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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2512.10306
|
Takuya Katayama
|
Takuya Katayama, Erika Kuno
|
The bicorn curves on closed surfaces
|
14 pages with 1 fugure
| null | null | null |
math.GT math.GR math.MG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper focuses on using the theory of bicorn curves in the context of closed surfaces to understand hyperbolic phenomena of the curve graphs of those surfaces. We prove that the curve graph of any closed surface is 15-hyperbolic with one exception. Furthermore, we provide significantly tighter bounds for the bounded geodesic image theorem, originally proven by Masur--Minsky.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 05:51:42 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Katayama",
"Takuya",
""
],
[
"Kuno",
"Erika",
""
]
] |
2512.10307
|
Duane Loh
|
Ying Chen Lim and Rakesh Das and Tetsuya Hiraiwa and N. Duane Loh
|
Motifs in self-organising cells
|
26 pages, 24 figures
| null | null | null |
physics.bio-ph
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In complex systems, groups of interacting objects may form prevalent and persistent spatiotemporal patterns, which we refer to as motifs. These motifs can exhibit features that reveal how individual objects interact with one another. Simultaneously, the motifs can also interact, causing new coarse-grained properties to emerge in the system.
In this paper, we found motifs in a simulated system of Dynamically Self-Organising cells. We also found that quantifying these motifs with a set of physically interpretable structural and dynamic features efficiently captures the interaction dynamics of the motifs' underlying cells. Using these motif features, we revealed packing strain and defects in large compact aggregates, semi-periodicity in motif ensembles, and phase space classes with unsupervised machine learning. Additionally, we trained neural networks to infer the critical hidden microscopic interaction parameters within each motif from coarse-grained motif features extracted from snapshots of the system. Furthermore, we uncovered emergent features that can predict the movement of cell collectives by hierarchically coarse-graining smaller motifs into larger ones (e.g. motif clusters). We speculate that this concept of motif hierarchies may be applied broadly to many-body interacting systems that are otherwise too complex to understand.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 05:52:25 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lim",
"Ying Chen",
""
],
[
"Das",
"Rakesh",
""
],
[
"Hiraiwa",
"Tetsuya",
""
],
[
"Loh",
"N. Duane",
""
]
] |
2512.10308
|
Vasiliki Stoumpou
|
Vasiliki Stoumpou, Maciej Tysarowski, Talhat Azemi, Jawad Haider, Howard L. Haronian, Robert C. Hagberg, and Dimitris Bertsimas
|
An Interpretable AI Tool for SAVR vs TAVR in Low to Intermediate Risk Patients with Severe Aortic Stenosis
| null | null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Background. Treatment selection for low to intermediate risk patients with severe aortic stenosis between surgical (SAVR) and transcatheter (TAVR) aortic valve replacement remains variable in clinical practice, driven by patient heterogeneity and institutional preferences. While existing models predict postprocedural risk, there is a lack of interpretable, individualized treatment recommendations that directly optimize long-term outcomes.
Methods. We introduce an interpretable prescriptive framework that integrates prognostic matching, counterfactual outcome modeling, and an Optimal Policy Tree (OPT) to recommend the treatment minimizing expected 5-year mortality. Using data from Hartford Hospital and St. Vincent's Hospital, we emulate randomization via prognostic matching and sample weighting and estimate counterfactual mortality under both SAVR and TAVR. The policy model, trained on these counterfactual predictions, partitions patients into clinically coherent subgroups and prescribes the treatment associated with lower estimated risk.
Findings. If the OPT prescriptions are applied, counterfactual evaluation showed an estimated reduction in 5-year mortality of 20.3\% in Hartford and 13.8\% in St. Vincent's relative to real-life prescriptions, showing promising generalizability to unseen data from a different institution. The learned decision boundaries aligned with real-world outcomes and clinical observations.
Interpretation. Our interpretable prescriptive framework is, to the best of our knowledge, the first to provide transparent, data-driven recommendations for TAVR versus SAVR that improve estimated long-term outcomes both in an internal and external cohort, while remaining clinically grounded and contributing toward a more systematic and evidence-based approach to precision medicine in structural heart disease.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 05:54:22 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Stoumpou",
"Vasiliki",
""
],
[
"Tysarowski",
"Maciej",
""
],
[
"Azemi",
"Talhat",
""
],
[
"Haider",
"Jawad",
""
],
[
"Haronian",
"Howard L.",
""
],
[
"Hagberg",
"Robert C.",
""
],
[
"Bertsimas",
"Dimitris",
""
]
] |
2512.10309
|
Jiayu Weng
|
Jiayu Weng, Xinyi Zhu, Jing Liu, Linyuan L\"u, Pan Zhang, Ying Tang
|
Tracking large chemical reaction networks and rare events by neural networks
| null | null | null | null |
q-bio.MN cs.LG physics.bio-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Chemical reaction networks are widely used to model stochastic dynamics in chemical kinetics, systems biology and epidemiology. Solving the chemical master equation that governs these systems poses a significant challenge due to the large state space exponentially growing with system sizes. The development of autoregressive neural networks offers a flexible framework for this problem; however, its efficiency is limited especially for high-dimensional systems and in scenarios with rare events. Here, we push the frontier of neural-network approach by exploiting faster optimizations such as natural gradient descent and time-dependent variational principle, achieving a 5- to 22-fold speedup, and by leveraging enhanced-sampling strategies to capture rare events. We demonstrate reduced computational cost and higher accuracy over the previous neural-network method in challenging reaction networks, including the mitogen-activated protein kinase (MAPK) cascade network, the hitherto largest biological network handled by the previous approaches of solving the chemical master equation. We further apply the approach to spatially extended reaction-diffusion systems, the Schl\"ogl model with rare events, on two-dimensional lattices, beyond the recent tensor-network approach that handles one-dimensional lattices. The present approach thus enables efficient modeling of chemical reaction networks in general.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 05:55:44 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Weng",
"Jiayu",
""
],
[
"Zhu",
"Xinyi",
""
],
[
"Liu",
"Jing",
""
],
[
"Lü",
"Linyuan",
""
],
[
"Zhang",
"Pan",
""
],
[
"Tang",
"Ying",
""
]
] |
2512.10310
|
Duo Zheng
|
Duo Zheng, Shijia Huang, Yanyang Li, Liwei Wang
|
Efficient-VLN: A Training-Efficient Vision-Language Navigation Model
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Multimodal large language models (MLLMs) have shown promising potential in Vision-Language Navigation (VLN). However, their practical development is severely hindered by the substantial training overhead. We recognize two key issues that contribute to the overhead: (1) the quadratic computational burden from processing long-horizon historical observations as massive sequences of tokens, and (2) the exploration-efficiency trade-off in DAgger, i.e., a data aggregation process of collecting agent-explored trajectories. While more exploration yields effective error-recovery trajectories for handling test-time distribution shifts, it comes at the cost of longer trajectory lengths for both training and inference. To address these challenges, we propose Efficient-VLN, a training-efficient VLN model. Specifically, to mitigate the token processing burden, we design two efficient memory mechanisms: a progressive memory that dynamically allocates more tokens to recent observations, and a learnable recursive memory that utilizes the key-value cache of learnable tokens as the memory state. Moreover, we introduce a dynamic mixed policy to balance the exploration-efficiency trade-off. Extensive experiments show that Efficient-VLN achieves state-of-the-art performance on R2R-CE (64.2% SR) and RxR-CE (67.0% SR). Critically, our model consumes merely 282 H800 GPU hours, demonstrating a dramatic reduction in training overhead compared to state-of-the-art methods.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 05:57:48 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zheng",
"Duo",
""
],
[
"Huang",
"Shijia",
""
],
[
"Li",
"Yanyang",
""
],
[
"Wang",
"Liwei",
""
]
] |
2512.10311
|
Huijie Qiao
|
Huijie Qiao
|
Large deviation principles for fully coupled multiscale multivalued stochastic systems
|
29 pages
| null | null | null |
math.PR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This study focuses on large deviation principles for fully coupled multiscale multivalued stochastic systems, in which the slow component is governed by a multivalued stochastic differential equation and the fast component is described by a general stochastic differential equation. First, we establish the large deviation principle for the slow component at any fixed time by leveraging viscosity solutions of second-order Hamilton-Jacobi-Bellman equations involving multivalued operators. Subsequently, we illustrate the theoretical results through a concrete example.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:01:23 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Qiao",
"Huijie",
""
]
] |
2512.10312
|
Julian Rodr\'iguez
|
Julian Rodriguez, Piotr Lopez, Emiliano Lerma, Rafael Medrano and Jacobo Hernandez
|
High-Dimensional Data Processing: Benchmarking Machine Learning and Deep Learning Architectures in Local and Distributed Environments
|
8 pages, 3 figures
| null | null | null |
cs.DC cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
This document reports the sequence of practices and methodologies implemented during the Big Data course. It details the workflow beginning with the processing of the Epsilon dataset through group and individual strategies, followed by text analysis and classification with RestMex and movie feature analysis with IMDb. Finally, it describes the technical implementation of a distributed computing cluster with Apache Spark on Linux using Scala.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:02:13 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Rodriguez",
"Julian",
""
],
[
"Lopez",
"Piotr",
""
],
[
"Lerma",
"Emiliano",
""
],
[
"Medrano",
"Rafael",
""
],
[
"Hernandez",
"Jacobo",
""
]
] |
2512.10313
|
Kangkun Mao
|
Kangkun Mao, Fang Xu, Jinru Ding, Yidong Jiang, Yujun Yao, Yirong Chen, Junming Liu, Xiaoqin Wu, Qian Wu, Xiaoyan Huang, Jie Xu
|
EpiPlanAgent: Agentic Automated Epidemic Response Planning
| null | null | null | null |
cs.AI cs.CY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Epidemic response planning is essential yet traditionally reliant on labor-intensive manual methods. This study aimed to design and evaluate EpiPlanAgent, an agent-based system using large language models (LLMs) to automate the generation and validation of digital emergency response plans. The multi-agent framework integrated task decomposition, knowledge grounding, and simulation modules. Public health professionals tested the system using real-world outbreak scenarios in a controlled evaluation. Results demonstrated that EpiPlanAgent significantly improved the completeness and guideline alignment of plans while drastically reducing development time compared to manual workflows. Expert evaluation confirmed high consistency between AI-generated and human-authored content. User feedback indicated strong perceived utility. In conclusion, EpiPlanAgent provides an effective, scalable solution for intelligent epidemic response planning, demonstrating the potential of agentic AI to transform public health preparedness.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:03:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mao",
"Kangkun",
""
],
[
"Xu",
"Fang",
""
],
[
"Ding",
"Jinru",
""
],
[
"Jiang",
"Yidong",
""
],
[
"Yao",
"Yujun",
""
],
[
"Chen",
"Yirong",
""
],
[
"Liu",
"Junming",
""
],
[
"Wu",
"Xiaoqin",
""
],
[
"Wu",
"Qian",
""
],
[
"Huang",
"Xiaoyan",
""
],
[
"Xu",
"Jie",
""
]
] |
2512.10314
|
Mai Anh Vu
|
Anh M. Vu (equal contribution), Khang P. Le (equal contribution), Trang T. K. Vo (equal contribution), Ha Thach, Huy Hung Nguyen, David Yang, Han H. Huynh, Quynh Nguyen, Tuan M. Pham, Tuan-Anh Le, Minh H. N. Le, Thanh-Huy Nguyen, Akash Awasthi, Chandra Mohan, Zhu Han, Hien Van Nguyen
|
DualProtoSeg: Simple and Efficient Design with Text- and Image-Guided Prototype Learning for Weakly Supervised Histopathology Image Segmentation
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Weakly supervised semantic segmentation (WSSS) in histopathology seeks to reduce annotation cost by learning from image-level labels, yet it remains limited by inter-class homogeneity, intra-class heterogeneity, and the region-shrinkage effect of CAM-based supervision. We propose a simple and effective prototype-driven framework that leverages vision-language alignment to improve region discovery under weak supervision. Our method integrates CoOp-style learnable prompt tuning to generate text-based prototypes and combines them with learnable image prototypes, forming a dual-modal prototype bank that captures both semantic and appearance cues. To address oversmoothing in ViT representations, we incorporate a multi-scale pyramid module that enhances spatial precision and improves localization quality. Experiments on the BCSS-WSSS benchmark show that our approach surpasses existing state-of-the-art methods, and detailed analyses demonstrate the benefits of text description diversity, context length, and the complementary behavior of text and image prototypes. These results highlight the effectiveness of jointly leveraging textual semantics and visual prototype learning for WSSS in digital pathology.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:03:28 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Vu",
"Anh M.",
"",
"equal contribution"
],
[
"Le",
"Khang P.",
"",
"equal contribution"
],
[
"Vo",
"Trang T. K.",
"",
"equal contribution"
],
[
"Thach",
"Ha",
""
],
[
"Nguyen",
"Huy Hung",
""
],
[
"Yang",
"David",
""
],
[
"Huynh",
"Han H.",
""
],
[
"Nguyen",
"Quynh",
""
],
[
"Pham",
"Tuan M.",
""
],
[
"Le",
"Tuan-Anh",
""
],
[
"Le",
"Minh H. N.",
""
],
[
"Nguyen",
"Thanh-Huy",
""
],
[
"Awasthi",
"Akash",
""
],
[
"Mohan",
"Chandra",
""
],
[
"Han",
"Zhu",
""
],
[
"Van Nguyen",
"Hien",
""
]
] |
2512.10315
|
Shaoliang Guo
|
Shaoliang Guo and Ziping Yang
|
Infusing Experimental Reality into Complex Many-Body Hamiltonians: The Observable-Constrained Variational Framework (OCVF)
| null | null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Deep learning potentials for complex many-body systems often face challenges of insufficient accuracy and a lack of physical realism. This paper proposes an "Observable-Constrained Variational Framework" (OCVF), a general top-down correction paradigm designed to infuse physical realism into theoretical "skeleton" models (H_o) by imposing constraints from macroscopic experimental observables (\mathfrak{O}_{\text{exp},s}). We theoretically derive OCVF as a numerically tractable extension of the "Constrained-Ensemble Variational Method" (CEVM), wherein a neural network (\Delta H_\theta) learns the correction functional required to match the experimental data. We apply OCVF to BaTiO3 (BTO) to validate the framework: a neural network potential trained on DFT data serves as H_o, and experimental PDF data at various temperatures are used as constraints (\mathfrak{O}{\text{exp},s}). The final model, H_o + \Delta H_\theta, successfully predicts the complete phase transition sequence accurately (s', s \neq s'). Compared to the prior model, the accuracy of the Cubic-Tetragonal (C-T) phase transition temperature is improved by 95.8\% , and the Orthorhombic-Rhombohedral (O-R) T_c accuracy is improved by 36.1\%. Furthermore, the lattice structure accuracy in the Rhombohedral (R) phase is improved by 55.6\%, validating the efficacy of the OCVF framework in calibrating theoretical models via observational constraints.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:07:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Guo",
"Shaoliang",
""
],
[
"Yang",
"Ziping",
""
]
] |
2512.10316
|
Mai Anh Vu
|
Khang Le (equal contribution), Ha Thach (equal contribution), Anh M. Vu (equal contribution), Trang T. K. Vo, Han H. Huynh, David Yang, Minh H. N. Le, Thanh-Huy Nguyen, Akash Awasthi, Chandra Mohan, Zhu Han, Hien Van Nguyen
|
ConStruct: Structural Distillation of Foundation Models for Prototype-Based Weakly Supervised Histopathology Segmentation
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Weakly supervised semantic segmentation (WSSS) in histopathology relies heavily on classification backbones, yet these models often localize only the most discriminative regions and struggle to capture the full spatial extent of tissue structures. Vision-language models such as CONCH offer rich semantic alignment and morphology-aware representations, while modern segmentation backbones like SegFormer preserve fine-grained spatial cues. However, combining these complementary strengths remains challenging, especially under weak supervision and without dense annotations. We propose a prototype learning framework for WSSS in histopathological images that integrates morphology-aware representations from CONCH, multi-scale structural cues from SegFormer, and text-guided semantic alignment to produce prototypes that are simultaneously semantically discriminative and spatially coherent. To effectively leverage these heterogeneous sources, we introduce text-guided prototype initialization that incorporates pathology descriptions to generate more complete and semantically accurate pseudo-masks. A structural distillation mechanism transfers spatial knowledge from SegFormer to preserve fine-grained morphological patterns and local tissue boundaries during prototype learning. Our approach produces high-quality pseudo masks without pixel-level annotations, improves localization completeness, and enhances semantic consistency across tissue types. Experiments on BCSS-WSSS datasets demonstrate that our prototype learning framework outperforms existing WSSS methods while remaining computationally efficient through frozen foundation model backbones and lightweight trainable adapters.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:08:29 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Le",
"Khang",
"",
"equal contribution"
],
[
"Thach",
"Ha",
"",
"equal contribution"
],
[
"Vu",
"Anh M.",
"",
"equal contribution"
],
[
"Vo",
"Trang T. K.",
""
],
[
"Huynh",
"Han H.",
""
],
[
"Yang",
"David",
""
],
[
"Le",
"Minh H. N.",
""
],
[
"Nguyen",
"Thanh-Huy",
""
],
[
"Awasthi",
"Akash",
""
],
[
"Mohan",
"Chandra",
""
],
[
"Han",
"Zhu",
""
],
[
"Van Nguyen",
"Hien",
""
]
] |
2512.10317
|
Ziyu Wang
|
Ziyu Wang, Bowen Yang, Shihao Zhou, Chenyi Li, Yuan Zhang, Bin Dong, Zaiwen Wen
|
Translating Informal Proofs into Formal Proofs Using a Chain of States
|
31 pages, 5 figures
| null | null | null |
cs.LO cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We address the problem of translating informal mathematical proofs expressed in natural language into formal proofs in Lean4 under a constrained computational budget. Our approach is grounded in two key insights. First, informal proofs tend to proceed via a sequence of logical transitions - often implications or equivalences - without explicitly specifying intermediate results or auxiliary lemmas. In contrast, formal systems like Lean require an explicit representation of each proof state and the tactics that connect them. Second, each informal reasoning step can be viewed as an abstract transformation between proof states, but identifying the corresponding formal tactics often requires nontrivial domain knowledge and precise control over proof context. To bridge this gap, we propose a two stage framework. Rather than generating formal tactics directly, we first extract a Chain of States (CoS), a sequence of intermediate formal proof states aligned with the logical structure of the informal argument. We then generate tactics to transition between adjacent states in the CoS, thereby constructing the full formal proof. This intermediate representation significantly reduces the complexity of tactic generation and improves alignment with informal reasoning patterns. We build dedicated datasets and benchmarks for training and evaluation, and introduce an interactive framework to support tactic generation from formal states. Empirical results show that our method substantially outperforms existing baselines, achieving higher proof success rates.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:08:34 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wang",
"Ziyu",
""
],
[
"Yang",
"Bowen",
""
],
[
"Zhou",
"Shihao",
""
],
[
"Li",
"Chenyi",
""
],
[
"Zhang",
"Yuan",
""
],
[
"Dong",
"Bin",
""
],
[
"Wen",
"Zaiwen",
""
]
] |
2512.10318
|
Aniruddh Mishra
|
Aniruddh Mishra, Benjamin Oommen, Jimmy Liang
|
L2 Ethernet Switch VLSI Implementation
|
9 pages, 21 figures
| null | null | null |
cs.NI cs.SY eess.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Ethernet switches are foundational to the global internet infrastructure. These devices route packets of data on a local area network between source addresses to destination media access control addresses. On the L2 layer of the Open Systems Interconnections model, Ethernet switches take in digitized data from a Media Independent Interface and send it to the corresponding output port for the destination address. Switches need to handle parallel input and output streams from each port, prioritizing throughput, efficiency, and packet integrity. Due to the confidential nature of the networking device industry, there do not exist many open source implementations of switching fabrics. We propose an open source design for an L2 Ethernet switch along with the power, performance, and area tradeoffs for architecture decisions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:08:55 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mishra",
"Aniruddh",
""
],
