id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2503.23604 | Jamie Dunsmore | J. Dunsmore, L. M. Arthur, R. S. Kemp | The optimum mix of storage and backup in a highly renewable, highly reliable European electricity grid | null | null | null | null | econ.GN q-fin.EC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We estimate the variability of solar and wind energy generation potential in Europe over a 43 year period between 1980-2022 with the MERRA-2 reanalysis datasets. We compare the estimated supply potential to hourly demand data from 36 European countries to calculate the reliability of a highly renewable electricity grid in Europe. We find that in cost-optimised scenarios with onshore wind, solar and storage, but no natural gas, reliably meeting the last 1% of demand represents 36% of the entire system cost. Including small amounts of dispatchable natural gas drastically reduces the cost of a renewable, highly reliable grid: overall system costs fall by 31% when just 1% of total generation is permitted to come from natural gas. Large renewable overbuild factors (greater than $\times$4 peak demand) are required to meet modern grid reliability standards in all scenarios, and wind, rather than solar, dominates the generation mix. | [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 22:04:30 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:19:20 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dunsmore",
"J.",
""
],
[
"Arthur",
"L. M.",
""
],
[
"Kemp",
"R. S.",
""
]
] |
2503.23809 | Eric Bonnet | E. Bonnet, B. Borderie, R. Bougault, A. Chbihi, Q. Fable, J.D. Frankland, D. Gruyer, M. La Commara, A. Le F\`evre, N. Le Neindre, I. Lombardo, J. {\L}ukasik, M. P\^arlog, G. Verde, J.P. Wieleczko | Nuclear clustering process in heavy-ion collisions : experimental constraints on the low-temperature region of the QCD phase diagram | 10 pages, 3 figure panels, accepted for publication in PRC | null | null | null | nucl-th hep-ph nucl-ex | http://creativecommons.org/licenses/by/4.0/ | In this article, we study the production of Hydrogen and Helium isotopes in heavy-ion collisions in the incident energy range between 80 and 150 MeV/nucleon. We compare their inclusive multiplicities emitted in the transverse plane of the reaction with the predictions given by the thermal model. As a first step, we validate the choice of this approach to describe the experimental measurements. We also show that the transient states have to be explicitly taken into account for a good statistical description of the experimental multiplicities. From the thermodynamical parameter values obtained we complete the existing database built with the use of thermal-statistical models to reproduce particle production in the (ultra-)relativistic-energy measurements. We then proposed a new constraint on the so-called freeze-out region in the temperature (T) versus baryonic chemical potential (muB) phase diagram of the quantum chromodynamics. These new results indicate that there is a common framework to describe the hadron production and nuclear clustering processes in heavy-ion collisions. | [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 07:43:48 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:37:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bonnet",
"E.",
""
],
[
"Borderie",
"B.",
""
],
[
"Bougault",
"R.",
""
],
[
"Chbihi",
"A.",
""
],
[
"Fable",
"Q.",
""
],
[
"Frankland",
"J. D.",
""
],
[
"Gruyer",
"D.",
""
],
[
"La Commara",
"M.",
""
],
[
"Fèvre",
"A. Le",
""
],
[
"Neindre",
"N. Le",
""
],
[
"Lombardo",
"I.",
""
],
[
"Łukasik",
"J.",
""
],
[
"Pârlog",
"M.",
""
],
[
"Verde",
"G.",
""
],
[
"Wieleczko",
"J. P.",
""
]
] |
2503.23899 | Gabrielle Gaudeau | Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi gu, Zheng Yuan, Keisuke Sakaguchi, Paula Buttery | Rubrik's Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset | 10 main pages (24 appendix pages), 9 figures, accepted to ACL 2025 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The performance and usability of Large-Language Models (LLMs) are driving their use in explanation generation tasks. However, despite their widespread adoption, LLM explanations have been found to be unreliable, making it difficult for users to distinguish good from bad explanations. To address this issue, we present Rubrik's CUBE, an education-inspired rubric and a dataset of 26k explanations, written and later quality-annotated using the rubric by both humans and six open- and closed-source LLMs. The CUBE dataset focuses on two reasoning and two language tasks, providing the necessary diversity for us to effectively test our proposed rubric. Using Rubrik, we find that explanations are influenced by both task and perceived difficulty. Low quality stems primarily from a lack of conciseness in LLM-generated explanations, rather than cohesion and word choice. The full dataset, rubric, and code are available at https://github.com/RubriksCube/rubriks_cube. | [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 09:48:59 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:23:54 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Galvan-Sosa",
"Diana",
""
],
[
"Gaudeau",
"Gabrielle",
""
],
[
"Kavumba",
"Pride",
""
],
[
"Li",
"Yunmeng",
""
],
[
"gu",
"Hongyi",
""
],
[
"Yuan",
"Zheng",
""
],
[
"Sakaguchi",
"Keisuke",
""
],
[
"Buttery",
"Paula",
""
]
] |
2504.00030 | Susav Shrestha | Aayush Gautam, Susav Shrestha, Narasimha Reddy | Token-Driven GammaTune: Adaptive Calibration for Enhanced Speculative Decoding | 6 pages, 2 figures, 1 table | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speculative decoding accelerates large language model (LLM) inference by using a smaller draft model to propose tokens, which are then verified by a larger target model. However, selecting an optimal speculation length is critical for maximizing speedup while minimizing wasted computation. We introduce \textit{GammaTune} and \textit{GammaTune+}, training-free adaptive algorithms that dynamically adjust speculation length based on token acceptance rates using a heuristic-based switching mechanism. Evaluated on SpecBench across multiple tasks and model pairs, our method outperforms other heuristic-based approaches and fixed-length speculative decoding, achieving an average speedup of 15\% ($\pm$5\%) with \textit{GammaTune} and 16\% ($\pm$3\%) with \textit{GammaTune+}, while reducing performance variance. This makes \textit{GammaTune} a robust and efficient solution for real-world deployment. | [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 23:41:55 GMT"
},
{
"version": "v2",
"created": "Thu, 03 Apr 2025 12:31:40 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 06:07:17 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gautam",
"Aayush",
""
],
[
"Shrestha",
"Susav",
""
],
[
"Reddy",
"Narasimha",
""
]
] |
2504.00084 | R. K. L. Lo | Rico K. L. Lo, Leart Sabani, Vitor Cardoso | Quasinormal modes and excitation factors of Kerr black holes | 20 pages, 14 figures; for tabulated values of QNM frequencies and QNEFs, see https://the-center-of-gravity.com/resources/ and for interactive plots, see https://ricokaloklo.github.io/kerr-qnm-qnef/ | Phys. Rev. D 111, 124002 (2025) | 10.1103/PhysRevD.111.124002 | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Theoretical understanding of the characteristic oscillations of a perturbed black hole, also referred to as quasinormal modes (QNMs), is crucial to interpreting the late stage of binary black hole mergers that we now routinely observe in gravitational wave detectors. In this work, we introduce a new approach, based on the generalized Sasaki-Nakamura formalism, to compute the QNM spectra and their excitation factors (QNEFs), for scalar, electromagnetic and gravitational perturbations. Using this approach, QNM wavefunctions remain finite at the horizon and spatial infinity. Our results in general agree with previous calculations that were performed using different methods, though we further clarify that QNEFs and their scaling with the mass of the black hole depend on the spin-weight of the perturbation. We show that avoided crossing is not a general phenomenon: the real or the imaginary part of the eigenvalues can cross each other but not both simultaneously, and when crossing occurs in one part, repulsion follows in the another part. Eigenvalue repulsion, originating from branch point singularities, still plays an important role in the QNM spectra, despite the fact that the spectra depend only on the real-valued black hole angular momentum. | [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 18:00:01 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:52:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Lo",
"Rico K. L.",
""
],
[
"Sabani",
"Leart",
""
],
[
"Cardoso",
"Vitor",
""
]
] |
2504.00330 | Ghao Wang | Gehao Wang and Yuexin Yu | Stability analysis of Runge-Kutta methods for nonlinear delay-integro-differential-algebraic equations | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper is devoted to examining the stability of Runge-Kutta methods for solving nonlinear Volterra delay-integro-differential-algebraic equations (DIDAEs) with constant delay. Hybrid numerical schemes combining Runge-Kutta methods and compound quadrature rules are analyzed for nonlinear DIDAEs. Criteria for ensuring the global and asymptotic stability of the proposed schemes are established. Several numerical examples are provided to validate the theoretical findings. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 01:27:25 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 01:32:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Gehao",
""
],
[
"Yu",
"Yuexin",
""
]
] |
2504.00466 | Sergey Troshin | S.M. Troshin, N.E. Tyurin | Multiplicity distribution under the reflective scattering mode | 9 pages, 2 figures | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | Transition to the reflective scattering mode which has emerged at the highest LHC energy of $\sqrt{s}=13$ TeV results in a relative shrinkage with the energy of the impact parameter region responsible for the inelastic hadron collisions. Respective increasing role of the multiplicity fluctuations of quantum origin is emphasized. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 06:44:03 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 07:54:34 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Troshin",
"S. M.",
""
],
[
"Tyurin",
"N. E.",
""
]
] |
2504.00731 | Dhanika Mahipala | Dhanika Mahipala, Trym Tengesdal, B{\o}rge Rokseth, Tor Arne Johansen | Design and Validation of an Intention-Aware Probabilistic Framework for Trajectory Prediction: Integrating COLREGS, Grounding Hazards, and Planned Routes | null | null | 10.1016/j.oceaneng.2025.121564 | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Collision avoidance capability is an essential component in an autonomous vessel navigation system. To this end, an accurate prediction of dynamic obstacle trajectories is vital. Traditional approaches to trajectory prediction face limitations in generalizability and often fail to account for the intentions of other vessels. While recent research has considered incorporating the intentions of dynamic obstacles, these efforts are typically based on the own-ship's interpretation of the situation. The current state-of-the-art in this area is a Dynamic Bayesian Network (DBN) model, which infers target vessel intentions by considering multiple underlying causes and allowing for different interpretations of the situation by different vessels. However, since its inception, there have not been any significant structural improvements to this model. In this paper, we propose enhancing the DBN model by incorporating considerations for grounding hazards and vessel waypoint information. The proposed model is validated using real vessel encounters extracted from historical Automatic Identification System (AIS) data. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 12:43:33 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mahipala",
"Dhanika",
""
],
[
"Tengesdal",
"Trym",
""
],
[
"Rokseth",
"Børge",
""
],
[
"Johansen",
"Tor Arne",
""
]
] |
2504.00816 | Yeqi Fang | Yeqi Fang, Rong Zhou | Two-stage deep learning framework for the restoration of incomplete-ring PET images | 20 pages, 7 figures | null | null | null | cs.CV physics.med-ph | http://creativecommons.org/licenses/by/4.0/ | Positron Emission Tomography (PET) is an important molecular imaging tool widely used in medicine. Traditional PET systems rely on complete detector rings for full angular coverage and reliable data collection. However, incomplete-ring PET scanners have emerged due to hardware failures, cost constraints, or specific clinical needs. Standard reconstruction algorithms often suffer from performance degradation with these systems because of reduced data completeness and geometric inconsistencies. We present a two-stage deep-learning framework that, without incorporating any time-of-flight (TOF) information, restores high-quality images from data with about 50% missing coincidences - double the loss levels previously addressed by CNN-based methods. The pipeline operates in two stages: a projection-domain Attention U-Net first predicts the missing sections of the sinogram by leveraging spatial context from neighbouring slices, after which the completed data are reconstructed with OSEM algorithm and passed to a U-Net-diffusion module that removes residual artefacts while reinstating high-frequency detail. Using 206 brain volumes from a public dataset, the result shows that our model successfully preserves most anatomical structures and tracer distribution features with PSNR of 30.92 dB and SSIM of 0.9708. We also achieve higher inference speed, thus providing an effective solution for incomplete-ring PET imaging. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 14:05:32 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Apr 2025 11:57:11 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 13:46:00 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fang",
"Yeqi",
""
],
[
"Zhou",
"Rong",
""
]
] |
2504.01462 | Alberto Rodriguez | Pilar Mart\'in Clavero, Alberto Rodr\'iguez | Characterization of the chaotic phase in the tilted Bose-Hubbard model | 10 pages, 7 figures | null | null | null | quant-ph cond-mat.quant-gas | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The chaotic phase of the tilted Bose-Hubbard model is identified as a function of energy, tilt strength and particle interaction, from the eigenstate structure and the statistical features of the energy spectrum. Our analysis reveals that the chaotic phase of the bare Bose-Hubbard Hamiltonian can actually be enhanced by the presence of a moderate tilt. We further unveil the development and scaling of the chaotic regime from the perspective of a homogeneous density configuration typically used in cold atom experiments, providing a valuable phase diagram for future theoretical and experimental studies of this system. | [
{
"version": "v1",
"created": "Wed, 02 Apr 2025 08:18:33 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:47:35 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Clavero",
"Pilar Martín",
""
],
[
"Rodríguez",
"Alberto",
""
]
] |
2504.01525 | Anzori Shalvovich Georgadze | Anzori Sh. Georgadze | Rapid Muon Tomography for Border Security | null | null | null | null | physics.app-ph physics.ins-det | http://creativecommons.org/licenses/by/4.0/ | Cosmic-ray muon tomography is a promising technique for border security applications, leveraging highly penetrating cosmic-ray muons and their interactions with various materials to generate 3D images of large and dense objects, such as shipping containers. Using scattering and absorption of muons as they pass through dense cargo materials, muon tomography provides a viable solution for customs and border security by enabling the verification of shipping container declarations and preventing illegal trafficking. In this study, we utilized Monte Carlo simulations to evaluate the effectiveness of muon tomography for cargo characterization and contraband detection in various smuggling scenarios. Our results demonstrate that muon tomography can offers a novel approach to cargo inspection, moving beyond traditional 3D image reconstruction. Instead, it analyzes muon scattering and absorption rates in real time during scanning, enabling the prompt detection of discrepancies between actual cargo contents and declared goods within just 10 to 20 seconds. This method is particularly effective for cargo consisting of uniform loads composed of a single material or product, a common practice in shipping. Unlike traditional X-ray radiography, which analyzes detailed 2D images, muon tomography begins evaluating scatter-absorption rates within the first few seconds of scanning. This early assessment enables cargo evaluation long before a statistically reliable 3D image is formed, significantly improving scanning throughput without disrupting trade flow. | [
{
"version": "v1",
"created": "Wed, 02 Apr 2025 09:12:31 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 13:01:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Georgadze",
"Anzori Sh.",
""
]
] |
2504.03980 | Roberta Mota | Roberta Mota, Ehud Sharlin, Usman Alim | Virtual Reality Lensing for Surface Approximation in Feature-driven DVR | null | null | null | null | cs.GR | http://creativecommons.org/licenses/by/4.0/ | We present a novel lens technique to support the identification of heterogeneous features in direct volume rendering (DVR) visualizations. In contrast to data-centric transfer function (TF) design, our image-driven approach enables users to specify target features directly within the visualization using deformable quadric surfaces. The lens leverages quadrics for their expressive yet simple parametrization, enabling users to sculpt feature approximations by composing multiple quadric lenses. By doing so, the lens offers greater versatility than traditional rigid-shape lenses for selecting and bringing into focus features with irregular geometry. We discuss the lens visualization and interaction design, advocating for bimanual spatial virtual reality (VR) input for reducing cognitive and physical strain. We also report findings from a pilot qualitative evaluation with a domain specialist using a public asteroid impact dataset. These insights not only shed light on the benefits and pitfalls of using deformable lenses but also suggest directions for future research. | [
{
"version": "v1",
"created": "Fri, 04 Apr 2025 22:47:05 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 21:54:52 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 20:24:24 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mota",
"Roberta",
""
],
[
"Sharlin",
"Ehud",
""
],
[
"Alim",
"Usman",
""
]
] |
2504.05154 | Geyang Guo | Geyang Guo, Tarek Naous, Hiromi Wakaki, Yukiko Nishimura, Yuki Mitsufuji, Alan Ritter, Wei Xu | CARE: Assessing the Impact of Multilingual Human Preference Learning on Cultural Awareness | 27 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Language Models (LMs) are typically tuned with human preferences to produce helpful responses, but the impact of preference tuning on the ability to handle culturally diverse queries remains understudied. In this paper, we systematically analyze how native human cultural preferences can be incorporated into the preference learning process to train more culturally aware LMs. We introduce \textbf{CARE}, a multilingual resource containing 3,490 culturally specific questions and 31.7k responses with native judgments. We demonstrate how a modest amount of high-quality native preferences improves cultural awareness across various LMs, outperforming larger generic preference data. Our analyses reveal that models with stronger initial cultural performance benefit more from alignment, leading to gaps among models developed in different regions with varying access to culturally relevant data. CARE will be made publicly available at https://github.com/Guochry/CARE. | [
{
"version": "v1",
"created": "Mon, 07 Apr 2025 14:57:06 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 05:34:48 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 05:22:27 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Guo",
"Geyang",
""
],
[
"Naous",
"Tarek",
""
],
[
"Wakaki",
"Hiromi",
""
],
[
"Nishimura",
"Yukiko",
""
],
[
"Mitsufuji",
"Yuki",
""
],
[
"Ritter",
"Alan",
""
],
[
"Xu",
"Wei",
""
]
] |
2504.05276 | Yucheng Chu | Yucheng Chu, Peng He, Hang Li, Haoyu Han, Kaiqi Yang, Yu Xue, Tingting Li, Joseph Krajcik and Jiliang Tang | Enhancing LLM-Based Short Answer Grading with Retrieval-Augmented Generation | EDM 2025 Short Paper | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Short answer assessment is a vital component of science education, allowing evaluation of students' complex three-dimensional understanding. Large language models (LLMs) that possess human-like ability in linguistic tasks are increasingly popular in assisting human graders to reduce their workload. However, LLMs' limitations in domain knowledge restrict their understanding in task-specific requirements and hinder their ability to achieve satisfactory performance. Retrieval-augmented generation (RAG) emerges as a promising solution by enabling LLMs to access relevant domain-specific knowledge during assessment. In this work, we propose an adaptive RAG framework for automated grading that dynamically retrieves and incorporates domain-specific knowledge based on the question and student answer context. Our approach combines semantic search and curated educational sources to retrieve valuable reference materials. Experimental results in a science education dataset demonstrate that our system achieves an improvement in grading accuracy compared to baseline LLM approaches. The findings suggest that RAG-enhanced grading systems can serve as reliable support with efficient performance gains. | [
