id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 3
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-08-29 00:00:00
| authors_parsed
listlengths 1
3.31k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2412.12404
|
Mohammed Chamma
|
Mohammed A. Chamma, Victor Pop and Fereshteh Rajabi
|
Spectro-temporal analysis of ultra-fast radio bursts using per-channel arrival times
|
19 pages, 10 figures. Submitted to MNRAS. Conclusions on methods comparison updated in this version. Tests of robustness and other improvements also added
| null | null | null |
astro-ph.HE
|
http://creativecommons.org/licenses/by/4.0/
|
Fast radio bursts (FRBs), especially those from repeating sources, exhibit a rich variety of morphologies in their dynamic spectra (or waterfalls). Characterizing these morphologies and spectro-temporal properties is a key strategy in investigating the underlying unknown emission mechanism of FRBs. This type of analysis has been typically accomplished using two-dimensional Gaussian techniques and the autocorrelation function (ACF) of the waterfall. These techniques are effective and precise at all duration scales, but can be limited in the presence of scattered tails, complex morphologies, or recently observed microshot forests. Here, we present a technique that involves the tagging of per-channel arrival times of an FRB to perform spectro-temporal measurements using a Gaussian profile model for each channel. While scattering and dispersion remain important and often dominating sources of uncertainty in measurements, this technique provides an adaptable and firm foundation for obtaining spectro-temporal properties from all types of FRB morphologies. We present measurements using this technique of several hundred bursts across 12 repeating sources, including over 400 bursts from the repeating sources FRB 20121102A, FRB 20220912A, and FRB 20200120E, all of which exhibit recently observed microsecond-long ultra-FRBs, as well as 143 drift rates. In addition to retrieving the known relationship between sub-burst slope and duration, we explore other correlations between burst properties. We find that the sub-burst slope law extends smoothly to ultra-FRBs, and that ultra-FRBs appear to form a distinct population in the duration-frequency relation.
|
[
{
"version": "v1",
"created": "Mon, 16 Dec 2024 23:15:52 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 18:53:23 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Chamma",
"Mohammed A.",
""
],
[
"Pop",
"Victor",
""
],
[
"Rajabi",
"Fereshteh",
""
]
] |
2412.13329
|
Lin Zhang
|
Lin Zhang, Igor Tyulnev, Lenard Vamos, Julita Poborska, Utso Bhattacharya, Ravindra W. Chhajlany, Tobias Grass, Jens Biegert, Maciej Lewenstein
|
Revealing the anisotropic charge-density-wave order of TiSe$_2$ through high harmonic generation
|
published version
|
Phys. Rev. B 112, 085147 (2025)
|
10.1103/812z-nl4m
| null |
cond-mat.str-el
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Titanium diselenide (TiSe$_{2}$) is a transition-metal dichalcogenide material that undergoes a charge-density-wave (CDW) transition at $T_{c}\approx 200\,\mathrm{K}$. In a recent experiment [I. Tyulnev {\it et al.}, Commun. Mater. 6, 152 (2025)], the high harmonic generation (HHG) spectra of this material has been studied, which exhibits asymmetric behavior with respect to the polarization angle of the incident light and provides a new perspective to the CDW phase transition. In this work, we work out a theoretical explanation for the experimentally observed behavior of HHG spectra. We propose a simplified phenomenological mean-field model for this material, based on which the HHG spectra is calculated through the time-dependent Schr{\" o}dinger equation. This model correctly describes the measured intensity distribution of the third-, fifth-, and seventh-order harmonic generation as a function of polarization direction and reveals a strong asymmetry due to the anisotropic CDW order in the low temperature phase. Our work provides a basis for applying high harmonic spectroscopy to reveal a new perspective on the nature of CDW orders.
|
[
{
"version": "v1",
"created": "Tue, 17 Dec 2024 20:58:56 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 20:45:20 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Zhang",
"Lin",
""
],
[
"Tyulnev",
"Igor",
""
],
[
"Vamos",
"Lenard",
""
],
[
"Poborska",
"Julita",
""
],
[
"Bhattacharya",
"Utso",
""
],
[
"Chhajlany",
"Ravindra W.",
""
],
[
"Grass",
"Tobias",
""
],
[
"Biegert",
"Jens",
""
],
[
"Lewenstein",
"Maciej",
""
]
] |
2412.13426
|
Zhifeng Jiang
|
Zhifeng Jiang, Zhihua Jin, Guoliang He
|
PromptKeeper: Safeguarding System Prompts for LLMs
|
Accepted to the Findings of EMNLP 2025. 17 pages, 6 figures, 3 tables
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
System prompts are widely used to guide the outputs of large language models (LLMs). These prompts often contain business logic and sensitive information, making their protection essential. However, adversarial and even regular user queries can exploit LLM vulnerabilities to expose these hidden prompts. To address this issue, we propose PromptKeeper, a defense mechanism designed to safeguard system prompts by tackling two core challenges: reliably detecting leakage and mitigating side-channel vulnerabilities when leakage occurs. By framing detection as a hypothesis-testing problem, PromptKeeper effectively identifies both explicit and subtle leakage. Upon leakage detected, it regenerates responses using a dummy prompt, ensuring that outputs remain indistinguishable from typical interactions when no leakage is present. PromptKeeper ensures robust protection against prompt extraction attacks via either adversarial or regular queries, while preserving conversational capability and runtime efficiency during benign user interactions.
|
[
{
"version": "v1",
"created": "Wed, 18 Dec 2024 01:43:25 GMT"
},
{
"version": "v2",
"created": "Thu, 09 Jan 2025 14:33:25 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 06:38:03 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Jiang",
"Zhifeng",
""
],
[
"Jin",
"Zhihua",
""
],
[
"He",
"Guoliang",
""
]
] |
2412.13750
|
Jesper Byggm\"astar
|
Jesper Byggm\"astar, Damian Sobieraj, Jan S. Wr\'obel, Daniel K. Schreiber, Osman El-Atwani, Enrique Martinez, Duc Nguyen-Manh
|
Segregation, ordering, and precipitation in WTaV-based concentrated refractory alloys
| null |
Acta Materialia 296 (2025) 121276
|
10.1016/j.actamat.2025.121276
| null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Tungsten-based low-activation high-entropy alloys are possible candidates for next-generation fusion reactors due to their exceptional tolerance to irradiation, thermal loads, and stress. We develop an accurate and efficient machine-learned interatomic potential for the W-Ta-Cr-V system and use it in hybrid Monte Carlo molecular dynamics simulations of ordering and segregation to all common types of defects in WTaCrV. The predictions are compared to atom probe tomography analysis of segregation and precipitation in WTaCrV thin films. By also considering two other alloys, WTaV and MoNbTaVW, we are able to draw general conclusions about preferred segregation in refractory alloys and the reasons behind it, guiding future alloy design and elucidating experimental observations. We show that the experimentally observed CrV precipitates in WTaCrV form semicoherent bcc-to-bcc interfaces with the surrounding matrix, as coherent precipitates are not thermodynamically stable due to excessive lattice mismatch. The predictions from simulations align well with our atom probe tomography analysis as well as previous experimental observations.