[
"Oommen",
"Benjamin",
""
],
[
"Liang",
"Jimmy",
""
]
] |
2512.10319
|
Muhammad Usama
|
Muhammad Usama, Muhammad Ibrahim Khan, Ahmad Hasan, Muhammad Shaaf Nadeem, Khawaja Fahad Iqbal, Jawad Aslam, Mian Ashfaq Ali, Asad Nisar Awan
|
Design of a six wheel suspension and a three-axis linear actuation mechanism for a laser weeding robot
|
15 Pages, 10 figures
| null | null | null |
cs.RO cs.CV cs.SY eess.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Mobile robots are increasingly utilized in agriculture to automate labor-intensive tasks such as weeding, sowing, harvesting and soil analysis. Recently, agricultural robots have been developed to detect and remove weeds using mechanical tools or precise herbicide sprays. Mechanical weeding is inefficient over large fields, and herbicides harm the soil ecosystem. Laser weeding with mobile robots has emerged as a sustainable alternative in precision farming. In this paper, we present an autonomous weeding robot that uses controlled exposure to a low energy laser beam for weed removal. The proposed robot is six-wheeled with a novel double four-bar suspension for higher stability. The laser is guided towards the detected weeds by a three-dimensional linear actuation mechanism. Field tests have demonstrated the robot's capability to navigate agricultural terrains effectively by overcoming obstacles up to 15 cm in height. At an optimal speed of 42.5 cm/s, the robot achieves a weed detection rate of 86.2\% and operating time of 87 seconds per meter. The laser actuation mechanism maintains a minimal mean positional error of 1.54 mm, combined with a high hit rate of 97\%, ensuring effective and accurate weed removal. This combination of speed, accuracy, and efficiency highlights the robot's potential for significantly enhancing precision farming practices.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:11:05 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Usama",
"Muhammad",
""
],
[
"Khan",
"Muhammad Ibrahim",
""
],
[
"Hasan",
"Ahmad",
""
],
[
"Nadeem",
"Muhammad Shaaf",
""
],
[
"Iqbal",
"Khawaja Fahad",
""
],
[
"Aslam",
"Jawad",
""
],
[
"Ali",
"Mian Ashfaq",
""
],
[
"Awan",
"Asad Nisar",
""
]
] |
2512.10320
|
Jin Tang
|
Yongsen Zhang, Wei Liu, Meng Shi, Yaodong Wu, Jialiang Jiang, Sheng Qiu, Huanhuan Zhang, Hui Han, Mingliang Tian, Haifeng Du, Shouguo Wang, Jin Tang
|
Fresnel Magnetic Imaging of Ultrasmall Skyrmion Lattices
|
Published in Advanced Science
| null |
10.1002/advs.202509786
| null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Magnetic skyrmions with ultrasmall nanometric dimensions hold significant promise for next-generation high-density spintronic devices. Direct real-space imaging of these topological spin textures is critical for elucidating their emergent properties at the nanoscale. Here, we present Lorentz transmission electron microscopy studies of nanometric skyrmion lattices in B20-structured Mn0.5Fe0.5Ge crystals using Fresnel mode. According to conventional chiral discrimination methods relying on static bright-dark contrast, we demonstrate an abnormal periodic chiral-reversal phenomenon retrieved through the transport of intensity equation analysis of defocus-dependent Fresnel images. Through systematic off-axis electron holography experiments and numerical simulations, we attribute these chiral misinterpretations to the sinusoidal modulation mechanism of the contrast transfer functionthat correlates with both defocus values and skyrmion dimensions. Our findings establish quantitative limitations of conventional Fresnel contrast analysis for ultrasmall skyrmions while revealing fundamental insights into defocus-mediated phase-to-intensity conversion processes in nanoscale magnetic imaging.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:11:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Yongsen",
""
],
[
"Liu",
"Wei",
""
],
[
"Shi",
"Meng",
""
],
[
"Wu",
"Yaodong",
""
],
[
"Jiang",
"Jialiang",
""
],
[
"Qiu",
"Sheng",
""
],
[
"Zhang",
"Huanhuan",
""
],
[
"Han",
"Hui",
""
],
[
"Tian",
"Mingliang",
""
],
[
"Du",
"Haifeng",
""
],
[
"Wang",
"Shouguo",
""
],
[
"Tang",
"Jin",
""
]
] |
2512.10321
|
Hyunsoo Lee
|
Hyunsoo Lee, Daeum Jeon, Hyeokjae Oh
|
Point2Pose: A Generative Framework for 3D Human Pose Estimation with Multi-View Point Cloud Dataset
|
WACV 2026 camera ready
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose a novel generative approach for 3D human pose estimation. 3D human pose estimation poses several key challenges due to the complex geometry of the human body, self-occluding joints, and the requirement for large-scale real-world motion datasets. To address these challenges, we introduce Point2Pose, a framework that effectively models the distribution of human poses conditioned on sequential point cloud and pose history. Specifically, we employ a spatio-temporal point cloud encoder and a pose feature encoder to extract joint-wise features, followed by an attention-based generative regressor. Additionally, we present a large-scale indoor dataset MVPose3D, which contains multiple modalities, including IMU data of non-trivial human motions, dense multi-view point clouds, and RGB images. Experimental results show that the proposed method outperforms the baseline models, demonstrating its superior performance across various datasets.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:11:24 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lee",
"Hyunsoo",
""
],
[
"Jeon",
"Daeum",
""
],
[
"Oh",
"Hyeokjae",
""
]
] |
2512.10322
|
Yongqiang Yu
|
Yongqiang Yu, Xuhui Li, Hazza Mahmood, Jinxing Zhou, Haodong Hong, Longtao Jiang, Zhiqiang Xu, Qi Wu, and Xiaojun Chang
|
User-Feedback-Driven Continual Adaptation for Vision-and-Language Navigation
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Vision-and-Language Navigation (VLN) requires agents to navigate complex environments by following natural-language instructions. General Scene Adaptation for VLN (GSA-VLN) shifts the focus from zero-shot generalization to continual, environment-specific adaptation, narrowing the gap between static benchmarks and real-world deployment. However, current GSA-VLN frameworks exclude user feedback, relying solely on unsupervised adaptation from repeated environmental exposure. In practice, user feedback offers natural and valuable supervision that can significantly enhance adaptation quality. We introduce a user-feedback-driven adaptation framework that extends GSA-VLN by systematically integrating human interactions into continual learning. Our approach converts user feedback-navigation instructions and corrective signals-into high-quality, environment-aligned training data, enabling efficient and realistic adaptation. A memory-bank warm-start mechanism further reuses previously acquired environmental knowledge, mitigating cold-start degradation and ensuring stable redeployment. Experiments on the GSA-R2R benchmark show that our method consistently surpasses strong baselines such as GR-DUET, improving navigation success and path efficiency. The memory-bank warm start stabilizes early navigation and reduces performance drops after updates. Results under both continual and hybrid adaptation settings confirm the robustness and generality of our framework, demonstrating sustained improvement across diverse deployment conditions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:11:45 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yu",
"Yongqiang",
""
],
[
"Li",
"Xuhui",
""
],
[
"Mahmood",
"Hazza",
""
],
[
"Zhou",
"Jinxing",
""
],
[
"Hong",
"Haodong",
""
],
[
"Jiang",
"Longtao",
""
],
[
"Xu",
"Zhiqiang",
""
],
[
"Wu",
"Qi",
""
],
[
"Chang",
"Xiaojun",
""
]
] |
2512.10323
|
Longju Yu
|
Yixun Zhang, Longju Yu, Yizhou Tong, Ying Sun, Xu Li, Hong Jian Zhao, and Yanming Ma
|
Spin Orientation Driven Polarization in Collinear Magnets
|
6 pages and 2 figures
| null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In a collinear magnet, the predominant magnetic moments are collectively aligned along a specific spatial orientation, and this alignment may yield intriguing phenomena such as spin orientation driven polarization. It is well known that spin orientation driven polarization is a relativistic effect that widely occurs in various type-II multiferroics. However, a universal theory that describes such a phenomenon and directs the corresponding materials discovery is lacking. Here, we revisit the magnetic structures of collinear magnets and explore the spin-orientation-dependent phenomena therein. Based on symmetry principles, we analyze the spin point groups (SPGs) that are associated with collinear magnets in the non-relativistic regime, demonstrate how relativistic spin-orbit interaction reduces each SPG to various magnetic point groups that are associated with different magnetic alignments, and classify the SPGs with respect to spin orientation driven polarization. We employ our theory to elucidate the mechanisms of spin orientation driven polarization in a variety of type-II multiferroics. Combined with first-principles simulations, we further show that polarization may be driven in nonpolar collinear antiferromagnets (e.g., CuFeS$_2$) by reorienting their magnetic alignments. Our theory provides guidelines for designing and discovering materials with spin orientation driven polarization, which will benefit the development of spintronics based on type-II multiferroics and related materials.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:14:46 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Yixun",
""
],
[
"Yu",
"Longju",
""
],
[
"Tong",
"Yizhou",
""
],
[
"Sun",
"Ying",
""
],
[
"Li",
"Xu",
""
],
[
"Zhao",
"Hong Jian",
""
],
[
"Ma",
"Yanming",
""
]
] |
2512.10324
|
Chao Gong
|
Chao Gong, Depeng Wang, Zhipeng Wei, Ya Guo, Huijia Zhu, Jingjing Chen
|
EchoingPixels: Cross-Modal Adaptive Token Reduction for Efficient Audio-Visual LLMs
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Audio-Visual Large Language Models (AV-LLMs) face prohibitive computational overhead from massive audio and video tokens. Token reduction, while extensively explored for video-only LLMs, is insufficient for the audio-visual domain, as these unimodal methods cannot leverage audio-visual cross-modal synergies. Furthermore, the distinct and dynamic information densities of audio and video render static budgets per modality suboptimal. How to perform token reduction on a joint audio-visual stream thus remains an unaddressed bottleneck. To fill this gap, we introduce EchoingPixels, a framework inspired by the coexistence and interaction of visuals and sound in real-world scenes. The core of our framework is the Cross-Modal Semantic Sieve (CS2), a module enabling early audio-visual interaction. Instead of compressing modalities independently, CS2 co-attends to the joint multimodal stream and reduces tokens from an entire combined pool of audio-visual tokens rather than using fixed budgets per modality. This single-pool approach allows it to adaptively allocate the token budget across both modalities and dynamically identify salient tokens in concert. To ensure this aggressive reduction preserves the vital temporal modeling capability, we co-design a Synchronization-Augmented RoPE (Sync-RoPE) to maintain critical temporal relationships for the sparsely selected tokens. Extensive experiments demonstrate that EchoingPixels achieves performance comparable to strong baselines using only 5-20% of the original tokens, with a 2-3x speedup and memory reduction.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:18:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Gong",
"Chao",
""
],
[
"Wang",
"Depeng",
""
],
[
"Wei",
"Zhipeng",
""
],
[
"Guo",
"Ya",
""
],
[
"Zhu",
"Huijia",
""
],
[
"Chen",
"Jingjing",
""
]
] |
2512.10325
|
Francesco Alemanno
|
Francesco Alemanno
|
Residual subspace evolution strategies for nonlinear inverse problems
| null | null | null | null |
math.OC cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Nonlinear inverse problems often feature noisy, non-differentiable, or expensive residual evaluations that make Jacobian-based solvers unreliable. Popular derivative-free optimizers such as natural evolution strategies (NES) or Powell's NEWUOA still assume smoothness or expend many evaluations to maintain stability. Ensemble Kalman inversion (EKI) relies on empirical covariances that require preconditioning and scale poorly with residual dimension.
We introduce residual subspace evolution strategies (RSES), a derivative-free solver that samples Gaussian probes around the current iterate, builds a residual-only surrogate from their differences, and recombines the probes through a least-squares solve yielding an optimal update without forming Jacobians or covariances. Each iteration costs $k+1$ residual evaluations, where $k \ll n$ for $n$-dimensional problems, with $O(k^3)$ linear algebra overhead.
Benchmarks on calibration, regression, and deconvolution problems demonstrate consistent misfit reduction in both deterministic and stochastic settings. RSES matches or surpasses xNES and NEWUOA while staying competitive with EKI under matched evaluation budgets, particularly when smoothness or covariance assumptions fail.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:20:13 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Alemanno",
"Francesco",
""
]
] |
2512.10326
|
Jiawen Li
|
Jiawen Li, Jiali Hu, Xitong Ling, Yongqiang Lv, Yuxuan Chen, Yizhi Wang, Tian Guan, Yifei Liu, Yonghong He
|
StainNet: A Special Staining Self-Supervised Vision Transformer for Computational Pathology
|
15 pages, 6 figures
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Foundation models trained with self-supervised learning (SSL) on large-scale histological images have significantly accelerated the development of computational pathology. These models can serve as backbones for region-of-interest (ROI) image analysis or patch-level feature extractors in whole-slide images (WSIs) based on multiple instance learning (MIL). Existing pathology foundation models (PFMs) are typically pre-trained on Hematoxylin-Eosin (H&E) stained pathology images. However, images with special stains, such as immunohistochemistry, are also frequently used in clinical practice. PFMs pre-trained mainly on H\&E-stained images may be limited in clinical applications involving special stains. To address this issue, we propose StainNet, a specialized foundation model for special stains based on the vision transformer (ViT) architecture. StainNet adopts a self-distillation SSL approach and is trained on over 1.4 million patch images cropping from 20,231 publicly available special staining WSIs in the HISTAI database. To evaluate StainNet, we conduct experiments on an in-house slide-level liver malignancy classification task and two public ROI-level datasets to demonstrate its strong ability. We also perform few-ratio learning and retrieval evaluations, and compare StainNet with recently larger PFMs to further highlight its strengths. We have released the StainNet model weights at: https://huggingface.co/JWonderLand/StainNet.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:21:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Jiawen",
""
],
[
"Hu",
"Jiali",
""
],
[
"Ling",
"Xitong",
""
],
[
"Lv",
"Yongqiang",
""
],
[
"Chen",
"Yuxuan",
""
],
[
"Wang",
"Yizhi",
""
],
[
"Guan",
"Tian",
""
],
[
"Liu",
"Yifei",
""
],
[
"He",
"Yonghong",
""
]
] |
2512.10327
|
Jinlong Liu
|
Cai Xu, Jinlong Liu, Yilin Zhang, Ziyu Guan and Wei Zhao
|
Simple Yet Effective Selective Imputation for Incomplete Multi-view Clustering
| null | null | null | null |
cs.CV cs.MM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Incomplete multi-view data, where different views suffer from missing and unbalanced observations, pose significant challenges for clustering. Existing imputation-based methods attempt to estimate missing views to restore data associations, but indiscriminate imputation often introduces noise and bias, especially when the available information is insufficient. Imputation-free methods avoid this risk by relying solely on observed data, but struggle under severe incompleteness due to the lack of cross-view complementarity. To address this issue, we propose Informativeness-based Selective imputation Multi-View Clustering (ISMVC). Our method evaluates the imputation-relevant informativeness of each missing position based on intra-view similarity and cross-view consistency, and selectively imputes only when sufficient support is available. Furthermore, we integrate this selection with a variational autoencoder equipped with a mixture-of-Gaussians prior to learn clustering-friendly latent representations. By performing distribution-level imputation, ISMVC not only stabilizes the aggregation of posterior distributions but also explicitly models imputation uncertainty, enabling robust fusion and preventing overconfident reconstructions. Compared with existing cautious imputation strategies that depend on training dynamics or model feedback, our method is lightweight, data-driven, and model-agnostic. It can be readily integrated into existing IMC models as a plug-in module. Extensive experiments on multiple benchmark datasets under a more realistic and challenging unbalanced missing scenario demonstrate that our method outperforms both imputation-based and imputation-free approaches.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:22:23 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Xu",
"Cai",
""
],
[
"Liu",
"Jinlong",
""
],
[
"Zhang",
"Yilin",
""
],
[
"Guan",
"Ziyu",
""
],
[
"Zhao",
"Wei",
""
]
] |
2512.10328
|
Xuan Lu
|
Xuan Lu, Yulin Yu
|
Balancing Turnover and Promotion Outcomes: Evidence on the Optimal Hybrid-Work Frequency
| null | null | null | null |
cs.SI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Hybrid work policy, especially return-to-office requirements, remains a globally salient topic as workers, companies, and governments continue to debate and disagree. Despite extensive discussions on the benefits and drawbacks of remote and hybrid arrangements, the optimal number of remote days that jointly considers multiple organizational outcomes has not been empirically established. Focusing on two critical career outcomes -- turnover risk and promotion -- we examine how remote work frequency shapes employee trajectories using large-scale observational activity data from a company with over one million employees. We find that increased remote-work frequency is associated with an initial decrease and then an increase in turnover, while promotion likelihood initially rises and then declines. Accordingly, we identify approximately two remote days per week as an optimal balance -- maximizing promotion, a positive outcome for employees, while minimizing turnover, which is undesirable for organizations and may indicate negative employee experiences. These patterns vary across subgroups defined by gender, role type, and leadership status. Several notable results emerge. First, male employees derive greater promotion benefits from remote work than female employees. Second, support workers (non-core business roles) do not experience promotion gains, and the reduction in turnover at their optimal remote-work frequency is marginal compared with employees in core business roles. Third, organizational leaders face greater challenges in remote settings than individual contributors: their turnover risk increases substantially at higher remote frequencies, and their likelihood of promotion decreases as remote frequency rises. We further show that time-allocation patterns partly explain how remote-work frequency influences these career outcomes.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:22:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lu",
"Xuan",
""
],
[
"Yu",
"Yulin",
""
]
] |
2512.10329
|
Xi Guo
|
Dong An and Xi Guo
|
Improved gap dependence in adiabatic state preparation by adaptive schedule
| null | null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Adiabatic quantum computing is a powerful framework for state preparation, while its evolution time often scales quadratically in the inverse Hamiltonian spectral gap, leading to sub-optimal computational complexity. In this work, we introduce a nonlinear adaptive strategy for finding the time scheduling function, and show that the gap dependence can be quadratically improved to be inverse linear for a wide range of systems under a mild gap measure condition. Through variational analysis, we further demonstrate the optimality of our schedule for systems with linear gap and the partial optimality for general systems, while we also rigorously show that the commonly used linear schedule is never optimal.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:27:04 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"An",
"Dong",
""
],
[
"Guo",
"Xi",
""
]
] |
2512.10330
|
Elina Shishkina
|
V. N. Kolokoltsov and E. L. Shishkina
|
Matrix approach to the fractional calculus
| null | null | null | null |
math.NA cs.NA math.FA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we introduce the new construction of fractional derivatives and integrals with respect to a function, based on a matrix approach. We believe that this is a powerful tool in both analytical and numerical calculations. We begin with the differential operator with respect to a function that generates a semigroup. By discretizing this operator, we obtain a matrix approximation. Importantly, this discretization provides not only an approximating operator but also an approximating semigroup. This point motivates our approach, as we then apply Balakrishnan's representations of fractional powers of operators, which are based on semigroups. Using estimates of the semigroup norm and the norm of the difference between the operator and its matrix approximation, we derive the convergence rate for the approximation of the fractional power of operators with the fractional power of correspondings matrix operators. In addition, an explicit formula for calculating an arbitrary power of a two-band matrix is obtained, which is indispensable in the numerical solution of fractional differential and integral equations.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:27:08 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kolokoltsov",
"V. N.",
""
],
[
"Shishkina",
"E. L.",
""
]
] |
2512.10331
|
Marc\'ilio Santos
|
Marcilio Ferreira dos Santos
|
Curvature-Weighted Contact Networks: Spectral Reduction and Global Stability in a Markovian SIR Model
|
6 figures, 12 pages
| null | null | null |
math.DS math.PR q-bio.PE
|
http://creativecommons.org/licenses/by/4.0/
|