{
"version": "v1",
"created": "Mon, 07 Apr 2025 17:17:41 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 21:02:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chu",
"Yucheng",
""
],
[
"He",
"Peng",
""
],
[
"Li",
"Hang",
""
],
[
"Han",
"Haoyu",
""
],
[
"Yang",
"Kaiqi",
""
],
[
"Xue",
"Yu",
""
],
[
"Li",
"Tingting",
""
],
[
"Krajcik",
"Joseph",
""
],
[
"Tang",
"Jiliang",
""
]
] |
2504.05704 | Cong Wang | Xingguo Huang, Li Han, Stewart Greenhalgh and Ru-Shan Wu | Wave propagation and scattering in time dependent media: Lippmann-Schwinger equations, multiple scattering theory, Kirchhoff Helmholtz integrals, Green's functions, reciprocity theorems and Huygens' principle | null | null | null | null | physics.optics physics.comp-ph physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | Wave scattering plays a central role for the modeling of complex wave propagation across all corners of science and engineering applications, including electromagnetic, acoustics, seismic and scattering physics. Wave control using time interfaces, where the properties of the medium through with the wave travels rapidly change in time, has opened further opportunities to control wave propagation in both space and time. For acoustic waves, studies on time modulated media have not been reported. In this context, full numerical solution of the wave equation using time interfaces is key to fully understand their potential. When applying time interfaces, the underlying physics of acoustic wave propagation and scattering and their similar roles on time and space, are still being explored. In this work, we introduce a mathematical formulation of the Lippmann-Schwinger integral equations for acoustic wave scattering when time interfaces are induced via a change of the velocity of the medium. We demonstrate that space-time duality for acoustic wave propagation with time interfaces and derive the Lippmann-Schwinger integral equations for wave scattering in time-dependent media, multiple scattering theory, Kirchhoff Helmholtz integrals, Green's functions, reciprocity theorems. We experimentally verify our theoretical derivation by studying and measuring the acoustic wave scattering in strongly scattering media. We illustrate the proposed framework and present results of acoustic wave scattering without prior knowledge of the background wave-fields. This improves the understanding of the generation and wave scattering and opens previously inaccessible research directions, potentially facilitating practical applications for acoustic, geophysical and optical imaging. | [
{
"version": "v1",
"created": "Tue, 08 Apr 2025 05:56:20 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 03:20:41 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Huang",
"Xingguo",
""
],
[
"Han",
"Li",
""
],
[
"Greenhalgh",
"Stewart",
""
],
[
"Wu",
"Ru-Shan",
""
]
] |
2504.06212 | Thomas Mulc | Thomas Mulc, Mike Anderson, Paul Cubre, Huikun Zhang, Ivy Liu, and Saket Kumar | NNN: Next-Generation Neural Networks for Marketing Measurement | The title was updated to reflect broader scope. We clarified that our method is not an MMM and emphasized its experimental nature. R2 values in Tables 1 and 2 were corrected after fixing a bug that inflated NNN scores. Notation in the unrolling section now uses 0-indexing. We noted NNN has not been tested with many channels, added an acknowledgment, and improved grammar | null | null | null | cs.LG stat.AP | http://creativecommons.org/licenses/by/4.0/ | We present NNN, an experimental Transformer-based neural network approach to marketing measurement. Unlike Marketing Mix Models (MMMs) which rely on scalar inputs and parametric decay functions, NNN uses rich embeddings to capture both quantitative and qualitative aspects of marketing and organic channels (e.g., search queries, ad creatives). This, combined with its attention mechanism, potentially enables NNN to model complex interactions, capture long-term effects, and improve sales attribution accuracy. We show that L1 regularization permits the use of such expressive models in typical data-constrained settings. Evaluating NNN on simulated and real-world data demonstrates its efficacy, particularly through considerable improvement in predictive power. In addition to marketing measurement, the NNN framework can provide valuable, complementary insights through model probing, such as evaluating keyword or creative effectiveness. | [
{
"version": "v1",
"created": "Tue, 08 Apr 2025 16:57:11 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 22:23:07 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 22:19:31 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mulc",
"Thomas",
""
],
[
"Anderson",
"Mike",
""
],
[
"Cubre",
"Paul",
""
],
[
"Zhang",
"Huikun",
""
],
[
"Liu",
"Ivy",
""
],
[
"Kumar",
"Saket",
""
]
] |
2504.06910 | Sheng Lu | Sheng Lu, Ilia Kuznetsov, Iryna Gurevych | Identifying Aspects in Peer Reviews | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Peer review is central to academic publishing, but the growing volume of submissions is straining the process. This motivates the development of computational approaches to support peer review. While each review is tailored to a specific paper, reviewers often make assessments according to certain aspects such as Novelty, which reflect the values of the research community. This alignment creates opportunities for standardizing the reviewing process, improving quality control, and enabling computational support. While prior work has demonstrated the potential of aspect analysis for peer review assistance, the notion of aspect remains poorly formalized. Existing approaches often derive aspects from review forms and guidelines, yet data-driven methods for aspect identification are underexplored. To address this gap, our work takes a bottom-up approach: we propose an operational definition of aspect and develop a data-driven schema for deriving aspects from a corpus of peer reviews. We introduce a dataset of peer reviews augmented with aspects and show how it can be used for community-level review analysis. We further show how the choice of aspects can impact downstream applications, such as LLM-generated review detection. Our results lay a foundation for a principled and data-driven investigation of review aspects, and pave the path for new applications of NLP to support peer review. | [
{
"version": "v1",
"created": "Wed, 09 Apr 2025 14:14:42 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:50:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Lu",
"Sheng",
""
],
[
"Kuznetsov",
"Ilia",
""
],
[
"Gurevych",
"Iryna",
""
]
] |
2504.07879 | Ilhan Aslan | Daniel Hove Paludan, Julie Fredsg{\aa}rd, Kasper Patrick B\"ahrentz, Ilhan Aslan, Niels van Berkel | Towards Sustainable Creativity Support: An Exploratory Study on Prompt Based Image Generation | 20 pages, 8 figures | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Creativity is a valuable human skill that has long been augmented through both analog and digital tools. Recent progress in generative AI, such as image generation, provides a disruptive technological solution to supporting human creativity further and helping humans generate solutions faster. While AI image generators can help to rapidly visualize ideas based on user prompts, the use of such AI systems has also been critiqued due to their considerable energy usage. In this paper, we report on a user study (N = 24) to understand whether energy consumption can be reduced without impeding on the tool's perceived creativity support. Our results highlight that, for example, a main effect of (image generation) condition on energy consumption, and index of creativity support per prompt but not per task, which seem mainly attributed to image quantity per prompt. We provide details of our analysis on the relation between energy usage, creativity support, and prompting behavior, including attitudes towards designing with AI and its environmental impact. | [
{
"version": "v1",
"created": "Thu, 10 Apr 2025 15:55:14 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 10:11:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Paludan",
"Daniel Hove",
""
],
[
"Fredsgård",
"Julie",
""
],
[
"Bährentz",
"Kasper Patrick",
""
],
[
"Aslan",
"Ilhan",
""
],
[
"van Berkel",
"Niels",
""
]
] |
2504.08612 | Hernan Gonzalez | Mat\'ias Brice\~no, Hern\'an A. Gonz\'alez and Alfredo P\'erez | Scalar subleading soft theorems from an infinite tower of charges | 26 pages, one figure; V2: Minor modifications. New references added | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We investigate the emergence of infinite-dimensional symmetries in the absence of gauge invariance by analyzing massless scalar theories. We construct an infinite tower of charges that arise from the subleading equations of motion at null infinity and are built from specific combinations of asymptotic field coefficients. Interestingly, these expressions are finite from the outset, requiring no holographic renormalization. By carefully analyzing the dynamics at spatial infinity, we show that this tower of surface integrals commutes with the S-matrix of the interacting model. As an application, we demonstrate that these symmetries lead to an infinite set of subleading soft relations, valid at leading order in a cubic interaction with massive scalar fields. | [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:19:50 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 04:23:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Briceño",
"Matías",
""
],
[
"González",
"Hernán A.",
""
],
[
"Pérez",
"Alfredo",
""
]
] |
2504.09132 | Matthew Webster | Matthew B. Webster, Dongheon Lee, Joonnyong Lee | Self-Supervised Autoencoder Network for Robust Heart Rate Extraction from Noisy Photoplethysmogram: Applying Blind Source Separation to Biosignal Analysis | 12 pages, 5 figures, 1 table, preprint | null | null | null | cs.LG eess.SP | http://creativecommons.org/licenses/by/4.0/ | Biosignals can be viewed as mixtures measuring particular physiological events, and blind source separation (BSS) aims to extract underlying source signals from mixtures. This paper proposes a self-supervised multi-encoder autoencoder (MEAE) to separate heartbeat-related source signals from photoplethysmogram (PPG), enhancing heart rate (HR) detection in noisy PPG data. The MEAE is trained on PPG signals from a large open polysomnography database without any pre-processing or data selection. The trained network is then applied to a noisy PPG dataset collected during the daily activities of nine subjects. The extracted heartbeat-related source signal significantly improves HR detection as compared to the original PPG. The absence of pre-processing and the self-supervised nature of the proposed method, combined with its strong performance, highlight the potential of MEAE for BSS in biosignal analysis. | [
{
"version": "v1",
"created": "Sat, 12 Apr 2025 08:47:53 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 05:48:13 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Webster",
"Matthew B.",
""
],
[
"Lee",
"Dongheon",
""
],
[
"Lee",
"Joonnyong",
""
]
] |
2504.09460 | Tieguang Zi | Shailesh Kumar, Tieguang Zi and Arpan Bhattacharyya | Imprints of extra dimensions in eccentric extreme-mass-ratio inspirals gravitational waveforms | 31 pages, 9 figures, minor changes, results remain the same, version to appear in PRD | null | null | null | gr-qc hep-th | http://creativecommons.org/licenses/by/4.0/ | Studies regarding extra-dimensions have been of great interest in modern theoretical physics, including their observational consequences from future gravitational wave (GW) observatories. In this direction, extreme-mass-ratio inspirals (EMRIs), attracting considerable interest in GW astronomy and fundamental physics, can potentially provide a useful platform for the search of extra dimensions. In this paper, we examine a rotating braneworld black hole in the context of equatorial eccentric EMRI and attempt to provide an order of magnitude analysis for the extra-dimensional parameter termed "tidal charge". We estimate GW fluxes for the dominant mode and determine the impact of the tidal charge parameter on the orbital evolution. We further evaluate the prospects of detecting such a parameter through mismatch computation. We observe a significant enhancement in the mismatch as the value of orbital eccentricity or tidal charge parameter increases; the phenomenon becomes more obvious for rapidly rotating massive black holes. Thus, the study suggests that eccentric EMRI can potentially probe the existence of extra dimensions with future low-frequency detectors such as the Laser Interferometer Space Antenna (LISA). | [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 07:15:58 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 15:52:16 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 02:07:05 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kumar",
"Shailesh",
""
],
[
"Zi",
"Tieguang",
""
],
[
"Bhattacharyya",
"Arpan",
""
]
] |
2504.09564 | Dario Kieffer | Dario Kieffer and Angelika Rohde | The weak-feature-impact effect on the NPMLE in monotone binary regression | Added lower bounds | null | null | null | math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | The nonparametric maximum likelihood estimator (NPMLE) in monotone binary regression models is studied when the impact of the features on the labels is weak. Here, weakness is colloquially understood as "close to flatness" of the feature-label relationship $x \mapsto \mathbb{P}(Y=1 | X=x)$. Statistical literature provides limit distributions of the NPMLE for the two extremal cases: If the feature-label relation is strictly monotone and sufficiently smooth, then it converges at a nonparametric rate pointwise and in $L^1$ with scaled Chernoff-type and Gaussian limit distribution, respectively, and it converges at the parametric $\sqrt{n}$-rate if the underlying relation is flat. To explore the distributional transition of the NPMLE from the nonparametric to the parametric regime, we introduce a novel mathematical scenario. New restricted minimax lower bounds and matching pointwise and $L^1$-rates of convergence of the NPMLE in the weak-feature-impact scenario together with corresponding limit distributions are derived. They are shown to exhibit an elbow and a phase transition respectively, solely characterized by the level of feature impact. | [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 13:33:15 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:16:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kieffer",
"Dario",
""
],
[
"Rohde",
"Angelika",
""
]
] |
2504.09567 | Jian Huang | Chenxuan He, Yuan Gao, Liping Zhu, and Jian Huang | Conditional Independence Test Based on Transport Maps | 48 pages | null | null | null | stat.ML cs.LG stat.ME | http://creativecommons.org/licenses/by/4.0/ | Testing conditional independence between two random vectors given a third is a fundamental and challenging problem in statistics, particularly in multivariate nonparametric settings due to the complexity of conditional structures. We propose a innovative framework for testing conditional independence using transport maps. At the population level, we show that two well-defined transport maps can transform the conditional independence test into an unconditional independence test, this substantially simplifies the problem. These transport maps are estimated from data using conditional continuous normalizing flow models. Within this framework, we derive a test statistic and prove its asymptotic validity under both the null and alternative hypotheses. A permutation-based procedure is employed to evaluate the significance of the test. We validate the proposed method through extensive simulations and real-data analysis. Our numerical studies demonstrate the practical effectiveness of the proposed method for conditional independence testing. | [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 13:38:25 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 02:17:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"He",
"Chenxuan",
""
],
[
"Gao",
"Yuan",
""
],
[
"Zhu",
"Liping",
""
],
[
"Huang",
"Jian",
""
]
] |
2504.09748 | Zachary Wegert Mr | Zachary J. Wegert, Jordi Manyer, Connor Mallon, Santiago Badia, Vivien J. Challis | Level-set topology optimisation with unfitted finite elements and automatic shape differentiation | null | null | null | null | math.OC cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | In this paper we develop automatic shape differentiation techniques for unfitted discretisations and link these to recent advances in shape calculus for unfitted methods. We extend existing analytic shape calculus results to the case where the domain boundary intersects with the boundary of the background domain. We further show that we can recover these analytic derivatives to machine precision regardless of the mesh size using the developed automatic shape differentiation techniques, drastically reducing the burden associated with the analytic derivation of these quantities. In addition, we show that we can also recover the symmetric shape Hessian. We implement these techniques for both serial and distributed computing frameworks in the Julia package GridapTopOpt and the wider Gridap ecosystem. As part of this implementation we propose a novel graph-based approach for isolated volume detection. We demonstrate the applicability of the unfitted automatic shape differentiation framework and our implementation by considering the three-dimensional minimum compliance topology optimisation of a linear elastic wheel and of a linear elastic structure in a fluid-structure interaction problem with Stokes flow. The implementation is general and allows GridapTopOpt to solve a wider range of problems on unstructured meshes without analytic calculation of shape derivatives and avoiding issues that arise when material properties are smoothed at the domain boundary. The software is open source and available at https://github.com/zjwegert/GridapTopOpt.jl. | [
{
"version": "v1",
"created": "Sun, 13 Apr 2025 22:55:35 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:14:12 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wegert",
"Zachary J.",
""
],
[
"Manyer",
"Jordi",
""
],
[
"Mallon",
"Connor",
""
],
[
"Badia",
"Santiago",
""
],
[
"Challis",
"Vivien J.",
""
]
] |
2504.10938 | Dirk Heimann | Dirk Heimann and Felix Wiebe and Tahereh Abad and Elie Mounzer and Tangyou Huang and Frank Kirchner and Shivesh Kumar | Iterative Linear Quadratic Regulator for Quantum Optimal Control | 8 pages, 4 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum optimal control for gate optimization aims to provide accurate, robust, and fast pulse sequences to achieve gate fidelities on quantum systems below the error correction threshold. Many methods have been developed and successfully applied in simulation and on quantum hardware. In this paper, we establish a connection between the iterative linear quadratic regulator and quantum optimal control by adapting it to gate optimization of quantum systems. We include constraints on the controls and their derivatives to enable smoother pulses. We achieve high-fidelity simulation results for X and cross-resonance gates on one- and two-qubit fixed-frequency transmons simulated with two and three levels. | [
{
"version": "v1",
"created": "Tue, 15 Apr 2025 07:36:32 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:36:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Heimann",
"Dirk",
""
],
[
"Wiebe",
"Felix",
""
],
[
"Abad",
"Tahereh",
""
],
[
"Mounzer",
"Elie",
""
],
[
"Huang",
"Tangyou",
""
],
[
"Kirchner",
"Frank",
""
],
[
"Kumar",
"Shivesh",
""
]
] |
2504.12002 | Armando Martino | Harry Iveson and Armando Martino and Wagner Sgobbi and Peter Wong with an appendix by Francesco Fournier-Facio | Property $R_{\infty}$ for groups with infinitely many ends | 28 pages. After submitting the first version on arXiv, we were contacted by Francesco Fournier-Facio who observed that the results could also be obtained via the use of quasimorphisms. He has written us an appendix explaining this and extending the results | null | null | null | math.GR math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that an accessible group with infinitely many ends has property $R_{\infty}$. That is, it has infinitely many twisted conjugacy classes for any twisting automorphism. We deduce that having property $R_{\infty}$ is undecidable amongst finitely presented groups.
We also show that the same is true for a wide class of relatively hyperbolic groups, filling in some of the gaps in the literature. Specifically, we show that a non-elementary, finitely presented relatively hyperbolic group with finitely generated peripheral subgroups which are not themselves relatively hyperbolic, has property $R_{\infty}$.