|
[
{
"version": "v1",
"created": "Wed, 18 Dec 2024 11:36:33 GMT"
},
{
"version": "v2",
"created": "Thu, 19 Jun 2025 07:12:44 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Byggmästar",
"Jesper",
""
],
[
"Sobieraj",
"Damian",
""
],
[
"Wróbel",
"Jan S.",
""
],
[
"Schreiber",
"Daniel K.",
""
],
[
"El-Atwani",
"Osman",
""
],
[
"Martinez",
"Enrique",
""
],
[
"Nguyen-Manh",
"Duc",
""
]
] |
2412.14044
|
Mar\'ia Dias Astros
|
Maria Dias Astros and Stefan Vogl
|
Freeze-in leptogenesis with sterile neutrino self-interactions
|
19 pages, 6 figures
| null | null | null |
hep-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Sterile neutrinos are a simple yet compelling addition to the Standard Model. For right-handed neutrinos with masses below the electroweak scale, leptogenesis can proceed through CP-violating oscillations of the sterile neutrinos. This is known as ARS or freeze-in leptogenesis. However, the ARS scenario requires the right-handed neutrinos to have a high degree of mass degeneracy. In this work, we study an extension of the SM that introduces a scalar singlet in addition to the two sterile neutrinos required to generate the baryon asymmetry. The new scalar interacts with the sterile neutrinos via a Yukawa interaction. This leads to an additional rate for the production and destruction of the sterile neutrinos and to a novel contribution to the effective potential. For the case in which the mass and the new Yukawa matrices are not diagonal in the same basis, we find that the effective potential can boost the baryon asymmetry of the universe by several orders of magnitude. This significantly alleviates the fine-tuned mass condition required in vanilla ARS leptogenesis.
|
[
{
"version": "v1",
"created": "Wed, 18 Dec 2024 17:00:30 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 12:34:01 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Astros",
"Maria Dias",
""
],
[
"Vogl",
"Stefan",
""
]
] |
2412.14207
|
Weizi Li
|
Supriya Sarker, Brent Maples, Iftekharul Islam, Muyang Fan, Christos Papadopoulos, and Weizi Li
|
A Comprehensive Review on Traffic Datasets and Simulators for Autonomous Vehicles
|
This manuscript has been withdrawn due to the need for substantial updates and revisions
| null | null | null |
cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Autonomous driving has rapidly evolved through synergistic developments in hardware and artificial intelligence. This comprehensive review investigates traffic datasets and simulators as dual pillars supporting autonomous vehicle (AV) development. Unlike prior surveys that examine these resources independently, we present an integrated analysis spanning the entire AV pipeline-perception, localization, prediction, planning, and control. We evaluate annotation practices and quality metrics while examining how geographic diversity and environmental conditions affect system reliability. Our analysis includes detailed characterizations of datasets organized by functional domains and an in-depth examination of traffic simulators categorized by their specialized contributions to research and development. The paper explores emerging trends, including novel architecture frameworks, multimodal AI integration, and advanced data generation techniques that address critical edge cases. By highlighting the interconnections between real-world data collection and simulation environments, this review offers researchers a roadmap for developing more robust and resilient autonomous systems equipped to handle the diverse challenges encountered in real-world driving environments.
|
[
{
"version": "v1",
"created": "Tue, 17 Dec 2024 07:43:23 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:17:24 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 14:23:23 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Sarker",
"Supriya",
""
],
[
"Maples",
"Brent",
""
],
[
"Islam",
"Iftekharul",
""
],
[
"Fan",
"Muyang",
""
],
[
"Papadopoulos",
"Christos",
""
],
[
"Li",
"Weizi",
""
]
] |
2412.14565
|
Takeru Utsugi
|
Takeru Utsugi, Nobuhiro Kusuno, Takuma Kuno, Noriyuki Lee, Itaru Yanagi, Toshiyuki Mine, Shinichi Saito, Digh Hisamoto, Ryuta Tsuchiya, and Hiroyuki Mizuno
|
Thermal circuit model for silicon quantum-dot array structures
|
10 pages, 9 figures
| null | null | null |
cond-mat.mes-hall physics.app-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Temperature rise of qubits due to heating is a critical issue in large-scale quantum computers based on quantum-dot (QD) arrays. This leads to shorter coherence times, induced readout errors, and increased charge noise. Here, we propose a simple thermal circuit model to describe the heating effect on silicon QD array structures. Noting that the QD array is a periodic structure, we represent it as a thermal distributed-element circuit, forming a thermal transmission line. We validate this model by measuring the electron temperature in a QD array device using Coulomb blockade thermometry, finding that the model effectively reproduces experimental results. This simple and scalable model can be used to develop the thermal design of large-scale silicon-based quantum computers.