We propose a new network-based SIR epidemic model in which transmission is modulated by a curvature-weighted contact matrix that encodes structural and geometric features of the underlying graph. The formulation encompasses both adjacency-driven and Markovian mixing, allowing heterogeneous interactions to be shaped by curvature-sensitive topological properties. We prove that the basic reproduction number satisfies \[ R_0=\frac{\beta}{\gamma}\lambda_{\max}(M), \] where $M$ is the curvature-weighted transmission operator. Using Perron--Frobenius theory together with linear and nonlinear Lyapunov functionals, we establish: (i) global asymptotic stability of the disease-free equilibrium when $R_0<1$, and (ii) existence and global asymptotic stability of a unique endemic equilibrium when $R_0>1$. Our results show that curvature acts as a geometric regularizer of connectivity, lowering spectral radii, raising effective epidemic thresholds, and organizing the long-term dynamics through monotone contraction toward the endemic state. This framework generalizes classical network epidemiology by integrating geometric information directly into transmission operators, providing a rigorous foundation for epidemic dynamics on structurally heterogeneous networks.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:28:41 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Santos",
"Marcilio Ferreira dos",
""
]
] |
2512.10332
|
Jing Wang
|
Xiaodong Liu and Qingxiang Shi and Jing Wang
|
The Radon Transform-Based Sampling Methods for Biharmonic Sources from the Scattered Fields
| null | null | null | null |
math-ph cs.NA math.MP math.NA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents three quantitative sampling methods for reconstructing extended sources of the biharmonic wave equation using scattered field data. The first method employs an indicator function that solely relies on scattered fields $ u^s$ measured on a single circle, eliminating the need for Laplacian or derivative data. Its theoretical foundation lies in an explicit formula for the source function, which also serves as a constructive proof of uniqueness. To improve computational efficiency, we introduce a simplified double integral formula for the source function, at the cost of requiring additional measurements $\Delta u^s$. This advancement motivates the second indicator function, which outperforms the first method in both computational speed and reconstruction accuracy. The third indicator function is proposed to reconstruct the support boundary of extended sources from the scattered fields $ u^s$ at a finite number of sensors. By analyzing singularities induced by the source boundary, we establish the uniqueness of annulus and polygon-shaped sources. A key characteristic of the first and third indicator functions is their link between scattered fields and the Radon transform of the source function. Numerical experiments demonstrate that the proposed sampling methods achieve high-resolution imaging of the source support or the source function itself.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:28:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Xiaodong",
""
],
[
"Shi",
"Qingxiang",
""
],
[
"Wang",
"Jing",
""
]
] |
2512.10333
|
Shaima Farhan ShfEng
|
Shaima Farhana, Dong Yua, Amirhossein Karamoozianc, Ali Al-shawafid, Amar N. Alsheavif
|
AI-Enhanced TOE Framework for Sustainable Industrial Performance in Fragile and Transforming Economies: Evidence from Yemen and Saudi Arabia
| null | null | null | null |
econ.EM
|
http://creativecommons.org/licenses/by/4.0/
|
Using an integrated framework rooted in the TOE model enhanced with AI, this study looks at ways to improve industrial performance and environmental sustainability in fragile and rapidly transforming contexts such as those found in Yemen and Saudi Arabia. Data for the research are field-based and were obtained from a total of 600 SMEs operating in both countries. Based on the questionnaires' responses by 294 managers, results from the partial least squares structural equation modeling (PLS-SEM) have indicated significant positive effects of AI-TOE on environmental performance (beta = 0.487) and manufacturing performance (beta = 0.759). Results indicate that AI acts as a transformative force, though its impact differs based on the maturity of infrastructure and organizational readiness. The Saudi SMEs gain from their institutional support and advanced technologies, while those in Yemen are dependent on the low-cost adoption of AI and organizational flexibility to accept structural challenges. PLS-SEM analysis of the study showed that integrating AI into the TOE dimensions accelerates operational efficiency in order to support environmental performance. Industrial performance was found to be a very important mediator in this relationship. This study responds to the call for digital transformation literature by providing an actionable framework of AI adoption in resource-constrained environments. These findings offer insights that might guide policymakers and organizations toward more resilient and sustainable operational strategies. These findings provide valuable guidance for engineering managers within the context of negotiating digital transformation and sustainability trade-offs in fragile and resource-constrained contexts.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:35:15 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Farhana",
"Shaima",
""
],
[
"Yua",
"Dong",
""
],
[
"Karamoozianc",
"Amirhossein",
""
],
[
"Al-shawafid",
"Ali",
""
],
[
"Alsheavif",
"Amar N.",
""
]
] |
2512.10334
|
Yichi Zhang
|
Yi Liu and Yichi Zhang
|
A Conditional Generative Framework for Synthetic Data Augmentation in Segmenting Thin and Elongated Structures in Biological Images
| null | null | null | null |
cs.CV
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Thin and elongated filamentous structures, such as microtubules and actin filaments, often play important roles in biological systems. Segmenting these filaments in biological images is a fundamental step for quantitative analysis. Recent advances in deep learning have significantly improved the performance of filament segmentation. However, there is a big challenge in acquiring high quality pixel-level annotated dataset for filamentous structures, as the dense distribution and geometric properties of filaments making manual annotation extremely laborious and time-consuming. To address the data shortage problem, we propose a conditional generative framework based on the Pix2Pix architecture to generate realistic filaments in microscopy images from binary masks. We also propose a filament-aware structural loss to improve the structure similarity when generating synthetic images. Our experiments have demonstrated the effectiveness of our approach and outperformed existing model trained without synthetic data.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:36:44 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Yi",
""
],
[
"Zhang",
"Yichi",
""
]
] |
2512.10335
|
Chen-Hsuan Hsu
|
Anna Ohorodnyk and Chen-Hsuan Hsu
|
Cascade of topological phase transitions and revival of topological zero modes in imperfect double helical liquids
|
25 pages, 19 figures
| null | null | null |
cond-mat.mes-hall
|
http://creativecommons.org/licenses/by/4.0/
|
Two parallel helical edge channels hosting interacting electrons, when proximitized by local and nonlocal pairings, can host time-reversal-invariant pairs of topological zero modes at the system corners. Here we show that realistic imperfections substantially enrich the physics of such proximitized double helical liquids. Specifically, we analyze this platform and its fractional counterparts in the presence of pairing and interaction asymmetries between the two channels, as well as random spin-flip terms arising from either magnetic disorder or coexisting charge disorder and external magnetic fields. Using renormalization-group analysis, we determine how Coulomb interactions, pairings, and magnetic disorder collectively influence the transport behavior and topological properties of the double helical liquid. As the system transitions from class DIII to class BDI, an additional topological phase supporting a single Majorana zero mode per corner emerges. We further show how additional pairing or Coulomb asymmetry influences the stability of various topological phases and uncovers a revival of Majorana zero modes and cascades of transitions through topological phases characterized by a $\mathbb {Z}$ invariant, which are accessible through controlling the electrical screening effect. In contrast to conventional understanding, disorder is not merely detrimental, as it in general allows for a tuning knob that qualitatively reshapes the topological superconductivity in imperfect helical liquids.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:36:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ohorodnyk",
"Anna",
""
],
[
"Hsu",
"Chen-Hsuan",
""
]
] |
2512.10336
|
Alexis Roger
|
Jules Lahmi, Alexis Roger
|
Multilingual VLM Training: Adapting an English-Trained VLM to French
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Artificial intelligence has made great progress in recent years, particularly in the development of Vision--Language Models (VLMs) that understand both visual and textual data. However, these advancements remain largely limited to English, reducing their accessibility for non--English speakers. It is essential to extend these capabilities to a broader range of languages. This paper explores the challenges of adapting an English-trained VLM to different languages. To this end, we will explore and compare different methods for their performance and computational cost. We consider a translation-based pipeline, LoRA finetuning, and a two-stage finetuning strategy that separates vision adaptation from language adaptation. To evaluate these methods, we use a combination of standard multimodal benchmarks translated into the target language and manual assessments by native experts. The results reveal that dataset translation remains a major bottleneck in multilingual VLM performance, with data quality limiting the effectiveness of training and evaluation. These findings suggest that future efforts should focus on native-language dataset collection and improved translation strategies.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:38:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lahmi",
"Jules",
""
],
[
"Roger",
"Alexis",
""
]
] |
2512.10338
|
Jie Li
|
Zi-Xu Lu, Huai-Bing Zhu, Xuan Zuo, Jie Li
|
Optomagnonic generation of entangled travelling fields with different polarizations
| null | null | null | null |
quant-ph cond-mat.mes-hall physics.app-ph physics.optics
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The optomagnonic coupling between magnons and optical photons is an essential component for building remote quantum networks based on magnonics. Here we show that such a coupling, manifested as the magnon-induced Brillouin light scattering, can be exploited to entangle two propagating optical fields. The protocol employs two pairs of the whispering gallery modes coupled to the same magnon mode in a YIG sphere. In each pair a strong pump field is applied to activate either Stokes or anti-Stokes scattering. Due to the magnon mode involving in the two scattering processes and as a mediation, Stokes and anti-Stokes photons of different polarizations get entangled. The entanglement can be extracted by filtering the travelling output fields centered at the Stokes and anti-Stokes sidebands. Optimal conditions are identified under which strong output entanglement can be achieved.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:42:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lu",
"Zi-Xu",
""
],
[
"Zhu",
"Huai-Bing",
""
],
[
"Zuo",
"Xuan",
""
],
[
"Li",
"Jie",
""
]
] |
2512.10339
|
Ziseok Lee
|
Ziseok Lee, Minyeong Hwang, Sanghyun Jo, Wooyeol Lee, Jihyung Ko, Young Bin Park, Jae-Mun Choi, Eunho Yang, Kyungsu Kim
|
On the Collapse of Generative Paths: A Criterion and Correction for Diffusion Steering
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Inference-time steering enables pretrained diffusion/flow models to be adapted to new tasks without retraining. A widely used approach is the ratio-of-densities method, which defines a time-indexed target path by reweighting probability-density trajectories from multiple models with positive, or in some cases, negative exponents. This construction, however, harbors a critical and previously unformalized failure mode: Marginal Path Collapse, where intermediate densities become non-normalizable even though endpoints remain valid. Collapse arises systematically when composing heterogeneous models trained on different noise schedules or datasets, including a common setting in molecular design where de-novo, conformer, and pocket-conditioned models must be combined for tasks such as flexible-pose scaffold decoration. We provide a novel and complete solution for the problem. First, we derive a simple path existence criterion that predicts exactly when collapse occurs from noise schedules and exponents alone. Second, we introduce Adaptive path Correction with Exponents (ACE), which extends Feynman-Kac steering to time-varying exponents and guarantees a valid probability path. On a synthetic 2D benchmark and on flexible-pose scaffold decoration, ACE eliminates collapse and enables high-guidance compositional generation, improving distributional and docking metrics over constant-exponent baselines and even specialized task-specific scaffold decoration models. Our work turns ratio-of-densities steering with heterogeneous experts from an unstable heuristic into a reliable tool for controllable generation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:44:08 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lee",
"Ziseok",
""
],
[
"Hwang",
"Minyeong",
""
],
[
"Jo",
"Sanghyun",
""
],
[
"Lee",
"Wooyeol",
""
],
[
"Ko",
"Jihyung",
""
],
[
"Park",
"Young Bin",
""
],
[
"Choi",
"Jae-Mun",
""
],
[
"Yang",
"Eunho",
""
],
[
"Kim",
"Kyungsu",
""
]
] |
2512.10340
|
Yi-Cheng Liao
|
Yi-Cheng Liao, Shyang-En Weng, Yu-Syuan Xu, Chi-Wei Hsiao, Wei-Chen Chiu, Ching-Chun Huang
|
Zero-shot Adaptation of Stable Diffusion via Plug-in Hierarchical Degradation Representation for Real-World Super-Resolution
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Real-World Image Super-Resolution (Real-ISR) aims to recover high-quality images from low-quality inputs degraded by unknown and complex real-world factors. Real-world scenarios involve diverse and coupled degradations, making it necessary to provide diffusion models with richer and more informative guidance. However, existing methods often assume known degradation severity and rely on CLIP text encoders that cannot capture numerical severity, limiting their generalization ability. To address this, we propose \textbf{HD-CLIP} (\textbf{H}ierarchical \textbf{D}egradation CLIP), which decomposes a low-quality image into a semantic embedding and an ordinal degradation embedding that captures ordered relationships and allows interpolation across unseen levels. Furthermore, we integrated it into diffusion models via classifier-free guidance (CFG) and proposed classifier-free projection guidance (CFPG). HD-CLIP leverages semantic cues to guide generative restoration while using degradation cues to suppress undesired hallucinations and artifacts. As a \textbf{plug-and-play module}, HD-CLIP can be seamlessly integrated into various super-resolution frameworks without training, significantly improving detail fidelity and perceptual realism across diverse real-world datasets.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:45:28 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liao",
"Yi-Cheng",
""
],
[
"Weng",
"Shyang-En",
""
],
[
"Xu",
"Yu-Syuan",
""
],
[
"Hsiao",
"Chi-Wei",
""
],
[
"Chiu",
"Wei-Chen",
""
],
[
"Huang",
"Ching-Chun",
""
]
] |
2512.10342
|
Shresth Grover
|
Shresth Grover, Priyank Pathak, Akash Kumar, Vibhav Vineet, Yogesh S Rawat
|
CoSPlan: Corrective Sequential Planning via Scene Graph Incremental Updates
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Large-scale Vision-Language Models (VLMs) exhibit impressive complex reasoning capabilities but remain largely unexplored in visual sequential planning, i.e., executing multi-step actions towards a goal. Additionally, practical sequential planning often involves non-optimal (erroneous) steps, challenging VLMs to detect and correct such steps. We propose Corrective Sequential Planning Benchmark (CoSPlan) to evaluate VLMs in error-prone, vision-based sequential planning tasks across 4 domains: maze navigation, block rearrangement, image reconstruction,and object reorganization. CoSPlan assesses two key abilities: Error Detection (identifying non-optimal action) and Step Completion (correcting and completing action sequences to reach the goal). Despite using state-of-the-art reasoning techniques such as Chain-of-Thought and Scene Graphs, VLMs (e.g. Intern-VLM and Qwen2) struggle on CoSPlan, failing to leverage contextual cues to reach goals. Addressing this, we propose a novel training-free method, Scene Graph Incremental updates (SGI), which introduces intermediate reasoning steps between the initial and goal states. SGI helps VLMs reason about sequences, yielding an average performance gain of 5.2%. In addition to enhancing reliability in corrective sequential planning, SGI generalizes to traditional planning tasks such as Plan-Bench and VQA.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:46:51 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Grover",
"Shresth",
""
],
[
"Pathak",
"Priyank",
""
],
[
"Kumar",
"Akash",
""
],
[
"Vineet",
"Vibhav",
""
],
[
"Rawat",
"Yogesh S",
""
]
] |
2512.10343
|
Nitesh Dubey
|
Nitesh K. Dubey, Sanved Kolekar, and Sudipta Sarkar
|
Stationary Stars Are Axisymmetric in Higher Curvature Gravity
| null | null | null | null |
gr-qc astro-ph.HE astro-ph.SR hep-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The final equilibrium stage of stellar evolution can result in either a black hole or a compact object such as a white dwarf or neutron star. In general relativity, both stationary black holes and stationary stellar configurations are known to be axisymmetric, and black hole rigidity has been extended to several higher curvature modifications of gravity. In contrast, no comparable result had previously been established for stationary stars beyond general relativity. In this work we extend the stellar axisymmetry theorem to a broad class of diffeomorphism invariant metric theories. Assuming asymptotic flatness and standard smoothness requirements, we show that the Killing symmetry implied by thermodynamic equilibrium inside the star uniquely extends to the exterior region, thereby enforcing rotational invariance. This demonstrates that axisymmetry of stationary stellar configurations is not a feature peculiar to Einstein gravity but a universal property of generally covariant gravitational theories, persisting even in the presence of higher curvature corrections.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:53:21 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Dubey",
"Nitesh K.",
""
],
[
"Kolekar",
"Sanved",
""
],
[
"Sarkar",
"Sudipta",
""
]
] |
2512.10344
|
Liu Tonghua
|
Tonghua Liu and Kai Liao
|
Improved Identification of Strongly Lensed Gravitational Waves with Host Galaxy Locations
|
8 pages, 4 figures, comments are welcome
| null | null | null |
astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present a Bayesian framework that enhances the identification of strongly lensed gravitational waves (GWs) by incorporating informative positional priors from the Euclid galaxy lens catalog. The core of our method introduces a two-step reweighting scheme: first, gravitational wave parameter estimation is performed under a uniform sky prior; the resulting posterior is then used to reweight galaxy positions within the Euclid catalog, constructing an astrophysically informed positional prior. Comparing this Euclid-informed prior against a uniform prior within our framework reveals distinct behaviors. While the posterior estimates of the intrinsic waveform parameters show little sensitivity to the prior change, the Bayes factor for lensing identification exhibits significant prior dependence. Crucially, for truly lensed event pairs, the Bayes factor systematically increases, whereas for unlensed pairs it decreases. This dual effect is vital for robust discrimination. Our analysis demonstrates that this multi-messenger approach significantly improves the confidence of lensing searches. For lensed pairs, the method boosts the Bayes factor by an average factor of $\sim 10$, while effectively suppressing false positives for unlensed coincidences. This underscores the critical importance of prior specification and showcases the substantial gains achievable by synergizing gravitational-wave data with electromagnetic survey information.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:55:56 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Tonghua",
""
],
[
"Liao",
"Kai",
""
]
] |
2512.10346
|
Qiang Zhao
|
Ye Cao, Ming-Xiao Duan, Zhong Tao, and Qiang Zhao
|
A unified approach for the hadronic weak decays of $\Lambda$ and $\Sigma^{\pm}\to N\pi$
|
Revtex, 35 pages, 6 Figures, 19 Tables
| null | null | null |
hep-ph hep-ex nucl-th
|
http://creativecommons.org/licenses/by/4.0/
|
We provide a unified approach for the two-body hadronic weak decays of hyperons with $S=-1$, i.e. $\Lambda$ and $\Sigma^\pm$, in the framework of the non-relativistic constitute quark model (NRCQM). A combined analysis shows that the branching ratios and asymmetry parameters of the decay channels $\Lambda\to p\pi^-$ and $\Sigma^\pm\to N\pi$ can be well described in the same framework with the direct pion emission, color suppressed internal $W$ emission, and pole terms included. However, the channel $\Lambda\to n\pi^0$ indicates significant deviations from the experimental data based on these mentioned transition mechanism. We demonstrate that the final state interactions (FSIs) via the coupled-channel rescatterings play a crucial role in $\Lambda\to n\pi^0$. Namely, the dominant decay channel of $\Lambda\to p\pi^-$ can contribute to $\Lambda\to n\pi^0$ via the $p\pi^-\to n\pi^0$ rescatterings. This is a leading correction effect for $\Lambda\to n\pi^0$ at the one-loop level. We find that such FSIs only become the leading effects in $\Lambda\to n\pi^0$, but contribute as subleading contributions in other channels. We also demonstrate that the pole terms are indispensable in these hyperon decays. In particular, for $\Sigma^-\to n\pi^-$ and $\Sigma^+\to n\pi^+$, it shows that $\Lambda(1405)$ as the intermediate state in the pole term amplitude is necessary for reproducing the experimental data. We also find that a dynamic selection rule forbids the radial excitation state $N(1710)$ of the quark model multiplet $|70, ^28,2,0^+,1/2^+\rangle$ from contribution. To some extent, the hyperon hadronic weak decays serve as a special probe for the underlying transition mechanisms and can provide some constraints on the intermediate $J^P=1/2^\pm$ baryon resonances.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 06:59:03 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Cao",