In an appendix, Francesco Fournier-Facio shows that a group with a non-zero Aut-invariant homogeneous quasimorphism has property $R_\infty$, which applies to many groups with hyperbolic features. | [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 11:55:40 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 07:15:03 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Iveson",
"Harry",
""
],
[
"Martino",
"Armando",
""
],
[
"Sgobbi",
"Wagner",
""
],
[
"Fournier-Facio",
"Peter Wong with an appendix by Francesco",
""
]
] |
2504.12423 | Haohan Shi | Haohan Shi, Xiyu Shi, Safak Dogan, Saif Alzubi, Tianjin Huang, Yunxiao Zhang | Benchmarking Audio Deepfake Detection Robustness in Real-world Communication Scenarios | Accepted by EUSIPCO 2025 | null | null | null | eess.AS eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing Audio Deepfake Detection (ADD) systems often struggle to generalise effectively due to the significantly degraded audio quality caused by audio codec compression and channel transmission effects in real-world communication scenarios. To address this challenge, we developed a rigorous benchmark to evaluate the performance of the ADD system under such scenarios. We introduced ADD-C, a new test dataset to evaluate the robustness of ADD systems under diverse communication conditions, including different combinations of audio codecs for compression and packet loss rates. Benchmarking three baseline ADD models on the ADD-C dataset demonstrated a significant decline in robustness under such conditions. A novel Data Augmentation (DA) strategy was proposed to improve the robustness of ADD systems. Experimental results demonstrated that the proposed approach significantly enhances the performance of ADD systems on the proposed ADD-C dataset. Our benchmark can assist future efforts towards building practical and robustly generalisable ADD systems. | [
{
"version": "v1",
"created": "Wed, 16 Apr 2025 18:44:05 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 20:05:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shi",
"Haohan",
""
],
[
"Shi",
"Xiyu",
""
],
[
"Dogan",
"Safak",
""
],
[
"Alzubi",
"Saif",
""
],
[
"Huang",
"Tianjin",
""
],
[
"Zhang",
"Yunxiao",
""
]
] |
2504.13197 | Takayuki Ishitobi | Takayuki Ishitobi | Superzone gap formation induced by ferroic orders | 2 pages, 2 figures | J. Phys. Soc. Jpn. 94, 075001 (2025) | 10.7566/JPSJ.94.075001 | null | cond-mat.str-el cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We demonstrate that a superzone gap, typically associated with antiferroic ordering, can also emerge from ferroic orders in systems with sublattice degrees of freedom. By analyzing a $p$-orbital tight-binding model on a zigzag chain, we show that a Su--Schrieffer--Heeger-type gap is induced by ferroquadrupolar or ferromagnetic order or by applying an external magnetic field. | [
{
"version": "v1",
"created": "Mon, 14 Apr 2025 11:22:50 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 05:14:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ishitobi",
"Takayuki",
""
]
] |
2504.13658 | Grigory Uskov | Grigory Uskov, Sergey Sazonov, Igor Lapshov, Alexander Mikhailov, Ekaterina Filippova, Alexander Lutovinov, Ilya Mereminskiy, Maria Mochalina, Andrey Semena, Alexey Tkachenko | SRGAJ230631.0+155633: an extremely X-ray luminous, heavily obscured, radio-loud quasar at z=0.44 discovered by SRG/ART-XC | Accepted to MNRAS, 16 pages, 14 figures | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report on a detailed study of a luminous, heavily obscured ($N_{\rm H} \sim 2 \times 10^{23}$ cm$^{-2}$), radio-loud quasar SRGAJ230631.0+155633, discovered in the 4--12 keV energy band by the Mikhail Pavlinsky ART-XC telescope aboard the SRG observatory during the first two years of its all-sky X-ray survey in 2020--2021. The object is located at $z=0.4389$ and is a type 2 AGN according to optical spectroscopy (SDSS, confirmed by DESI). We combine radio-to-X-ray data, including near-simultaneous ART-XC and Swift/XRT observations conducted in June 2023. During these follow-up observations, the source was found in a significantly fainter but still very luminous state ($L_{\rm X}=1.0^{+0.8}_{-0.3} \times 10^{45}$ erg s$^{-1}$, absorption corrected, 2--10 keV) compared to its discovery during the all-sky survey ($L_{\rm X}=6^{+6}_{-3}\times10^{45}$ erg s$^{-1}$), which indicates significant intrinsic variability on a rest-frame time scale of $\sim 1$ year. The radio data show a complex morphology with a core and two extended radio lobes, indicating a giant FRII radio galaxy. From multi-wavelength photometry and the black hole-bulge relation we infer a bolometric luminosity of $\sim 6\times10^{46}$ erg s$^{-1}$ and a black hole mass of $\sim1.4\times10^{9}\,M_\odot$, implying accretion at $\sim30$\% of the Eddington limit. SRGAJ230631.0+155633 proves to be one of the most luminous obscured quasars out to $z=0.5$. As such, it can serve as a valuable testbed for in-depth exploration of the physics of such objects, which were much more abundant in the younger Universe. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 12:38:23 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:58:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Uskov",
"Grigory",
""
],
[
"Sazonov",
"Sergey",
""
],
[
"Lapshov",
"Igor",
""
],
[
"Mikhailov",
"Alexander",
""
],
[
"Filippova",
"Ekaterina",
""
],
[
"Lutovinov",
"Alexander",
""
],
[
"Mereminskiy",
"Ilya",
""
],
[
"Mochalina",
"Maria",
""
],
[
"Semena",
"Andrey",
""
],
[
"Tkachenko",
"Alexey",
""
]
] |
2504.13677 | Andrea Santilli | Andrea Santilli, Adam Golinski, Michael Kirchhof, Federico Danieli, Arno Blaas, Miao Xiong, Luca Zappella, Sinead Williamson | Revisiting Uncertainty Quantification Evaluation in Language Models: Spurious Interactions with Response Length Bias Results | Accepted at ACL 2025 (Main) | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Uncertainty Quantification (UQ) in Language Models (LMs) is key to improving their safety and reliability. Evaluations often use metrics like AUROC to assess how well UQ methods (e.g., negative sequence probabilities) correlate with task correctness functions (e.g., ROUGE-L). We show that mutual biases--when both UQ methods and correctness functions are biased by the same factors--systematically distort evaluation. First, we formally prove that any mutual bias non-randomly skews AUROC rankings, compromising benchmark integrity. Second, we confirm this happens empirically by testing 7 widely used correctness functions, from lexical-based and embedding-based metrics to LM-as-a-judge approaches, across 4 datasets x 4 models x 8 UQ methods. Our analysis shows that length biases in correctness functions distort UQ assessments by interacting with length biases in UQ methods. We identify LM-as-a-judge methods as the least length-biased, offering a promising path for a fairer UQ evaluation. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 13:13:42 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:25:43 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Santilli",
"Andrea",
""
],
[
"Golinski",
"Adam",
""
],
[
"Kirchhof",
"Michael",
""
],
[
"Danieli",
"Federico",
""
],
[
"Blaas",
"Arno",
""
],
[
"Xiong",
"Miao",
""
],
[
"Zappella",
"Luca",
""
],
[
"Williamson",
"Sinead",
""
]
] |
2504.13865 | Fei Tang | Fei Tang, Haolei Xu, Hang Zhang, Siqi Chen, Xingyu Wu, Yongliang Shen, Wenqi Zhang, Guiyang Hou, Zeqi Tan, Yuchen Yan, Kaitao Song, Jian Shao, Weiming Lu, Jun Xiao, Yueting Zhuang | A Survey on (M)LLM-Based GUI Agents | null | null | null | null | cs.HC cs.AI cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | Graphical User Interface (GUI) Agents have emerged as a transformative paradigm in human-computer interaction, evolving from rule-based automation scripts to sophisticated AI-driven systems capable of understanding and executing complex interface operations. This survey provides a comprehensive examination of the rapidly advancing field of LLM-based GUI Agents, systematically analyzing their architectural foundations, technical components, and evaluation methodologies. We identify and analyze four fundamental components that constitute modern GUI Agents: (1) perception systems that integrate text-based parsing with multimodal understanding for comprehensive interface comprehension; (2) exploration mechanisms that construct and maintain knowledge bases through internal modeling, historical experience, and external information retrieval; (3) planning frameworks that leverage advanced reasoning methodologies for task decomposition and execution; and (4) interaction systems that manage action generation with robust safety controls. Through rigorous analysis of these components, we reveal how recent advances in large language models and multimodal learning have revolutionized GUI automation across desktop, mobile, and web platforms. We critically examine current evaluation frameworks, highlighting methodological limitations in existing benchmarks while proposing directions for standardization. This survey also identifies key technical challenges, including accurate element localization, effective knowledge retrieval, long-horizon planning, and safety-aware execution control, while outlining promising research directions for enhancing GUI Agents' capabilities. Our systematic review provides researchers and practitioners with a thorough understanding of the field's current state and offers insights into future developments in intelligent interface automation. | [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 17:58:31 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:29:13 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tang",
"Fei",
""
],
[
"Xu",
"Haolei",
""
],
[
"Zhang",
"Hang",
""
],
[
"Chen",
"Siqi",
""
],
[
"Wu",
"Xingyu",
""
],
[
"Shen",
"Yongliang",
""
],
[
"Zhang",
"Wenqi",
""
],
[
"Hou",
"Guiyang",
""
],
[
"Tan",
"Zeqi",
""
],
[
"Yan",
"Yuchen",
""
],
[
"Song",
"Kaitao",
""
],
[
"Shao",
"Jian",
""
],
[
"Lu",
"Weiming",
""
],
[
"Xiao",
"Jun",
""
],
[
"Zhuang",
"Yueting",
""
]
] |
2504.14009 | Samaporn Tinyanont | Samaporn Tinyanont, Ori D. Fox, Melissa Shahbandeh, Tea Temim, Robert Williams, Kittipong Wangnok, Armin Rest, Ryan M. Lau, Keiichi Maeda, Jacob E. Jencson, Katie Auchettl, Alexei V. Filippenko, Conor Larison, Christopher Ashall, Thomas Brink, Kyle W. Davis, Luc Dessart, Ryan J. Foley, Llu\'is Galbany, Matthew Grayling, Joel Johansson, Mansi M. Kasliwal, Zachary G. Lane, Natalie LeBaron, Dan Milisavljevic, Jeonghee Rho, Itsuki Sakon, Arkaprabha Sarangi, Tamas Szalai, Kirsty Taggart, Schuyler D. Van Dyk, Qinan Wang, Yi Yang, WeiKang Zheng, Szanna Zs\'iros | Large Cold Dust Reservoir Revealed in Transitional SN Ib 2014C by James Webb Space Telescope Mid-Infrared Spectroscopy | Published in ApJ | null | 10.3847/1538-4357/adccc0 | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Supernova (SN) 2014C is a rare transitional event that exploded as a hydrogen-poor, helium-rich Type Ib SN and subsequently interacted with a hydrogen-rich circumstellar medium (CSM) a few months post-explosion. This unique interacting object provides an opportunity to probe the mass-loss history of a stripped-envelope SN progenitor. Using the James Webb Space Telescope (JWST), we observed SN 2014C with the Mid-Infrared Instrument Medium Resolution Spectrometer at 3477 days post-explosion (rest frame), and the Near-Infrared Spectrograph Integral Field Unit at 3568 days post-explosion, covering 1.7 to 25 $\mu$m. The bolometric luminosity indicates that the SN is still interacting with the same CSM that was observed with the Spitzer Space Telescope 40--1920 days post-explosion. JWST spectra and near-contemporaneous optical and near-infrared spectra show strong [Ne II] 12.831 $\mu$m, He 1.083 $\mu$m, H$\alpha$, and forbidden oxygen ([O I] $\lambda$$\lambda$6300, 6364, [O II] $\lambda$$\lambda$7319, 7330, and [O III] $\lambda$$\lambda$4959, 5007) emission lines with asymmetric profiles, suggesting a highly asymmetric CSM. The mid-IR continuum can be explained by ~$0.036 \ M_\odot$ of carbonaceous dust at ~300 K and ~0.043 $M_\odot$ of silicate dust at ~200 K. The observed dust mass has increased tenfold since the last Spitzer observation 4 yr ago, with evidence suggesting that new grains have condensed in the cold dense shell between the forward and reverse shocks. This dust mass places SN 2014C among the dustiest SNe in the mid-IR and supports the emerging observational trend that SN explosions produce enough dust to explain the observed dust mass at high redshifts. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 18:01:54 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:37:52 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tinyanont",
"Samaporn",
""
],
[
"Fox",
"Ori D.",
""
],
[
"Shahbandeh",
"Melissa",
""
],
[
"Temim",
"Tea",
""
],
[
"Williams",
"Robert",
""
],
[
"Wangnok",
"Kittipong",
""
],
[
"Rest",
"Armin",
""
],
[
"Lau",
"Ryan M.",
""
],
[
"Maeda",
"Keiichi",
""
],
[
"Jencson",
"Jacob E.",
""
],
[
"Auchettl",
"Katie",
""
],
[
"Filippenko",
"Alexei V.",
""
],
[
"Larison",
"Conor",
""
],
[
"Ashall",
"Christopher",
""
],
[
"Brink",
"Thomas",
""
],
[
"Davis",
"Kyle W.",
""
],
[
"Dessart",
"Luc",
""
],
[
"Foley",
"Ryan J.",
""
],
[
"Galbany",
"Lluís",
""
],
[
"Grayling",
"Matthew",
""
],
[
"Johansson",
"Joel",
""
],
[
"Kasliwal",
"Mansi M.",
""
],
[
"Lane",
"Zachary G.",
""
],
[
"LeBaron",
"Natalie",
""
],
[
"Milisavljevic",
"Dan",
""
],
[
"Rho",
"Jeonghee",
""
],
[
"Sakon",
"Itsuki",
""
],
[
"Sarangi",
"Arkaprabha",
""
],
[
"Szalai",
"Tamas",
""
],
[
"Taggart",
"Kirsty",
""
],
[
"Van Dyk",
"Schuyler D.",
""
],
[
"Wang",
"Qinan",
""
],
[
"Yang",
"Yi",
""
],
[
"Zheng",
"WeiKang",
""
],
[
"Zsíros",
"Szanna",
""
]
] |
2504.14021 | Unnati Kashyap | Unnati Kashyap (TTU, USA), Thomas J. Maccarone (TTU, USA), Thomas D. Russell (INAF, Italy), Mason Ng (McGill University, Canada), Swati Ravi (MIT, USA), Eliot C. Pattie (TTU, USA), and Herman L. Marshall (MIT, USA) | X-ray and radio polarimetry of the neutron star low-mass X-ray binary 4U 1728-34 | 13 pages, 8 figures, 5 tables. Accepted for publication in APJ | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report the first X-ray and radio polarimetric results of the neutron star (NS) low-mass X-ray binary (LMXB) atoll-source 4U 1728-34 using the Imaging X-ray Polarimetry Explorer (IXPE) and Australia Telescope Compact Array (ATCA). We discovered that the X-ray source was polarized at PD = 1.9 +/- 1.0% with a polarization angle of PA = -41 +/- 16 degree. Simultaneous Neutron Star Interior Composition Explorer (NICER) observations show that the source was in a relatively hard state, marking it as the first IXPE observation of an NS atoll source in the hard state. We do not detect any significant linear polarization (LP) in the radio band, with a 3-sigma upper limit of 2% at 5.5 GHz and 1.8% at 9 GHz. Combining the radio datasets provides the deepest upper limits on the radio polarization at < 1.5% on the linear and circular polarization (measured at 7.25 GHz). The X-ray polarimetric results suggest a source geometry with a Comptonization component possibly attributed to a boundary layer (BL) emission or BL emission reflected off the disk, consistent with the other NS atoll sources. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 18:19:59 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 20:29:22 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kashyap",
"Unnati",
"",
"TTU, USA"
],
[
"Maccarone",
"Thomas J.",
"",
"TTU, USA"
],
[
"Russell",
"Thomas D.",
"",
"INAF, Italy"
],
[
"Ng",
"Mason",
"",
"McGill University, Canada"
],
[
"Ravi",
"Swati",
"",
"MIT, USA"
],
[
"Pattie",
"Eliot C.",
"",
"TTU, USA"
],
[
"Marshall",
"Herman L.",
"",
"MIT, USA"
]
] |
2504.14175 | Kunwoo Park | Yejun Yoon, Jaeyoon Jung, Seunghyun Yoon and Kunwoo Park | Hypothetical Documents or Knowledge Leakage? Rethinking LLM-based Query Expansion | ACL 2025 (Findings) | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | Query expansion methods powered by large language models (LLMs) have demonstrated effectiveness in zero-shot retrieval tasks. These methods assume that LLMs can generate hypothetical documents that, when incorporated into a query vector, enhance the retrieval of real evidence. However, we challenge this assumption by investigating whether knowledge leakage in benchmarks contributes to the observed performance gains. Using fact verification as a testbed, we analyze whether the generated documents contain information entailed by ground-truth evidence and assess their impact on performance. Our findings indicate that, on average, performance improvements consistently occurred for claims whose generated documents included sentences entailed by gold evidence. This suggests that knowledge leakage may be present in fact-verification benchmarks, potentially inflating the perceived performance of LLM-based query expansion methods. | [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 04:32:38 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:32:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yoon",
"Yejun",
""
],
[
"Jung",
"Jaeyoon",
""
],
[
"Yoon",
"Seunghyun",
""
],
[
"Park",
"Kunwoo",
""
]
] |
2504.14194 | Xinlin Zhuang | Xinlin Zhuang, Jiahui Peng, Ren Ma, Yinfan Wang, Tianyi Bai, Xingjian Wei, Jiantao Qiu, Chi Zhang, Ying Qian, Conghui He | Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models | Accepted by ACL 2025 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The composition of pre-training datasets for large language models (LLMs) remains largely undisclosed, hindering transparency and efforts to optimize data quality, a critical driver of model performance. Current data selection methods, such as natural language quality assessments, diversity-based filters, and classifier-based approaches, are limited by single-dimensional evaluation or redundancy-focused strategies. To address these gaps, we propose four dimensions to evaluate data quality: professionalism, readability, reasoning, and cleanliness. We further introduce Meta-rater,a multi-dimensional data selection method that integrates these dimensions with existing quality metrics through learned optimal weightings. Meta-rater employs proxy models to train a regression model that predicts validation loss, enabling the identification of optimal combinations of quality scores. Experiments demonstrate that Meta-rater doubles convergence speed for 1.3B parameter models and improves downstream task performance by 3.23, with advantages that scale to models as large as 7.2B parameters. Our work establishes that holistic, multi-dimensional quality integration significantly outperforms conventional single-dimension approaches, offering a scalable paradigm for enhancing pre-training efficiency and model capability. To advance future research, we release scripts, data, and models at https://github.com/opendatalab/Meta-rater. | [
{
"version": "v1",
"created": "Sat, 19 Apr 2025 06:12:33 GMT"
},
{
"version": "v2",
"created": "Thu, 01 May 2025 02:37:14 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 15:35:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhuang",
"Xinlin",
""
],
[
"Peng",
"Jiahui",
""
],
[
"Ma",
"Ren",
""
],
[
"Wang",
"Yinfan",
""
],
[
"Bai",
"Tianyi",
""
],
[
"Wei",
"Xingjian",
""
],
[
"Qiu",
"Jiantao",
""
],
[
"Zhang",
"Chi",
""
],
[
"Qian",
"Ying",
""
],
[
"He",
"Conghui",
""
]
] |
2504.15307 | Saarik Kalia | Hao-Ran Cui, Saarik Kalia, Zhen Liu | Modeling frequency instability in high-quality resonant experiments | 20 pages, 7 figures; v2 updates sensitivity analysis in light of non-Gaussian statistics of the system response | null | null | UMN-TH-4420/25, FERMILAB-PUB-25-0260-SQMS | physics.ins-det hep-ex hep-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Modern resonant sensing tools can achieve increasingly high quality factors, which correspond to extremely narrow linewidths. In such systems, time-variation of the resonator's natural frequency can potentially impact its ability to accumulate power and its resulting sensitivity. One such example is the Dark SRF experiment, which utilizes superconducting radio frequency (SRF) cavities with quality factors of $Q\sim10^{10}$. Microscopic deformations of the cavity lead to stochastic jittering of its resonant frequency with amplitude 20 times its linewidth. Naively, one may expect this to lead to a large suppression in accumulated power. In this work, we study in detail the effects of frequency instability on high-quality resonant systems, utilizing the Dark SRF experiment as a case study. We show that the timescale of jittering is crucial to determining its effect on power accumulation. Namely, when the resonant frequency varies sufficiently quickly, the system accumulates power as if there were no jittering at all. This implies that the sensitivity of a jittering resonator is comparable to that of a stable resonator. In the case of Dark SRF, we find that jittering only induces a $\sim 10\%$ loss in power. Our results imply that the dark-photon exclusion bound from Dark SRF's pathfinder run can potentially be refined to be an order of magnitude stronger. | [
{
"version": "v1",
"created": "Fri, 18 Apr 2025 19:22:27 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 19:54:57 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cui",
"Hao-Ran",
""
],
[
"Kalia",
"Saarik",
""
],
[
"Liu",
"Zhen",
""
]
] |
2504.15462 | Fidele Twagirayezu | Fidele J. Twagirayezu | Light front holographic QCD theory in the generalized uncertainty principle framework | 8 pages, 3 figures, 1 table(s) | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | In this article, we develop the framework of light-front holographic QCD in the presence of a minimal length scale by incorporating the Generalized Uncertainty Principle (GUP) into the QCD Lagrangian. From this modified theory, we derive a GUP-corrected light-front holographic QCD (LFH QCD) equation and obtain the corresponding hadronic mass spectrum. Our results show that the hadronic mass spectrum acquires an additional GUP-dependent term that increases the masses. This mass enhancement leads to significantly improved agreement between the theoretical predictions and experimental data. | [
{
"version": "v1",
"created": "Mon, 21 Apr 2025 21:59:28 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Apr 2025 15:54:22 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 19:00:46 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Twagirayezu",
"Fidele J.",
""
]
] |
2504.16656 | Tianyidan Xie | Chris, Yichen Wei, Yi Peng, Xiaokun Wang, Weijie Qiu, Wei Shen, Tianyidan Xie, Jiangbo Pei, Jianhao Zhang, Yunzhuo Hao, Xuchen Song, Yang Liu, Yahui Zhou | Skywork R1V2: Multimodal Hybrid Reinforcement Learning for Reasoning | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present Skywork R1V2, a next-generation multimodal reasoning model and a major leap forward from its predecessor, Skywork R1V. At its core, R1V2 introduces a hybrid reinforcement learning paradigm that jointly leverages the Mixed Preference Optimization (MPO) and the Group Relative Policy Optimization (GRPO), which harmonizes reward-model guidance with rule-based strategies, thereby addressing the long-standing challenge of balancing sophisticated reasoning capabilities with broad generalization. To further enhance training efficiency, we propose the Selective Sample Buffer (SSB) mechanism, which effectively addresses the vanishing advantages dilemma inherent in GRPO by prioritizing high-value samples throughout the optimization process. Notably, we observe that excessive reinforcement signals can induce visual hallucinations--a phenomenon we systematically monitor and mitigate through calibrated reward thresholds throughout the training process. Empirical results affirm the exceptional capability of R1V2, with benchmark-leading performances such as 62.6 on OlympiadBench, 78.9 on AIME2024, 63.6 on LiveCodeBench, and 73.6 on MMMU. These results underscore R1V2's superiority over existing open-source models and demonstrate significant progress in closing the performance gap with premier proprietary systems, including Gemini 2.5 and OpenAI-o4-mini. The Skywork R1V2 model weights have been publicly released to promote openness and reproducibility https://huggingface.co/Skywork/Skywork-R1V2-38B. | [