|
[
{
"version": "v1",
"created": "Thu, 19 Dec 2024 06:34:12 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 02:34:13 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Utsugi",
"Takeru",
""
],
[
"Kusuno",
"Nobuhiro",
""
],
[
"Kuno",
"Takuma",
""
],
[
"Lee",
"Noriyuki",
""
],
[
"Yanagi",
"Itaru",
""
],
[
"Mine",
"Toshiyuki",
""
],
[
"Saito",
"Shinichi",
""
],
[
"Hisamoto",
"Digh",
""
],
[
"Tsuchiya",
"Ryuta",
""
],
[
"Mizuno",
"Hiroyuki",
""
]
] |
2412.14747
|
Giulia Maniccia
|
Giovanni Montani, Giulia Maniccia, Elisa Fazzari and Alessandro Melchiorri
|
Running Einstein Constant and a Possible Vacuum State of the Universe
|
11 pages, 3 figures; discussion expanded, results updated and references added; matches the version published in EPJC available at https://link.springer.com/article/10.1140/epjc/s10052-025-14618-8
|
Eur. Phys. J. C 85, 881 (2025)
|
10.1140/epjc/s10052-025-14618-8
| null |
gr-qc astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
We propose a revised formulation of General Relativity for cosmological settings, in which the Einstein constant varies with the energy density of the Universe. We demonstrate that this modification has only phenomenological impact of providing an effective dark energy density expression. Assuming a state close to vacuum, here defined by the vanishing product of the Einstein coupling constant and the Universe's energy density, we perform a Taylor expansion of the theory and hence extend it to the whole domain. In this framework, the (renormalized) vacuum energy problem is studied, and an additional constant pressure term, which induces a Chaplygin-like contribution to the dark energy sector, arises in the late-time dynamics. The correction to the late-time Hubble parameter is investigated by comparing theoretical predictions with the late Universe observational data. Our findings indicate that the current value of the stated vacuum energy is consistent with zero within 1$\sigma$. Implications of the modified $\Lambda$CDM model with respect to the Hubble tension are also discussed.
|
[
{
"version": "v1",
"created": "Thu, 19 Dec 2024 11:28:25 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 17:01:33 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Montani",
"Giovanni",
""
],
[
"Maniccia",
"Giulia",
""
],
[
"Fazzari",
"Elisa",
""
],
[
"Melchiorri",
"Alessandro",
""
]
] |
2412.15196
|
Theodore McKeever
|
Theodore McKeever, Owen Diba, Ahsan Nazir
|
A Finite-Time Quantum Otto Engine subject to Control Noise and Enhancement Techniques
|
12 pages, 6 figures
| null |
10.1103/ct9q-xff8
|
Phys. Rev. Research 7, 033172 (2025)
|
quant-ph cond-mat.stat-mech
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
With the development of any quantum technology comes a need for precise control of quantum systems. Here, we evaluate the impact of control noise on a quantum Otto cycle. Whilst it is postulated that noiseless quantum engines can approach maximal Otto efficiency in finite times, the existence of white noise on the controls is shown to negatively affect average engine performance. Two methods of quantum enhancement, counterdiabatic driving and quantum lubrication, are implemented and found to improve the performance of the noisy cycle only in specified parameter regimes. To gain insight into performance fluctuations, projective energy measurements are used to construct a noise-averaged probability distribution without assuming full thermalisation or adiabaticity. From this, the variances in thermodynamic currents are observed to increase as average power and efficiency improve, and are also shown to be consistent with known bounds from thermodynamic uncertainty relations. Lastly, by comparing the average functioning of the unmonitored engine to a projectively-measured engine cycle, the role of coherence in work extraction for this quantum engine model is investigated.
|
[
{
"version": "v1",
"created": "Thu, 19 Dec 2024 18:58:14 GMT"
},
{
"version": "v2",
"created": "Tue, 01 Jul 2025 15:59:53 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"McKeever",
"Theodore",
""
],
[
"Diba",
"Owen",
""
],
[
"Nazir",
"Ahsan",
""
]
] |
2412.15579
|
Chengyi Liu
|
Chengyi Liu and Jiahao Zhang and Shijie Wang and Wenqi Fan and Qing Li
|
Score-based Generative Diffusion Models for Social Recommendations
|
Accepted by IEEE Transactions on Knowledge and Data Engineering
| null | null | null |
cs.SI cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
With the prevalence of social networks on online platforms, social recommendation has become a vital technique for enhancing personalized recommendations. The effectiveness of social recommendations largely relies on the social homophily assumption, which presumes that individuals with social connections often share similar preferences. However, this foundational premise has been recently challenged due to the inherent complexity and noise present in real-world social networks. In this paper, we tackle the low social homophily challenge from an innovative generative perspective, directly generating optimal user social representations that maximize consistency with collaborative signals. Specifically, we propose the Score-based Generative Model for Social Recommendation (SGSR), which effectively adapts the Stochastic Differential Equation (SDE)-based diffusion models for social recommendations. To better fit the recommendation context, SGSR employs a joint curriculum training strategy to mitigate challenges related to missing supervision signals and leverages self-supervised learning techniques to align knowledge across social and collaborative domains. Extensive experiments on real-world datasets demonstrate the effectiveness of our approach in filtering redundant social information and improving recommendation performance.
|
[
{
"version": "v1",
"created": "Fri, 20 Dec 2024 05:23:45 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 08:40:47 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Liu",
"Chengyi",
""
],
[
"Zhang",
"Jiahao",
""
],
[
"Wang",
"Shijie",
""
],
[
"Fan",
"Wenqi",
""
],
[
"Li",
"Qing",
""
]
] |
2412.17103
|
Nicholas Vlamis
|
Megha Bhat, Rongdao Chen, Adityo Mamun, Ariana Verbanac, Eric Vergo, and Nicholas G. Vlamis
|
Algebraic and geometric properties of homeomorphism groups of ordinals
|
v2: incorporates referee's comments, particularly improving the introduction and preliminaries. To appear in Michigan Math. J., 30 pages. v1:28 pages
| null | null | null |
math.GR math.GN math.GT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study the homeomorphism groups of ordinals equipped with their order topology, focusing on successor ordinals whose limit capacity is also a successor. This is a rich family of groups that has connections to both permutation groups and homeomorphism groups of manifolds.
For ordinals of Cantor--Bendixson degree one, we prove that the homeomorphism group is strongly distorted and uniformly perfect, and we classify its normal generators. As a corollary, we recover and provide a new proof of the classical result that the subgroup of finite permutations in the symmetric group on a countably infinite set is the maximal proper normal subgroup.
