategy effectively leverages offline data for initial stability while progressively focusing learning on the most relevant, high-rewarding online experiences. Our extensive experiments on D4RL benchmarks demonstrate that ARB consistently mitigates early performance degradation and significantly improves the final performance of various O2O RL algorithms, highlighting the importance of an adaptive, behavior-aware replay buffer design.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:30:04 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Song",
"Chihyeon",
""
],
[
"Lee",
"Jaewoo",
""
],
[
"Park",
"Jinkyoo",
""
]
] |
2512.10511
|
Anatolii Kravets
|
O. Kozlov, V. Kalita, S. Reshetniak, A. Kravets, D. Polishchuk, V. Korenivski
|
Thermally-controlled interlayer exchange and field-induced anisotropy in synthetic antiferromagnets
|
12 pages, 4 figures
| null | null | null |
cond-mat.mes-hall
|
http://creativecommons.org/licenses/by/4.0/
|
Interlayer exchange in synthetic antiferromagnets incorporating thin paramagnetic spacers can be controlled thermally. The spacer provides an additional ferromagnetic contribution that renormalizes the otherwise temperature-independent interlayer coupling. As a result, the system shows antiferromagnetic alignment at high temperatures and ferromagnetic alignment at low temperatures. This behavior is observed in Fe(2 nm)/Cr(0.4 nm)/Fe$_{17.5}$Cr$_{82.5}$(0.9 nm)/Cr(0.4 nm)/Fe(2 nm) multilayers with the inner spacer Fe$_{17.5}$Cr$_{82.5}$ paramagnetic at and above room temperature, and is shown to be due to the spacer being significantly magnetically polarized on lowering the temperature toward its Curie point. Although the Fe layers lack intrinsic magnetocrystalline anisotropy, the magnetization reversal demonstrates a field-induced uniaxial anisotropy of antiferromagnetic character. The resulting reversal process resembles that of a metamagnet with a spin-flip transition.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:30:19 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kozlov",
"O.",
""
],
[
"Kalita",
"V.",
""
],
[
"Reshetniak",
"S.",
""
],
[
"Kravets",
"A.",
""
],
[
"Polishchuk",
"D.",
""
],
[
"Korenivski",
"V.",
""
]
] |
2512.10512
|
Chongyang Tian
|
Qing Guo, Chongyang Tian
|
Asymptotic Sphere Concentration at Infinity for NLS with L^2 Constraint
|
31 pages
| null | null | null |
math.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We consider the nonlinear Schr\"odinger equation$$-\Delta u + V(x)\,u = a\,u^p + \mu u \quad \text{in }\mathbb{R}^n,\qquad \int_{\mathbb{R}^n} u^2 = 1,$$modeling attractive Bose--Einstein condensates. For all dimensions $n\ge 2$ and all exponents $p>1$, we prove the existence of normalized solutions whose $L^2$-mass concentrates on spheres with radii diverging to infinity. In particular, the concentration set escapes to infinity rather than remaining on a fixed compact hypersurface, which makes our regime qualitatively different both from classical point-concentration phenomena and from concentrating profiles in unconstrained problems. Our approach combines a tailored finite-dimensional reduction with a blow-up analysis based on Pohozaev identities and, in this way, extends the two-dimensional mass-critical result for $(n,p)=(2,3)$ obtained in Guo--Tian--Zhou (Calc.\ Var.\ Partial Differential Equations, 2022). The proof in that paper relies in an essential way on the two-dimensional structure and does not directly apply in higher dimensions, whereas here we develop a different approximation scheme and functional setting adapted to the high-dimensional sphere-at-infinity concentration regime.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 10:33:57 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Guo",
"Qing",
""
],
[
"Tian",
"Chongyang",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.