{
"version": "v1",
"created": "Wed, 23 Apr 2025 12:24:10 GMT"
},
{
"version": "v2",
"created": "Fri, 25 Apr 2025 15:28:34 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 10:46:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chris",
"",
""
],
[
"Wei",
"Yichen",
""
],
[
"Peng",
"Yi",
""
],
[
"Wang",
"Xiaokun",
""
],
[
"Qiu",
"Weijie",
""
],
[
"Shen",
"Wei",
""
],
[
"Xie",
"Tianyidan",
""
],
[
"Pei",
"Jiangbo",
""
],
[
"Zhang",
"Jianhao",
""
],
[
"Hao",
"Yunzhuo",
""
],
[
"Song",
"Xuchen",
""
],
[
"Liu",
"Yang",
""
],
[
"Zhou",
"Yahui",
""
]
] |
2504.17108 | Matthew C. Sullivan | Yash Mohod and Matthew C. Sullivan | Computational Physics in the Advanced Lab: Experiment and Simulation of Thermal Diffusion in Metal Rods | 12 pages, 5 figures | null | null | null | physics.ed-ph cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a simple experiment on thermal diffusion in metal rods that can be used to integrate computation into the advanced lab. An analytical solution exists for the transient heat conduction in an infinite rod with a delta function heat input, but no analytical solution exists for short rods or for long duration heat inputs. Our apparatus is a copper rod with a heater and thermometers attached to the rod. The temperature of the metal rod due to transient heat conduction can be modeled using a simple numerical simulation and the finite centered difference method. Using a 22 cm long copper rod with the ends thermally sunk in aluminum blocks, we show poor agreement between the experimental results and the infinite-rod analytical model, but excellent agreement between the experimental results and our numerical simulation. This experiment shows the power of a numerical simulation but also the limitations of the chosen model, which can be used as motivation for further exploration. | [
{
"version": "v1",
"created": "Wed, 23 Apr 2025 21:38:54 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:48:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mohod",
"Yash",
""
],
[
"Sullivan",
"Matthew C.",
""
]
] |
2504.17959 | Yinlong Dai | Yinlong Dai, Robert Ramirez Sanchez, Ryan Jeronimus, Shahabedin Sagheb, Cara M. Nunez, Heramb Nemlekar and Dylan P. Losey | CIVIL: Causal and Intuitive Visual Imitation Learning | null | null | null | null | cs.RO cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Today's robots learn new tasks by imitating human examples. However, this standard approach to visual imitation learning is fundamentally limited: the robot observes what the human does, but not why the human chooses those behaviors. Without understanding the features that factor into the human's decisions, robot learners often misinterpret the data and fail to perform the task when the environment changes. We therefore propose a shift in perspective: instead of asking human teachers just to show what actions the robot should take, we also enable humans to indicate task-relevant features using markers and language prompts. Our proposed algorithm, CIVIL, leverages this augmented data to filter the robot's visual observations and extract a feature representation that causally informs human actions. CIVIL then applies these causal features to train a transformer-based policy that emulates human behaviors without being confused by visual distractors. Our simulations, real-world experiments, and user study demonstrate that robots trained with CIVIL can learn from fewer human demonstrations and perform better than state-of-the-art baselines, especially in previously unseen scenarios. See videos at our project website: https://civil2025.github.io | [
{
"version": "v1",
"created": "Thu, 24 Apr 2025 22:08:29 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:05:00 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dai",
"Yinlong",
""
],
[
"Sanchez",
"Robert Ramirez",
""
],
[
"Jeronimus",
"Ryan",
""
],
[
"Sagheb",
"Shahabedin",
""
],
[
"Nunez",
"Cara M.",
""
],
[
"Nemlekar",
"Heramb",
""
],
[
"Losey",
"Dylan P.",
""
]
] |
2504.18073 | Jimin Kim | Jimin Kim, Jeremy T. Florman, Julia A. Santos, Mark J. Alkema, Eli Shlizerman | Modular integration of neural connectomics, dynamics and biomechanics for identification of behavioral sensorimotor pathways in Caenorhabditis elegans | null | null | null | null | q-bio.QM q-bio.NC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computational approaches which emulate in-vivo nervous system are needed to investigate mechanisms of the brain to orchestrate behavior. Such approaches must integrate a series of biophysical models encompassing the nervous system, muscles, biomechanics to allow observing the system in its entirety while supporting model variations. Here we develop modWorm: a modular modeling framework for the nematode C. elegans. modWorm allows for construction of a model as an integrated series of configurable, exchangeable modules each describing specific biophysical processes across different modalities. Utilizing modWorm, we propose a base neuro-mechanical model for C. elegans built upon the complete connectome. The model integrates a series of 7 modules: i) intra-cellular dynamics, ii) electrical and iii) chemical extra-cellular neural dynamics, iv) translation of neural activity to muscle calcium dynamics, v) muscle calcium dynamics to muscle forces, vi) muscle forces to body postures and vii) proprioceptive feedback. We validate the base model by in-silico injection of constant currents into neurons known to be associated with locomotion behaviors and by applying external forces to the body. Applications of in-silico neural stimuli experimentally known to modulate locomotion show that the model can recapitulate natural behavioral responses such as forward and backward locomotion as well as mid-locomotion responses such as avoidance and turns. Furthermore, through in-silico ablation surveys, the model can infer novel neural circuits involved in sensorimotor behaviors. To further dissect mechanisms of locomotion, we utilize modWorm to introduce empirical based model variations and model optimizations to elucidate their effects on simulated locomotion. Our results show that modWorm can be utilized to identify neural circuits which control, mediate and generate natural behavior. | [
{
"version": "v1",
"created": "Fri, 25 Apr 2025 05:04:03 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 21:52:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kim",
"Jimin",
""
],
[
"Florman",
"Jeremy T.",
""
],
[
"Santos",
"Julia A.",
""
],
[
"Alkema",
"Mark J.",
""
],
[
"Shlizerman",
"Eli",
""
]
] |
2504.18268 | Ana Matoso | Ana Matoso, Catarina Passarinho, Marta P. Loureiro, Jos\'e Maria Moreira, Patr\'icia Figueiredo, Rita G. Nunes | Towards a deep learning approach for classifying treatment response in glioblastomas | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Glioblastomas are the most aggressive type of glioma, having a 5-year survival rate of 6.9%. Treatment typically involves surgery, followed by radiotherapy and chemotherapy, and frequent magnetic resonance imaging (MRI) scans to monitor disease progression. To assess treatment response, radiologists use the Response Assessment in Neuro-Oncology (RANO) criteria to categorize the tumor into one of four labels based on imaging and clinical features: complete response, partial response, stable disease, and progressive disease. This assessment is very complex and time-consuming. Since deep learning (DL) has been widely used to tackle classification problems, this work aimed to implement the first DL pipeline for the classification of RANO criteria based on two consecutive MRI acquisitions. The models were trained and tested on the open dataset LUMIERE. Five approaches were tested: 1) subtraction of input images, 2) different combinations of modalities, 3) different model architectures, 4) different pretraining tasks, and 5) adding clinical data. The pipeline that achieved the best performance used a Densenet264 considering only T1-weighted, T2-weighted, and Fluid Attenuated Inversion Recovery (FLAIR) images as input without any pretraining. A median Balanced Accuracy of 50.96% was achieved. Additionally, explainability methods were applied. Using Saliency Maps, the tumor region was often successfully highlighted. In contrast, Grad-CAM typically failed to highlight the tumor region, with some exceptions observed in the Complete Response and Progressive Disease classes, where it effectively identified the tumor region. These results set a benchmark for future studies on glioblastoma treatment response assessment based on the RANO criteria while emphasizing the heterogeneity of factors that might play a role when assessing the tumor's response to treatment. | [
{
"version": "v1",
"created": "Fri, 25 Apr 2025 11:27:05 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:53:41 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Matoso",
"Ana",
""
],
[
"Passarinho",
"Catarina",
""
],
[
"Loureiro",
"Marta P.",
""
],
[
"Moreira",
"José Maria",
""
],
[
"Figueiredo",
"Patrícia",
""
],
[
"Nunes",
"Rita G.",
""
]
] |
2504.18638 | Simone Michele Mazza Dr | A. Molnar, Y. Zhao, S.M. Mazza, G. Oregan, M. Davis, S. Beringer, A. Tiernan, J. Ott, H.F.-W. Sadrozinski, A. Seiden, B. Schumm, F. McKinney-Martinez, A. Bisht, M. Centis-Vignali, G. Paternoster, M. Boscardin | Synchrotron light source focused X-ray detection with LGADs, AC-LGADs and TI-LGADs | null | null | null | null | physics.ins-det hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The response of Low Gain Avalanche Diodes (LGADs), a type of thin silicon detector with internal gain, to X-rays of energies between 6-16~keV was characterized at the Stanford Synchrotron Radiation Lightsource (SSRL). The utilized beamline at SSRL was 7-2, with a nominal beam size of 30~$\mu$m, repetition rate of 500~MHz, and with an energy dispersion $\Delta E/E$ of $10^{-4}$. Multi-channel LGADs, AC-LGADs, and TI-LGADs of different thicknesses and gain layer configurations from Hamamatsu Photonics (HPK) and Fondazione Bruno Kessler (FBK) were tested. The sensors were read out with a discrete component board and digitized with a fast oscilloscope or a CAEN fast digitizer. The devices' energy response, energy resolution, and time resolution were measured as a function of X-ray energy and position. The charge collection and multiplication mechanism were simulated using TCAD Sentaurus, and the results were compared with the collected data. | [
{
"version": "v1",
"created": "Fri, 25 Apr 2025 18:33:15 GMT"
},
{
"version": "v2",
"created": "Fri, 09 May 2025 22:44:33 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 20:56:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Molnar",
"A.",
""
],
[
"Zhao",
"Y.",
""
],
[
"Mazza",
"S. M.",
""
],
[
"Oregan",
"G.",
""
],
[
"Davis",
"M.",
""
],
[
"Beringer",
"S.",
""
],
[
"Tiernan",
"A.",
""
],
[
"Ott",
"J.",
""
],
[
"Sadrozinski",
"H. F. -W.",
""
],
[
"Seiden",
"A.",
""
],
[
"Schumm",
"B.",
""
],
[
"McKinney-Martinez",
"F.",
""
],
[
"Bisht",
"A.",
""
],
[
"Centis-Vignali",
"M.",
""
],
[
"Paternoster",
"G.",
""
],
[
"Boscardin",
"M.",
""
]
] |
2504.18791 | Uday Kiran Reddy Tadipatri | Uday Kiran Reddy Tadipatri, Benjamin D. Haeffele, Joshua Agterberg, Ingvar Ziemann, and Ren\'e Vidal | Nonconvex Linear System Identification with Minimal State Representation | Accepted to the 7th Annual Conference on Learning for Dynamics and Control (L4DC) 2025. The full version including appendix | null | null | null | eess.SY cs.LG cs.SY eess.SP stat.ML | http://creativecommons.org/licenses/by/4.0/ | Low-order linear System IDentification (SysID) addresses the challenge of estimating the parameters of a linear dynamical system from finite samples of observations and control inputs with minimal state representation. Traditional approaches often utilize Hankel-rank minimization, which relies on convex relaxations that can require numerous, costly singular value decompositions (SVDs) to optimize. In this work, we propose two nonconvex reformulations to tackle low-order SysID (i) Burer-Monterio (BM) factorization of the Hankel matrix for efficient nuclear norm minimization, and (ii) optimizing directly over system parameters for real, diagonalizable systems with an atomic norm style decomposition. These reformulations circumvent the need for repeated heavy SVD computations, significantly improving computational efficiency. Moreover, we prove that optimizing directly over the system parameters yields lower statistical error rates, and lower sample complexities that do not scale linearly with trajectory length like in Hankel-nuclear norm minimization. Additionally, while our proposed formulations are nonconvex, we provide theoretical guarantees of achieving global optimality in polynomial time. Finally, we demonstrate algorithms that solve these nonconvex programs and validate our theoretical claims on synthetic data. | [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 04:11:02 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 23:12:37 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tadipatri",
"Uday Kiran Reddy",
""
],
[
"Haeffele",
"Benjamin D.",
""
],
[
"Agterberg",
"Joshua",
""
],
[
"Ziemann",
"Ingvar",
""
],
[
"Vidal",
"René",
""
]
] |
2504.18820 | Cheng Jia | Cheng Jia, Enci Wang, Cheqiu Lyu, Chengyu Ma, Jie Song, Yangyao Chen, Kai Wang, Haoran Yu, Zeyu Chen, Jinyang Wang, Yifan Wang, Xu Kong | Potential-Driven Metal Cycling: JADES Census of Gas-Phase Metallicity for galaxies at 1 < z < 7 | Accepted for publication in ApJL, 16 pages, 8 figures | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The gravitational potential is established as a critical determinant of gas-phase metallicity (12+log(O/H)) in low-redshift galaxies, whereas its influence remains unconfirmed at high redshifts. We investigate the correlation between gas-phase metallicity and effective radius ($R_{\rm e}$) for a sample of galaxies with redshifts ranging from 1 to 7, drawn from JADES (JWST Advanced Deep Extragalactic Survey) Data Release 3. We calculate the metallicities using four strong-line methods: ${\rm N2S2H\alpha}$, ${\rm R23}$, ${\rm N2}$, and ${\rm O3N2}$, respectively. After taking out the evolution of size, we find that the offsets of mass-size relation ($\Delta \log R_{\rm e}$) are significantly negatively correlated with the offset of mass-metallicity relation ($\Delta \log({\rm O/H})$) for the four metallicity tracers. Regardless of the metallicity tracer used, we obtain Spearman rank $p-$values much less than 0.01, rejecting the null hypothesis that the observed correlation is statistically nonsignificant and attributable to random chance. This is also true for galaxies with $z>3$, with $p-$values less than 0.05 for the four metallicity tracers. We for the first time find evidence of size playing a key role in determining gas-phase metallicity towards cosmic dawn, suggesting that the gravitational potential influences their material-exchange processes with the surrounding environment at very early universe. | [
{
"version": "v1",
"created": "Sat, 26 Apr 2025 06:35:00 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 01:42:11 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jia",
"Cheng",
""
],
[
"Wang",
"Enci",
""
],
[
"Lyu",
"Cheqiu",
""
],
[
"Ma",
"Chengyu",
""
],
[
"Song",
"Jie",
""
],
[
"Chen",
"Yangyao",
""
],
[
"Wang",
"Kai",
""
],
[
"Yu",
"Haoran",
""
],
[
"Chen",
"Zeyu",
""
],
[
"Wang",
"Jinyang",
""
],
[
"Wang",
"Yifan",
""
],
[
"Kong",
"Xu",
""
]
] |
2504.19223 | Alexander Baumann | Alexander Baumann, Leonardo Ayala, Silvia Seidlitz, Jan Sellner, Alexander Studier-Fischer, Berkin \"Ozdemir, Lena Maier-Hein, Slobodan Ilic | CARL: Camera-Agnostic Representation Learning for Spectral Image Analysis | null | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Spectral imaging offers promising applications across diverse domains, including medicine and urban scene understanding, and is already established as a critical modality in remote sensing. However, variability in channel dimensionality and captured wavelengths among spectral cameras impede the development of AI-driven methodologies, leading to camera-specific models with limited generalizability and inadequate cross-camera applicability. To address this bottleneck, we introduce $\textbf{CARL}$, a model for $\textbf{C}$amera-$\textbf{A}$gnostic $\textbf{R}$epresentation $\textbf{L}$earning across RGB, multispectral, and hyperspectral imaging modalities. To enable the conversion of a spectral image with any channel dimensionality to a camera-agnostic embedding, we introduce wavelength positional encoding and a self-attention-cross-attention mechanism to compress spectral information into learned query representations. Spectral-spatial pre-training is achieved with a novel spectral self-supervised JEPA-inspired strategy tailored to CARL. Large-scale experiments across the domains of medical imaging, autonomous driving, and satellite imaging demonstrate our model's unique robustness to spectral heterogeneity, outperforming on datasets with simulated and real-world cross-camera spectral variations. The scalability and versatility of the proposed approach position our model as a backbone for future spectral foundation models. | [
{
"version": "v1",
"created": "Sun, 27 Apr 2025 13:06:40 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 13:25:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Baumann",
"Alexander",
""
],
[
"Ayala",
"Leonardo",
""
],
[
"Seidlitz",
"Silvia",
""
],
[
"Sellner",
"Jan",
""
],
[
"Studier-Fischer",
"Alexander",
""
],
[
"Özdemir",
"Berkin",
""
],
[
"Maier-Hein",
"Lena",
""
],
[
"Ilic",
"Slobodan",
""
]
] |
2504.20227 | Zhe-Fei Yu | Zhe-fei Yu | On the CFT dual of superstring on AdS$_3$ | 55 pages; v2: references added, typos corrected | null | null | null | hep-th math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we study the proposed duality between superstring on AdS$_3\times$S$^3\times\mathbb{T}^4$ with pure NS-NS flux and a deformed symmetric orbifold CFT. We extract the residues of 3-point string correlators at their next-to-leading poles, which can be expressed as integrals. Because of picture changing, the integrand is not unique. We make an ansatz for its proper form which linearly combines three different picture choices. On the CFT side, we obtain the corresponding residues by doing a conformal perturbation computation at the second order. The result can also be expressed as integrals, with the integrand being a five point function in the symmetric orbifold theory, which is calculated using covering maps. After unifying the integration variables of the two sides, we find that the two integrands match precisely, due to some novel mathematical identities of covering maps. Our calculation verifies the proposed marginal operator in the dual CFT of superstring on AdS$_3\times$S$^3\times\mathbb{T}^4$ and could be generalized to the CFT duals of superstrings on general backgrounds AdS$_3\times X$ with few modifications. | [
{
"version": "v1",
"created": "Mon, 28 Apr 2025 19:54:08 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:01:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yu",
"Zhe-fei",
""
]
] |
2504.20618 | Qijun Jiang | Qijun Jiang, Xiaodan Shao, Rui Zhang | Statistical Channel Based Low-Complexity Rotation and Position Optimization for 6D Movable Antennas Enabled Wireless Communication | 13 pages, 13 figures, submitted for journal publication | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Six-dimensional movable antenna (6DMA) is a promising technology to fully exploit spatial variation in wireless channels by allowing flexible adjustment of three-dimensional (3D) positions and rotations of antennas at the transceiver. In this paper, we investigate the practical low-complexity design of 6DMA-enabled communication systems, including transmission protocol, statistical channel information (SCI) acquisition, and joint position and rotation optimization of 6DMA surfaces based on the SCI of users. Specifically, an orthogonal matching pursuit (OMP)-based algorithm is proposed for the estimation of SCI of users at all possible position-rotation pairs of 6DMA surfaces based on the channel measurements at a small subset of position-rotation pairs. Then, the average sum logarithmic rate of all users is maximized by jointly designing the positions and rotations of 6DMA surfaces based on their SCI acquired. Different from prior works on 6DMA which adopt alternating optimization to design 6DMA positions/rotations with iterations, we propose a new sequential optimization approach that first determines 6DMA rotations and then finds their feasible positions to realize the optimized rotations subject to practical antenna placement constraints. Simulation results show that the proposed sequential optimization significantly reduces the computational complexity of conventional alternating optimization, while achieving comparable communication performance. It is also shown that the proposed SCI-based 6DMA design can effectively enhance the communication throughput of wireless networks over existing fixed (position and rotation) antenna arrays, yet with a practically appealing low-complexity implementation. | [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 10:37:34 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 02:10:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jiang",
"Qijun",
""
],
[
"Shao",
"Xiaodan",
""
],
[
"Zhang",
"Rui",
""
]
] |
2504.21002 | Nhat Hoang Le | Nhat Hoang Le | The local twisted Gan-Gross-Prasad conjecture for $\text{GL(V)/U(V)}$ | 75 pages, change template, correct minor typos. arXiv admin note: text overlap with arXiv:1506.01452 by other authors | null | null | null | math.RT math.NT | http://creativecommons.org/licenses/by/4.0/ | In this paper, following the method developed by J.-L. Waldspurger and R. Beuzart-Plessis for Bessel models, we study two local relative trace formulas for the local twisted Gan-Gross-Prasad conjecture. By obtaining spectral expansions and a partial geometric comparision between the two formulas, we prove the local twisted Gan-Gross-Prasad conjecture for $\text{GL(V)/U(V)}$ over nonarchimedean fields for tempered representations. | [
{
"version": "v1",
"created": "Tue, 01 Apr 2025 07:27:43 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:34:27 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Le",
"Nhat Hoang",
""
]
] |