For ordinals of higher Cantor--Bendixson degree, we establish a semi-direct product decomposition of the (pure) homeomorphism group. When the limit capacity is one, we further compute the abelianizations and determine normal generating sets of minimal cardinality for these groups.
|
[
{
"version": "v1",
"created": "Sun, 22 Dec 2024 17:30:30 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 16:46:37 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bhat",
"Megha",
""
],
[
"Chen",
"Rongdao",
""
],
[
"Mamun",
"Adityo",
""
],
[
"Verbanac",
"Ariana",
""
],
[
"Vergo",
"Eric",
""
],
[
"Vlamis",
"Nicholas G.",
""
]
] |
2412.17706
|
Eric Brunner
|
Eric Brunner (1), Luuk Coopmans (1), Gabriel Matos (1), Matthias Rosenkranz (1), Frederic Sauvage (1), Yuta Kikuchi (2 and 3) ((1) Quantinuum UK, (2) Quantinuum Japan, (3) RIKEN (iTHEMS) Japan)
|
Lindblad engineering for quantum Gibbs state preparation under the eigenstate thermalization hypothesis
|
62 pages, 22 figures
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Building upon recent progress in Lindblad engineering for quantum Gibbs state preparation algorithms, we propose a simplified protocol that is shown to be efficient under the eigenstate thermalization hypothesis (ETH). The ETH reduces circuit overheads of the Lindblad simulation algorithm and ensures a fast convergence toward the target Gibbs state. Moreover, we show that the realized Lindblad dynamics exhibits an inherent resilience against stochastic noise, opening up the path to a first demonstration on quantum computers. We complement our claims with numerical studies of the algorithm's convergence in various regimes of the mixed-field Ising model. In line with our predictions, we observe a mixing time scaling polynomially with system size when the ETH is satisfied. In addition, we assess the impact of algorithmic and hardware-induced errors on the algorithm's performance by carrying out quantum circuit simulations of our Lindblad simulation protocol with a local depolarizing noise model. This work bridges the gap between recent theoretical advances in dissipative Gibbs state preparation algorithms and their eventual quantum hardware implementation.
|
[
{
"version": "v1",
"created": "Mon, 23 Dec 2024 16:35:12 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Mar 2025 13:57:30 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 11:07:40 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Brunner",
"Eric",
"",
"Quantinuum UK"
],
[
"Coopmans",
"Luuk",
"",
"Quantinuum UK"
],
[
"Matos",
"Gabriel",
"",
"Quantinuum UK"
],
[
"Rosenkranz",
"Matthias",
"",
"Quantinuum UK"
],
[
"Sauvage",
"Frederic",
"",
"Quantinuum UK"
],
[
"Kikuchi",
"Yuta",
"",
"2 and 3"
]
] |
2412.17923
|
Mohammad M. Amirian
|
Mohammad M. Amirian, Zoe V. Finkel, Emmanuel Devred, Andrew J. Irwin
|
Parameterization of Photoinhibition For Phytoplankton
| null | null | null | null |
q-bio.OT
|
http://creativecommons.org/licenses/by/4.0/
|
Mathematical models of photosynthesis-irradiance relationships in phytoplankton are used to compute integrated water-column photosynthetic rates and predict primary production. Models typically ignore an important phenomenon observed in most experiments: photosynthetic rate remains constant over a range of irradiances before declining due to photoinhibition. Here we develop an approach that captures both photoinhibition and this plateau. We test six new models of photoinhibition and ten more photoinhibition models from the literature against a database of 1808 photosynthesis-irradiance curves exhibiting photoinhibition. The best model improves adjusted R2 by 6-37% and reduces RMSE by 15-70% compared to existing models. In the best model, photoinhibition is phenomenologically described by multiplication by a saturating function of the reciprocal of irradiance, simplifies to the widely-used Jassby & Platt photosynthesis-irradiance curve in the absence of photoinhibition, and only requires one new parameter. This photoinhibition parameter identifies the onset of photoinhibition and is the rate of decrease in photosynthetic rate at that irradiance. Simulations show that while parameter values vary up to 40% across existing models, particularly with and without photoinhibition, our representation of photoinhibition does not affect the interpretation or numerical values of parameters compared to the corresponding model without photoinhibition.
|
[
{
"version": "v1",
"created": "Mon, 23 Dec 2024 19:28:10 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 12:46:05 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Amirian",
"Mohammad M.",
""
],
[
"Finkel",
"Zoe V.",
""
],
[
"Devred",
"Emmanuel",
""
],
[
"Irwin",
"Andrew J.",
""
]
] |
2412.18164
|
Yinbin Han
|
Yinbin Han, Meisam Razaviyayn, Renyuan Xu
|
Stochastic Control for Fine-tuning Diffusion Models: Optimality, Regularity, and Convergence
|
31 pages
|
Forty-Second International Conference on Machine Learning 2025
| null | null |
cs.LG math.OC
|
http://creativecommons.org/licenses/by/4.0/
|
Diffusion models have emerged as powerful tools for generative modeling, demonstrating exceptional capability in capturing target data distributions from large datasets. However, fine-tuning these massive models for specific downstream tasks, constraints, and human preferences remains a critical challenge. While recent advances have leveraged reinforcement learning algorithms to tackle this problem, much of the progress has been empirical, with limited theoretical understanding. To bridge this gap, we propose a stochastic control framework for fine-tuning diffusion models. Building on denoising diffusion probabilistic models as the pre-trained reference dynamics, our approach integrates linear dynamics control with Kullback-Leibler regularization. We establish the well-posedness and regularity of the stochastic control problem and develop a policy iteration algorithm (PI-FT) for numerical solution. We show that PI-FT achieves global convergence at a linear rate. Unlike existing work that assumes regularities throughout training, we prove that the control and value sequences generated by the algorithm maintain the regularity. Additionally, we explore extensions of our framework to parametric settings and continuous-time formulations, and demonstrate the practical effectiveness of the proposed PI-FT algorithm through numerical experiments. Our code is available at https://github.com/yinbinhan/fine-tuning-of-diffusion-models.
|
[
{
"version": "v1",
"created": "Tue, 24 Dec 2024 04:55:46 GMT"
},
{
"version": "v2",
"created": "Sun, 03 Aug 2025 08:59:30 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 00:47:37 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Han",
"Yinbin",
""
],
[
"Razaviyayn",
"Meisam",
""
],
[
"Xu",
"Renyuan",
""
]
] |
2412.18328
|
Abdul Hadi
|
Abdul Hadi, Uha Isnaini, Indah Emilia Wijayanti, and Martianus Frederic Ezerman
|
On Codes over Eisenstein Integers
|
17 Pages, 4 figures
| null |
10.1109/TIT.2025.3586264
| null |
cs.IT math.IT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose constructions of codes over quotient rings of Eisenstein integers equipped with the Euclidean, square Euclidean, and hexagonal distances as a generalization of codes over Eisenstein integer fields. By set partitioning, we effectively divide the ring of Eisenstein integers into equal-sized subsets for distinct encoding. Unlike in Eisenstein integer fields of prime size, where partitioning is not feasible due to structural limitations, we partition the quotient rings into additive subgroups in such a way that the minimum square Euclidean and hexagonal distances of each subgroup are strictly larger than in the original set. This technique facilitates multilevel coding and enhances signal constellation efficiency.