"Ye",
""
],
[
"Duan",
"Ming-Xiao",
""
],
[
"Tao",
"Zhong",
""
],
[
"Zhao",
"Qiang",
""
]
] |
2512.10347
|
Jie Li
|
Hao-Tian Li, Hong-Bin Wang, Zi-Xu Lu, Jie Li
|
Generation of mechanical cat-like states via optomagnomechanics
|
Invited contribution to Quantum Review Letters
| null | null | null |
quant-ph cond-mat.mes-hall physics.optics
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose an optomagnomechanical approach for preparing a cat-like superposition state of mechanical motion. Our protocol consists of two steps and is based on the magnomechanical system where the magnetostrictively induced displacement further couples to an optical cavity mode via radiation pressure. We first prepare a squeezed mechanical state by driving the magnomechanical system with a two-tone microwave field. We then switch off the microwave drives and send a weak red-detuned optical pulse to the optical cavity to weakly activate the optomechanical anti-Stokes scattering. We show that $k$ phonons can be subtracted from the prepared squeezed state, conditioned on the detection of $k$ anti-Stokes photons from the cavity output field, which prepares the mechanical motion in a cat-like state. The work provides a new avenue for preparing mechanical superposition states by combining opto- and magnomechanics and may find applications in the study of macroscopic quantum states and the test of collapse theories.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:03:14 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Hao-Tian",
""
],
[
"Wang",
"Hong-Bin",
""
],
[
"Lu",
"Zi-Xu",
""
],
[
"Li",
"Jie",
""
]
] |
2512.10348
|
Wenhan Wu
|
Wenhan Wu, Zhili He, Huanghuang Liang, Yili Gong, Jiawei Jiang, Chuang Hu, Dazhao Cheng
|
REMISVFU: Vertical Federated Unlearning via Representation Misdirection for Intermediate Output Feature
|
The 40th Annual AAAI Conference on Artificial Intelligence
| null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Data-protection regulations such as the GDPR grant every participant in a federated system a right to be forgotten. Federated unlearning has therefore emerged as a research frontier, aiming to remove a specific party's contribution from the learned model while preserving the utility of the remaining parties. However, most unlearning techniques focus on Horizontal Federated Learning (HFL), where data are partitioned by samples. In contrast, Vertical Federated Learning (VFL) allows organizations that possess complementary feature spaces to train a joint model without sharing raw data. The resulting feature-partitioned architecture renders HFL-oriented unlearning methods ineffective. In this paper, we propose REMISVFU, a plug-and-play representation misdirection framework that enables fast, client-level unlearning in splitVFL systems. When a deletion request arrives, the forgetting party collapses its encoder output to a randomly sampled anchor on the unit sphere, severing the statistical link between its features and the global model. To maintain utility for the remaining parties, the server jointly optimizes a retention loss and a forgetting loss, aligning their gradients via orthogonal projection to eliminate destructive interference. Evaluations on public benchmarks show that REMISVFU suppresses back-door attack success to the natural class-prior level and sacrifices only about 2.5% points of clean accuracy, outperforming state-of-the-art baselines.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:05:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wu",
"Wenhan",
""
],
[
"He",
"Zhili",
""
],
[
"Liang",
"Huanghuang",
""
],
[
"Gong",
"Yili",
""
],
[
"Jiang",
"Jiawei",
""
],
[
"Hu",
"Chuang",
""
],
[
"Cheng",
"Dazhao",
""
]
] |
2512.10349
|
Quan Yuan
|
Quan Yuan, Zhenting Du, Daqian Cao and Weibang Bai
|
Design and Validation of an Under-actuated Robotic Finger with Synchronous Tendon Routing
|
7 pages and 11 figures
| null | null | null |
cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Tendon-driven under-actuated robotic fingers provide advantages for dexterous manipulation through reduced actuator requirements and simplified mechanical design. However, achieving both high load capacity and adaptive compliance in a compact form remains challenging. This paper presents an under-actuated tendon-driven robotic finger (UTRF) featuring a synchronous tendon routing that mechanically couples all joints with fixed angular velocity ratios, enabling the entire finger to be actuated by a single actuator. This approach significantly reduces the number of actuators required in multi-finger hands, resulting in a lighter and more compact structure without sacrificing stiffness or compliance. The kinematic and static models of the finger are derived, incorporating tendon elasticity to predict structural stiffness. A single-finger prototype was fabricated and tested under static loading, showing an average deflection prediction error of 1.0 mm (0.322% of total finger length) and a measured stiffness of 1.2x10^3 N/m under a 3 kg tip load. Integration into a five-finger robotic hand (UTRF-RoboHand) demonstrates effective object manipulation across diverse scenarios, confirming that the proposed routing achieves predictable stiffness and reliable grasping performance with a minimal actuator count.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:05:43 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yuan",
"Quan",
""
],
[
"Du",
"Zhenting",
""
],
[
"Cao",
"Daqian",
""
],
[
"Bai",
"Weibang",
""
]
] |
2512.10350
|
Nicolas Tacheny
|
Nicolas Tacheny
|
Dynamics of Agentic Loops in Large Language Models: A Geometric Theory of Trajectories
| null | null | null | null |
cs.LG cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Agentic systems built on large language models operate through recursive feedback loops, where each output becomes the next input. Yet the geometric behavior of these agentic loops (whether they converge, diverge, or exhibit more complex dynamics) remains poorly understood. This paper introduces a geometric framework for analyzing agentic trajectories in semantic embedding space, treating iterative transformations as discrete dynamical systems. We distinguish the artifact space, where linguistic transformations occur, from the embedding space, where geometric measurements are performed. Because cosine similarity is biased by embedding anisotropy, we introduce an isotonic calibration that eliminates systematic bias and aligns similarities with human semantic judgments while preserving high local stability. This enables rigorous measurement of trajectories, clusters and attractors. Through controlled experiments on singular agentic loops, we identify two fundamental regimes. A contractive rewriting loop converges toward a stable attractor with decreasing dispersion, while an exploratory summarize and negate loop produces unbounded divergence with no cluster formation. These regimes display qualitatively distinct geometric signatures of contraction and expansion. Our results show that prompt design directly governs the dynamical regime of an agentic loop, enabling systematic control of convergence, divergence and trajectory structure in iterative LLM transformations.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:06:14 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Tacheny",
"Nicolas",
""
]
] |
2512.10351
|
Georgi Bebrov PhD
|
Georgi Bebrov
|
On the Optimality of a Quantum Key Distribution
|
10 pages, 2 figures, 2 tables, uses IEEEtran document class
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Quantum key distribution (QKD) systems require optimal performance of both quantum and classical channels - utilizing as few as possible qubits and bits for establishing as many as possible key bits. Here we report a way to determine if a quantum key distribution model (or protocol) operates in an optimal behavior. This is accomplished by introducing a quantity, called optimality, which is the maximum over the total efficiency of a QKD under any circumstances (any values of QKD parameters). The optimality definition is given for the asymptotic operation of a QKD system - when infinitely many quantum systems are transferred/used in a quantum key distribution protocol or a quantum key distribution system is used infinitely many times. A way to attain the optimality is considered\textemdash implementation of a completely efficient QKD system (a combination of capacity-reaching quantum channel and a completely compressed classical channel) is presented. Optimal versions of BB84-QKD and twin-field QKD are introduced.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:07:27 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Bebrov",
"Georgi",
""
]
] |
2512.10352
|
Keyi Chen
|
Keyi Chen, Mingze Sun, Zhenyu Liu, Zhangquan Chen, Ruqi Huang
|
Topology-Agnostic Animal Motion Generation from Text Prompt
|
10 pages, 7 figures.Conference submission
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Motion generation is fundamental to computer animation and widely used across entertainment, robotics, and virtual environments. While recent methods achieve impressive results, most rely on fixed skeletal templates, which prevent them from generalizing to skeletons with different or perturbed topologies. We address the core limitation of current motion generation methods - the combined lack of large-scale heterogeneous animal motion data and unified generative frameworks capable of jointly modeling arbitrary skeletal topologies and textual conditions. To this end, we introduce OmniZoo, a large-scale animal motion dataset spanning 140 species and 32,979 sequences, enriched with multimodal annotations. Building on OmniZoo, we propose a generalized autoregressive motion generation framework capable of producing text-driven motions for arbitrary skeletal topologies. Central to our model is a Topology-aware Skeleton Embedding Module that encodes geometric and structural properties of any skeleton into a shared token space, enabling seamless fusion with textual semantics. Given a text prompt and a target skeleton, our method generates temporally coherent, physically plausible, and semantically aligned motions, and further enables cross-species motion style transfer.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:08:29 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chen",
"Keyi",
""
],
[
"Sun",
"Mingze",
""
],
[
"Liu",
"Zhenyu",
""
],
[
"Chen",
"Zhangquan",
""
],
[
"Huang",
"Ruqi",
""
]
] |
2512.10353
|
Yiheng Lyu
|
Yiheng Lyu, Lian Xu, Mohammed Bennamoun, Farid Boussaid, Coen Arrow, Girish Dwivedi
|
Hybrid Transformer-Mamba Architecture for Weakly Supervised Volumetric Medical Segmentation
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Weakly supervised semantic segmentation offers a label-efficient solution to train segmentation models for volumetric medical imaging. However, existing approaches often rely on 2D encoders that neglect the inherent volumetric nature of the data. We propose TranSamba, a hybrid Transformer-Mamba architecture designed to capture 3D context for weakly supervised volumetric medical segmentation. TranSamba augments a standard Vision Transformer backbone with Cross-Plane Mamba blocks, which leverage the linear complexity of state space models for efficient information exchange across neighboring slices. The information exchange enhances the pairwise self-attention within slices computed by the Transformer blocks, directly contributing to the attention maps for object localization. TranSamba achieves effective volumetric modeling with time complexity that scales linearly with the input volume depth and maintains constant memory usage for batch processing. Extensive experiments on three datasets demonstrate that TranSamba establishes new state-of-the-art performance, consistently outperforming existing methods across diverse modalities and pathologies. Our source code and trained models are openly accessible at: https://github.com/YihengLyu/TranSamba.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:09:32 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lyu",
"Yiheng",
""
],
[
"Xu",
"Lian",
""
],
[
"Bennamoun",
"Mohammed",
""
],
[
"Boussaid",
"Farid",
""
],
[
"Arrow",
"Coen",
""
],
[
"Dwivedi",
"Girish",
""
]
] |
2512.10354
|
Jihoon Jang
|
Jihoon Jang, Yehyun Nam, Kunsoo Park, Hyunjoon Kim
|
Efficient Defective Clique Enumeration and Search with Worst-Case Optimal Search Space
|
Accepted at SIGMOD 2026. This is the full version
| null |
10.1145/3769787
| null |
cs.DS cs.DB
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
A $k$-defective clique is a relaxation of the traditional clique definition, allowing up to $k$ missing edges. This relaxation is crucial in various real-world applications such as link prediction, community detection, and social network analysis. Although the problems of enumerating maximal $k$-defective cliques and searching a maximum $k$-defective clique have been extensively studied, existing algorithms suffer from limitations such as the combinatorial explosion of small partial solutions and sub-optimal search spaces. To address these limitations, we propose a novel clique-first branch-and-bound framework that first generates cliques and then adds missing edges. Furthermore, we introduce a new pivoting technique that achieves a search space size of $\mathcal{O}(3^{\frac{n}{3}} \cdot n^k)$, where $n$ is the number of vertices in the input graph. We prove that the worst-case number of maximal $k$-defective cliques is $\Omega(3^{\frac{n}{3}} \cdot n^k)$ when $k$ is a constant, establishing that our algorithm's search space is worst-case optimal. Leveraging the diameter-two property of defective cliques, we further reduce the search space size to $\mathcal{O}(n \cdot 3^{\frac{\delta}{3}} \cdot (\delta \Delta)^k)$, where $\delta$ is the degeneracy and $\Delta$ is the maximum degree of the input graph. We also propose an efficient framework for maximum $k$-defective clique search based on our branch-and-bound, together with practical techniques to reduce the search space. Experiments on real-world benchmark datasets with more than 1 million edges demonstrate that each of our proposed algorithms for maximal $k$-defective clique enumeration and maximum $k$-defective clique search outperforms the respective state-of-the-art algorithms by up to four orders of magnitude in terms of processing time.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:11:54 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Jang",
"Jihoon",
""
],
[
"Nam",
"Yehyun",
""
],
[
"Park",
"Kunsoo",
""
],
[
"Kim",
"Hyunjoon",
""
]
] |
2512.10355
|
Hyunsung Kim
|
Hyunsung Kim, Sangwoo Seo, Hoyoung Choi, Tom Boomstra, Jinsung Yoon, Chanyoung Park
|
Better Prevent than Tackle: Valuing Defense in Soccer Based on Graph Neural Networks
| null | null | null | null |
cs.LG cs.MA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Evaluating defensive performance in soccer remains challenging, as effective defending is often expressed not through visible on-ball actions such as interceptions and tackles, but through preventing dangerous opportunities before they arise. Existing approaches have largely focused on valuing on-ball actions, leaving much of defenders' true impact unmeasured. To address this gap, we propose DEFCON (DEFensive CONtribution evaluator), a comprehensive framework that quantifies player-level defensive contributions for every attacking situation in soccer. Leveraging Graph Attention Networks, DEFCON estimates the success probability and expected value of each attacking option, along with each defender's responsibility for stopping it. These components yield an Expected Possession Value (EPV) for the attacking team before and after each action, and DEFCON assigns positive or negative credits to defenders according to whether they reduced or increased the opponent's EPV. Trained on 2023-24 and evaluated on 2024-25 Eredivisie event and tracking data, DEFCON's aggregated player credits exhibit strong positive correlations with market valuations. Finally, we showcase several practical applications, including in-game timelines of defensive contributions, spatial analyses across pitch zones, and pairwise summaries of attacker-defender interactions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:12:23 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kim",
"Hyunsung",
""
],
[
"Seo",
"Sangwoo",
""
],
[
"Choi",
"Hoyoung",
""
],
[
"Boomstra",
"Tom",
""
],
[
"Yoon",
"Jinsung",
""
],
[
"Park",
"Chanyoung",
""
]
] |
2512.10356
|
Jia Zhang
|
Yizhuo Song, Jianting Dong, Jiahao Shentu and Jia Zhang
|
Spin-orbit torques in bulk collinear antiferromagnets:complete classifications and the induced spin dynamics
|
24 pages, 6 figures,
| null | null | null |
cond-mat.mtrl-sci cond-mat.other
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Electric field induced spin-orbit torques are the crucial mechanism for electric regulations of antiferromagnetic order. However, the spin-orbit torques in antiferromagnets and the induced spin dynamics remain largely unexplored. In this work, the full classifications of SOTs in bulk collinear AFMs have been achieved based on magnetic point group. Dependent on the symmetries connecting the opposite spin sublattices, the SOTs are classified into six distinct types. Among them, the SOTs and the induced Neel vector dynamics in three representative AFMs have been investigated, where the spin sublattices are connected by fractional translation, spatial inversion, and neither by translation nor inversion symmetry respectively. The SOTs on spin sublattices have been calculated by first-principles calculations based on Kubo linear response theory, and then the induced spin dynamics are simulated by LLG equations. In typical PT symmetric AFM and the inversion symmetry breaking altermagnet, the simulations indicate that the deterministic switching of Neel vectors can be driven by field like torques. What's more, the fully electric writing of multiple antiferromagnetic domains into single domain state with preset Neel vector direction and 180 deterministic switching may also be realized. Our work may shed light on the current control of antiferromagnetic orders in collinear AFMs. Especially, for inversion symmetry breaking altermagnet, the electric writing and reading of Neel vector are highly desirable for antiferromagnetic memory applications.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:13:34 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Song",
"Yizhuo",
""
],
[
"Dong",
"Jianting",
""
],
[
"Shentu",
"Jiahao",
""
],
[
"Zhang",
"Jia",
""
]
] |
2512.10357
|
Tarik Reza Toha
|
Tarik Reza Toha, Shao-Jung (Louie) Lu, and Shahriar Nirjon
|
mmCounter: Static People Counting in Dense Indoor Scenarios Using mmWave Radar
|
Accepted at the 22nd International Conference on Embedded Wireless Systems and Networks (EWSN 2025)
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
mmWave radars struggle to detect or count individuals in dense, static (non-moving) groups due to limitations in spatial resolution and reliance on movement for detection. We present mmCounter, which accurately counts static people in dense indoor spaces (up to three people per square meter). mmCounter achieves this by extracting ultra-low frequency (< 1 Hz) signals, primarily from breathing and micro-scale body movements such as slight torso shifts, and applying novel signal processing techniques to differentiate these subtle signals from background noise and nearby static objects. Our problem differs significantly from existing studies on breathing rate estimation, which assume the number of people is known a priori. In contrast, mmCounter utilizes a novel multi-stage signal processing pipeline to extract relevant low-frequency sources along with their spatial information and map these sources to individual people, enabling accurate counting. Extensive evaluations in various environments demonstrate that mmCounter delivers an 87% average F1 score and 0.6 mean absolute error in familiar environments, and a 60% average F1 score and 1.1 mean absolute error in previously untested environments. It can count up to seven individuals in a three square meter space, such that there is no side-by-side spacing and only a one-meter front-to-back distance.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:16:47 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Toha",
"Tarik Reza",
"",
"Louie"
],
[
"Shao-Jung",
"",
"",
"Louie"
],
[
"Lu",
"",
""
],
[
"Nirjon",
"Shahriar",
""
]
] |
2512.10358
|
Runhao Liu
|
Runhao Liu, Ziming Chen, You Li, Zequn Xie, Peng Zhang
|
Integrated Planning and Machine-Level Scheduling for High-Mix Discrete Manufacturing: A Profit-Driven Heuristic Framework
| null | null | null | null |
cs.CE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Modern manufacturing enterprises struggle to create efficient and reliable production schedules under multi-variety, small-batch, and rush-order conditions. High-mix discrete manufacturing systems require jointly optimizing mid-term production planning and machine-level scheduling under heterogeneous resources and stringent delivery commitments. We address this problem with a profit-driven integrated framework that couples a mixed-integer planning model with a machine-level scheduling heuristic. The planning layer allocates production, accessory co-production, and outsourcing under aggregate economic and capacity constraints, while the scheduling layer refines these allocations using a structure-aware procedure that enforces execution feasibility and stabilizes daily machine behavior. This hierarchical design preserves the tractability of aggregated optimization while capturing detailed operational restrictions. Evaluations are conducted on a real industrial scenario. A flexible machine-level execution scheme yields 73.3% on-time completion and significant outsourcing demand, revealing bottleneck congestion. In contrast, a stability-enforcing execution policy achieves 100% on-time completion, eliminates all outsourcing, and maintains balanced machine utilization with only 1.9 to 4.6% capacity loss from changeovers. These results show that aligning planning decisions with stability-oriented execution rules enables practical and interpretable profit-maximizing decisions in complex manufacturing environments.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:17:29 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Runhao",