2504.21131 | Remo Christen | Remo Christen, Florian Pommerening, Clemens B\"uchner, Malte Helmert | A Formalism for Optimal Search with Dynamic Heuristics (Extended Version) | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | While most heuristics studied in heuristic search depend only on the state, some accumulate information during search and thus also depend on the search history. Various existing approaches use such dynamic heuristics in $\mathrm{A}^*$-like algorithms and appeal to classic results for $\mathrm{A}^*$ to show optimality. However, doing so ignores the complexities of searching with a mutable heuristic. In this paper we formalize the idea of dynamic heuristics and use them in a generic algorithm framework. We study a particular instantiation that models $\mathrm{A}^*$ with dynamic heuristics and show general optimality results. Finally we show how existing approaches from classical planning can be viewed as special cases of this instantiation, making it possible to directly apply our optimality results. | [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 19:25:31 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:22:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Christen",
"Remo",
""
],
[
"Pommerening",
"Florian",
""
],
[
"Büchner",
"Clemens",
""
],
[
"Helmert",
"Malte",
""
]
] |
2504.21223 | Katsunori Kubo | Katsunori Kubo | Surface Structure and Surface State of a Tight-Binding Model on a Diamond Lattice | 6 pages, 12 figures | J. Phys. Soc. Jpn. 94, 074701 (2025) | 10.7566/JPSJ.94.074701 | null | cond-mat.str-el cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Surface states of a tight-binding model with nearest-neighbor hopping on a diamond lattice of finite thickness are investigated. We consider systems with (001), (110), and (111) surfaces. Even if the surface direction is fixed, there is freedom to choose the surface structure due to the two-sublattice nature of a diamond lattice. The existence of surface states is governed by the topology of the matrix elements of the bulk Hamiltonian. In this sense, the existence of surface states is determined by the bulk Hamiltonian. However, the matrix elements depend on the choice of the unit cell, which should be chosen to conform to the surface structure. Thus, the surface states depend on the surface structure. We find that for each surface direction, there are two choices of surface structure, and depending on the chosen structure, the corresponding surface states emerge in distinct regions of the surface Brillouin zone. | [
{
"version": "v1",
"created": "Tue, 29 Apr 2025 23:26:04 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 02:53:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kubo",
"Katsunori",
""
]
] |
2504.21386 | Roshna K | Roshna K and V. Sreenath | Power Suppression and Lensing Anomaly -- A phenomenological investigation | 25 pages, 7 figures and 16 tables, new section added | null | null | null | astro-ph.CO gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Primordial power spectra with low power at long wavelengths can alleviate lensing anomaly. However the extent to which data favours such a primordial spectra is not clear. In this work, we investigate power suppression and related mitigation of lensing anomaly with the help of phenomenological models which are valid over scales of interest. We consider simple extensions to nearly scale invariant power spectra such as those which includes running and running of running of spectral index. We perform Bayesian analysis of these models, which are agnostic about power suppression, with Planck legacy data and show that data tend to choose parameters which leads to power suppression at low multipoles. We then investigate the connection between power suppression and alleviation of lensing anomaly and show that lensing anomaly is mitigated the most in models with maximum suppression of power at low multipoles. We also analyse the significance of these findings using information criteria. These results are further analyzed in the light of Planck Release 4 data using \texttt{CamSpec}, \texttt{HiLLiPoP} and \texttt{LoLLiPoP} likelihoods in which departure of lensing parameter from one is significantly reduced. Furthermore, we investigate the ability of near-ultimate future CMB missions such as ECHO to put tighter constraints on these models and to settle the issue. We conclude that we can make stronger conclusions about the presence of power suppression in the future by studying such simple phenomenological models. | [
{
"version": "v1",
"created": "Wed, 30 Apr 2025 07:37:11 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:47:37 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"K",
"Roshna",
""
],
[
"Sreenath",
"V.",
""
]
] |
2505.00070 | Nadie LiTenn | Nadie Yiluo LiTenn, Tianci Zhou, Brian Swingle | Scrambling Dynamics with Imperfections in a Solvable Model | 33 pages, 10 figures | null | null | null | quant-ph cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study how probes of quantum scrambling dynamics respond to two kinds of imperfections -- unequal forward and backward evolutions and decoherence -- in a solvable Brownian circuit model. We calculate a ``renormalized'' out-of-time-order correlator (ROTOC) in the model with $N$ qubits, and we show that the circuit-averaged ROTOC is controlled by an effective probability distribution in operator weight space which obeys a system of $N$ non-linear equations of motion. These equations can be easily solved numerically for large system sizes which are beyond the reach of exact methods. Moreover, for an operator initially concentrated on weight one $w_0=1$, we provide an exact solution to the equations in the thermodynamic limit of many qubits that is valid for all times, all non-vanishing perturbation strengths $p\gtrsim 1/\sqrt{N}$, and all decoherence strengths. We also show that a generic initial condition $w_0 >1$ leads to a metastable state that eventually collapses to the $w_0=1$ case after a lifetime $\sim \log(N/w_0)$. Our results highlight situations where it is still possible to extract the unperturbed chaos exponent even in the presence of imperfections, and we comment on the applications of our results to existing experiments with nuclear spins and to future scrambling experiments. | [
{
"version": "v1",
"created": "Wed, 30 Apr 2025 18:00:01 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:02:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"LiTenn",
"Nadie Yiluo",
""
],
[
"Zhou",
"Tianci",
""
],
[
"Swingle",
"Brian",
""
]
] |
2505.00237 | Ze Zhang | Ze Zhang, Georg Hess, Junjie Hu, Emmanuel Dean, Lennart Svensson, Knut {\AA}kesson | Future-Oriented Navigation: Dynamic Obstacle Avoidance with One-Shot Energy-Based Multimodal Motion Prediction | Published in IEEE Robotics and Automation Letters (RA-L) | null | null | null | cs.RO cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes an integrated approach for the safe and efficient control of mobile robots in dynamic and uncertain environments. The approach consists of two key steps: one-shot multimodal motion prediction to anticipate motions of dynamic obstacles and model predictive control to incorporate these predictions into the motion planning process. Motion prediction is driven by an energy-based neural network that generates high-resolution, multi-step predictions in a single operation. The prediction outcomes are further utilized to create geometric shapes formulated as mathematical constraints. Instead of treating each dynamic obstacle individually, predicted obstacles are grouped by proximity in an unsupervised way to improve performance and efficiency. The overall collision-free navigation is handled by model predictive control with a specific design for proactive dynamic obstacle avoidance. The proposed approach allows mobile robots to navigate effectively in dynamic environments. Its performance is accessed across various scenarios that represent typical warehouse settings. The results demonstrate that the proposed approach outperforms other existing dynamic obstacle avoidance methods. | [
{
"version": "v1",
"created": "Thu, 01 May 2025 01:13:56 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 08:53:16 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 14:05:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhang",
"Ze",
""
],
[
"Hess",
"Georg",
""
],
[
"Hu",
"Junjie",
""
],
[
"Dean",
"Emmanuel",
""
],
[
"Svensson",
"Lennart",
""
],
[
"Åkesson",
"Knut",
""
]
] |
2505.00270 | Yijie Xia | Yijie Xia, Xiaohan Lin, Zicheng Ma, Jinyuan Hu, Yanheng Li, Zhaoxin Xie, Hao Li, Li Yang, Zhiqiang Zhao, Lijiang Yang, Zhenyu Chen, Yi Qin Gao | Large Language Models as AI Agents for Digital Atoms and Molecules: Catalyzing a New Era in Computational Biophysics | 26 pages, 3 figures, 2 tables | null | null | null | physics.comp-ph physics.bio-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In computational biophysics, where molecular data is expanding rapidly and system complexity is increasing exponentially, large language models (LLMs) and agent-based systems are fundamentally reshaping the field. This perspective article examines the recent advances at the intersection of LLMs, intelligent agents, and scientific computation, with a focus on biophysical computation. Building on these advancements, we introduce ADAM (Agent for Digital Atoms and Molecules), an innovative multi-agent LLM-based framework. ADAM employs cutting-edge AI architectures to reshape scientific workflows through a modular design. It adopts a hybrid neural-symbolic architecture that combines LLM-driven semantic tools with deterministic symbolic computations. Moreover, its ADAM Tool Protocol (ATP) enables asynchronous, database-centric tool orchestration, fostering community-driven extensibility. Despite the significant progress made, ongoing challenges call for further efforts in establishing benchmarking standards, optimizing foundational models and agents, building an open collaborative ecosystem and developing personalized memory modules. ADAM is accessible at https://sidereus-ai.com. | [
{
"version": "v1",
"created": "Thu, 01 May 2025 03:33:57 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:22:57 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Xia",
"Yijie",
""
],
[
"Lin",
"Xiaohan",
""
],
[
"Ma",
"Zicheng",
""
],
[
"Hu",
"Jinyuan",
""
],
[
"Li",
"Yanheng",
""
],
[
"Xie",
"Zhaoxin",
""
],
[
"Li",
"Hao",
""
],
[
"Yang",
"Li",
""
],
[
"Zhao",
"Zhiqiang",
""
],
[
"Yang",
"Lijiang",
""
],
[
"Chen",
"Zhenyu",
""
],
[
"Gao",
"Yi Qin",
""
]
] |
2505.00467 | David Pellow | Vahid Balazadeh, Michael Cooper, David Pellow, Atousa Assadi, Jennifer Bell, Mark Coastworth, Kaivalya Deshpande, Jim Fackler, Gabriel Funingana, Spencer Gable-Cook, Anirudh Gangadhar, Abhishek Jaiswal, Sumanth Kaja, Christopher Khoury, Amrit Krishnan, Randy Lin, Kaden McKeen, Sara Naimimohasses, Khashayar Namdar, Aviraj Newatia, Allan Pang, Anshul Pattoo, Sameer Peesapati, Diana Prepelita, Bogdana Rakova, Saba Sadatamin, Rafael Schulman, Ajay Shah, Syed Azhar Shah, Syed Ahmar Shah, Babak Taati, Balagopal Unnikrishnan, I\~nigo Urteaga, Stephanie Williams, and Rahul G Krishnan | Red Teaming Large Language Models for Healthcare | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large language model (LLM) outputs a response that could cause clinical harm. Red-teaming with clinicians enables the identification of LLM vulnerabilities that may not be recognised by LLM developers lacking clinical expertise. We report the vulnerabilities found, categorise them, and present the results of a replication study assessing the vulnerabilities across all LLMs provided. | [
{
"version": "v1",
"created": "Thu, 01 May 2025 11:43:27 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Balazadeh",
"Vahid",
""
],
[
"Cooper",
"Michael",
""
],
[
"Pellow",
"David",
""
],
[
"Assadi",
"Atousa",
""
],
[
"Bell",
"Jennifer",
""
],
[
"Coastworth",
"Mark",
""
],
[
"Deshpande",
"Kaivalya",
""
],
[
"Fackler",
"Jim",
""
],
[
"Funingana",
"Gabriel",
""
],
[
"Gable-Cook",
"Spencer",
""
],
[
"Gangadhar",
"Anirudh",
""
],
[
"Jaiswal",
"Abhishek",
""
],
[
"Kaja",
"Sumanth",
""
],
[
"Khoury",
"Christopher",
""
],
[
"Krishnan",
"Amrit",
""
],
[
"Lin",
"Randy",
""
],
[
"McKeen",
"Kaden",
""
],
[
"Naimimohasses",
"Sara",
""
],
[
"Namdar",
"Khashayar",
""
],
[
"Newatia",
"Aviraj",
""
],
[
"Pang",
"Allan",
""
],
[
"Pattoo",
"Anshul",
""
],
[
"Peesapati",
"Sameer",
""
],
[
"Prepelita",
"Diana",
""
],
[
"Rakova",
"Bogdana",
""
],
[
"Sadatamin",
"Saba",
""
],
[
"Schulman",
"Rafael",
""
],
[
"Shah",
"Ajay",
""
],
[
"Shah",
"Syed Azhar",
""
],
[
"Shah",
"Syed Ahmar",
""
],
[
"Taati",
"Babak",
""
],
[
"Unnikrishnan",
"Balagopal",
""
],
[
"Urteaga",
"Iñigo",
""
],
[
"Williams",
"Stephanie",
""
],
[
"Krishnan",
"Rahul G",
""
]
] |
2505.01874 | Youssef Allouah | Youssef Allouah, Rachid Guerraoui, John Stephan | Towards Trustworthy Federated Learning with Untrusted Participants | ICML 2025 conference paper | null | null | null | cs.LG cs.CR cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Resilience against malicious participants and data privacy are essential for trustworthy federated learning, yet achieving both with good utility typically requires the strong assumption of a trusted central server. This paper shows that a significantly weaker assumption suffices: each pair of participants shares a randomness seed unknown to others. In a setting where malicious participants may collude with an untrusted server, we propose CafCor, an algorithm that integrates robust gradient aggregation with correlated noise injection, using shared randomness between participants. We prove that CafCor achieves strong privacy-utility trade-offs, significantly outperforming local differential privacy (DP) methods, which do not make any trust assumption, while approaching central DP utility, where the server is fully trusted. Empirical results on standard benchmarks validate CafCor's practicality, showing that privacy and robustness can coexist in distributed systems without sacrificing utility or trusting the server. | [
{
"version": "v1",
"created": "Sat, 03 May 2025 17:35:47 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:22:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Allouah",
"Youssef",
""
],
[
"Guerraoui",
"Rachid",
""
],
[
"Stephan",
"John",
""
]
] |
2505.02945 | Egil Diau | Egil Diau | The Cognitive Foundations of Economic Exchange: A Modular Framework Grounded in Behavioral Evidence | This version updates the position paper with clearer language and improved structure. It also corrects minor mistakes in wording and formatting. There is no change in framing, scope, or modeling domain. The core contribution remains a simulateable, agent-based framework intended for cs.CE / cs.MA | null | null | null | cs.CY cs.AI cs.MA | http://creativecommons.org/licenses/by/4.0/ | The origins of economic behavior remain unresolved-not only in the social sciences but also in AI, where dominant theories often rely on predefined incentives or institutional assumptions. Contrary to the longstanding myth of barter as the foundation of exchange, converging evidence from early human societies suggests that reciprocity-not barter-was the foundational economic logic, enabling communities to sustain exchange and social cohesion long before formal markets emerged. Yet despite its centrality, reciprocity lacks a simulateable and cognitively grounded account. Here, we introduce a minimal behavioral framework based on three empirically supported cognitive primitives-individual recognition, reciprocal credence, and cost--return sensitivity-that enable agents to participate in and sustain reciprocal exchange, laying the foundation for scalable economic behavior. These mechanisms scaffold the emergence of cooperation, proto-economic exchange, and institutional structure from the bottom up. By bridging insights from primatology, developmental psychology, and economic anthropology, this framework offers a unified substrate for modeling trust, coordination, and economic behavior in both human and artificial systems. | [
{
"version": "v1",
"created": "Mon, 05 May 2025 18:21:53 GMT"
},
{
"version": "v2",
"created": "Fri, 23 May 2025 08:31:57 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 09:19:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Diau",
"Egil",
""
]
] |
2505.04627 | Jean-Michel Tucny Ph.D. | Jean-Michel Tucny, Mihir Durve, Sauro Succi | Is the end of Insight in Sight ? | 15 pages, 2 figures | null | null | null | physics.comp-ph cs.LG physics.data-an | http://creativecommons.org/licenses/by/4.0/ | The rise of deep learning challenges the longstanding scientific ideal of insight - the human capacity to understand phenomena by uncovering underlying mechanisms. In many modern applications, accurate predictions no longer require interpretable models, prompting debate about whether explainability is a realistic or even meaningful goal. From our perspective in physics, we examine this tension through a concrete case study: a physics-informed neural network (PINN) trained on a rarefied gas dynamics problem governed by the Boltzmann equation. Despite the system's clear structure and well-understood governing laws, the trained network's weights resemble Gaussian-distributed random matrices, with no evident trace of the physical principles involved. This suggests that deep learning and traditional simulation may follow distinct cognitive paths to the same outcome - one grounded in mechanistic insight, the other in statistical interpolation. Our findings raise critical questions about the limits of explainable AI and whether interpretability can - or should-remain a universal standard in artificial reasoning. | [
{
"version": "v1",
"created": "Wed, 07 May 2025 19:57:36 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:57:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tucny",
"Jean-Michel",
""
],
[
"Durve",
"Mihir",
""
],
[
"Succi",
"Sauro",
""
]
] |
2505.04670 | Charly Reux | Charly Reux (DiverSe), Mathieu Acher (DiverSe), Djamel Eddine Khelladi (DiverSe), Olivier Barais (DiverSe), Cl\'ement Quinton (SPIRALS) | LLM Code Customization with Visual Results: A Benchmark on TikZ | null | EASE 2025 - Evaluation and Assessment in Software Engineering, Jun 2025, Istanbul, Turkey. pp.1-10 | null | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rise of AI-based code generation, customizing existing code out of natural language instructions to modify visual results -such as figures or images -has become possible, promising to reduce the need for deep programming expertise. However, even experienced developers can struggle with this task, as it requires identifying relevant code regions (feature location), generating valid code variants, and ensuring the modifications reliably align with user intent. In this paper, we introduce vTikZ, the first benchmark designed to evaluate the ability of Large Language Models (LLMs) to customize code while preserving coherent visual outcomes. Our benchmark consists of carefully curated vTikZ editing scenarios, parameterized ground truths, and a reviewing tool that leverages visual feedback to assess correctness. Empirical evaluation with stateof-the-art LLMs shows that existing solutions struggle to reliably modify code in alignment with visual intent, highlighting a gap in current AI-assisted code editing approaches. We argue that vTikZ opens new research directions for integrating LLMs with visual feedback mechanisms to improve code customization tasks in various domains beyond TikZ, including image processing, art creation, Web design, and 3D modeling. | [
{
"version": "v1",
"created": "Wed, 07 May 2025 08:26:54 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:57:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Reux",
"Charly",
"",
"DiverSe"
],
[
"Acher",
"Mathieu",
"",
"DiverSe"
],
[
"Khelladi",
"Djamel Eddine",
"",
"DiverSe"
],
[
"Barais",
"Olivier",
"",
"DiverSe"
],
[
"Quinton",
"Clément",
"",
"SPIRALS"
]
] |
2505.04947 | Hongliang Zhang | Hongliang Zhang, Fenghua Xu, Zhongyuan Yu, Chunqiang Hu, Shanchen Pang, Xiaofen Wang, Jiguo Yu | DFPL: Decentralized Federated Prototype Learning Across Heterogeneous Data Distributions | null | null | null | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | Federated learning is a distributed machine learning paradigm through centralized model aggregation. However, standard federated learning relies on a centralized server, making it vulnerable to server failures. While existing solutions utilize blockchain technology to implement Decentralized Federated Learning (DFL), the statistical heterogeneity of data distributions among clients severely degrades the performance of DFL. Driven by this issue, this paper proposes a decentralized federated prototype learning framework, named DFPL, which significantly improves the performance of DFL across heterogeneous data distributions. Specifically, DFPL introduces prototype learning into DFL to mitigate the impact of statistical heterogeneity and reduces the amount of parameters exchanged between clients. Additionally, blockchain is embedded into our framework, enabling the training and mining processes to be implemented locally on each client. From a theoretical perspective, we provide convergence guarantee of DFPL by modeling the resource allocation between training and mining. The experiments highlight the superiority of our DFPL framework in model performance and communication efficiency across four benchmark datasets with heterogeneous data distributions. | [
{
"version": "v1",
"created": "Thu, 08 May 2025 04:58:45 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:57:22 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhang",
"Hongliang",
""
],
[
"Xu",
"Fenghua",
""
],
[
"Yu",
"Zhongyuan",
""
],
[
"Hu",
"Chunqiang",
""
],
[
"Pang",
"Shanchen",
""
],
[
"Wang",
"Xiaofen",
""
],
[
"Yu",
"Jiguo",
""
]
] |
2505.05470 | Gongye Liu | Jie Liu, Gongye Liu, Jiajun Liang, Yangguang Li, Jiaheng Liu, Xintao Wang, Pengfei Wan, Di Zhang, Wanli Ouyang | Flow-GRPO: Training Flow Matching Models via Online RL | Code: https://github.com/yifan123/flow_grpo | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose Flow-GRPO, the first method integrating online reinforcement learning (RL) into flow matching models. Our approach uses two key strategies: (1) an ODE-to-SDE conversion that transforms a deterministic Ordinary Differential Equation (ODE) into an equivalent Stochastic Differential Equation (SDE) that matches the original model's marginal distribution at all timesteps, enabling statistical sampling for RL exploration; and (2) a Denoising Reduction strategy that reduces training denoising steps while retaining the original inference timestep number, significantly improving sampling efficiency without performance degradation. Empirically, Flow-GRPO is effective across multiple text-to-image tasks. For complex compositions, RL-tuned SD3.5 generates nearly perfect object counts, spatial relations, and fine-grained attributes, boosting GenEval accuracy from 63% to 95%. In visual text rendering, its accuracy improves from 59% to 92%, significantly enhancing text generation. Flow-GRPO also achieves substantial gains in human preference alignment. Notably, very little reward hacking occurred, meaning rewards did not increase at the cost of appreciable image quality or diversity degradation. | [
{
"version": "v1",
"created": "Thu, 08 May 2025 17:58:45 GMT"
},
{
"version": "v2",
"created": "Sun, 11 May 2025 14:44:28 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 13:31:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Jie",
""
],
[
"Liu",
"Gongye",
""
],
[
"Liang",
"Jiajun",
""
],
[
"Li",
"Yangguang",
""