|
[
{
"version": "v1",
"created": "Tue, 24 Dec 2024 10:30:55 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hadi",
"Abdul",
""
],
[
"Isnaini",
"Uha",
""
],
[
"Wijayanti",
"Indah Emilia",
""
],
[
"Ezerman",
"Martianus Frederic",
""
]
] |
2412.18830
|
Jos\'e Ignacio Y\'a\~nez
|
Joaqu\'in Moraga and Jos\'e Ignacio Y\'a\~nez
|
Calabi-Yau pairs of complexity two
|
31 pages v2: New title and new introduction. Added Section 6 on Gorenstein del Pezzo surfaces
| null | null | null |
math.AG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
A Calabi-Yau pair of index one and complexity zero is toric. Furthermore, a Calabi-Yau pair of index one and complexity one is of cluster type. In this article, we study Calabi-Yau pairs of index one and complexity two. We develop machinery to decide whether a Calabi-Yau of complexity two is of cluster type. This approach reduces the problem to studying del Pezzo fibrations over toric varieties. We apply this to the setting of Gorenstein del Pezzo surfaces of Picard rank one. We prove that such a surface $X$ is cluster type if and only if $X$ has only $A$-type singularities and either $\mathrm{vol}(X)>1$ or $|X^{\rm sing}|\leq 3$.
|
[
{
"version": "v1",
"created": "Wed, 25 Dec 2024 08:36:37 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 18:52:16 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Moraga",
"Joaquín",
""
],
[
"Yáñez",
"José Ignacio",
""
]
] |
2501.00507
|
Nermin \v{C}ovi\'c
|
Nermin Covic, Bakir Lacevic, Dinko Osmankovic, Tarik Uzunovic
|
Real-Time Sampling-Based Safe Motion Planning for Robotic Manipulators in Dynamic Environments
| null |
The IEEE Transactions on Robotics, 2025
|
10.1109/TRO.2025.3598119
| null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we present the main features of Dynamic Rapidly-exploring Generalized Bur Tree (DRGBT) algorithm, a sampling-based planner for dynamic environments. We provide a detailed time analysis and appropriate scheduling to facilitate a real-time operation. To this end, an extensive analysis is conducted to identify the time-critical routines and their dependence on the number of obstacles. Furthermore, information about the distance to obstacles is used to compute a structure called dynamic expanded bubble of free configuration space, which is then utilized to establish sufficient conditions for a guaranteed safe motion of the robot while satisfying all kinematic constraints. An extensive randomized simulation trial is conducted to compare the proposed algorithm to a competing state-of-the-art method. Finally, an experimental study on a real robot is carried out covering a variety of scenarios including those with human presence. The results show the effectiveness and feasibility of real-time execution of the proposed motion planning algorithm within a typical sensor-based arrangement, using cheap hardware and sequential architecture, without the necessity for GPUs or heavy parallelization.
|
[
{
"version": "v1",
"created": "Tue, 31 Dec 2024 15:38:21 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 09:42:45 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Covic",
"Nermin",
""
],
[
"Lacevic",
"Bakir",
""
],
[
"Osmankovic",
"Dinko",
""
],
[
"Uzunovic",
"Tarik",
""
]
] |
2501.00543
|
Brett D. Wick
|
Fabrizio Colombo, Elodie Pozzi, Irene Sabadini, Brett D. Wick
|
The Corona Problem for Slice Hyperholomorphic Functions
|
v2; 38 pages, typos corrected, hypothesis made more clear
| null | null | null |
math.CV
|
http://creativecommons.org/licenses/by/4.0/
|
This paper addresses the Corona problem for slice hyperholomorphic functions for a single quaternionic variable. While the Corona problem is well-understood in the context of one complex variable, it remains highly challenging in the case of several complex variables. The extension of the theory of one complex variable to several complex variables is not the only possible extension to multi-dimensional complex analysis. Instead of functions holomorphic in each variable separately, in this paper we will consider functions in some hypercomplex algebras, in particular in the algebra of quaternions. Previously, the Corona problem had not been studied within the hypercomplex framework because of challenges posed by pointwise multiplication, which is not closed for hypercomplex-valued analytic functions. Alternative notions of multiplication that are closed often compromise other desirable properties, further complicating the analysis. In this work, we resolve the Corona problem within the quaternionic slice hyperholomorphic setting. Our approach involves reformulating the quaternionic Bezout equation with respect to the appropriate multiplication into a new system of Bezout equations on the unit disc. We solve this system by adapting Wolff's proof of the Corona theorem for bounded analytic functions. As the number of generators increases, the associated algebra grows increasingly intricate.
|
[
{
"version": "v1",
"created": "Tue, 31 Dec 2024 17:05:39 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 00:09:02 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Colombo",
"Fabrizio",
""
],
[
"Pozzi",
"Elodie",
""
],
[
"Sabadini",
"Irene",
""
],
[
"Wick",
"Brett D.",
""
]
] |
2501.01963
|
Daniel Andr\'es D\'iaz-Pach\'on
|
Daniel Andr\'es D\'iaz-Pach\'on and H. Renata Gallegos and Ola H\"ossjer and J. Sunil Rao
|
Statistical learning does not always entail knowledge
|
Main file (33 pages) and supplement (16 pages). 1 figure
| null | null | null |
cs.LG cs.AI cs.IT math.IT math.PR math.ST stat.ML stat.TH
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In this paper, we study learning and knowledge acquisition (LKA) of an agent about a proposition that is either true or false. We use a Bayesian approach, where the agent receives data to update his beliefs about the proposition according to a posterior distribution. The LKA is formulated in terms of active information, with data representing external or exogenous information that modifies the agent's beliefs. It is assumed that data provide details about a number of features that are relevant to the proposition. We show that this leads to a Gibbs distribution posterior, which is in maximum entropy relative to the prior, conditioned on the side constraints that the data provide in terms of the features. We demonstrate that full learning is sometimes not possible and full knowledge acquisition is never possible when the number of extracted features is too small. We also distinguish between primary learning (receiving data about features of relevance for the proposition) and secondary learning (receiving data about the learning of another agent). We argue that this type of secondary learning does not represent true knowledge acquisition. Our results have implications for statistical learning algorithms, and we claim that such algorithms do not always generate true knowledge. The theory is illustrated with several examples.