""
],
[
"Chen",
"Ziming",
""
],
[
"Li",
"You",
""
],
[
"Xie",
"Zequn",
""
],
[
"Zhang",
"Peng",
""
]
] |
2512.10359
|
Sunqi Fan
|
Sunqi Fan, Jiashuo Cui, Meng-Hao Guo, Shuojin Yang
|
Tool-Augmented Spatiotemporal Reasoning for Streamlining Video Question Answering Task
|
Accepted by NeurIPS 2025 main track
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Video Question Answering (VideoQA) task serves as a critical playground for evaluating whether foundation models can effectively perceive, understand, and reason about dynamic real-world scenarios. However, existing Multimodal Large Language Models (MLLMs) struggle with simultaneously modeling spatial relationships within video frames and understanding the causal dynamics of temporal evolution on complex and reasoning-intensive VideoQA task. In this work, we equip MLLM with a comprehensive and extensible Video Toolkit, to enhance MLLM's spatiotemporal reasoning capabilities and ensure the harmony between the quantity and diversity of tools. To better control the tool invocation sequence and avoid toolchain shortcut issues, we propose a Spatiotemporal Reasoning Framework (STAR) that strategically schedules temporal and spatial tools, thereby progressively localizing the key area in the video. Our STAR framework enhances GPT-4o using lightweight tools, achieving an 8.2% gain on VideoMME and 4.6% on LongVideoBench. We believe that our proposed Video Toolkit and STAR framework make an important step towards building autonomous and intelligent video analysis assistants. The code is publicly available at https://github.com/fansunqi/VideoTool.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:17:57 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Fan",
"Sunqi",
""
],
[
"Cui",
"Jiashuo",
""
],
[
"Guo",
"Meng-Hao",
""
],
[
"Yang",
"Shuojin",
""
]
] |
2512.10360
|
Liuyi Wang
|
Liuyi Wang, Zongtao He, Jinlong Li, Xiaoyan Qi, Mengxian Hu, Chenpeng Yao, Chengju Liu, Qijun Chen
|
CLASH: Collaborative Large-Small Hierarchical Framework for Continuous Vision-and-Language Navigation
| null | null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vision-and-Language Navigation (VLN) requires robots to follow natural language instructions and navigate complex environments without prior maps. While recent vision-language large models demonstrate strong reasoning abilities, they often underperform task-specific panoramic small models in VLN tasks. To address this, we propose CLASH (Collaborative Large-Small Hierarchy), a VLN-CE framework that integrates a reactive small-model planner (RSMP) with a reflective large-model reasoner (RLMR). RSMP adopts a causal-learning-based dual-branch architecture to enhance generalization, while RLMR leverages panoramic visual prompting with chain-of-thought reasoning to support interpretable spatial understanding and navigation. We further introduce an uncertainty-aware collaboration mechanism (UCM) that adaptively fuses decisions from both models. For obstacle avoidance, in simulation, we replace the rule-based controller with a fully learnable point-goal policy, and in real-world deployment, we design a LiDAR-based clustering module for generating navigable waypoints and pair it with an online SLAM-based local controller. CLASH achieves state-of-the-art (SoTA) results (ranking 1-st) on the VLN-CE leaderboard, significantly improving SR and SPL on the test-unseen set over the previous SoTA methods. Real-world experiments demonstrate CLASH's strong robustness, validating its effectiveness in both simulation and deployment scenarios.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:20:06 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wang",
"Liuyi",
""
],
[
"He",
"Zongtao",
""
],
[
"Li",
"Jinlong",
""
],
[
"Qi",
"Xiaoyan",
""
],
[
"Hu",
"Mengxian",
""
],
[
"Yao",
"Chenpeng",
""
],
[
"Liu",
"Chengju",
""
],
[
"Chen",
"Qijun",
""
]
] |
2512.10361
|
Wei Shao
|
Wei Shao, Najmeh Nazari, Behnam Omidi, Setareh Rafatirad, Houman Homayoun, Khaled N. Khasawneh, Chongzhou Fang
|
Bit of a Close Talker: A Practical Guide to Serverless Cloud Co-Location Attacks
|
In the proceedings of Network and Distributed System Security (NDSS) Symposium 2026
| null | null | null |
cs.CR cs.DC
|
http://creativecommons.org/licenses/by/4.0/
|
Serverless computing has revolutionized cloud computing by offering an efficient and cost-effective way for users to develop and deploy applications without managing infrastructure details. However, serverless cloud users remain vulnerable to various types of attacks, including micro-architectural side-channel attacks. These attacks typically rely on the physical co-location of victim and attacker instances, and attackers will need to exploit cloud schedulers to achieve co-location with victims. Therefore, it is crucial to study vulnerabilities in serverless cloud schedulers and assess the security of different serverless scheduling algorithms. This study addresses the gap in understanding and constructing co-location attacks in serverless clouds. We present a comprehensive methodology to uncover exploitable features in serverless scheduling algorithms and devise strategies for constructing co-location attacks through normal user interfaces. In our experiments, we successfully reveal exploitable vulnerabilities and achieve instance co-location on prevalent open-source infrastructures and Microsoft Azure Functions. We also present a mitigation strategy to defend against co-location attacks in serverless clouds. Our work highlights critical areas for security enhancements in current cloud schedulers, offering insights to fortify serverless computing environments against potential co-location attacks.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:22:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Shao",
"Wei",
""
],
[
"Nazari",
"Najmeh",
""
],
[
"Omidi",
"Behnam",
""
],
[
"Rafatirad",
"Setareh",
""
],
[
"Homayoun",
"Houman",
""
],
[
"Khasawneh",
"Khaled N.",
""
],
[
"Fang",
"Chongzhou",
""
]
] |
2512.10362
|
Woojun Jung
|
Woojun Jung, Jaehoon Go, Mingyu Jeon, Sunjae Yoon, Junyeong Kim
|
Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language Models
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Multimodal Large Language Models (MLLMs) demonstrate impressive reasoning capabilities, but often fail to perceive fine-grained visual details, limiting their applicability in precision-demanding tasks. While methods that crop salient regions of an image offer a partial solution, we identify a critical limitation they introduce: "Contextual Blindness". This failure occurs due to structural disconnect between high-fidelity details (from the crop) and the broader global context (from the original image), even when all necessary visual information is present. We argue that this limitation stems not from a lack of information 'Quantity', but from a lack of 'Structural Diversity' in the model's input. To resolve this, we propose Visual Funnel, a training-free, two-step approach. Visual Funnel first performs Contextual Anchoring to identify the region of interest in a single forward pass. It then constructs an Entropy-Scaled Portfolio that preserves the hierarchical context - ranging from focal detail to broader surroundings - by dynamically determining crop sizes based on attention entropy and refining crop centers. Through extensive experiments, we demonstrate that Visual Funnel significantly outperforms naive single-crop and unstructured multi-crop baselines. Our results further validate that simply adding more unstructured crops provides limited or even detrimental benefits, confirming that the hierarchical structure of our portfolio is key to resolving Contextual Blindness.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:22:54 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Jung",
"Woojun",
""
],
[
"Go",
"Jaehoon",
""
],
[
"Jeon",
"Mingyu",
""
],
[
"Yoon",
"Sunjae",
""
],
[
"Kim",
"Junyeong",
""
]
] |
2512.10363
|
Mingyu Jeon
|
Mingyu Jeon, Jisoo Yang, Sungjin Han, Jinkwon Hwang, Sunjae Yoon, Jonghee Kim, Junyeoung Kim
|
Point to Span: Zero-Shot Moment Retrieval for Navigating Unseen Hour-Long Videos
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Zero-shot Long Video Moment Retrieval (ZLVMR) is the task of identifying temporal segments in hour-long videos using a natural language query without task-specific training. The core technical challenge of LVMR stems from the computational infeasibility of processing entire lengthy videos in a single pass. This limitation has established a 'Search-then-Refine' approach, where candidates are rapidly narrowed down, and only those portions are analyzed, as the dominant paradigm for LVMR. However, existing approaches to this paradigm face severe limitations. Conventional supervised learning suffers from limited scalability and poor generalization, despite substantial resource consumption. Yet, existing zero-shot methods also fail, facing a dual challenge: (1) their heuristic strategies cause a 'search' phase candidate explosion, and (2) the 'refine' phase, which is vulnerable to semantic discrepancy, requires high-cost VLMs for verification, incurring significant computational overhead. We propose \textbf{P}oint-\textbf{to}-\textbf{S}pan (P2S), a novel training-free framework to overcome this challenge of inefficient 'search' and costly 'refine' phases. P2S overcomes these challenges with two key innovations: an 'Adaptive Span Generator' to prevent the search phase candidate explosion, and 'Query Decomposition' to refine candidates without relying on high-cost VLM verification. To our knowledge, P2S is the first zero-shot framework capable of temporal grounding in hour-long videos, outperforming supervised state-of-the-art methods by a significant margin (e.g., +3.7\% on R5@0.1 on MAD).
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:25:48 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Jeon",
"Mingyu",
""
],
[
"Yang",
"Jisoo",
""
],
[
"Han",
"Sungjin",
""
],
[
"Hwang",
"Jinkwon",
""
],
[
"Yoon",
"Sunjae",
""
],
[
"Kim",
"Jonghee",
""
],
[
"Kim",
"Junyeoung",
""
]
] |
2512.10364
|
Lu Lu
|
Yueli Han, Lu Lu
|
Spectral bounds for vertex-weighted Laplacians of simplicial complexes
| null | null | null | null |
math.CO
|
http://creativecommons.org/publicdomain/zero/1.0/
|
The vertex-weighted Laplacian naturally extends the combinatorial Laplacian for simplicial complexes. Inspired by Lew's foundational techniques for vertex-weighted Laplacians, we present a comprehensive spectral analysis of this operator. First, we determine how basic operations, including joins, complements, and Alexander duals, affect its spectrum. This yields a sharp upper bound on the spectral radius in terms of vertex weights, along with a lower bound on the multiplicity at which this bound is attained. Second, we establish a sharp lower bound for the spectral gap and characterize when the equality holds. Third, explicit lower bounds for the remaining eigenvalues are derived, linking the vertex-weighted Laplacian spectrum to that of a related weighted graph. Finally, we reveal new spectral relations between a simplicial complex and its subcomplexes. These results not only generalize numerous known theorems on combinatorial Laplacians but also provide deeper spectral insights into simplicial structures, ultimately unifying and extending a broad range of earlier work in this field.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:27:18 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Han",
"Yueli",
""
],
[
"Lu",
"Lu",
""
]
] |
2512.10365
|
Hangyu Mao
|
Hangyu Mao, Guangting Dong, and Zhicheng Dou
|
GPG: Generalized Policy Gradient Theorem for Transformer-based Policies
| null | null | null | null |
cs.LG cs.AI cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present the Generalized Policy Gradient (GPG) Theorem, specifically designed for Transformer-based policies. Notably, we demonstrate that both standard Policy Gradient Theorem and GRPO emerge as special cases within our GPG framework. Furthermore, we explore its practical applications in training Large Language Models (LLMs), offering new insights into efficient policy optimization.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:30:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Mao",
"Hangyu",
""
],
[
"Dong",
"Guangting",
""
],
[
"Dou",
"Zhicheng",
""
]
] |
2512.10366
|
Minh N. Dao
|
Minh N. Dao, Hung M. Phan, Matthew K. Tam, and Thang D. Truong
|
Primal-dual splitting for structured composite monotone inclusions with or without cocoercivity
| null | null | null | null |
math.OC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we propose a primal-dual splitting algorithm for a broad class of structured composite monotone inclusions that involve finitely many set-valued operators, compositions of set-valued operators with bounded linear operators, and single-valued operators possibly without cocoercivity. The proposed algorithm is not only a unification for several contemporary algorithms but also a blueprint to generate new algorithms with graph-based structures. Our approach reduces dimensionality compared with the standard product space technique, which typically reformulates the original problem as the sum of two maximally monotone operators in order to apply splitting methods. It accommodates different cocoercive or Lipschitz constants as well as different resolvent parameters, and yields a larger allowable step-size range than in product space reformulations. We demonstrate the practicality of the approach by a numerical experiment on the decentralized fused lasso problem.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:31:09 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Dao",
"Minh N.",
""
],
[
"Phan",
"Hung M.",
""
],
[
"Tam",
"Matthew K.",
""
],
[
"Truong",
"Thang D.",
""
]
] |
2512.10367
|
Naoki Ogawa
|
Hideo Furugori, Naoki Ogawa, Sotaro Sugishita, Takahiro Waki
|
Bridging dS/CFT and Celestial Holography via Ward-Takahashi Identities
|
37 pages, 1 figure
| null | null |
KUNS-3081, YITP-25-179, RIKEN-iTHEMS-Report-25
|
hep-th gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In 2507.17558, we provide a map from a scalar theory on $(D+2)$-dimensional Minkowski spacetime to a scalar theory with a continuous mass spectrum on $(D+1)$-dimensional de Sitter spacetime, and propose a link between celestial amplitudes and cosmological correlators (the cosmological-celestial dictionary). We extend the construction to fields with spin 1 and 2, and find that massless spin fields map to spin fields with continuous mass spectra. In this construction, we identify the de Sitter counterparts of the Nambu-Goldstone modes associated with the asymptotic symmetries in Minkowski spacetime. For $U(1)$ gauge theories, the counterpart is restricted to the massless sector within the continuous Proca spectrum, while for linearized gravity supertranslations are encoded in the partially massless sector and superrotations in the strictly massless sector. Using the identification, we reveal that the associated Ward-Takahashi identities of the cosmological correlators reproduce the conformally soft photon and graviton theorems via the cosmological-celestial dictionary. In particular, the celestial stress tensor is derived from the asymptotic limit of gravitons in de Sitter spacetime.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:31:21 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Furugori",
"Hideo",
""
],
[
"Ogawa",
"Naoki",
""
],
[
"Sugishita",
"Sotaro",
""
],
[
"Waki",
"Takahiro",
""
]
] |
2512.10368
|
Michio Seto
|
Michio Seto
|
L\"{o}wner equations and de Branges--Rovnyak spaces
| null | null | null | null |
math.FA math.CV
|
http://creativecommons.org/licenses/by/4.0/
|
We study de Branges--Rovnyak spaces parametrized by L\"{o}wner equations. A new approach based on the L\"{o}wner theory to problem ``Find concrete elements in de Branges--Rovnyak spaces'' is given.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:35:06 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Seto",
"Michio",
""
]
] |
2512.10369
|
Chaoran Feng
|
Zhankuo Xu, Chaoran Feng, Yingtao Li, Jianbin Zhao, Jiashu Yang, Wangbo Yu, Li Yuan, and Yonghong Tian
|
Breaking the Vicious Cycle: Coherent 3D Gaussian Splatting from Sparse and Motion-Blurred Views
|
20 pages, 14 figures
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
3D Gaussian Splatting (3DGS) has emerged as a state-of-the-art method for novel view synthesis. However, its performance heavily relies on dense, high-quality input imagery, an assumption that is often violated in real-world applications, where data is typically sparse and motion-blurred. These two issues create a vicious cycle: sparse views ignore the multi-view constraints necessary to resolve motion blur, while motion blur erases high-frequency details crucial for aligning the limited views. Thus, reconstruction often fails catastrophically, with fragmented views and a low-frequency bias. To break this cycle, we introduce CoherentGS, a novel framework for high-fidelity 3D reconstruction from sparse and blurry images. Our key insight is to address these compound degradations using a dual-prior strategy. Specifically, we combine two pre-trained generative models: a specialized deblurring network for restoring sharp details and providing photometric guidance, and a diffusion model that offers geometric priors to fill in unobserved regions of the scene. This dual-prior strategy is supported by several key techniques, including a consistency-guided camera exploration module that adaptively guides the generative process, and a depth regularization loss that ensures geometric plausibility. We evaluate CoherentGS through both quantitative and qualitative experiments on synthetic and real-world scenes, using as few as 3, 6, and 9 input views. Our results demonstrate that CoherentGS significantly outperforms existing methods, setting a new state-of-the-art for this challenging task. The code and video demos are available at https://potatobigroom.github.io/CoherentGS/.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:36:35 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Xu",
"Zhankuo",
""
],
[
"Feng",
"Chaoran",
""
],
[
"Li",
"Yingtao",
""
],
[
"Zhao",
"Jianbin",
""
],
[
"Yang",
"Jiashu",
""
],
[
"Yu",
"Wangbo",
""
],
[
"Yuan",
"Li",
""
],
[
"Tian",
"Yonghong",
""
]
] |
2512.10370
|
Ziying Zhang
|
Ziying Zhang, Quanming Yao, Yaqing Wang
|
LLM-Empowered Representation Learning for Emerging Item Recommendation
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
In this work, we tackle the challenge of recommending emerging items, whose interactions gradually accumulate over time. Existing methods often overlook this dynamic process, typically assuming that emerging items have few or even no historical interactions. Such an assumption oversimplifies the problem, as a good model must preserve the uniqueness of emerging items while leveraging their shared patterns with established ones. To address this challenge, we propose EmerFlow, a novel LLM-empowered representation learning framework that generates distinctive embeddings for emerging items. It first enriches the raw features of emerging items through LLM reasoning, then aligns these representations with the embedding space of the existing recommendation model. Finally, new interactions are incorporated through meta-learning to refine the embeddings. This enables EmerFlow to learn expressive embeddings for emerging items from only limited interactions. Extensive experiments across diverse domains, including movies and pharmaceuticals, show that EmerFlow consistently outperforms existing methods.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:36:44 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Ziying",
""
],
[
"Yao",
"Quanming",
""
],
[
"Wang",
"Yaqing",
""
]
] |
2512.10371
|
Shizuo Tian
|
Shizuo Tian, Hao Wen, Yuxuan Chen, Jiacheng Liu, Shanhui Zhao, Guohong Liu, Ju Ren, Yunxin Liu, Yuanchun Li
|
AgentProg: Empowering Long-Horizon GUI Agents with Program-Guided Context Management
|
16 pages, 8 figures
| null | null | null |
cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The rapid development of mobile GUI agents has stimulated growing research interest in long-horizon task automation. However, building agents for these tasks faces a critical bottleneck: the reliance on ever-expanding interaction history incurs substantial context overhead. Existing context management and compression techniques often fail to preserve vital semantic information, leading to degraded task performance. We propose AgentProg, a program-guided approach for agent context management that reframes the interaction history as a program with variables and control flow. By organizing information according to the structure of program, this structure provides a principled mechanism to determine which information should be retained and which can be discarded. We further integrate a global belief state mechanism inspired by Belief MDP framework to handle partial observability and adapt to unexpected environmental changes. Experiments on AndroidWorld and our extended long-horizon task suite demonstrate that AgentProg has achieved the state-of-the-art success rates on these benchmarks. More importantly, it maintains robust performance on long-horizon tasks while baseline methods experience catastrophic degradation. Our system is open-sourced at https://github.com/MobileLLM/AgentProg.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:37:38 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Tian",
"Shizuo",
""
],
[
"Wen",
"Hao",
""
],
[
"Chen",
"Yuxuan",
""
],
[
"Liu",
"Jiacheng",
""
],
[
"Zhao",
"Shanhui",
""
],
[
"Liu",
"Guohong",
""
],
[
"Ren",
"Ju",
""
],
[
"Liu",
"Yunxin",
""
],
[
"Li",
"Yuanchun",
""
]
] |
2512.10372
|
Nitin Awathare
|
Yash Srivastava, Shalin Jain, Sneha Awathare and Nitin Awathare
|
D2M: A Decentralized, Privacy-Preserving, Incentive-Compatible Data Marketplace for Collaborative Learning
| null | null | null | null |
cs.CR cs.AI cs.DC cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The rising demand for collaborative machine learning and data analytics calls for secure and decentralized data sharing frameworks that balance privacy, trust, and incentives. Existing approaches, including federated learning (FL) and blockchain-based data markets, fall short: FL often depends on trusted aggregators and lacks Byzantine robustness, while blockchain frameworks struggle with computation-intensive training and incentive integration.