],
[
"Liu",
"Jiaheng",
""
],
[
"Wang",
"Xintao",
""
],
[
"Wan",
"Pengfei",
""
],
[
"Zhang",
"Di",
""
],
[
"Ouyang",
"Wanli",
""
]
] |
2505.05822 | Haotian Hang | Haotian Hang, Chenchen Huang, Alex Barnett, Eva Kanso | Self-reorganization and Information Transfer in Massive Schools of Fish | null | null | null | null | physics.bio-ph cond-mat.stat-mech physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | The remarkable cohesion and coordination observed in moving animal groups and their collective responsiveness to threats are thought to be mediated by scale-free correlations, where changes in the behavior of one animal influence others in the group, regardless of the distance between them. But are these features independent of group size? Here, we investigate group cohesiveness and collective responsiveness in computational models of massive schools of fish of up to 50,000 individuals. We show that as the number of swimmers increases, flow interactions destabilize the school, creating clusters that constantly fragment, disperse, and regroup, similar to their biological counterparts. We calculate the spatial correlation and speed of information propagation in these dynamic clusters. Spatial correlations in cohesive and polarized clusters are indeed scale free, much like in natural animal groups, but fragmentation events are preceded by a decrease in correlation length, thus diminishing the group's collective responsiveness, leaving it more vulnerable to predation events. Importantly, in groups undergoing collective turns, the information about the change in direction propagates linearly in time among group members, thanks to the non-reciprocal nature of the visual interactions between individuals. Merging speeds up the transfer of information within each cluster by several fold, while fragmentation slows it down. Our findings suggest that flow interactions may have played an important role in group size regulation, behavioral adaptations, and dispersion in living animal groups. | [
{
"version": "v1",
"created": "Fri, 09 May 2025 06:47:52 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 23:40:50 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hang",
"Haotian",
""
],
[
"Huang",
"Chenchen",
""
],
[
"Barnett",
"Alex",
""
],
[
"Kanso",
"Eva",
""
]
] |
2505.06369 | Grigory Tarnopolsky | Erick Arguello Cruz, Igor R. Klebanov, Grigory Tarnopolsky, Yuan Xin | Yang-Lee Quantum Criticality in Various Dimensions | 44 pages, 20 figures. v2: minor improvements, references added. v3: minor improvements | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | The Yang-Lee universality class arises when imaginary magnetic field is tuned to its critical value in the paramagnetic phase of the $d<6$ Ising model. In $d=2$, this non-unitary Conformal Field Theory (CFT) is exactly solvable via the $M(2,5)$ minimal model. As found long ago by von Gehlen using Exact Diagonalization, the corresponding real-time, quantum critical behavior arises in the periodic Ising spin chain when the imaginary longitudinal magnetic field is tuned to its critical value from below. Even though the Hamiltonian is not Hermitian, the energy levels are real due to the $PT$ symmetry. In this paper, we explore the analogous quantum critical behavior in higher dimensional non-Hermitian Hamiltonians on regularized spheres $S^{d-1}$. For $d=3$, we use the recently invented, powerful fuzzy sphere method, as well as discretization by the platonic solids cube, icosahedron and dodecaherdron. The low-lying energy levels and structure constants we find are in agreement with expectations from the conformal symmetry. The energy levels are in good quantitative agreement with the high-temperature expansions and with Pad\'{e} extrapolations of the $6-\epsilon$ expansions in Fisher's $i\phi^3$ Euclidean field theory for the Yang-Lee criticality. In the course of this work, we clarify some aspects of matching between operators in this field theory and quasiprimary fields in the $M(2,5)$ minimal model. For $d=4$, we obtain new results by replacing the $S^3$ with the self-dual polytope called the $24$-cell. | [
{
"version": "v1",
"created": "Fri, 09 May 2025 18:24:15 GMT"
},
{
"version": "v2",
"created": "Fri, 16 May 2025 17:56:53 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 17:59:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cruz",
"Erick Arguello",
""
],
[
"Klebanov",
"Igor R.",
""
],
[
"Tarnopolsky",
"Grigory",
""
],
[
"Xin",
"Yuan",
""
]
] |
2505.06459 | Pablo Flores | Pablo Flores, Olga Graf, Pavlos Protopapas, Karim Pichara | Improved Uncertainty Quantification in Physics-Informed Neural Networks Using Error Bounds and Solution Bundles | null | null | null | null | cs.LG cs.AI physics.comp-ph stat.ML | http://creativecommons.org/licenses/by/4.0/ | Physics-Informed Neural Networks (PINNs) have been widely used to obtain solutions to various physical phenomena modeled as Differential Equations. As PINNs are not naturally equipped with mechanisms for Uncertainty Quantification, some work has been done to quantify the different uncertainties that arise when dealing with PINNs. In this paper, we use a two-step procedure to train Bayesian Neural Networks that provide uncertainties over the solutions to differential equation systems provided by PINNs. We use available error bounds over PINNs to formulate a heteroscedastic variance that improves the uncertainty estimation. Furthermore, we solve forward problems and utilize the obtained uncertainties when doing parameter estimation in inverse problems in cosmology. | [
{
"version": "v1",
"created": "Fri, 09 May 2025 22:40:39 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 00:12:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Flores",
"Pablo",
""
],
[
"Graf",
"Olga",
""
],
[
"Protopapas",
"Pavlos",
""
],
[
"Pichara",
"Karim",
""
]
] |
2505.06787 | Emir Cem Gezer | Emir Cem Gezer, Mael Korentin Ivan Moreau, Anders Sandneseng H{\o}gden, Dong Trong Nguyen, Roger Skjetne, Asgeir S{\o}rensen | Digital-physical testbed for ship autonomy studies in the Marine Cybernetics Laboratory basin | null | null | null | null | cs.RO cs.SY eess.SY | http://creativecommons.org/licenses/by-sa/4.0/ | The algorithms developed for Maritime Autonomous Surface Ships (MASS) are often challenging to test on actual vessels due to high operational costs and safety considerations. Simulations offer a cost-effective alternative and eliminate risks, but they may not accurately represent real-world dynamics for the given tasks. Utilizing small-scale model ships and robotic vessels in conjunction with a laboratory basin provides an accessible testing environment for the early stages of validation processes. However, designing and developing a model vessel for a single test can be costly and cumbersome, and researchers often lack access to such infrastructure. To address these challenges and enable streamlined testing, we have developed an in-house testbed that facilitates the development, testing, verification, and validation of MASS algorithms in a digital-physical laboratory. This infrastructure includes a set of small-scale model vessels, a simulation environment for each vessel, a comprehensive testbed environment, and a digital twin in Unity. With this, we aim to establish a full design and verification pipeline that starts with high-fidelity simulation models of each model vessel, to the model-scale testing in the laboratory basin, allowing possibilities for moving towards semi-fullscale validation with R/V milliAmpere1 and full-scale validation with R/V Gunnerus. In this work, we present our progress on the development of this testbed environment and its components, demonstrating its effectiveness in enabling ship guidance, navigation, and control (GNC), including autonomy. | [
{
"version": "v1",
"created": "Sat, 10 May 2025 23:48:32 GMT"
},
{
"version": "v2",
"created": "Wed, 28 May 2025 12:45:34 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 08:30:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gezer",
"Emir Cem",
""
],
[
"Moreau",
"Mael Korentin Ivan",
""
],
[
"Høgden",
"Anders Sandneseng",
""
],
[
"Nguyen",
"Dong Trong",
""
],
[
"Skjetne",
"Roger",
""
],
[
"Sørensen",
"Asgeir",
""
]
] |
2505.07960 | Alexander Schenkel | Marco Benini, Victor Carmona, Pieter Naaijkens and Alexander Schenkel | $C^\ast$-categorical prefactorization algebras for superselection sectors and topological order | v2: 39 pages. Added reference and minor clarifications | null | null | null | math-ph cond-mat.str-el math.MP math.QA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a conceptual and efficient geometric framework to encode the algebraic structures on the category of superselection sectors of an algebraic quantum field theory on the $n$-dimensional lattice $\mathbb{Z}^n$. It is shown that, under certain assumptions which are implied by Haag duality, the monoidal $C^\ast$-categories of localized superselection sectors carry the structure of a locally constant prefactorization algebra over the category of cone-shaped subsets of $\mathbb{Z}^n$. Employing techniques from higher algebra, one extracts from this datum an underlying locally constant prefactorization algebra defined on open disks in the cylinder $\mathbb{R}^1\times\mathbb{S}^{n-1}$. While the sphere $\mathbb{S}^{n-1}$ arises geometrically as the angular coordinates of cones, the origin of the line $\mathbb{R}^1$ is analytic and rooted in Haag duality. The usual braided (for $n=2$) or symmetric (for $n\geq 3$) monoidal $C^\ast$-categories of superselection sectors are recovered by removing a point of the sphere $\mathbb{R}^1\times(\mathbb{S}^{n-1}\setminus\mathrm{pt}) \cong\mathbb{R}^n$ and using the equivalence between $\mathbb{E}_n$-algebras and locally constant prefactorization algebras defined on open disks in $\mathbb{R}^n$. The non-trivial homotopy groups of spheres induce additional algebraic structures on these $\mathbb{E}_n$-monoidal $C^\ast$-categories, which in the case of $\mathbb{Z}^2$ is given by a braided monoidal self-equivalence arising geometrically as a kind of `holonomy' around the circle $\mathbb{S}^1$. The locally constant prefactorization algebra structures discovered in this work generalize, under some mild geometric conditions, to other discrete spaces and thereby provide a clear link between the geometry of the localization regions and the algebraic structures on the category of superselection sectors. | [
{
"version": "v1",
"created": "Mon, 12 May 2025 18:02:59 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:33:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Benini",
"Marco",
""
],
[
"Carmona",
"Victor",
""
],
[
"Naaijkens",
"Pieter",
""
],
[
"Schenkel",
"Alexander",
""
]
] |
2505.08834 | Muhammad Junaid Asif | Muhammad Junaid Asif | Crowd Scene Analysis using Deep Learning Techniques | MS Graduate Research Thesis | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Our research is focused on two main applications of crowd scene analysis crowd counting and anomaly detection In recent years a large number of researches have been presented in the domain of crowd counting We addressed two main challenges in this domain 1 Deep learning models are datahungry paradigms and always need a large amount of annotated data for the training of algorithm It is timeconsuming and costly task to annotate such large amount of data Selfsupervised training is proposed to deal with this challenge 2 MCNN consists of multicolumns of CNN with different sizes of filters by presenting a novel approach based on a combination of selfsupervised training and MultiColumn CNN This enables the model to learn features at different levels and makes it effective in dealing with challenges of occluded scenes nonuniform density complex backgrounds and scale invariation The proposed model was evaluated on publicly available data sets such as ShanghaiTech and UCFQNRF by means of MAE and MSE A spatiotemporal model based on VGG19 is proposed for crowd anomaly detection addressing challenges like lighting environmental conditions unexpected objects and scalability The model extracts spatial and temporal features allowing it to be generalized to realworld scenes Spatial features are learned using CNN while temporal features are learned using LSTM blocks The model works on binary classification and can detect normal or abnormal behavior The models performance is improved by replacing fully connected layers with dense residual blocks Experiments on the Hockey Fight dataset and SCVD dataset show our models outperform other stateoftheart approaches | [
{
"version": "v1",
"created": "Tue, 13 May 2025 05:29:30 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 03:58:31 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Asif",
"Muhammad Junaid",
""
]
] |
2505.10018 | Lijie Wang | Lijie Wang, Xiaoyi Zhong, Ziyi Xu, Kaixin Chai, Anke Zhao, Tianyu Zhao, Changjian Jiang, Qianhao Wang, Fei Gao | LEMON-Mapping: Loop-Enhanced Large-Scale Multi-Session Point Cloud Merging and Optimization for Globally Consistent Mapping | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-robot collaboration is becoming increasingly critical and presents significant challenges in modern robotics, especially for building a globally consistent, accurate map. Traditional multi-robot pose graph optimization (PGO) methods ensure basic global consistency but ignore the geometric structure of the map, and only use loop closures as constraints between pose nodes, leading to divergence and blurring in overlapping regions. To address this issue, we propose LEMON-Mapping, a loop-enhanced framework for large-scale, multi-session point cloud fusion and optimization. We re-examine the role of loops for multi-robot mapping and introduce three key innovations. First, we develop a robust loop processing mechanism that rejects outliers and a loop recall strategy to recover mistakenly removed but valid loops. Second, we introduce spatial bundle adjustment for multi-robot maps, reducing divergence and eliminating blurring in overlaps. Third, we design a PGO-based approach that leverages refined bundle adjustment constraints to propagate local accuracy to the entire map. We validate LEMON-Mapping on several public datasets and a self-collected dataset. The experimental results show superior mapping accuracy and global consistency of our framework compared to traditional merging methods. Scalability experiments also demonstrate its strong capability to handle scenarios involving numerous robots. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 06:59:48 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:07:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Lijie",
""
],
[
"Zhong",
"Xiaoyi",
""
],
[
"Xu",
"Ziyi",
""
],
[
"Chai",
"Kaixin",
""
],
[
"Zhao",
"Anke",
""
],
[
"Zhao",
"Tianyu",
""
],
[
"Jiang",
"Changjian",
""
],
[
"Wang",
"Qianhao",
""
],
[
"Gao",
"Fei",
""
]
] |
2505.10035 | Bi-Heng Liu | Xiao-Min Hu, Cen-Xiao Huang, Nicola d'Alessandro, Gabriele Cobucci, Chao Zhang, Yu Guo, Yun-Feng Huang, Chuan-Feng Li, Guang-Can Guo, Xiaoqin Gao, Marcus Huber, Armin Tavakoli, Bi-Heng Liu | Observation of Genuine High-dimensional Multi-partite Non-locality in Entangled Photon States | null | Nature Communications 16:5017 (2025) | 10.1038/s41467-025-59717-y | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Quantum information science has leaped forward with the exploration of high-dimensional quantum systems, offering greater potential than traditional qubits in quantum communication and quantum computing. To advance the field of high-dimensional quantum technology, a significant effort is underway to progressively enhance the entanglement dimension between two particles. An alternative effective strategy involves not only increasing the dimensionality but also expanding the number of particles that are entangled. We present an experimental study demonstrating multi-partite quantum non-locality beyond qubit constraints, thus moving into the realm of strongly entangled high-dimensional multi-particle quantum systems. In the experiment, quantum states were encoded in the path degree of freedom (DoF) and controlled via polarization, enabling efficient operations in a two-dimensional plane to prepare three- and four-particle Greenberger-Horne-Zeilinger (GHZ) states in three-level systems. Our experimental results reveal ways in which high-dimensional systems can surpass qubits in terms of violating local-hidden-variable theories. Our realization of multiple complex and high-quality entanglement technologies is an important primary step for more complex quantum computing and communication protocols. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 07:33:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hu",
"Xiao-Min",
""
],
[
"Huang",
"Cen-Xiao",
""
],
[
"d'Alessandro",
"Nicola",
""
],
[
"Cobucci",
"Gabriele",
""
],
[
"Zhang",
"Chao",
""
],
[
"Guo",
"Yu",
""
],
[
"Huang",
"Yun-Feng",
""
],
[
"Li",
"Chuan-Feng",
""
],
[
"Guo",
"Guang-Can",
""
],
[
"Gao",
"Xiaoqin",
""
],
[
"Huber",
"Marcus",
""
],
[
"Tavakoli",
"Armin",
""
],
[
"Liu",
"Bi-Heng",
""
]
] |
2505.10052 | Yuhang Guo | Yuhang Guo, Qiang Li, Yu Bao, Ziwen Pan, You Lv, Rhea Stewart, Adrian Hillier, Stephen Cottrell, Peter Baker, James Lord, Lei Liu, Zhijun Liang, Mengzhao Li, Mei Zhao, Gaobo Xu, Meichan Feng | Test of LGAD as Potential Next-Generation {\mu}SR Spectrometer Detectors | null | null | null | null | physics.ins-det hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Muon Spin Rotation/Relaxation/Resonance ($\mu$SR) is a versatile and powerful non-destructive technology for investigating the magnetic properties of materials at the microscopic level. The $\mu$SR technique typically utilizes fully spin polarized beams of positive muons generated at particle accelerator facilities and measures the evolution of the muon spin polarization inside a sample to extract information about the local magnetic environment in materials. With the development of accelerator technologies, intensities of muon beams are being continuously improved, which will cause a pile-up problem to the $\mu$SR spectrometer. The first muon source in China, named MELODY, is currently under construction and will be a pulsed source of muons operated at a repetition frequency of only 1 Hz due to limitations of the accelerator system at CSNS. Consequently, there is a strong motivation to operate MELODY at significantly higher muon intensities. This necessitates an upgrade of the detector system inside the spectrometer, which should be smaller and faster to accommodate the increased intensity per pulse of muons. The Low Gain Avalanche Diode (LGAD), characterized by a typical pulse width of 2 ns and a segmentation size in the centimeters range, has the potential to significantly improve the counting rates of $\mu$SR spectrometers that utilize a high intensity pulsed muon source. Thus, it is expected that the LGAD detector is a promising candidate to enhance the performance of $\mu$SR spectrometers at the new MELODY muon source.To validate this, tests on the LGAD were conducted at the ISIS pulsed muon source at the Rutherford Appleton Laboratory, UK. This paper will describe the setup of the candidate LGAD devices and the subsequent analysis of the experiment data. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 07:55:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 02:31:17 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Guo",
"Yuhang",
""
],
[
"Li",
"Qiang",
""
],
[
"Bao",
"Yu",
""
],
[
"Pan",
"Ziwen",
""
],
[
"Lv",
"You",
""
],
[
"Stewart",
"Rhea",
""
],
[
"Hillier",
"Adrian",
""
],
[
"Cottrell",
"Stephen",
""
],
[
"Baker",
"Peter",
""
],
[
"Lord",
"James",
""
],
[
"Liu",
"Lei",
""
],
[
"Liang",
"Zhijun",
""
],
[
"Li",
"Mengzhao",
""
],
[
"Zhao",
"Mei",
""
],
[
"Xu",
"Gaobo",
""
],
[
"Feng",
"Meichan",
""
]
] |
2505.10152 | Yikang Wei | Yikang Wei | Multi-Source Collaborative Style Augmentation and Domain-Invariant Learning for Federated Domain Generalization | IJCAI 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated domain generalization aims to learn a generalizable model from multiple decentralized source domains for deploying on the unseen target domain. The style augmentation methods have achieved great progress on domain generalization. However, the existing style augmentation methods either explore the data styles within isolated source domain or interpolate the style information across existing source domains under the data decentralization scenario, which leads to limited style space. To address this issue, we propose a Multi-source Collaborative Style Augmentation and Domain-invariant learning method (MCSAD) for federated domain generalization. Specifically, we propose a multi-source collaborative style augmentation module to generate data in the broader style space. Furthermore, we conduct domain-invariant learning between the original data and augmented data by cross-domain feature alignment within the same class and classes relation ensemble distillation between different classes to learn a domain-invariant model. By alternatively conducting collaborative style augmentation and domain-invariant learning, the model can generalize well on unseen target domain. Extensive experiments on multiple domain generalization datasets indicate that our method significantly outperforms the state-of-the-art federated domain generalization methods. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 10:26:17 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 10:41:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wei",
"Yikang",
""
]
] |
2505.10229 | Kun Yin | Kun Yin | Strong and weak convergence rates for fully coupled multiscale stochastic differential equations driven by $\alpha$-stable processes | 33 pages | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We first establish strong convergence rates for multiscale systems driven by $\alpha$-stable processes, with analyses constructed in two distinct scaling regimes. When addressing weak convergence rates of this system, we derive four averaged equations with respect to four scaling regimes. Notably, under sufficient H\"{o}lder regularity conditions on the time-dependent drifts of slow process, the strong convergence orders are related to the known optimal strong convergence order $1-\frac{1}{\alpha}$, and the weak convergence orders are 1. Our primary approach involves employing nonlocal Poisson equations to construct ``corrector equations" that effectively eliminate inhomogeneous terms. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 12:38:50 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:01:32 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yin",
"Kun",
""
]
] |
2505.10360 | Lars Maal{\o}e | Victor Petr\'en Bach Hansen and Lasse Krogsb{\o}ll and Jonas Lyngs{\o} and Mathias Baltzersen and Andreas Motzfeldt and Kevin Pelgrims and Lars Maal{\o}e | FactsR: A Safer Method for Producing High Quality Healthcare Documentation | null | null | null | null | cs.LG cs.AI stat.AP | http://creativecommons.org/licenses/by/4.0/ | There are now a multitude of AI-scribing solutions for healthcare promising the utilization of large language models for ambient documentation. However, these AI scribes still rely on one-shot, or few-shot prompts for generating notes after the consultation has ended, employing little to no reasoning. This risks long notes with an increase in hallucinations, misrepresentation of the intent of the clinician, and reliance on the proofreading of the clinician to catch errors. A dangerous combination for patient safety if vigilance is compromised by workload and fatigue. In this paper, we introduce a method for extracting salient clinical information in real-time alongside the healthcare consultation, denoted Facts, and use that information recursively to generate the final note. The FactsR method results in more accurate and concise notes by placing the clinician-in-the-loop of note generation, while opening up new use cases within real-time decision support. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 14:51:22 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:39:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hansen",