|
[
{
"version": "v1",
"created": "Tue, 17 Dec 2024 19:51:32 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 19:08:52 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Díaz-Pachón",
"Daniel Andrés",
""
],
[
"Gallegos",
"H. Renata",
""
],
[
"Hössjer",
"Ola",
""
],
[
"Rao",
"J. Sunil",
""
]
] |
2501.02142
|
Caroline Cardinale
|
Caroline Cardinale, Steven L. Brunton, Tim Colonius
|
Spectral proper orthogonal decomposition of rapid snapshot pairs sampled at sub-Nyquist intervals
|
22 pages, 9 figures
| null | null | null |
physics.flu-dyn
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Modal decomposition methods are important for characterizing the low-dimensional dynamics of complex systems, including turbulent flows. Different methods have varying data requirements and produce modes with different properties. Spectral proper orthogonal decomposition (SPOD) produces orthogonal, energy-ranked spatial modes at discrete temporal frequencies for statistically stationary flows. However, SPOD requires long stretches of sequential, uniformly sampled, time-resolved data. These data requirements limit SPOD's use in experimental settings where the maximum capture rate of a camera is often slower than the Nyquist sampling rate required to resolve the highest turbulent frequencies. However, if two PIV systems operate in tandem, pairs of data can be acquired that are arbitrarily close in time. The dynamic mode decomposition (DMD) uses this pairwise data to resolve frequencies up to the Nyquist frequency associated with the small time step within a pair. However, these modes do not form a basis and have no set ranking. The present work attempts to compute SPOD modes from pairwise data with a small time step but with large gaps between pairs. We use DMD on pairwise data to estimate segment-wise, uniformly sampled series that can then be used to estimate the SPOD modes, intending to resolve frequencies between the gap and pair Nyquist limits. The method is tested on numerically obtained data of the linearized complex Ginzburg-Landau equation, as well as a Mach 0.4 isothermal turbulent jet. For the jet, pairwise SPOD can accurately de-alias the SPOD spectrum and estimate mode shapes at frequencies up to St = 1.0, while using over 90% less data.
|
[
{
"version": "v1",
"created": "Fri, 03 Jan 2025 23:35:55 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 19:52:29 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Cardinale",
"Caroline",
""
],
[
"Brunton",
"Steven L.",
""
],
[
"Colonius",
"Tim",
""
]
] |
2501.02306
|
Krista Soderlund
|
Krista M. Soderlund, Paula Wulff, Petri K\"apyl\"a, Jonathan M. Aurnou
|
Lorentz force mediation of turbulent dynamo transitions
| null |
Mon Not R Astron Soc (2025) 1816-1830
|
10.1093/mnras/staf1081
| null |
astro-ph.SR physics.plasm-ph
|
http://creativecommons.org/licenses/by/4.0/
|
We investigate how the strength of the Lorentz force alters stellar convection zone dynamics in a suite of buoyancy-dominated, three-dimensional, spherical shell convective dynamo models. This is done by varying only the magnetic Prandtl number, $Pm$, the non-dimensional form of the fluid's electrical conductivity $\sigma$. Because the strength of the dynamo magnetic field and the Lorentz force scale with $Pm$, it is found that the fluid motions, the pattern of convective heat transfer, and the mode of dynamo generation all differ across the $0.25 \leq Pm \leq 10$ range investigated here. For example, we show that strong magnetohydrodynamic effects cause a fundamental change in the surface zonal flows: differential rotation switches from solar-like (prograde equatorial zonal flow) for larger electrical conductivities to an anti-solar differential rotation (retrograde equatorial zonal flow) at lower electrical conductivities. This study shows that the value of the bulk electrical conductivity is important not only for sustaining dynamo action, but can also drive first-order changes in the characteristics of the magnetic, velocity, and temperature fields. It is also associated with the relative strength of the Lorentz force in the system as measured by the local magnetic Rossby number, $Ro_\ell^M$, which we show is crucial in setting the characteristics of the large-scale convection regime that generates those dynamo fields.
|
[
{
"version": "v1",
"created": "Sat, 04 Jan 2025 14:59:51 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Soderlund",
"Krista M.",
""
],
[
"Wulff",
"Paula",
""
],
[
"Käpylä",
"Petri",
""
],
[
"Aurnou",
"Jonathan M.",
""
]
] |
2501.02531
|
Ljubisa Bojic
|
Ljubisa Bojic, Dylan Seychell and Milan Cabarkapa
|
Towards New Benchmark for AI Alignment & Sentiment Analysis in Socially Important Issues: A Comparative Study of Human and LLMs in the Context of AGI
|
34 pages, 3 figures
| null | null | null |
cs.CY cs.CL
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
As general-purpose artificial intelligence systems become increasingly integrated into society and are used for information seeking, content generation, problem solving, textual analysis, coding, and running processes, it is crucial to assess their long-term impact on humans. This research explores the sentiment of large language models (LLMs) and humans toward artificial general intelligence (AGI) using a Likert-scale survey. Seven LLMs, including GPT-4 and Bard, were analyzed and compared with sentiment data from three independent human sample populations. Temporal variations in sentiment were also evaluated over three consecutive days. The results show a diversity in sentiment scores among LLMs, ranging from 3.32 to 4.12 out of 5. GPT-4 recorded the most positive sentiment toward AGI, while Bard leaned toward a neutral sentiment. In contrast, the human samples showed a lower average sentiment of 2.97. The analysis outlines potential conflicts of interest and biases in the sentiment formation of LLMs, and indicates that LLMs could subtly influence societal perceptions. To address the need for regulatory oversight and culturally grounded assessments of AI systems, we introduce the Societal AI Alignment and Sentiment Benchmark (SAAS-AI), which leverages multidimensional prompts and empirically validated societal value frameworks to evaluate language model outputs across temporal, model, and multilingual axes. This benchmark is designed to guide policymakers and AI agencies, including within frameworks such as the EU AI Act, by providing robust, actionable insights into AI alignment with human values, public sentiment, and ethical norms at both national and international levels. Future research should further refine the operationalization of the SAAS-AI benchmark and systematically evaluate its effectiveness through comprehensive empirical testing.