We present \prot, a decentralized data marketplace that unifies federated learning, blockchain arbitration, and economic incentives into a single framework for privacy-preserving data sharing. \prot\ enables data buyers to submit bid-based requests via blockchain smart contracts, which manage auctions, escrow, and dispute resolution. Computationally intensive training is delegated to \cone\ (\uline{Co}mpute \uline{N}etwork for \uline{E}xecution), an off-chain distributed execution layer. To safeguard against adversarial behavior, \prot\ integrates a modified YODA protocol with exponentially growing execution sets for resilient consensus, and introduces Corrected OSMD to mitigate malicious or low-quality contributions from sellers. All protocols are incentive-compatible, and our game-theoretic analysis establishes honesty as the dominant strategy.
We implement \prot\ on Ethereum and evaluate it over benchmark datasets -- MNIST, Fashion-MNIST, and CIFAR-10 -- under varying adversarial settings. \prot\ achieves up to 99\% accuracy on MNIST and 90\% on Fashion-MNIST, with less than 3\% degradation up to 30\% Byzantine nodes, and 56\% accuracy on CIFAR-10 despite its complexity. Our results show that \prot\ ensures privacy, maintains robustness under adversarial conditions, and scales efficiently with the number of participants, making it a practical foundation for real-world decentralized data sharing.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:38:05 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Srivastava",
"Yash",
""
],
[
"Jain",
"Shalin",
""
],
[
"Awathare",
"Sneha",
""
],
[
"Awathare",
"Nitin",
""
]
] |
2512.10373
|
Torsten Djurhuus PhD.
|
Torsten Djurhuus, Viktor Krozer
|
A Novel Phase-Noise Module for the QUCS Circuit Simulator. Part I : the Periodic Steady-State
|
13 pages, 5 figures
| null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The paper discusses work done to expand and extend the capabilities of the open-source QUCS circuit simulator through the implementation of a computationally efficient time-domain steady-state analysis module, supporting simulation of autonomous circuits. To our knowledge, this represents the first time such an analysis module has been implemented in the QUCS environment. Hitherto, the only available option was a harmonic-balance module which was strictly limited to non-autonomous (driven) circuits. The research has several important scientific and industrial applications in the area of large-signal steady-state analysis of autonomous circuits e.g. free-running and coupled oscillator circuit networks. The reported results will have great impact w.r.t. analyzing, synthesizing and optimizing oscillatory behavior of various important industrial circuits and systems. The developed tool, furthermore, introduces support for simulating noise performance of circuits operating under large-signal conditions. This paper is the first part of a two-part series documenting the implementation of a novel (coupled)-oscillator phase-noise simulator engine in the QUCS environment. The goal of this undertaking is the advancement of the open-source QUCS project towards becoming a viable competitor to the commercial simulators currently on the market.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:39:42 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Djurhuus",
"Torsten",
""
],
[
"Krozer",
"Viktor",
""
]
] |
2512.10374
|
Pradeep M
|
Pradeep M, Twinkle Tripathy
|
Structural Sign Herdability in Temporally Switching Networks with Fixed Topology
| null | null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper investigates structural herdability in a special class of temporally switching networks with fixed topology. We show that when the underlying digraph remains unchanged across all snapshots, the network attains complete SS herdability even in the presence of signed or layer dilations, a condition not applicable to static networks. This reveals a fundamental structural advantage of temporal dynamics and highlights a novel mechanism through which switching can overcome classical obstructions to herdability. To validate these conclusions, we utilize a more relaxed form of sign matching within each snapshot of the temporal network. Furthermore, we show that when all snapshots share the same underlying topology, the temporally switching network achieves $\mathcal{SS}$ herdability within just two snapshots, which is fewer than the number required for structural controllability. Several examples are included to demonstrate these results.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:40:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"M",
"Pradeep",
""
],
[
"Tripathy",
"Twinkle",
""
]
] |
2512.10375
|
Wenye Zhu
|
Wenye Zhu, Jun Tang, Xiaofei Li
|
Neural personal sound zones with flexible bright zone control
| null | null | null | null |
cs.SD cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Personal sound zone (PSZ) reproduction system, which attempts to create distinct virtual acoustic scenes for different listeners at their respective positions within the same spatial area using one loudspeaker array, is a fundamental technology in the application of virtual reality. For practical applications, the reconstruction targets must be measured on the same fixed receiver array used to record the local room impulse responses (RIRs) from the loudspeaker array to the control points in each PSZ, which makes the system inconvenient and costly for real-world use. In this paper, a 3D convolutional neural network (CNN) designed for PSZ reproduction with flexible control microphone grid and alternative reproduction target is presented, utilizing the virtual target scene as inputs and the PSZ pre-filters as output. Experimental results of the proposed method are compared with the traditional method, demonstrating that the proposed method is able to handle varied reproduction targets on flexible control point grid using only one training session. Furthermore, the proposed method also demonstrates the capability to learn global spatial information from sparse sampling points distributed in PSZs.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:41:15 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhu",
"Wenye",
""
],
[
"Tang",
"Jun",
""
],
[
"Li",
"Xiaofei",
""
]
] |
2512.10376
|
Jingyun Fu
|
Jingyun Fu, Zhiyu Xiang, Na Zhao
|
RaLiFlow: Scene Flow Estimation with 4D Radar and LiDAR Point Clouds
|
Accepted by AAAI
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent multimodal fusion methods, integrating images with LiDAR point clouds, have shown promise in scene flow estimation. However, the fusion of 4D millimeter wave radar and LiDAR remains unexplored. Unlike LiDAR, radar is cheaper, more robust in various weather conditions and can detect point-wise velocity, making it a valuable complement to LiDAR. However, radar inputs pose challenges due to noise, low resolution, and sparsity. Moreover, there is currently no dataset that combines LiDAR and radar data specifically for scene flow estimation. To address this gap, we construct a Radar-LiDAR scene flow dataset based on a public real-world automotive dataset. We propose an effective preprocessing strategy for radar denoising and scene flow label generation, deriving more reliable flow ground truth for radar points out of the object boundaries. Additionally, we introduce RaLiFlow, the first joint scene flow learning framework for 4D radar and LiDAR, which achieves effective radar-LiDAR fusion through a novel Dynamic-aware Bidirectional Cross-modal Fusion (DBCF) module and a carefully designed set of loss functions. The DBCF module integrates dynamic cues from radar into the local cross-attention mechanism, enabling the propagation of contextual information across modalities. Meanwhile, the proposed loss functions mitigate the adverse effects of unreliable radar data during training and enhance the instance-level consistency in scene flow predictions from both modalities, particularly for dynamic foreground areas. Extensive experiments on the repurposed scene flow dataset demonstrate that our method outperforms existing LiDAR-based and radar-based single-modal methods by a significant margin.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:41:33 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Fu",
"Jingyun",
""
],
[
"Xiang",
"Zhiyu",
""
],
[
"Zhao",
"Na",
""
]
] |
2512.10377
|
Sheng-Ting Hung
|
Sheng-Ting Hung (1), Cheng Yan Lee (1), Chen-Yu Lien (1), Cheng-Hsuan Chan (1), Ya-Han Yang (2), Quark Yungsung Chen (3 and 4), Kuang-Hung Cheng (5), Kung-Kai Kuo (2 and 6), Li-Wei Tu (1) and Ching-Wen Chang (7) ((1) Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan (2) Department of Surgery, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan (3) Industry Academia Innovation School, National Yang Ming Chiao Tung University, Hsinchu, Taiwan (4) Department of Physics and Texas Center for Superconductivity, University of Houston, Houston, TX, USA (5) Institute of Biomedical Sciences, National Sun Yat-sen University, Kaohsiung, Taiwan (6) Department of Surgery, E-DA Healthcare Group E-DA Dachang Hospital, Kaohsiung, Taiwan (7) Institute and Undergraduate Program of Electro-Optical Engineering, National Taiwan Normal University, Taipei, Taiwan)
|
KRAS G12D protein screening for pancreatic cancer clinical trials using an AlGaN/GaN high electron mobility transistor biosensor
| null | null | null | null |
physics.med-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Clinical trials screening KRAS G12D protein for 30 pancreatic ductal adenocarcinoma (PDAC) patients and 30 healthy donors were conducted utilizing an AlGaN/GaN high electron mobility transistor (HEMT) biosensor. All resistance change ratios of PDAC patients are higher than the standard deviation above the mean resistance change ratio obtained from all healthy donors. The results demonstrate the effectiveness of the HEMT biosensor and reveal its potential for early detection of pancreatic cancer with KRAS G12D protein screening.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:41:42 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Hung",
"Sheng-Ting",
"",
"Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Lee",
"Cheng Yan",
"",
"Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Lien",
"Chen-Yu",
"",
"Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Chan",
"Cheng-Hsuan",
"",
"Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Yang",
"Ya-Han",
"",
"Department of Surgery, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan"
],
[
"Chen",
"Quark Yungsung",
"",
"3 and 4"
],
[
"Cheng",
"Kuang-Hung",
"",
"Institute of Biomedical Sciences, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Kuo",
"Kung-Kai",
"",
"2 and 6"
],
[
"Tu",
"Li-Wei",
"",
"Department of Physics, National Sun Yat-sen University, Kaohsiung, Taiwan"
],
[
"Chang",
"Ching-Wen",
"",
"Institute and Undergraduate Program of Electro-Optical Engineering, National Taiwan Normal University, Taipei, Taiwan"
]
] |
2512.10378
|
Peizhe Li
|
Pei-Zhe Li, Soumyakanti Bose, Hyunseok Jeong, William J. Munro, Kae Nemoto, Nicol\`o Lo Piparo
|
Loophole-free Bell-inequality violation between atomic states in cavity-QED systems mediated by hybrid atom-light entanglement
|
7 pages, 4 figures
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present a feasible and scalable approach to testing Bell nonlocality and implementing device-independent quantum key distribution (DI-QKD) between distant atomic states in cavity-based architectures, mediated by hybrid atom-light entanglement. We develop a full theoretical model that incorporates realistic sources of noise -- such as transmission loss, limited light-matter coupling efficiency, and imperfect detection. Our analysis shows that strong Bell-Clauser-Horne-Shimony-Holt (CHSH) violations and secure key generation over tens of kilometers are within reach using current or near-term technology. These results position cavity-based platforms with coherent-state encodings as a promising foundation for future scalable, DI quantum communication networks.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:42:30 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Li",
"Pei-Zhe",
""
],
[
"Bose",
"Soumyakanti",
""
],
[
"Jeong",
"Hyunseok",
""
],
[
"Munro",
"William J.",
""
],
[
"Nemoto",
"Kae",
""
],
[
"Piparo",
"Nicolò Lo",
""
]
] |
2512.10379
|
Alberto Rota
|
Alberto Rota and Elena De Momi
|
Self-Supervised Contrastive Embedding Adaptation for Endoscopic Image Matching
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Accurate spatial understanding is essential for image-guided surgery, augmented reality integration and context awareness. In minimally invasive procedures, where visual input is the sole intraoperative modality, establishing precise pixel-level correspondences between endoscopic frames is critical for 3D reconstruction, camera tracking, and scene interpretation. However, the surgical domain presents distinct challenges: weak perspective cues, non-Lambertian tissue reflections, and complex, deformable anatomy degrade the performance of conventional computer vision techniques. While Deep Learning models have shown strong performance in natural scenes, their features are not inherently suited for fine-grained matching in surgical images and require targeted adaptation to meet the demands of this domain. This research presents a novel Deep Learning pipeline for establishing feature correspondences in endoscopic image pairs, alongside a self-supervised optimization framework for model training. The proposed methodology leverages a novel-view synthesis pipeline to generate ground-truth inlier correspondences, subsequently utilized for mining triplets within a contrastive learning paradigm. Through this self-supervised approach, we augment the DINOv2 backbone with an additional Transformer layer, specifically optimized to produce embeddings that facilitate direct matching through cosine similarity thresholding. Experimental evaluation demonstrates that our pipeline surpasses state-of-the-art methodologies on the SCARED datasets improved matching precision and lower epipolar error compared to the related work. The proposed framework constitutes a valuable contribution toward enabling more accurate high-level computer vision applications in surgical endoscopy.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:44:00 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Rota",
"Alberto",
""
],
[
"De Momi",
"Elena",
""
]
] |
2512.10380
|
Zhi-Xi Wang
|
Yu Lu, Wen Zhou, Meng Su, Hong-Xing Wu, Shao-Ming Fei, Zhi-Xi Wang
|
Quantum Separability Criteria Based on Symmetric Measurements
|
17 pages, 6 figures
|
Phys. Rev. A 112(2025), 062424
|
10.1103/v9gs-6jvz
| null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose experimentally feasible separability criteria for bipartite systems based on local symmetric measurements. Through detailed examples, we demonstrate that our criteria can detect entanglement more effectively compared to existing counterparts. Furthermore,we demonstrate the potential for our results to be generalized to general multipartite systems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:44:31 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lu",
"Yu",
""
],
[
"Zhou",
"Wen",
""
],
[
"Su",
"Meng",
""
],
[
"Wu",
"Hong-Xing",
""
],
[
"Fei",
"Shao-Ming",
""
],
[
"Wang",
"Zhi-Xi",
""
]
] |
2512.10381
|
George Leontaris K
|
George K. Leontaris, George Prampromis
|
5D Rotating Black Holes as dark matter in Dark Dimension Scenario: Hawking Radiation versus the Memory Burden Effect
|
31 pages, 3 figures
| null | null | null |
hep-th gr-qc hep-ph
|
http://creativecommons.org/licenses/by/4.0/
|
This work explores the possibility that five-dimensional primordial rotating black holes could account for all, or a significant portion, of the dark matter in our universe. Our analysis is performed within the context of the ``dark dimension'' scenario, a theoretical consequence of the Swampland Program that predicts a single micron-scale extra dimension to explain the observed value of dark energy. We demonstrate that within this scenario, the mass loss of a primordial rotating black hole sensitive to the fifth dimension is significantly slower than that of its four-dimensional counterpart. Consequently, primordial black holes with an initial mass of $M\gtrsim 10^{10}$g can survive to the present day and potentially constitute the dominant form of dark matter. Finally, we investigate the memory burden effect and find that it dramatically prolongs the lifetime of five-dimensional rotating primordial black holes, making them compelling candidates for all the dark matter in the universe.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:45:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Leontaris",
"George K.",
""
],
[
"Prampromis",
"George",
""
]
] |
2512.10382
|
Liusha Yang
|
Liusha Yang, Ziru Ge, Gui Zhang, Junan Zhang, Zhizheng Wu
|
Investigating training objective for flow matching-based speech enhancement
| null | null | null | null |
cs.SD
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Speech enhancement(SE) aims to recover clean speech from noisy recordings. Although generative approaches such as score matching and Schrodinger bridge have shown strong effectiveness, they are often computationally expensive. Flow matching offers a more efficient alternative by directly learning a velocity field that maps noise to data. In this work, we present a systematic study of flow matching for SE under three training objectives: velocity prediction, $x_1$ prediction, and preconditioned $x_1$ prediction. We analyze their impact on training dynamics and overall performance. Moreover, by introducing perceptual(PESQ) and signal-based(SI-SDR) objectives, we further enhance convergence efficiency and speech quality, yielding substantial improvements across evaluation metrics.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:45:56 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Yang",
"Liusha",
""
],
[
"Ge",
"Ziru",
""
],
[
"Zhang",
"Gui",
""
],
[
"Zhang",
"Junan",
""
],
[
"Wu",
"Zhizheng",
""
]
] |
2512.10383
|
Angelos Nersesian Dr
|
Angelos Nersesian, Yasha Kaushal, Marco Martorano, Arjen van der Wel, Po-Feng Wu, Rachel Bezanson, Eric F. Bell, Francesco D'Eugenio, Anna R. Gallazzi, Joel Leja, Stefano Zibetti, Sandro Tacchella
|
The LEGA-C galaxy survey: multiple quenching channels for quiescent galaxies at $z\sim1$
|
16 pages, 12 figures, accepted
| null | null | null |
astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We analyzed the sizes and star-formation histories (SFHs) of 2908 galaxies with $M_\star \geq 10^9$ M$_\odot$ at $0.6 < z < 1.0$, drawn from the LEGA-C survey. The goal is to investigate the connection between galaxy sizes with SFH, stellar age, and metallicity. SFHs were derived with Prospector by fitting the high signal-to-noise, high spectral resolution spectroscopy drawn from the LEGA-C DR3 together with the broadband photometry from the UltraVISTA catalog. Galaxy sizes were measured by fitting a 2D S{\'e}rsic profile to the HST ACS~F814W images. We find diverse SFHs and quenching timescales ($\tau_\rm{q}$). The main quiescent population quenched over $\tau_\rm{q}=1.23\pm0.04$ Gyr, whereas compact post-starburst galaxies (PSBs) quenched much faster, $\tau_\rm{q}=0.13\pm0.03$ Gyr. At fixed stellar mass, smaller quiescent galaxies quenched more rapidly than larger ones; at fixed size, the dependence on stellar mass is weak. Larger quiescent galaxies are marginally younger, quenched more slowly, and have near-solar metallicities, while compact quiescent galaxies are older, metal-rich, and quenched faster. PSBs formed half their mass later ($z_\rm{form}\sim1.9$) and quenched on the shortest timescales. The general trends with galaxy size, $Z_\star$, and $z_\rm{form}$ for the quiescent populations remain consistent regardless of the method used to derive the stellar properties. We conclude that compact quiescent galaxies are consistent with both early, moderately fast quenching and with more rapid, late quenching. While this may suggest the existence of multiple quenching channels, our data are also compatible with a continuous distribution of quenching timescales. These findings suggest that different physical mechanisms may drive quenching across galaxy populations, potentially leading to similar morphological outcomes despite differing evolutionary histories.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:47:18 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Nersesian",
"Angelos",
""
],
[
"Kaushal",
"Yasha",
""
],
[
"Martorano",
"Marco",
""
],
[
"van der Wel",
"Arjen",
""
],
[
"Wu",
"Po-Feng",
""
],
[
"Bezanson",
"Rachel",
""
],
[
"Bell",
"Eric F.",
""
],
[
"D'Eugenio",
"Francesco",
""
],
[
"Gallazzi",
"Anna R.",
""
],
[
"Leja",
"Joel",
""
],
[
"Zibetti",
"Stefano",
""
],
[
"Tacchella",
"Sandro",
""
]
] |
2512.10384
|
Cong Pang
|
Cong Pang, Hongtao Yu, Zixuan Chen, Lewei Lu, Xin Lou
|
Towards Fine-Grained Recognition with Large Visual Language Models: Benchmark and Optimization Strategies
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Large Vision Language Models (LVLMs) have made remarkable progress, enabling sophisticated vision-language interaction and dialogue applications. However, existing benchmarks primarily focus on reasoning tasks, often neglecting fine-grained recognition, which is crucial for practical application scenarios. To address this gap, we introduce the Fine-grained Recognition Open World (FROW) benchmark, designed for detailed evaluation of LVLMs with GPT-4o. On the basis of that, we propose a novel optimization strategy from two perspectives: \textit{data construction} and \textit{training process}, to improve the performance of LVLMs. Our dataset includes mosaic data, which combines multiple short-answer responses, and open-world data, generated from real-world questions and answers using GPT-4o, creating a comprehensive framework for evaluating fine-grained recognition in LVLMs. Experiments show that mosaic data improves category recognition accuracy by 1\% and open-world data boosts FROW benchmark accuracy by 10\%-20\% and content accuracy by 6\%-12\%. Meanwhile, incorporating fine-grained data into the pre-training phase can improve the model's category recognition accuracy by up to 10\%. The benchmark will be available at https://github.com/pc-inno/FROW.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:48:34 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Pang",
"Cong",
""
],
[
"Yu",
"Hongtao",
""
],
[
"Chen",
"Zixuan",
""
],
[
"Lu",
"Lewei",
""
],
[
"Lou",
"Xin",
""
]
] |
2512.10385
|
David Mass
|
Tali Kaufman and David Mass
|
Improved Small Set Expansion in High Dimensional Expanders
| null | null | null | null |
cs.CC math.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Small set expansion in high dimensional expanders is of great importance, e.g., towards proving cosystolic expansion, local testability of codes and constructions of good quantum codes.