"Victor Petrén Bach",
""
],
[
"Krogsbøll",
"Lasse",
""
],
[
"Lyngsø",
"Jonas",
""
],
[
"Baltzersen",
"Mathias",
""
],
[
"Motzfeldt",
"Andreas",
""
],
[
"Pelgrims",
"Kevin",
""
],
[
"Maaløe",
"Lars",
""
]
] |
2505.10462 | Giacomo Marmorini | Yoshiyuki Kakihara, Daisuke Yamamoto, Giacomo Marmorini | Compressed sensing quantum state tomography for qudits: A comparison of Gell-Mann and Heisenberg-Weyl observable bases | 13 pages, 4 figures; v2 minor changes, typos fixed | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum state tomography (QST) is an essential technique for reconstructing the density matrix of an unknown quantum state from measurement data, crucial for quantum information processing. However, conventional QST requires an exponentially growing number of measurements as the system dimension increases, posing a significant challenge for high-dimensional systems. To mitigate this issue, compressed sensing quantum state tomography (CS-QST) has been proposed, significantly reducing the required number of measurements. In this study, we investigate the impact of basis selection in CS-QST for qudit systems, which are fundamental to high-dimensional quantum information processing. Specifically, we compare the efficiency of the generalized Gell-Mann (GGM) and Heisenberg-Weyl observable (HWO) bases by numerically reconstructing density matrices and evaluating reconstruction accuracy using fidelity and trace distance metrics. Our results demonstrate that, while both bases allow for successful density matrix reconstruction, the HWO basis becomes more efficient as the qudit dimension increases. Furthermore, we find the best fitting curves that estimate the number of measurement operators required to achieve a fidelity of at least 95%. These findings highlight the significance of basis selection in CS-QST and provide valuable insights for optimizing measurement strategies in high-dimensional quantum state tomography. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 16:17:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 10:07:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kakihara",
"Yoshiyuki",
""
],
[
"Yamamoto",
"Daisuke",
""
],
[
"Marmorini",
"Giacomo",
""
]
] |
2505.10519 | P. M. Aronow | P. M. Aronow, Austin Jang, Molly Offer-Westort | On the Foundations of the Design-Based Approach | null | null | null | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The design-based paradigm may be adopted in causal inference and survey sampling when we assume Rubin's stable unit treatment value assumption (SUTVA) or impose similar frameworks. While often taken for granted, such assumptions entail strong claims about the data generating process. We develop an alternative design-based approach: we first invoke a generalized, non-parametric model that allows for unrestricted forms of interference, such as spillover. We define a new set of inferential targets and discuss their interpretation under SUTVA and a weaker assumption that we call the No Unmodeled Revealable Variation Assumption (NURVA). We then reconstruct the standard paradigm, reconsidering SUTVA at the end rather than assuming it at the beginning. Despite its similarity to SUTVA, we demonstrate the practical insufficiency of NURVA for identifying substantively interesting quantities. In so doing, we provide clarity on the nature and importance of SUTVA for applied research. | [
{
"version": "v1",
"created": "Thu, 15 May 2025 17:26:43 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:02:36 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Aronow",
"P. M.",
""
],
[
"Jang",
"Austin",
""
],
[
"Offer-Westort",
"Molly",
""
]
] |
2505.10981 | Yexiang Liu | Yexiang Liu, Zekun Li, Zhi Fang, Nan Xu, Ran He, Tieniu Tan | Rethinking the Role of Prompting Strategies in LLM Test-Time Scaling: A Perspective of Probability Theory | ACL 2025 Main, 33 pages, 51 figures | null | null | null | cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recently, scaling test-time compute on Large Language Models (LLM) has garnered wide attention. However, there has been limited investigation of how various reasoning prompting strategies perform as scaling. In this paper, we focus on a standard and realistic scaling setting: majority voting. We systematically conduct experiments on 6 LLMs $\times$ 8 prompting strategies $\times$ 6 benchmarks. Experiment results consistently show that as the sampling time and computational overhead increase, complicated prompting strategies with superior initial performance gradually fall behind simple Chain-of-Thought. We analyze this phenomenon and provide theoretical proofs. Additionally, we propose a probabilistic method to efficiently predict scaling performance and identify the best prompting strategy under large sampling times, eliminating the need for resource-intensive inference processes in practical applications. Furthermore, we introduce two ways derived from our theoretical analysis to significantly improve the scaling performance. We hope that our research can promote to re-examine the role of complicated prompting, unleash the potential of simple prompting strategies, and provide new insights for enhancing test-time scaling performance. Code is available at https://github.com/MraDonkey/rethinking_prompting. | [
{
"version": "v1",
"created": "Fri, 16 May 2025 08:28:57 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:27:57 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Yexiang",
""
],
[
"Li",
"Zekun",
""
],
[
"Fang",
"Zhi",
""
],
[
"Xu",
"Nan",
""
],
[
"He",
"Ran",
""
],
[
"Tan",
"Tieniu",
""
]
] |
2505.11256 | Ruediger Klingeler | Ning Yuan, Erik Walendy, Nour Maraytta, Waldemar Hergett, Luca Bischof, Michael Merz, R\"udiger Klingeler | Anisotropic magnetic phase diagrams, tricriticality, and spin-reorientation in high-pressure grown SmCrO$_3$ single crystals | null | Phys. Rev. B 111, 174437 (2025) | 10.1103/PhysRevB.111.174437 | null | cond-mat.str-el cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-nd/4.0/ | SmCrO$_3$ single crystals were successfully grown utilizing the high-pressure optical floating-zone method and their crystal structure, magnetization behavior, and magnetic phase diagrams were thoroughly investigated. Magnetic studies were conducted for fields applied along all principal crystallographic directions, with measurements taken at temperatures as low as 0.4 K and magnetic fields up to 14 T. The single crystal growth parameters are reported and the orthorhombic structure with the centrosymmetric space group $Pbnm$ is confirmed. Long-range order of the Cr$^{3+}$ and Sm$^{3+}$ magnetic sublattices evolves at $T_{\rm N}$ = 192 K and $T_{\rm N2}$=3 K, respectively. In contrast to previous studies on polycrystals our single crystal data imply a discontinuous and one-step spin-reorientation (SR) of net magnetic moments $\tilde{M}$ from the $c$ axis into the $ab$ plane at zero magnetic field at $T_{\rm SR}$=33 K. Its discontinuous nature is maintained if $B$ is applied $||c$ axis but tricritical behavior and a triple point is found for $B||a$ axis. While our data are consistent with the magnetic representation $\Gamma_4$ for $T > T_{\mathrm {SR}}$, the size and in-plane direction of the observed net magnetic moment disagree to previously proposed spin configurations, i.e., $\Gamma_1$ and $\Gamma_2$, for the spin-reoriented phases. In general, our high-quality single crystals enable us to revisit the phase diagram and to clarify the complex magnetism in SmCrO3 arising from the interplay of anisotropic 3$d$ and 4$f$ magnetic sublattices. | [
{
"version": "v1",
"created": "Fri, 16 May 2025 13:49:36 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yuan",
"Ning",
""
],
[
"Walendy",
"Erik",
""
],
[
"Maraytta",
"Nour",
""
],
[
"Hergett",
"Waldemar",
""
],
[
"Bischof",
"Luca",
""
],
[
"Merz",
"Michael",
""
],
[
"Klingeler",
"Rüdiger",
""
]
] |
2505.11331 | Arindam Mitra | Arindam Mitra, Sumit Mukherjee, Changhyoup Lee | Distance-based measures and Epsilon-measures for measurement-based quantum resources | 15 pages, typos fixed | null | null | null | quant-ph math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | Quantum resource theories provide a structured and elegant framework for quantifying quantum resources. While state-based resource theories have been extensively studied, their measurement-based resource theories remain relatively underexplored. In practical scenarios where a quantum state or a set of measurements is only partially known, conventional resource measures often fall short in capturing the resource content. In such cases, \epsilon-measures offer a robust alternative, making them particularly valuable. In this work, we investigate the quantification of measurement-based resources using distance-based measures, followed by a detailed analysis of the mathematical properties of \epsilon-measures. We also extend our analysis by exploring the connections between \epsilon-measures and some key quantities relevant to resource manipulation tasks. Importantly, the analysis of resources based on sets of measurements are tedious compared to that of single measurements as the former allows more general transformations such as controlled implementation. Yet our framework applies not only to resources associated with individual measurements but also to those arising from sets of measurements. In short, our analysis is applicable to existing resource theories of measurements and has the potential to be useful for all resource theories of measurements that are yet to be developed. | [
{
"version": "v1",
"created": "Fri, 16 May 2025 14:56:28 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:20:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mitra",
"Arindam",
""
],
[
"Mukherjee",
"Sumit",
""
],
[
"Lee",
"Changhyoup",
""
]
] |
2505.11441 | Xianzhen Luo | Xianzhen Luo, Shijie Xuyang, Tianhao Cheng, Zheng Chu, Houyi Li, ziqi wang, Siming Huang, Qingfu Zhu, Qiufeng Wang, Xiangyu Zhang, Shuigeng Zhou, Wanxiang Che | Is Compression Really Linear with Code Intelligence? | work in progress | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding the relationship between data compression and the capabilities of Large Language Models (LLMs) is crucial, especially in specialized domains like code intelligence. Prior work posited a linear relationship between compression and general intelligence. However, it overlooked the multifaceted nature of code that encompasses diverse programming languages and tasks, and struggled with fair evaluation of modern Code LLMs. We address this by evaluating a diverse array of open-source Code LLMs on comprehensive multi-language, multi-task code benchmarks. To address the challenge of efficient and fair evaluation of pre-trained LLMs' code intelligence, we introduce \textit{Format Annealing}, a lightweight, transparent training methodology designed to assess the intrinsic capabilities of these pre-trained models equitably. Compression efficacy, measured as bits-per-character (BPC), is determined using a novel, large-scale, and previously unseen code validation set derived from GitHub. Our empirical results reveal a fundamental logarithmic relationship between measured code intelligence and BPC. This finding refines prior hypotheses of linearity, which we suggest are likely observations of the logarithmic curve's tail under specific, limited conditions. Our work provides a more nuanced understanding of compression's role in developing code intelligence and contributes a robust evaluation framework in the code domain. | [
{
"version": "v1",
"created": "Fri, 16 May 2025 16:59:14 GMT"
},
{
"version": "v2",
"created": "Sun, 25 May 2025 12:11:58 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 06:01:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Luo",
"Xianzhen",
""
],
[
"Xuyang",
"Shijie",
""
],
[
"Cheng",
"Tianhao",
""
],
[
"Chu",
"Zheng",
""
],
[
"Li",
"Houyi",
""
],
[
"wang",
"ziqi",
""
],
[
"Huang",
"Siming",
""
],
[
"Zhu",
"Qingfu",
""
],
[
"Wang",
"Qiufeng",
""
],
[
"Zhang",
"Xiangyu",
""
],
[
"Zhou",
"Shuigeng",
""
],
[
"Che",
"Wanxiang",
""
]
] |
2505.11626 | Parul Kalra | Udita Patel, Rutu Mulkar, Jay Roberts, Cibi Chakravarthy Senthilkumar, Sujay Gandhi, Xiaofei Zheng, Naumaan Nayyar, Parul Kalra, Rafael Castrillo | THELMA: Task Based Holistic Evaluation of Large Language Model Applications-RAG Question Answering | Added author | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We propose THELMA (Task Based Holistic Evaluation of Large Language Model Applications), a reference free framework for RAG (Retrieval Augmented generation) based question answering (QA) applications. THELMA consist of six interdependent metrics specifically designed for holistic, fine grained evaluation of RAG QA applications. THELMA framework helps developers and application owners evaluate, monitor and improve end to end RAG QA pipelines without requiring labelled sources or reference responses.We also present our findings on the interplay of the proposed THELMA metrics, which can be interpreted to identify the specific RAG component needing improvement in QA applications. | [
{
"version": "v1",
"created": "Fri, 16 May 2025 18:42:04 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 19:21:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Patel",
"Udita",
""
],
[
"Mulkar",
"Rutu",
""
],
[
"Roberts",
"Jay",
""
],
[
"Senthilkumar",
"Cibi Chakravarthy",
""
],
[
"Gandhi",
"Sujay",
""
],
[
"Zheng",
"Xiaofei",
""
],
[
"Nayyar",
"Naumaan",
""
],
[
"Kalra",
"Parul",
""
],
[
"Castrillo",
"Rafael",
""
]
] |
2505.12532 | Ahmet Bilican | Ahmet Bilican, M. Ak{\i}n Y{\i}lmaz, A. Murat Tekalp, R. G\"okberk Cinbi\c{s} | Exploring Sparsity for Parameter Efficient Fine Tuning Using Wavelets | null | null | null | null | cs.CV cs.AI cs.LG eess.IV eess.SP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Efficiently adapting large foundation models is critical, especially with tight compute and memory budgets. Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA offer limited granularity and effectiveness in few-parameter regimes. We propose Wavelet Fine-Tuning (WaveFT), a novel PEFT method that learns highly sparse updates in the wavelet domain of residual matrices. WaveFT allows precise control of trainable parameters, offering fine-grained capacity adjustment and excelling with remarkably low parameter count, potentially far fewer than LoRA's minimum, ideal for extreme parameter-efficient scenarios. Evaluated on personalized text-to-image generation using Stable Diffusion XL as baseline, WaveFT significantly outperforms LoRA and other PEFT methods, especially at low parameter counts; achieving superior subject fidelity, prompt alignment, and image diversity. | [
{
"version": "v1",
"created": "Sun, 18 May 2025 20:20:32 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 21:46:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bilican",
"Ahmet",
""
],
[
"Yılmaz",
"M. Akın",
""
],
[
"Tekalp",
"A. Murat",
""
],
[
"Cinbiş",
"R. Gökberk",
""
]
] |
2505.12894 | Le Cheng | Le Cheng, Peican Zhu, Yangming Guo, Keke Tang, Chao Gao, Zhen Wang | HyperDet: Source Detection in Hypergraphs via Interactive Relationship Construction and Feature-rich Attention Fusion | Accepted by IJCAI25 | null | null | null | cs.SI cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hypergraphs offer superior modeling capabilities for social networks, particularly in capturing group phenomena that extend beyond pairwise interactions in rumor propagation. Existing approaches in rumor source detection predominantly focus on dyadic interactions, which inadequately address the complexity of more intricate relational structures. In this study, we present a novel approach for Source Detection in Hypergraphs (HyperDet) via Interactive Relationship Construction and Feature-rich Attention Fusion. Specifically, our methodology employs an Interactive Relationship Construction module to accurately model both the static topology and dynamic interactions among users, followed by the Feature-rich Attention Fusion module, which autonomously learns node features and discriminates between nodes using a self-attention mechanism, thereby effectively learning node representations under the framework of accurately modeled higher-order relationships. Extensive experimental validation confirms the efficacy of our HyperDet approach, showcasing its superiority relative to current state-of-the-art methods. | [
{
"version": "v1",
"created": "Mon, 19 May 2025 09:27:46 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:44:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cheng",
"Le",
""
],
[
"Zhu",
"Peican",
""
],
[
"Guo",
"Yangming",
""
],
[
"Tang",
"Keke",
""
],
[
"Gao",
"Chao",
""
],
[
"Wang",
"Zhen",
""
]
] |
2505.12910 | Le Cheng | Le Cheng, Peican Zhu, Yangming Guo, Chao Gao, Zhen Wang, Keke Tang | SourceDetMamba: A Graph-aware State Space Model for Source Detection in Sequential Hypergraphs | Accepted by IJCAI25 | null | null | null | cs.SI cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Source detection on graphs has demonstrated high efficacy in identifying rumor origins. Despite advances in machine learning-based methods, many fail to capture intrinsic dynamics of rumor propagation. In this work, we present SourceDetMamba: A Graph-aware State Space Model for Source Detection in Sequential Hypergraphs, which harnesses the recent success of the state space model Mamba, known for its superior global modeling capabilities and computational efficiency, to address this challenge. Specifically, we first employ hypergraphs to model high-order interactions within social networks. Subsequently, temporal network snapshots generated during the propagation process are sequentially fed in reverse order into Mamba to infer underlying propagation dynamics. Finally, to empower the sequential model to effectively capture propagation patterns while integrating structural information, we propose a novel graph-aware state update mechanism, wherein the state of each node is propagated and refined by both temporal dependencies and topological context. Extensive evaluations on eight datasets demonstrate that SourceDetMamba consistently outperforms state-of-the-art approaches. | [
{
"version": "v1",
"created": "Mon, 19 May 2025 09:45:27 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:57:51 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cheng",
"Le",
""
],
[
"Zhu",
"Peican",
""
],
[
"Guo",
"Yangming",
""
],
[
"Gao",
"Chao",
""
],
[
"Wang",
"Zhen",
""
],
[
"Tang",
"Keke",
""
]
] |
2505.13066 | Yueyun Ouyang | Yueyun Ouyang, Feng Ding, Jun Yang | Retention of surface water on tidally locked rocky planets in the Venus zone around M dwarfs | 19 pages, 3 figures, 1 table, accepted for publication in ApJL | Astrophys. J. Lett. 985 (2025) L43 | 10.3847/2041-8213/adda4b | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Terrestrial planets within the Venus zone surrounding M dwarf stars can retain surface ice caps on the perpetual dark side if atmospheric heat transport is inefficient, {as suggested by previous global climate simulations \citep[e.g.,][]{leconte2013}.} This condition is {proposed} to play a role in the potential regional habitability of these planets. However, the amount of surface ice may be limited by considering the water condensed from the steam atmosphere in a runaway greenhouse state, and the physical mechanism for triggering the condensation process is not clear. Here, we use a two-column moist radiative-convective-subsiding model to investigate the water condensation process on tidally locked planets from the runaway greenhouse state. We find that the water condensation process is characterized by two distinct equilibrium states under the same {incoming stellar flux}. The initiation of condensation corresponds to a warm, unstable state exhibiting positive Planck feedback, whereas the termination phase corresponds to a cold, stable state exhibiting negative Planck feedback. We further show that the surface water mass in the collapsed state {decreases with the} incoming stellar flux, background surface pressure, and optical thickness of non-condensible greenhouse gases, with a global equivalent depth of less than $\sim 20$ cm. Our two-column approach provides a straightforward way to understand the water evolution on Venus zone planets around M dwarfs. | [
{
"version": "v1",
"created": "Mon, 19 May 2025 12:57:20 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 03:10:35 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ouyang",
"Yueyun",
""
],
[
"Ding",
"Feng",
""
],
[
"Yang",
"Jun",
""
]
] |
2505.13168 | Marc Kegel | Marc Kegel and Lisa Piccirillo | Knots that share four surgeries | 22 pages, 18 figures, 1 table; V2: Added supporting code and data to the arXiv | null | null | null | math.GT math.SG | http://creativecommons.org/licenses/by/4.0/ | Distinct knots K, K' can sometimes share a common p/q-framed Dehn surgery. A folk conjecture held that for a fixed pair of knots, this can occur for at most one value of p/q. We disprove this conjecture by constructing pairs of distinct knots K,K' that have common Dehn surgeries for four distinct slopes. We also construct non-isotopic Legendrian knots K,K' that have contactomorphic contact (+1)-and (-1)-surgeries, disproving an analogous conjecture in contact geometry. | [
{
"version": "v1",
"created": "Mon, 19 May 2025 14:26:19 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:26:06 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kegel",
"Marc",
""
],
[
"Piccirillo",
"Lisa",
""
]
] |
2505.13469 | Aayam Bansal | Aayam Bansal | Algorithmic Tradeoffs in Fair Lending: Profitability, Compliance, and Long-Term Impact | 8 pages | null | null | null | cs.CY cs.AI cs.CE cs.LG | http://creativecommons.org/licenses/by/4.0/ | As financial institutions increasingly rely on machine learning models to automate lending decisions, concerns about algorithmic fairness have risen. This paper explores the tradeoff between enforcing fairness constraints (such as demographic parity or equal opportunity) and maximizing lender profitability. Through simulations on synthetic data that reflects real-world lending patterns, we quantify how different fairness interventions impact profit margins and default rates. Our results demonstrate that equal opportunity constraints typically impose lower profit costs than demographic parity, but surprisingly, removing protected attributes from the model (fairness through unawareness) outperforms explicit fairness interventions in both fairness and profitability metrics. We further identify the specific economic conditions under which fair lending becomes profitable and analyze the feature-specific drivers of unfairness. These findings offer practical guidance for designing lending algorithms that balance ethical considerations with business objectives. | [