|
[
{
"version": "v1",
"created": "Sun, 05 Jan 2025 13:18:13 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Aug 2025 15:23:08 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 13:49:46 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bojic",
"Ljubisa",
""
],
[
"Seychell",
"Dylan",
""
],
[
"Cabarkapa",
"Milan",
""
]
] |
2501.02748
|
Yun Fang
|
Yun Fang and Rong-Gen Cai
|
Probing the Merger Rates of Supermassive Black Holes and Galaxies with Gravitational Waves
|
accepted by MNRAS, 16 pages, 13 figures
|
Mon Not R Astron Soc (2025) 1172-1187
|
10.1093/mnras/staf1278
| null |
astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The mergers of galaxies and supermassive black holes (SMBHs) are key drivers of galaxy evolution, contributing to the growth of both galaxies and their central black holes. Current and upcoming gravitational wave (GW) detectors -- Pulsar Timing Arrays (PTAs), LISA, Taiji, and Tianqin -- offer unique access to these processes by observing GW signals from SMBH binaries. We present a framework to infer galaxy and SMBH merger rates by combining mock LISA detections of SMBH mergers with PTA constraints on the stochastic GW background, while incorporating observational uncertainties in stellar mass functions and $M_\bullet$-$M_*$ relations. We find that the number of LISA-detected events and their joint distribution in mass and redshift are key to constraining merger rates -- datasets with around forty events yield results consistent with galaxy pair observations, whereas limited event counts lead to biases at high redshift. Including PTA data further reduces parameter uncertainties. Our method also effectively constrains the delay time between galaxy and SMBH mergers, with longer delays suppressing high-redshift SMBH merger rates and shifting mass growth from mergers to accretion. According to our mock analysis, the models with delay times longer than $0.5\text{Gyr}$ ($0.8\text{Gyr}$), accretion becomes the primary driver of SMBH mass growth beyond $z \sim 6$ ($4$). In contrast, the SMBH occupation fraction at $z>3$ remains poorly constrained due to its degeneracies with delay time and the galaxy merger rate. These findings highlight both the promise and limitations of using GW observations to probe the coevolution of galaxies and SMBHs.
|
[
{
"version": "v1",
"created": "Mon, 06 Jan 2025 03:52:16 GMT"
},
{
"version": "v2",
"created": "Fri, 01 Aug 2025 15:45:13 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Fang",
"Yun",
""
],
[
"Cai",
"Rong-Gen",
""
]
] |
2501.03738
|
Arghya Chakraborty
|
Arghya Chakraborty, Jessy Jose, Alex C. Carciofi
|
Exploring membership and variability in NGC 7419: An open cluster rich in super giants and Be type stars
| null |
Mon Not R Astron Soc (2025) 1866-1892
|
10.1093/mnras/staf1104
| null |
astro-ph.SR astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
NGC 7419 is a young open cluster notable for hosting five Red Supergiants and a high abundance of Classical Be (CBe) stars. CBe stars are main sequence non-supergiant B-type stars, which exhibit or have exhibited Balmer line emissions in their spectra. We perform a membership analysis using Gaia DR3 data and machine learning techniques like Gaussian Mixture Models (GMM) and Random Forest (RF) and determine the cluster's mean distance to be ${3.6^{+1.0}_{-0.6}}$ kpc. We identify 499 Gaia-based members with a mass above $\sim$ 1.2 M$_\odot$, and estimate the cluster's age to be $21.1 ^{+1.6}_{-0.6}$ Myr. Using our revised $H\alpha$ excess-based analysis, we find 42 CBe stars containing many known CBe stars, bringing the total number of CBe stars in NGC 7419 to 49 and the fraction of CBe to (B+CBe) members to 12.7%. We investigate the variability of the candidate members from ZTF and NEOWISE data using Standard Deviation, Median Absolute Deviation, and Stetson Index (J), and their periodicity using the Generalized Lomb Scargle Periodogram variability. We find that 66% of CBe stars are variable: 23% show periodic signals, typical of pulsation/rotation, 41% display variability characteristic of disk dynamics or binarity, and 14% exhibit long-term changes, consistent with disk dissipation/formation. We also find that all pulsating CBe stars are early-type, while 50% of stars with long-term variations are early-type, and the other 50% are mid-type. Our results agree with previous findings in the literature and confirm that CBe stars display variability through multiple mechanisms across different timescales.
|
[
{
"version": "v1",
"created": "Tue, 07 Jan 2025 12:29:38 GMT"
},
{
"version": "v2",
"created": "Fri, 04 Jul 2025 13:24:26 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Chakraborty",
"Arghya",
""
],
[
"Jose",
"Jessy",
""
],
[
"Carciofi",
"Alex C.",
""
]
] |
2501.04181
|
Abicumaran Uthamacumaran
|
Abicumaran Uthamacumaran
|
Deep Learning-based Feature Discovery for Decoding Phenotypic Plasticity in Pediatric High-Grade Gliomas Single-Cell Transcriptomics
| null |
Computers in Biology and Medicine. Volume 197, Part A, October 2025, 110971
|
10.1016/j.compbiomed.2025.110971
|
2025
|
q-bio.GN
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
By use of complex network dynamics and graph-based machine learning, we identified critical determinants of lineage-specific plasticity across the single-cell transcriptomics of pediatric high-grade glioma (pHGGs) subtypes: IDHWT glioblastoma and K27M-mutant glioma. Our study identified network interactions regulating glioma morphogenesis via the tumor-immune microenvironment, including neurodevelopmental programs, calcium dynamics, iron metabolism, metabolic reprogramming, and feedback loops between MAPK/ERK and WNT signaling. These relationships highlight the emergence of a hybrid spectrum of cellular states navigating a disrupted neuro-differentiation hierarchy. We identified transition genes such as DKK3, NOTCH2, GATAD1, GFAP, and SEZ6L in IDHWT glioblastoma, and H3F3A, ANXA6, HES6/7, SIRT2, FXYD6, PTPRZ1, MEIS1, CXXC5, and NDUFAB1 in K27M subtypes. We also identified MTRNR2L1, GAPDH, IGF2, FKBP variants, and FXYD7 as transition genes that influence cell fate decision-making across both subsystems. Our findings suggest pHGGs are developmentally trapped in states exhibiting maladaptive behaviors, and hybrid cellular identities. In effect, tumor heterogeneity (metastability) and plasticity emerge as stress-response patterns to immune-inflammatory microenvironments and oxidative stress. Furthermore, we show that pHGGs are steered by developmental trajectories from radial glia predominantly favoring neocortical cell fates, in telencephalon and prefrontal cortex (PFC) differentiation. By addressing underlying patterning processes and plasticity networks as therapeutic vulnerabilities, our findings provide precision medicine strategies aimed at modulating glioma cell fates and overcoming therapeutic resistance. We suggest transition therapy toward neuronal-like lineage differentiation as a potential therapy to help stabilize pHGG plasticity and aggressivity.