In this work we improve upon the state of the art results of small set expansion in high dimensional expanders. Our improvement is either on the expansion quality or on the size of sets for which expansion is guaranteed.
One line of previous works [KM22, DD24] has obtained weak expansion for small sets, which is sufficient for deducing cosystolic expansion of one dimension below. We improve upon their result by showing strong expansion for small sets.
Another line of works [KKL14, EK16, KM21] has shown strong expansion for small sets. However, they obtain it only for very small sets. We get an exponential improvement on the size of sets for which expansion is guaranteed by these prior works.
Interestingly, our result is obtained by bridging between these two lines of works. The works of [KM22, DD24] use global averaging operators in order to obtain expansion for larger sets. However, their method could be utilized only on sets that are cocycle-like. We show how to combine these global averaging operators with ideas from the so-called ``fat machinery'' of [KKL14, EK16, KM21] in order to apply them for general sets.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:48:52 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Kaufman",
"Tali",
""
],
[
"Mass",
"David",
""
]
] |
2512.10386
|
Ge Zhang
|
Ge Zhang, Chunyang Wang, Bo Xiao, Xuelian Liu, Bin Liu
|
Adaptive Dual-Weighted Gravitational Point Cloud Denoising Method
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
High-quality point cloud data is a critical foundation for tasks such as autonomous driving and 3D reconstruction. However, LiDAR-based point cloud acquisition is often affected by various disturbances, resulting in a large number of noise points that degrade the accuracy of subsequent point cloud object detection and recognition. Moreover, existing point cloud denoising methods typically sacrifice computational efficiency in pursuit of higher denoising accuracy, or, conversely, improve processing speed at the expense of preserving object boundaries and fine structural details, making it difficult to simultaneously achieve high denoising accuracy, strong edge preservation, and real-time performance. To address these limitations, this paper proposes an adaptive dual-weight gravitational-based point cloud denoising method. First, an octree is employed to perform spatial partitioning of the global point cloud, enabling parallel acceleration. Then, within each leaf node, adaptive voxel-based occupancy statistics and k-nearest neighbor (kNN) density estimation are applied to rapidly remove clearly isolated and low-density noise points, thereby reducing the effective candidate set. Finally, a gravitational scoring function that combines density weights with adaptive distance weights is constructed to finely distinguish noise points from object points. Experiments conducted on the Stanford 3D Scanning Repository, the Canadian Adverse Driving Conditions (CADC) dataset, and in-house FMCW LiDAR point clouds acquired in our laboratory demonstrate that, compared with existing methods, the proposed approach achieves consistent improvements in F1, PSNR, and Chamfer Distance (CD) across various noise conditions while reducing the single-frame processing time, thereby validating its high accuracy, robustness, and real-time performance in multi-noise scenarios.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:49:28 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Zhang",
"Ge",
""
],
[
"Wang",
"Chunyang",
""
],
[
"Xiao",
"Bo",
""
],
[
"Liu",
"Xuelian",
""
],
[
"Liu",
"Bin",
""
]
] |
2512.10387
|
Te Ba
|
Te Ba and Ze Zhou
|
A gradient descent algorithm for computing circle patterns
|
7 pages, 1 figure
| null | null | null |
cs.CG math.CO math.MG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a new algorithm for generating planar circle patterns. The algorithm employs gradient descent and conjugate gradient method to compute circle radii and centers separately. Compared with existing algorithms, the proposed method is more efficient in computing centers of circles and is applicable for realizing circle patterns with possible obtuse overlap angles.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:50:40 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Ba",
"Te",
""
],
[
"Zhou",
"Ze",
""
]
] |
2512.10388
|
Ziwei Liu
|
Ziwei Liu, Yejing Wang, Qidong Liu, Zijian Zhang, Chong Chen, Wei Huang, Xiangyu Zhao
|
The Best of the Two Worlds: Harmonizing Semantic and Hash IDs for Sequential Recommendation
| null | null | null | null |
cs.IR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Conventional Sequential Recommender Systems (SRS) typically assign unique Hash IDs (HID) to construct item embeddings. These HID embeddings effectively learn collaborative information from historical user-item interactions, making them vulnerable to situations where most items are rarely consumed (the long-tail problem). Recent methods that incorporate auxiliary information often suffer from noisy collaborative sharing caused by co-occurrence signals or semantic homogeneity caused by flat dense embeddings. Semantic IDs (SIDs), with their capability of code sharing and multi-granular semantic modeling, provide a promising alternative. However, the collaborative overwhelming phenomenon hinders the further development of SID-based methods. The quantization mechanisms commonly compromise the uniqueness of identifiers required for modeling head items, creating a performance seesaw between head and tail items. To address this dilemma, we propose \textbf{\name}, a novel framework that harmonizes the SID and HID. Specifically, we devise a dual-branch modeling architecture that enables the model to capture both the multi-granular semantics within SID while preserving the unique collaborative identity of HID. Furthermore, we introduce a dual-level alignment strategy that bridges the two representations, facilitating knowledge transfer and supporting robust preference modeling. Extensive experiments on three real-world datasets show that \name~ effectively balances recommendation quality for both head and tail items while surpassing the existing baselines. The implementation code can be found online\footnote{https://github.com/ziwliu8/H2Rec}.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:50:53 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Ziwei",
""
],
[
"Wang",
"Yejing",
""
],
[
"Liu",
"Qidong",
""
],
[
"Zhang",
"Zijian",
""
],
[
"Chen",
"Chong",
""
],
[
"Huang",
"Wei",
""
],
[
"Zhao",
"Xiangyu",
""
]
] |
2512.10389
|
Andrei Leonidov
|
Kovalenko Aleksandr, Andrey Leonidov
|
The $k$-flip Ising game
|
31 pages, 15 figures
| null | null | null |
cs.GT cond-mat.stat-mech physics.soc-ph
|
http://creativecommons.org/licenses/by/4.0/
|
A partially parallel dynamical noisy binary choice (Ising) game in discrete time of $N$ players on complete graphs with $k$ players having a possibility of changing their strategies at each time moment called $k$-flip Ising game is considered. Analytical calculation of the transition matrix of game as well as the first two moments of the distribution of $\varphi=N^+/N$, where $N^+$ is a number of players adhering to one of the two strategies, is presented. First two moments of the first hitting time distribution for sample trajectories corresponding to transition from a metastable and unstable states to a stable one are considered. A nontrivial dependence of these moments on $k$ for the decay of a metastable state is discussed. A presence of the minima at certain $k^*$ is attributed to a competition between $k$-dependent diffusion and restoring forces.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:51:36 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Aleksandr",
"Kovalenko",
""
],
[
"Leonidov",
"Andrey",
""
]
] |
2512.10390
|
Vijay Prakash S
|
Vijay Prakash S
|
Fitting magnetization data using continued fraction of straight lines
|
17 pages, 12 figures, 4 tables
| null | null | null |
cs.LG cond-mat.mtrl-sci physics.class-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Magnetization of a ferromagnetic substance in response to an externally applied magnetic field increases with the strength of the field. This is because at the microscopic level, magnetic moments in certain regions or domains of the substance increasingly align with the applied field, while the amount of misaligned domains decreases. The alignment of such magnetic domains with an applied magnetic field forms the physical basis for the nonlinearity of magnetization. In this paper, the nonlinear function is approximated as a combination of continued fraction of straight lines. The resulting fit is used to interpret the nonlinear behavior in both growing and shrinking magnetic domains. The continued fraction of straight lines used here is an algebraic expression which can be used to estimate parameters using nonlinear regression.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:57:17 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"S",
"Vijay Prakash",
""
]
] |
2512.10391
|
Liberato Manna
|
Stefano Toso, Derek Dardzinski, Liberato Manna, Noa Marom
|
Structure Prediction of Ionic Epitaxial Interfaces with Ogre Demonstrated for Colloidal Heterostructures of Lead Halide Perovskites
|
103 pages, 48 figures
|
ACS Nano 2025, 19, 5, 5326-5341
|
10.1021/acsnano.4c12713
| null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Colloidal epitaxial heterostructures are nanoparticles composed of two different materials connected at an interface, which can exhibit properties different from those of their individual components. Combining dissimilar materials offers opportunities to create several functional heterostructures. Yet, assessing structural compatibility (the main prerequisite for epitaxial growth) is challenging when pairing complex materials with different lattice parameters/crystal structures. This complicates both the selection of target heterostructures for synthesis and the assignment of interface models when new heterostructures are obtained. Here, we demonstrate Ogre as a powerful tool to accelerate the design and characterization of colloidal heterostructures. To this end we implemented developments tailored for the efficient prediction of epitaxial interfaces between ionic/polar materials, which encompass most colloidal semiconductors. These include pre-screening candidate models based on charge balance at the interface and using a classical potential for fast energy evaluations, with parameters automatically extracted from the input bulk structures. These developments are validated for CsPbBr3/Pb4S3Br2 heterostructures, where Ogre produces interface models in agreement with density functional theory and experiments. Furthermore, we use Ogre to rationalize the templating effect of CsPbCl3 on the growth of lead sulfochlorides, where perovskite seeds induce the formation of Pb4S3Cl2 rather than Pb3S2Cl2 due to better epitaxial compatibility. Combining Ogre simulations with experimental data enables us to unravel the structure and composition of the hitherto unsolved CsPbBr3/BixPbySz interface and assign a structure to many other reported metal halide/oxide based interfaces. The Ogre package is available on GitHub or via the OgreInterface desktop application, available for Windows, Linux and Mac.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:57:27 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Toso",
"Stefano",
""
],
[
"Dardzinski",
"Derek",
""
],
[
"Manna",
"Liberato",
""
],
[
"Marom",
"Noa",
""
]
] |
2512.10392
|
Sungjun Seo
|
Sungjun Seo and Kooktae Lee
|
Collision-Aware Density-Driven Control of Multi-Agent Systems via Control Barrier Functions
|
6 pages, 7 figures, Accepted for presentation at the 5th Modeling, Estimation and Control Conference (MECC 2025), Pittsburgh, USA
| null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper tackles the problem of safe and efficient area coverage using a multi-agent system operating in environments with obstacles. Applications such as environmental monitoring and search and rescue require robot swarms to cover large domains under resource constraints, making both coverage efficiency and safety essential. To address the efficiency aspect, we adopt the Density-Driven Control (D$^2$C) framework, which uses optimal transport theory to steer agents according to a reference distribution that encodes spatial coverage priorities. To ensure safety, we incorporate Control Barrier Functions (CBFs) into the framework. While CBFs are commonly used for collision avoidance, we extend their applicability by introducing obstacle-specific formulations for both circular and rectangular shapes. In particular, we analytically derive a unit normal vector based on the agent's position relative to the nearest face of a rectangular obstacle, improving safety enforcement in environments with non-smooth boundaries. Additionally, a velocity-dependent term is incorporated into the CBF to enhance collision avoidance. Simulation results validate the proposed method by demonstrating smoother navigation near obstacles and more efficient area coverage than the existing method, while still ensuring collision-free operation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:57:47 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Seo",
"Sungjun",
""
],
[
"Lee",
"Kooktae",
""
]
] |
2512.10393
|
Guoqiang Chen
|
Guoqiang Chen, Lingyun Ying, Ziyang Song, Daguang Liu, Qiang Wang, Zhiqi Wang, Li Hu, Shaoyin Cheng, Weiming Zhang, Nenghai Yu
|
Cross-modal Retrieval Models for Stripped Binary Analysis
| null | null | null | null |
cs.SE cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
LLM-agent based binary code analysis has demonstrated significant potential across a wide range of software security scenarios, including vulnerability detection, malware analysis, etc. In agent workflow, however, retrieving the positive from thousands of stripped binary functions based on user query remains under-studied and challenging, as the absence of symbolic information distinguishes it from source code retrieval. In this paper, we introduce, BinSeek, the first two-stage cross-modal retrieval framework for stripped binary code analysis. It consists of two models: BinSeekEmbedding is trained on large-scale dataset to learn the semantic relevance of the binary code and the natural language description, furthermore, BinSeek-Reranker learns to carefully judge the relevance of the candidate code to the description with context augmentation. To this end, we built an LLM-based data synthesis pipeline to automate training construction, also deriving a domain benchmark for future research. Our evaluation results show that BinSeek achieved the state-of-the-art performance, surpassing the the same scale models by 31.42% in Rec@3 and 27.17% in MRR@3, as well as leading the advanced general-purpose models that have 16 times larger parameters.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:58:10 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Chen",
"Guoqiang",
""
],
[
"Ying",
"Lingyun",
""
],
[
"Song",
"Ziyang",
""
],
[
"Liu",
"Daguang",
""
],
[
"Wang",
"Qiang",
""
],
[
"Wang",
"Zhiqi",
""
],
[
"Hu",
"Li",
""
],
[
"Cheng",
"Shaoyin",
""
],
[
"Zhang",
"Weiming",
""
],
[
"Yu",
"Nenghai",
""
]
] |
2512.10394
|
Weifan Guan
|
Weifan Guan, Huasen Xi, Chenxiao Zhang, Aosheng Li, Qinghao Hu, Jian Cheng
|
RoboNeuron: A Modular Framework Linking Foundation Models and ROS for Embodied AI
| null | null | null | null |
cs.RO cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Current embodied AI systems face severe engineering impediments, primarily characterized by poor cross-scenario adaptability, rigid inter-module coupling, and fragmented inference acceleration. To overcome these limitations, we propose RoboNeuron, a universal deployment framework for embodied intelligence. RoboNeuron is the first framework to deeply integrate the cognitive capabilities of Large Language Models (LLMs) and Vision-Language-Action (VLA) models with the real-time execution backbone of the Robot Operating System (ROS). We utilize the Model Context Protocol (MCP) as a semantic bridge, enabling the LLM to dynamically orchestrate underlying robotic tools. The framework establishes a highly modular architecture that strictly decouples sensing, reasoning, and control by leveraging ROS's unified communication interfaces. Crucially, we introduce an automated tool to translate ROS messages into callable MCP functions, significantly streamlining development. RoboNeuron significantly enhances cross-scenario adaptability and component flexibility, while establishing a systematic platform for horizontal performance benchmarking, laying a robust foundation for scalable real-world embodied applications.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 07:58:19 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Guan",
"Weifan",
""
],
[
"Xi",
"Huasen",
""
],
[
"Zhang",
"Chenxiao",
""
],
[
"Li",
"Aosheng",
""
],
[
"Hu",
"Qinghao",
""
],
[
"Cheng",
"Jian",
""
]
] |
2512.10395
|
Hiroto Tanaka
|
Hiroto Tanaka and Youichi Yanase
|
Surface acoustic wave-driven valley current generation in intervalley coherent states
|
7 pages, 4 figures
| null | null | null |
cond-mat.mes-hall
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent experiments have reported valley-gauge-symmetry-broken phases, identified as intervalley coherent (IVC) states. Exploration of anomalous responses, particularly those analogous to superconductivity, has become an urgent theoretical issue. In this study, we show that the IVC order gives rise to anomalous valley-current generation driven by surface acoustic waves (SAWs). The anomalous valley current exhibits a characteristic power-law dependence for low-frequency SAWs. Furthermore, we demonstrate by numerical analysis that the IVC order significantly enhances valley-current generation in rhombohedral graphene. These results open a pathway toward exploring exotic phenomena emerging from valley-gauge-symmetry breaking, in close analogy with gauge-symmetry breaking in superconductors.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:01:25 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Tanaka",
"Hiroto",
""
],
[
"Yanase",
"Youichi",
""
]
] |
2512.10396
|
Runhao Liu
|
Runhao Liu, Ziming Chen, You Li, Peng Zhang
|
Robust Crop Planning under Uncertainty: Aligning Economic Optimality with Agronomic Sustainability
| null | null | null | null |
cs.CE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Long-horizon agricultural planning requires optimizing crop allocation under complex spatial heterogeneity, temporal agronomic dependencies, and multi-source environmental uncertainty. Existing approaches often treat crop interactions, such as legume-cereal complementarity, which implicitly or rely on static deterministic formulations that fail to guarantee resilience against market and climate volatility. To address these challenges, we propose a Multi-Layer Robust Crop Planning Framework (MLRCPF) that integrates spatial reasoning, temporal dynamics, and robust optimization. Specifically, we formalize crop-to-crop relationships through a structured interaction matrix embedded within the state-transition logic, and employ a distributionally robust optimization layer to mitigate worst-case risks defined by a data-driven ambiguity set. Evaluations on a real-world high-mix farming dataset from North China demonstrate the effectiveness of the proposed approach. The framework autonomously generates sustainable checkerboard rotation patterns that restore soil fertility, significantly increasing the legume planting ratio compared to deterministic baselines. Economically, it successfully resolves the trade-off between optimality and stability. These results highlight the importance of explicitly encoding domain-specific structural priors into optimization models for resilient decision-making in complex agricultural systems.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:04:57 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Liu",
"Runhao",
""
],
[
"Chen",
"Ziming",
""
],
[
"Li",
"You",
""
],
[
"Zhang",
"Peng",
""
]
] |
2512.10397
|
Omololu Akin-Ojo
|
Ezekiel Oyeniyi and Omololu Akin-Ojo
|
Excitation energies and UV-Vis absorption spectra from INDO/s+ML
|
Submitted to JCTC (ACS)
| null | null | null |
cond-mat.mes-hall physics.atm-clus physics.chem-ph physics.comp-ph quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The semi-empirical INDO/s method is popular for studies of excitation energies and absorption of molecules due to its low computational requirement, making it possible to make predictions for large systems. However, its accuracy is generally low, particularly, when compared with the typical accuracy of other methods such as time-dependent density functional theory (TDDFT). Here, we present machine learning (ML) models that correct the INDO/s results with negligible increases in the amount of computing resources needed. While INDO/s excitations energies have an average error of about 1.1 eV relative to TDDFT energies, the added ML corrections reduce the error to 0.2 eV. Furthermore, this combination of INDO/s and ML produces UV-Vis absorption spectra that are in good agreement with the TDDFT predictions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:05:09 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Oyeniyi",
"Ezekiel",
""
],
[
"Akin-Ojo",
"Omololu",
""
]
] |
2512.10398
|
Zhenting Qi
|
Zhaodong Wang, Zhenting Qi, Sherman Wong, Nathan Hu, Samuel Lin, Jun Ge, Erwin Gao, Yining Yang, Ben Maurer, Wenlin Chen, David Recordon, Yilun Du, Minlan Yu, Ying Zhang
|
Confucius Code Agent: An Open-sourced AI Software Engineer at Industrial Scale
| null | null | null | null |