{
"version": "v1",
"created": "Thu, 08 May 2025 19:18:33 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:51:13 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bansal",
"Aayam",
""
]
] |
2505.14151 | Jiaming Li | Jiaming Li, Sheng Wang, Xin Wang, Yitao Zhu, Honglin Xiong, Zixu Zhuang and Qian Wang | ReactDiff: Latent Diffusion for Facial Reaction Generation | Accepted by Neural Networks | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given the audio-visual clip of the speaker, facial reaction generation aims to predict the listener's facial reactions. The challenge lies in capturing the relevance between video and audio while balancing appropriateness, realism, and diversity. While prior works have mostly focused on uni-modal inputs or simplified reaction mappings, recent approaches such as PerFRDiff have explored multi-modal inputs and the one-to-many nature of appropriate reaction mappings. In this work, we propose the Facial Reaction Diffusion (ReactDiff) framework that uniquely integrates a Multi-Modality Transformer with conditional diffusion in the latent space for enhanced reaction generation. Unlike existing methods, ReactDiff leverages intra- and inter-class attention for fine-grained multi-modal interaction, while the latent diffusion process between the encoder and decoder enables diverse yet contextually appropriate outputs. Experimental results demonstrate that ReactDiff significantly outperforms existing approaches, achieving a facial reaction correlation of 0.26 and diversity score of 0.094 while maintaining competitive realism. The code is open-sourced at \href{https://github.com/Hunan-Tiger/ReactDiff}{github}. | [
{
"version": "v1",
"created": "Tue, 20 May 2025 10:01:37 GMT"
},
{
"version": "v2",
"created": "Fri, 23 May 2025 08:51:27 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 04:30:30 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Jiaming",
""
],
[
"Wang",
"Sheng",
""
],
[
"Wang",
"Xin",
""
],
[
"Zhu",
"Yitao",
""
],
[
"Xiong",
"Honglin",
""
],
[
"Zhuang",
"Zixu",
""
],
[
"Wang",
"Qian",
""
]
] |
2505.14470 | Or Tal | Nadav Har-Tuv, Or Tal, Yossi Adi | PAST: Phonetic-Acoustic Speech Tokenizer | null | null | null | null | cs.SD cs.CL cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present PAST, a novel end-to-end framework that jointly models phonetic information alongside signal reconstruction, eliminating the need for external pretrained models. Unlike previous approaches that rely on pretrained self-supervised models, PAST employs supervised phonetic data, directly integrating domain knowledge into the tokenization process via auxiliary tasks. Additionally, we introduce a streamable, causal variant of PAST, enabling real-time speech applications. Results demonstrate that PAST surpasses existing evaluated baseline tokenizers across common evaluation metrics, including phonetic representation and speech reconstruction. Notably, PAST also achieves superior performance when serving as a speech representation for speech language models, further highlighting its effectiveness as a foundation for spoken language generation. To foster further research, we release the full implementation. For code, model checkpoints, and samples see: https://pages.cs.huji.ac.il/adiyoss-lab/PAST | [
{
"version": "v1",
"created": "Tue, 20 May 2025 15:05:14 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:23:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Har-Tuv",
"Nadav",
""
],
[
"Tal",
"Or",
""
],
[
"Adi",
"Yossi",
""
]
] |
2505.14590 | Huihao Jing | Huihao Jing, Haoran Li, Wenbin Hu, Qi Hu, Heli Xu, Tianshu Chu, Peizhao Hu, Yangqiu Song | MCIP: Protecting MCP Safety via Model Contextual Integrity Protocol | 17 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As Model Context Protocol (MCP) introduces an easy-to-use ecosystem for users and developers, it also brings underexplored safety risks. Its decentralized architecture, which separates clients and servers, poses unique challenges for systematic safety analysis. This paper proposes a novel framework to enhance MCP safety. Guided by the MAESTRO framework, we first analyze the missing safety mechanisms in MCP, and based on this analysis, we propose the Model Contextual Integrity Protocol (MCIP), a refined version of MCP that addresses these gaps. Next, we develop a fine-grained taxonomy that captures a diverse range of unsafe behaviors observed in MCP scenarios. Building on this taxonomy, we develop benchmark and training data that support the evaluation and improvement of LLMs' capabilities in identifying safety risks within MCP interactions. Leveraging the proposed benchmark and training data, we conduct extensive experiments on state-of-the-art LLMs. The results highlight LLMs' vulnerabilities in MCP interactions and demonstrate that our approach substantially improves their safety performance. | [
{
"version": "v1",
"created": "Tue, 20 May 2025 16:41:45 GMT"
},
{
"version": "v2",
"created": "Wed, 21 May 2025 14:48:40 GMT"
},
{
"version": "v3",
"created": "Fri, 30 May 2025 06:55:41 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 06:21:46 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jing",
"Huihao",
""
],
[
"Li",
"Haoran",
""
],
[
"Hu",
"Wenbin",
""
],
[
"Hu",
"Qi",
""
],
[
"Xu",
"Heli",
""
],
[
"Chu",
"Tianshu",
""
],
[
"Hu",
"Peizhao",
""
],
[
"Song",
"Yangqiu",
""
]
] |
2505.14613 | Jason Hartford | Emmanuel Noutahi, Jason Hartford, Prudencio Tossou, Shawn Whitfield, Alisandra K. Denton, Cas Wognum, Kristina Ulicna, Michael Craig, Jonathan Hsu, Michael Cuccarese, Emmanuel Bengio, Dominique Beaini, Christopher Gibson, Daniel Cohen, Berton Earnshaw | Virtual Cells: Predict, Explain, Discover | null | null | null | null | cs.LG q-bio.QM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Drug discovery is fundamentally a process of inferring the effects of treatments on patients, and would therefore benefit immensely from computational models that can reliably simulate patient responses, enabling researchers to generate and test large numbers of therapeutic hypotheses safely and economically before initiating costly clinical trials. Even a more specific model that predicts the functional response of cells to a wide range of perturbations would be tremendously valuable for discovering safe and effective treatments that successfully translate to the clinic. Creating such virtual cells has long been a goal of the computational research community that unfortunately remains unachieved given the daunting complexity and scale of cellular biology. Nevertheless, recent advances in AI, computing power, lab automation, and high-throughput cellular profiling provide new opportunities for reaching this goal. In this perspective, we present a vision for developing and evaluating virtual cells that builds on our experience at Recursion. We argue that in order to be a useful tool to discover novel biology, virtual cells must accurately predict the functional response of a cell to perturbations and explain how the predicted response is a consequence of modifications to key biomolecular interactions. We then introduce key principles for designing therapeutically-relevant virtual cells, describe a lab-in-the-loop approach for generating novel insights with them, and advocate for biologically-grounded benchmarks to guide virtual cell development. Finally, we make the case that our approach to virtual cells provides a useful framework for building other models at higher levels of organization, including virtual patients. We hope that these directions prove useful to the research community in developing virtual models optimized for positive impact on drug discovery outcomes. | [
{
"version": "v1",
"created": "Tue, 20 May 2025 16:59:24 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 16:15:27 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 08:07:44 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Noutahi",
"Emmanuel",
""
],
[
"Hartford",
"Jason",
""
],
[
"Tossou",
"Prudencio",
""
],
[
"Whitfield",
"Shawn",
""
],
[
"Denton",
"Alisandra K.",
""
],
[
"Wognum",
"Cas",
""
],
[
"Ulicna",
"Kristina",
""
],
[
"Craig",
"Michael",
""
],
[
"Hsu",
"Jonathan",
""
],
[
"Cuccarese",
"Michael",
""
],
[
"Bengio",
"Emmanuel",
""
],
[
"Beaini",
"Dominique",
""
],
[
"Gibson",
"Christopher",
""
],
[
"Cohen",
"Daniel",
""
],
[
"Earnshaw",
"Berton",
""
]
] |
2505.14884 | Susav Shrestha | Susav Shrestha, Brad Settlemyer, Nikoli Dryden, Narasimha Reddy | Polar Sparsity: High Throughput Batched LLM Inferencing with Scalable Contextual Sparsity | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Accelerating large language model (LLM) inference is critical for real-world deployments requiring high throughput and low latency. Contextual sparsity, where each token dynamically activates only a small subset of the model parameters, shows promise but does not scale to large batch sizes due to union of active neurons quickly approaching dense computation. We introduce Polar Sparsity, highlighting a key shift in sparsity importance from MLP to Attention layers as we scale batch size and sequence length. While MLP layers become more compute-efficient under batching, their sparsity vanishes. In contrast, attention becomes increasingly more expensive at scale, while their head sparsity remains stable and batch-invariant. We develop hardware-efficient, sparsity-aware GPU kernels for selective MLP and Attention computations, delivering up to \(2.2\times\) end-to-end speedups for models like OPT, LLaMA-2 \& 3, across various batch sizes and sequence lengths without compromising accuracy. To our knowledge, this is the first work to demonstrate that contextual sparsity can scale effectively to large batch sizes, delivering substantial inference acceleration with minimal changes, making Polar Sparsity practical for large-scale, high-throughput LLM deployment systems. Our code is available at: https://github.com/susavlsh10/Polar-Sparsity. | [
{
"version": "v1",
"created": "Tue, 20 May 2025 20:15:42 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:28:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shrestha",
"Susav",
""
],
[
"Settlemyer",
"Brad",
""
],
[
"Dryden",
"Nikoli",
""
],
[
"Reddy",
"Narasimha",
""
]
] |
2505.15053 | Tatsuya Abe | Tatsuya Abe | Owicki--Gries Logic for Timestamp Semantics | null | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Whereas an extension with non-interference of Hoare logic for sequential programs Owicki--Gries logic ensures the correctness of concurrent programs on strict consistency, it is unsound to weak memory models adopted by modern computer architectures and specifications of programming languages. This paper proposes a novel non-interference notion and provides concurrent program logic sound to timestamp semantics corresponding to a weak memory model that allows delays in the effects of store instructions. This paper reports three theoretically interesting techniques for modifying non-interference to support delays in the effects of store instructions. The techniques contribute to a better understanding of constructing concurrent program logic. | [
{
"version": "v1",
"created": "Wed, 21 May 2025 03:20:16 GMT"
},
{
"version": "v2",
"created": "Sun, 25 May 2025 11:01:54 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 00:37:33 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Abe",
"Tatsuya",
""
]
] |
2505.15965 | Gowtham Premananth | Gowtham Premananth, Vinith Kugathasan, Carol Espy-Wilson | Analyzing the Impact of Accent on English Speech: Acoustic and Articulatory Perspectives | Accepted to be presented at Interspeech 2025 | null | null | null | eess.AS eess.SP | http://creativecommons.org/licenses/by/4.0/ | Advancements in AI-driven speech-based applications have transformed diverse industries ranging from healthcare to customer service. However, the increasing prevalence of non-native accented speech in global interactions poses significant challenges for speech-processing systems, which are often trained on datasets dominated by native speech. This study investigates accented English speech through articulatory and acoustic analysis, identifying simpler coordination patterns and higher average pitch than native speech. Using eigenspectra and Vocal Tract Variable-based coordination features, we establish an efficient method for quantifying accent strength without relying on resource-intensive phonetic transcriptions. Our findings provide a new avenue for research on the impacts of accents on speech intelligibility and offer insights for developing inclusive, robust speech processing systems that accommodate diverse linguistic communities. | [
{
"version": "v1",
"created": "Wed, 21 May 2025 19:31:40 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:36:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Premananth",
"Gowtham",
""
],
[
"Kugathasan",
"Vinith",
""
],
[
"Espy-Wilson",
"Carol",
""
]
] |
2505.16044 | Gowtham Premananth | Gowtham Premananth, Philip Resnik, Sonia Bansal, Deanna L.Kelly, Carol Espy-Wilson | Multimodal Biomarkers for Schizophrenia: Towards Individual Symptom Severity Estimation | Accepted to be presented at Interspeech 2025 | null | null | null | eess.AS cs.LG eess.IV eess.SP | http://creativecommons.org/licenses/by/4.0/ | Studies on schizophrenia assessments using deep learning typically treat it as a classification task to detect the presence or absence of the disorder, oversimplifying the condition and reducing its clinical applicability. This traditional approach overlooks the complexity of schizophrenia, limiting its practical value in healthcare settings. This study shifts the focus to individual symptom severity estimation using a multimodal approach that integrates speech, video, and text inputs. We develop unimodal models for each modality and a multimodal framework to improve accuracy and robustness. By capturing a more detailed symptom profile, this approach can help in enhancing diagnostic precision and support personalized treatment, offering a scalable and objective tool for mental health assessment. | [
{
"version": "v1",
"created": "Wed, 21 May 2025 21:55:35 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:35:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Premananth",
"Gowtham",
""
],
[
"Resnik",
"Philip",
""
],
[
"Bansal",
"Sonia",
""
],
[
"Kelly",
"Deanna L.",
""
],
[
"Espy-Wilson",
"Carol",
""
]
] |
2505.16196 | Xuewu Lin | Xuewu Lin, Tianwei Lin, Lichao Huang, Hongyu Xie, Yiwei Jin, Keyu Li, Zhizhong Su | SEM: Enhancing Spatial Understanding for Robust Robot Manipulation | null | null | null | null | cs.RO cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | A key challenge in robot manipulation lies in developing policy models with strong spatial understanding, the ability to reason about 3D geometry, object relations, and robot embodiment. Existing methods often fall short: 3D point cloud models lack semantic abstraction, while 2D image encoders struggle with spatial reasoning. To address this, we propose SEM (Spatial Enhanced Manipulation model), a novel diffusion-based policy framework that explicitly enhances spatial understanding from two complementary perspectives. A spatial enhancer augments visual representations with 3D geometric context, while a robot state encoder captures embodiment-aware structure through graphbased modeling of joint dependencies. By integrating these modules, SEM significantly improves spatial understanding, leading to robust and generalizable manipulation across diverse tasks that outperform existing baselines. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 04:00:12 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:29:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Lin",
"Xuewu",
""
],
[
"Lin",
"Tianwei",
""
],
[
"Huang",
"Lichao",
""
],
[
"Xie",
"Hongyu",
""
],
[
"Jin",
"Yiwei",
""
],
[
"Li",
"Keyu",
""
],
[
"Su",
"Zhizhong",
""
]
] |
2505.16488 | Andrey Dymov | Andrey Dymov | Kinetic approximation for equations of discrete turbulence in the subcritical case | 46 pages | null | null | null | math-ph math.AP math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider a damped/driven cubic NLS equation on a torus under the limit when first the amplitude of solutions goes to zero and then the period of the torus goes to infinity. We suggest another proof of the kinetic approximation for the energy spectrum under a subcritical scaling, extending to the exact solutions result obtained in [Dymov, Kuksin, Maiocchi, Vladuts '2023] for quasisolutions which were defined as the second order truncations of decompositions for the solutions in amplitude. The proof does not involve Feynman diagrams, instead relying on a robust inductive analysis of cumulants. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 10:13:52 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:18:32 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dymov",
"Andrey",
""
]
] |
2505.16836 | Fanrui Zhang | Fanrui Zhang, Dian Li, Qiang Zhang, Chenjun, sinbadliu, Junxiong Lin, Jiahong Yan, Jiawei Liu, Zheng-Jun Zha | Fact-R1: Towards Explainable Video Misinformation Detection with Deep Reasoning | 28 pages, 27 figures | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | The rapid spread of multimodal misinformation on social media has raised growing concerns, while research on video misinformation detection remains limited due to the lack of large-scale, diverse datasets. Existing methods often overfit to rigid templates and lack deep reasoning over deceptive content. To address these challenges, we introduce FakeVV, a large-scale benchmark comprising over 100,000 video-text pairs with fine-grained, interpretable annotations. In addition, we further propose Fact-R1, a novel framework that integrates deep reasoning with collaborative rule-based reinforcement learning. Fact-R1 is trained through a three-stage process: (1) misinformation long-Chain-of-Thought (CoT) instruction tuning, (2) preference alignment via Direct Preference Optimization (DPO), and (3) Group Relative Policy Optimization (GRPO) using a novel verifiable reward function. This enables Fact-R1 to exhibit emergent reasoning behaviors comparable to those observed in advanced text-based reinforcement learning systems, but in the more complex multimodal misinformation setting. Our work establishes a new paradigm for misinformation detection, bridging large-scale video understanding, reasoning-guided alignment, and interpretable verification. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 16:05:06 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 03:33:32 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhang",
"Fanrui",
""
],
[
"Li",
"Dian",
""
],
[
"Zhang",
"Qiang",
""
],
[
"Chenjun",
"",
""
],
[
"sinbadliu",
"",
""
],
[
"Lin",
"Junxiong",
""
],
[
"Yan",
"Jiahong",
""
],
[
"Liu",
"Jiawei",
""
],
[
"Zha",
"Zheng-Jun",
""
]
] |
2505.16865 | Enze Liu | Enze Liu, Bowen Zheng, Xiaolei Wang, Wayne Xin Zhao, Jinpeng Wang, Sheng Chen, Ji-Rong Wen | LARES: Latent Reasoning for Sequential Recommendation | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential recommender systems have become increasingly important in real-world applications that model user behavior sequences to predict their preferences. However, existing sequential recommendation methods predominantly rely on non-reasoning paradigms, which may limit the model's computational capacity and result in suboptimal recommendation performance. To address these limitations, we present LARES, a novel and scalable LAtent REasoning framework for Sequential recommendation that enhances model's representation capabilities through increasing the computation density of parameters by depth-recurrent latent reasoning. Our proposed approach employs a recurrent architecture that allows flexible expansion of reasoning depth without increasing parameter complexity, thereby effectively capturing dynamic and intricate user interest patterns. A key difference of LARES lies in refining all input tokens at each implicit reasoning step to improve the computation utilization. To fully unlock the model's reasoning potential, we design a two-phase training strategy: (1) Self-supervised pre-training (SPT) with dual alignment objectives; (2) Reinforcement post-training (RPT). During the first phase, we introduce trajectory-level alignment and step-level alignment objectives, which enable the model to learn recommendation-oriented latent reasoning patterns without requiring supplementary annotated data. The subsequent phase utilizes reinforcement learning (RL) to harness the model's exploratory ability, further refining its reasoning capabilities. Comprehensive experiments on real-world benchmarks demonstrate our framework's superior performance. Notably, LARES exhibits seamless compatibility with existing advanced models, further improving their recommendation performance. Our code is available at https://anonymous.4open.science/r/LARES-E458/. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 16:22:54 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:31:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Enze",
""
],
[
"Zheng",
"Bowen",
""
],
[
"Wang",
"Xiaolei",
""
],
[
"Zhao",
"Wayne Xin",
""
],
[
"Wang",
"Jinpeng",
""
],
[
"Chen",
"Sheng",
""
],
[
"Wen",
"Ji-Rong",
""
]
] |
2505.16933 | Zebin You | Zebin You, Shen Nie, Xiaolu Zhang, Jun Hu, Jun Zhou, Zhiwu Lu, Ji-Rong Wen, Chongxuan Li | LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuning | Project page and codes: \url{https://ml-gsai.github.io/LLaDA-V-demo/} | null | null | null | cs.LG cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this work, we introduce LLaDA-V, a purely diffusion-based Multimodal Large Language Model (MLLM) that integrates visual instruction tuning with masked diffusion models, representing a departure from the autoregressive paradigms dominant in current multimodal approaches. Built upon LLaDA, a representative large language diffusion model, LLaDA-V incorporates a vision encoder and MLP connector that projects visual features into the language embedding space, enabling effective multimodal alignment. Our empirical investigation reveals several intriguing results: First, LLaDA-V demonstrates promising multimodal performance despite its language model being weaker on purely textual tasks than counterparts like LLaMA3-8B and Qwen2-7B. When trained on the same instruction data, LLaDA-V is highly competitive to LLaMA3-V across multimodal tasks with better data scalability. It also narrows the performance gap to Qwen2-VL, suggesting the effectiveness of its architecture for multimodal tasks. Second, LLaDA-V achieves state-of-the-art performance in multimodal understanding compared to existing hybrid autoregressive-diffusion and purely diffusion-based MLLMs. Our findings suggest that large language diffusion models show promise in multimodal contexts and warrant further investigation in future research. Project page and codes: https://ml-gsai.github.io/LLaDA-V-demo/. | [
{
"version": "v1",
"created": "Thu, 22 May 2025 17:23:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 05:52:10 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"You",
"Zebin",
""
],
[
"Nie",
"Shen",
""
],
[
"Zhang",
"Xiaolu",
""
],
[
"Hu",
"Jun",
""
],
[
"Zhou",
"Jun",
""
],
[
"Lu",
"Zhiwu",
""
],
[
"Wen",
"Ji-Rong",
""
],
[
"Li",
"Chongxuan",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.