cs.CL cs.AI cs.LG cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Real-world AI software engineering demands coding agents that can reason over massive repositories, maintain durable memory across and within long sessions, and robustly coordinate complex toolchains at test time. Existing open-source coding agents provide transparency but frequently fall short when pushed to these industrial-scale workloads, while proprietary coding agents offer strong practical performance but limited extensibility, interpretability, and controllability. We present the Confucius Code Agent (CCA), an open-sourced AI software engineer that can operate at an industrial scale. CCA is built atop the Confucius SDK, an open-sourced agent development platform designed around three complementary perspectives: Agent Experience (AX), User Experience (UX), and Developer Experience (DX). The SDK introduces a unified orchestrator with hierarchical working memory for long-context reasoning, a persistent note-taking system for cross-session continual learning, and a modular extension module for robust tool use. Moreover, a meta-agent automates the synthesis, evaluation, and refinement of agent configurations through a build-test-improve loop, enabling rapid agent development on new tasks, environments, and tool stacks. Instantiated on Confucius SDK with these mechanisms, CCA delivers strong performance on real-world software engineering tasks. On SWE-Bench-Pro, CCA achieves a state-of-the-art Resolve@1 performance of 54.3%, substantially improving over prior coding agents. Together, the Confucius SDK and CCA provide a transparent, extensible, and reproducible foundation for AI agents, bridge gaps between research prototypes and production-grade systems, and support agent development and deployment at industrial scale.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:05:58 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wang",
"Zhaodong",
""
],
[
"Qi",
"Zhenting",
""
],
[
"Wong",
"Sherman",
""
],
[
"Hu",
"Nathan",
""
],
[
"Lin",
"Samuel",
""
],
[
"Ge",
"Jun",
""
],
[
"Gao",
"Erwin",
""
],
[
"Yang",
"Yining",
""
],
[
"Maurer",
"Ben",
""
],
[
"Chen",
"Wenlin",
""
],
[
"Recordon",
"David",
""
],
[
"Du",
"Yilun",
""
],
[
"Yu",
"Minlan",
""
],
[
"Zhang",
"Ying",
""
]
] |
2512.10399
|
Lucas English
|
Lucas H. English, Sam Roberts, Stephen D. Bartlett, Andrew C. Doherty, Dominic J. Williamson
|
Ising on the donut: Regimes of topological quantum error correction from statistical mechanics
|
29 pages, comments welcome
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Utility-scale quantum computers require quantum error correcting codes with large numbers of physical qubits to achieve sufficiently low logical error rates. The performance of quantum error correction (QEC) is generally predicted through large-scale numerical simulations, used to estimate thresholds, finite-size scaling, and exponential suppression of logical errors below threshold. The connection of QEC to models from statistical mechanics provides an alternative tool for analysing QEC performance. However, predicting the behaviour of these models also requires large-scale numerical simulations, as analytic solutions are not generally known. Here we exploit an exact mapping, from a toric code under bit-flip noise that is post-selected on being syndrome free to the exactly-solvable two-dimensional Ising model on a torus, to derive an analytic solution for the logical failure rate across its full domain of physical error rates. In particular, this mapping provides closed-form expressions for the logical failure rate in four distinct regimes: the path-counting, below-threshold (ordered), near-threshold (critical), and above-threshold (disordered) regimes. Our framework places a number of familiar and long-standing numerical observations on firm theoretical ground. It also motivates explicit ans\"atze for the conventional QEC setting of non-post-selected codes whose statistical mechanics mappings involve random-bond disorder. Specifically, we introduce an effective surface tension model for the below-threshold regime, and a new scaling ansatz for the near-threshold regime, derived from an analysis of the ground state energy cost distribution. By bridging statistical mechanics theory and quantum error correction practice, our results offer a new toolkit for designing, benchmarking, and understanding topological codes beyond current computational limits.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:06:23 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"English",
"Lucas H.",
""
],
[
"Roberts",
"Sam",
""
],
[
"Bartlett",
"Stephen D.",
""
],
[
"Doherty",
"Andrew C.",
""
],
[
"Williamson",
"Dominic J.",
""
]
] |
2512.10400
|
Max Downing
|
Max Downing, Jesper Lykke Jacobsen, Rongvoram Nivesvivat, Sylvain Ribault, Hubert Saleur
|
Diagonal boundary conditions in critical loop models
|
23 pages
| null | null | null |
hep-th cond-mat.stat-mech math-ph math.MP
|
http://creativecommons.org/publicdomain/zero/1.0/
|
In critical loop models, we define diagonal boundaries as boundaries that couple to diagonal fields only. Using analytic bootstrap methods, we show that diagonal boundaries are characterised by one complex parameter, analogous to the boundary cosmological constant in Liouville theory. We determine disc 1-point functions, and write an explicit formula for disc 2-point functions as infinite combinations of conformal blocks. For a discrete subset of values of the boundary parameter, the boundary spectrum becomes discrete, and made of degenerate representations. In such cases, we check our results by numerically bootstrapping disc 2-point functions. We sketch the interpretation of diagonal and non-diagonal boundaries in lattice loop models. In particular, a loop can neither end on a diagonal boundary, nor change weight when it touches it. In bulk-to-boundary OPEs, numbers of legs can be conserved, or increase by even numbers.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:06:49 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Downing",
"Max",
""
],
[
"Jacobsen",
"Jesper Lykke",
""
],
[
"Nivesvivat",
"Rongvoram",
""
],
[
"Ribault",
"Sylvain",
""
],
[
"Saleur",
"Hubert",
""
]
] |
2512.10401
|
Zheng Zhao
|
Jennifer Rosina Andersson and Zheng Zhao
|
Diffusion differentiable resampling
| null | null | null | null |
stat.ML cs.LG math.ST stat.TH
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper is concerned with differentiable resampling in the context of sequential Monte Carlo (e.g., particle filtering). We propose a new informative resampling method that is instantly pathwise differentiable, based on an ensemble score diffusion model. We prove that our diffusion resampling method provides a consistent estimate to the resampling distribution, and we show by experiments that it outperforms the state-of-the-art differentiable resampling methods when used for stochastic filtering and parameter estimation.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:08:55 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Andersson",
"Jennifer Rosina",
""
],
[
"Zhao",
"Zheng",
""
]
] |
2512.10402
|
Zhou Feng
|
Zhou Feng, Jiahao Chen, Chunyi Zhou, Yuwen Pu, Tianyu Du, Jinbao Li, Jianhai Chen, Shouling Ji
|
The Eminence in Shadow: Exploiting Feature Boundary Ambiguity for Robust Backdoor Attacks
|
Accepted by KDD2026 Cycle 1 Research Track
| null |
10.1145/3770854.3780322
| null |
cs.LG cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Deep neural networks (DNNs) underpin critical applications yet remain vulnerable to backdoor attacks, typically reliant on heuristic brute-force methods. Despite significant empirical advancements in backdoor research, the lack of rigorous theoretical analysis limits understanding of underlying mechanisms, constraining attack predictability and adaptability. Therefore, we provide a theoretical analysis targeting backdoor attacks, focusing on how sparse decision boundaries enable disproportionate model manipulation. Based on this finding, we derive a closed-form, ambiguous boundary region, wherein negligible relabeled samples induce substantial misclassification. Influence function analysis further quantifies significant parameter shifts caused by these margin samples, with minimal impact on clean accuracy, formally grounding why such low poison rates suffice for efficacious attacks. Leveraging these insights, we propose Eminence, an explainable and robust black-box backdoor framework with provable theoretical guarantees and inherent stealth properties. Eminence optimizes a universal, visually subtle trigger that strategically exploits vulnerable decision boundaries and effectively achieves robust misclassification with exceptionally low poison rates (< 0.1%, compared to SOTA methods typically requiring > 1%). Comprehensive experiments validate our theoretical discussions and demonstrate the effectiveness of Eminence, confirming an exponential relationship between margin poisoning and adversarial boundary manipulation. Eminence maintains > 90% attack success rate, exhibits negligible clean-accuracy loss, and demonstrates high transferability across diverse models, datasets and scenarios.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:09:07 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Feng",
"Zhou",
""
],
[
"Chen",
"Jiahao",
""
],
[
"Zhou",
"Chunyi",
""
],
[
"Pu",
"Yuwen",
""
],
[
"Du",
"Tianyu",
""
],
[
"Li",
"Jinbao",
""
],
[
"Chen",
"Jianhai",
""
],
[
"Ji",
"Shouling",
""
]
] |
2512.10403
|
Hao Liang
|
Tianyu Guo, Hongyu Chen, Hao Liang, Meiyi Qiang, Bohan Zeng, Linzhuang Sun, Bin Cui, Wentao Zhang
|
BRACE: A Benchmark for Robust Audio Caption Quality Evaluation
| null | null | null | null |
cs.SD cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Automatic audio captioning is essential for audio understanding, enabling applications such as accessibility and content indexing. However, evaluating the quality of audio captions remains a major challenge, especially in reference-free settings where high-quality ground-truth captions are unavailable. While CLAPScore is currently the most widely used reference-free Audio Caption Evaluation Metric(ACEM), its robustness under diverse conditions has not been systematically validated.
To address this gap, we introduce BRACE, a new benchmark designed to evaluate audio caption alignment quality in a reference-free setting. BRACE is primarily designed for assessing ACEMs, and can also be extended to measure the modality alignment abilities of Large Audio Language Model(LALM). BRACE consists of two sub-benchmarks: BRACE-Main for fine-grained caption comparison and BRACE-Hallucination for detecting subtle hallucinated content. We construct these datasets through high-quality filtering, LLM-based corruption, and human annotation.
Given the widespread adoption of CLAPScore as a reference-free ACEM and the increasing application of LALMs in audio-language tasks, we evaluate both approaches using the BRACE benchmark, testing CLAPScore across various CLAP model variants and assessing multiple LALMs.
Notably, even the best-performing CLAP-based ACEM achieves only a 70.01 F1-score on the BRACE-Main benchmark, while the best LALM reaches just 63.19.
By revealing the limitations of CLAP models and LALMs, our BRACE benchmark offers valuable insights into the direction of future research.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:09:24 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Guo",
"Tianyu",
""
],
[
"Chen",
"Hongyu",
""
],
[
"Liang",
"Hao",
""
],
[
"Qiang",
"Meiyi",
""
],
[
"Zeng",
"Bohan",
""
],
[
"Sun",
"Linzhuang",
""
],
[
"Cui",
"Bin",
""
],
[
"Zhang",
"Wentao",
""
]
] |
2512.10404
|
Laszlo Wolf
|
Laszlo Wolf, Geoff L. Brennecka, Vladan Stevanovi\'c
|
First-principles study of the phase competition, mechanical and piezoelectric properties of pseudo-binary (SiC)(AlN) alloy
|
8 pages, 5 figures
| null | null | null |
cond-mat.mtrl-sci cond-mat.dis-nn
|
http://creativecommons.org/licenses/by/4.0/
|
The ongoing search for new piezoelectric materials offering adequate balance between piezoelectric response and other application-relevant properties has lead to the investigation of various alloy systems. In this work we study the alloy of the widely used AlN with SiC for their relative abundance, current use in other electronics applications and expected phase competition between wurtzite and other polymorphs, the kind of which has lead to some of the most interesting results notably between AlN and ScN. Here the pseudo-binary (SiC)(AlN) alloy is studied from first-principles over the entire composition range. Relevant crystalline phases are identified using the First-Principles Random Structure Sampling approach which, in accordance with previous bulk experiments, finds wurtzite, zincblende and rhombohedral phases to be the only statistically relevant phases of the alloy. Further study of these phases is done through Special Quasi-random Structures (SQS) and, in the case of the wurtzite phase, predictions of the stiffness, piezoelectric and dielectric tensors. Analysis of these tensors is done through the scope of a Bulk AcousticWave (BAW) filter application, where trends and trade-offs between the c-axis acoustic velocity and piezoelectric response enable identification of relevant compositions.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:10:24 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Wolf",
"Laszlo",
""
],
[
"Brennecka",
"Geoff L.",
""
],
[
"Stevanović",
"Vladan",
""
]
] |
2512.10405
|
Peixin Qin
|
Zhiyuan Duan, Peixin Qin, Chengyan Zhong, Shaoxuan Zhang, Li Liu, Guojian Zhao, Xiaoning Wang, Hongyu Chen, Ziang Meng, Jingyu Li, Sixu Jiang, Xiaoyang Tan, Qiong Wu, Yu Liu, Zhiqi Liu
|
Electric-Field-Controlled Altermagnetic Transition for Neuromorphic Computing
|
42 pages, 13 figures, published online at Journal of the American Chemical Society
| null |
10.1021/jacs.5c15276
| null |
cond-mat.mtrl-sci cond-mat.mes-hall physics.app-ph
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Altermagnets represent a novel magnetic phase with transformative potential for ultrafast spintronics, yet efficient control of their magnetic states remains challenging. We demonstrate an ultra-low-power electric-field control of altermagnetism in MnTe through strain-mediated coupling in MnTe/PMN-PT heterostructures with negligible Joule heating. Application of +6 kV/cm electric fields induces piezoelectric strain in PMN-PT, modulating the N\'eel temperature from 310 to 328 K. As a result, around the magnetic phase transition, the altermagnetic spin splitting of MnTe is reversibly switched "on" and "off" by the electric fields. Meanwhile, the piezoelectric strain generates lattice distortions and magnetic structure changes in MnTe, enabling up to 9.7% resistance modulation around the magnetic phase transition temperature. Leveraging this effect, we implement programmable resistance states in a Hopfield neuromorphic network, achieving 100% pattern recognition accuracy at <=40% noise levels. This approach establishes the electric-field control as a low-power strategy for altermagnetic manipulation while demonstrating the viability of altermagnetic materials for energy-efficient neuromorphic computing beyond conventional charge-based architectures.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:14:00 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Duan",
"Zhiyuan",
""
],
[
"Qin",
"Peixin",
""
],
[
"Zhong",
"Chengyan",
""
],
[
"Zhang",
"Shaoxuan",
""
],
[
"Liu",
"Li",
""
],
[
"Zhao",
"Guojian",
""
],
[
"Wang",
"Xiaoning",
""
],
[
"Chen",
"Hongyu",
""
],
[
"Meng",
"Ziang",
""
],
[
"Li",
"Jingyu",
""
],
[
"Jiang",
"Sixu",
""
],
[
"Tan",
"Xiaoyang",
""
],
[
"Wu",
"Qiong",
""
],
[
"Liu",
"Yu",
""
],
[
"Liu",
"Zhiqi",
""
]
] |
2512.10406
|
Rong Lin
|
Rong Lin, Fabio Bacchini, Jiansen He, Luca Pezzini, Jingyu Peng
|
Generation of proton beams at switchback boundary-like rotational discontinuities in the solar wind
|
7 figures
| null | null | null |
physics.space-ph astro-ph.SR physics.plasm-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Alfv\'enic rotational discontinuities (RDs) are abundant in the inner heliosphere and can be used to model the boundary of switchbacks, i.e. Alfv\'enic magnetic kinks. To investigate the effects of RDs on proton kinetics, we model a pair of switchback-boundary-like RDs with a hybrid Particle-In-Cell (PIC) approach in a 2D system. We find that, at one of the boundary RDs, a significant population of protons remains trapped over long times, creating a secondary beam-like component with temperature anisotropy $T_\perp/T_\|\gtrsim4$ in the proton velocity distribution function that excites ion cyclotron waves within the downstream portion of the transition layer. Further analysis suggests that the static electric field in the vicinity of the RD is the key factor in trapping the protons. This work indicates that switchback boundaries could represent a viable environment for the creation of proton beams in the heliosphere; it also highlights the need to investigate RD sub-structures, especially the embedded current systems of interplanetary RDs. Finally, this paper underscores the importance of high-resolution observations of the solar wind velocity distributions around RDs.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:16:22 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Lin",
"Rong",
""
],
[
"Bacchini",
"Fabio",
""
],
[
"He",
"Jiansen",
""
],
[
"Pezzini",
"Luca",
""
],
[
"Peng",
"Jingyu",
""
]
] |
2512.10407
|
Christian Soize
|
Christian Soize
|
Supervised Learning of Random Neural Architectures Structured by Latent Random Fields on Compact Boundaryless Multiply-Connected Manifolds
|
46 pages, 12 figures
| null | null | null |
stat.ML cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
This paper introduces a new probabilistic framework for supervised learning in neural systems. It is designed to model complex, uncertain systems whose random outputs are strongly non-Gaussian given deterministic inputs. The architecture itself is a random object stochastically generated by a latent anisotropic Gaussian random field defined on a compact, boundaryless, multiply-connected manifold. The goal is to establish a novel conceptual and mathematical framework in which neural architectures are realizations of a geometry-aware, field-driven generative process. Both the neural topology and synaptic weights emerge jointly from a latent random field. A reduced-order parameterization governs the spatial intensity of an inhomogeneous Poisson process on the manifold, from which neuron locations are sampled. Input and output neurons are identified via extremal evaluations of the latent field, while connectivity is established through geodesic proximity and local field affinity. Synaptic weights are conditionally sampled from the field realization, inducing stochastic output responses even for deterministic inputs. To ensure scalability, the architecture is sparsified via percentile-based diffusion masking, yielding geometry-aware sparse connectivity without ad hoc structural assumptions. Supervised learning is formulated as inference on the generative hyperparameters of the latent field, using a negative log-likelihood loss estimated through Monte Carlo sampling from single-observation-per-input datasets. The paper initiates a mathematical analysis of the model, establishing foundational properties such as well-posedness, measurability, and a preliminary analysis of the expressive variability of the induced stochastic mappings, which support its internal coherence and lay the groundwork for a broader theory of geometry-driven stochastic learning.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:17:12 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Soize",
"Christian",
""
]
] |
2512.10408
|
Qiyue Sun
|
Qiyue Sun, Tailin Chen, Yinghui Zhang, Yuchen Zhang, Jiangbei Yue, Jianbo Jiao, Zeyu Fu
|
MultiHateLoc: Towards Temporal Localisation of Multimodal Hate Content in Online Videos
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The rapid growth of video content on platforms such as TikTok and YouTube has intensified the spread of multimodal hate speech, where harmful cues emerge subtly and asynchronously across visual, acoustic, and textual streams. Existing research primarily focuses on video-level classification, leaving the practically crucial task of temporal localisation, identifying when hateful segments occur, largely unaddressed. This challenge is even more noticeable under weak supervision, where only video-level labels are available, and static fusion or classification-based architectures struggle to capture cross-modal and temporal dynamics. To address these challenges, we propose MultiHateLoc, the first framework designed for weakly-supervised multimodal hate localisation. MultiHateLoc incorporates (1) modality-aware temporal encoders to model heterogeneous sequential patterns, including a tailored text-based preprocessing module for feature enhancement; (2) dynamic cross-modal fusion to adaptively emphasise the most informative modality at each moment and a cross-modal contrastive alignment strategy to enhance multimodal feature consistency; (3) a modality-aware MIL objective to identify discriminative segments under video-level supervision. Despite relying solely on coarse labels, MultiHateLoc produces fine-grained, interpretable frame-level predictions. Experiments on HateMM and MultiHateClip show that our method achieves state-of-the-art performance in the localisation task.
|
[
{
"version": "v1",
"created": "Thu, 11 Dec 2025 08:18:22 GMT"
}
] | 2025-12-12T00:00:00
|
[
[
"Sun",
"Qiyue",
""
],
[
"Chen",
"Tailin",
""
],
[
"Zhang",
"Yinghui",
""
],
[
"Zhang",
"Yuchen",
""
],
[
"Yue",
"Jiangbei",
""
],
[
"Jiao",
"Jianbo",
""
],
[
"Fu",
"Zeyu",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.