|
[
{
"version": "v1",
"created": "Tue, 07 Jan 2025 23:20:59 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Uthamacumaran",
"Abicumaran",
""
]
] |
2501.04820
|
Vasudha Varadarajan
|
Allison Lahnala and Vasudha Varadarajan and Lucie Flek and H. Andrew Schwartz and Ryan L. Boyd
|
Unifying the Extremes: Developing a Unified Model for Detecting and Predicting Extremist Traits and Radicalization
|
17 pages, 7 figures, 4 tables
| null | null | null |
cs.SI cs.CL cs.CY
|
http://creativecommons.org/licenses/by/4.0/
|
The proliferation of ideological movements into extremist factions via social media has become a global concern. While radicalization has been studied extensively within the context of specific ideologies, our ability to accurately characterize extremism in more generalizable terms remains underdeveloped. In this paper, we propose a novel method for extracting and analyzing extremist discourse across a range of online community forums. By focusing on verbal behavioral signatures of extremist traits, we develop a framework for quantifying extremism at both user and community levels. Our research identifies 11 distinct factors, which we term ``The Extremist Eleven,'' as a generalized psychosocial model of extremism. Applying our method to various online communities, we demonstrate an ability to characterize ideologically diverse communities across the 11 extremist traits. We demonstrate the power of this method by analyzing user histories from members of the incel community. We find that our framework accurately predicts which users join the incel community up to 10 months before their actual entry with an AUC of $>0.6$, steadily increasing to AUC ~0.9 three to four months before the event. Further, we find that upon entry into an extremist forum, the users tend to maintain their level of extremism within the community, while still remaining distinguishable from the general online discourse. Our findings contribute to the study of extremism by introducing a more holistic, cross-ideological approach that transcends traditional, trait-specific models.
|
[
{
"version": "v1",
"created": "Wed, 08 Jan 2025 20:17:24 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 17:56:55 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Lahnala",
"Allison",
""
],
[
"Varadarajan",
"Vasudha",
""
],
[
"Flek",
"Lucie",
""
],
[
"Schwartz",
"H. Andrew",
""
],
[
"Boyd",
"Ryan L.",
""
]
] |
2501.05386
|
Ferdinand Kuemmeth
|
Fabrizio Berritta, Jacob Benestad, Lukas Pahl, Melvin Mathews, Jan A. Krzywda, R\'eouven Assouly, Youngkyu Sung, David K. Kim, Bethany M. Niedzielski, Kyle Serniak, Mollie E. Schwartz, Jonilyn L. Yoder, Anasua Chatterjee, Jeffrey A. Grover, Jeroen Danon, William D. Oliver, and Ferdinand Kuemmeth
|
Efficient Qubit Calibration by Binary-Search Hamiltonian Tracking
|
main text 10 pages, 4 figures, plus 10 supplementary pages, 6 supplementary figures
|
PRX Quantum 6, 030335 (2025)
|
10.1103/77qg-p68k
|
NBI QDEV 2025
|
quant-ph cond-mat.mes-hall
|
http://creativecommons.org/licenses/by/4.0/
|
We present and experimentally implement a real-time protocol for calibrating the frequency of a resonantly driven qubit, achieving exponential scaling in calibration precision with the number of measurements, up to the limit imposed by decoherence. The real-time processing capabilities of a classical controller dynamically generate adaptive probing sequences for qubit-frequency estimation. Each probing evolution time and drive frequency are calculated to divide the prior probability distribution into two branches, following a locally optimal strategy that mimics a conventional binary search. The scheme does not require repeated measurements at the same setting, as it accounts for state preparation and measurement errors. Its use of a parametrized probability distribution favors numerical accuracy and computational speed. We show the efficacy of the algorithm by stabilizing a flux-tunable transmon qubit, leading to improved coherence and gate fidelity. As benchmarked by gate-set tomography, the field-programmable gate array (FPGA) powered control electronics partially mitigates non-Markovian noise, which is detrimental to quantum error correction. The mitigation is achieved by dynamically updating and feeding forward the qubit frequency. Our protocol highlights the importance of feedback in improving the calibration and stability of qubits subject to drift and can be readily applied to other qubit platforms.
|
[
{
"version": "v1",
"created": "Thu, 09 Jan 2025 17:18:38 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 06:07:31 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Berritta",
"Fabrizio",
""
],
[
"Benestad",
"Jacob",
""
],
[
"Pahl",
"Lukas",
""
],
[
"Mathews",
"Melvin",
""
],
[
"Krzywda",
"Jan A.",
""
],
[
"Assouly",
"Réouven",
""
],
[
"Sung",
"Youngkyu",
""
],
[
"Kim",
"David K.",
""
],
[
"Niedzielski",
"Bethany M.",
""
],
[
"Serniak",
"Kyle",
""
],
[
"Schwartz",
"Mollie E.",
""
],
[
"Yoder",
"Jonilyn L.",
""
],
[
"Chatterjee",
"Anasua",
""
],
[
"Grover",
"Jeffrey A.",
""
],
[
"Danon",
"Jeroen",
""
],
[
"Oliver",
"William D.",
""
],
[
"Kuemmeth",
"Ferdinand",
""
]
] |
2501.05828
|
Felix Duelmer
|
Felix Duelmer, Mohammad Farid Azampour, Magdalena Wysocki, Nassir Navab
|
UltraRay: Introducing Full-Path Ray Tracing in Physics-Based Ultrasound Simulation
| null | null | null | null |
cs.CV cs.GR
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Traditional ultrasound simulators solve the wave equation to model pressure distribution fields, achieving high accuracy but requiring significant computational time and resources. To address this, ray tracing approaches have been introduced, modeling wave propagation as rays interacting with boundaries and scatterers. However, existing models simplify ray propagation, generating echoes at interaction points without considering return paths to the sensor. This can result in unrealistic artifacts and necessitates careful scene tuning for plausible results. We propose a novel ultrasound simulation pipeline that utilizes a ray tracing algorithm to generate echo data, tracing each ray from the transducer through the scene and back to the sensor. To replicate advanced ultrasound imaging, we introduce a ray emission scheme optimized for plane wave imaging, incorporating delay and steering capabilities. Furthermore, we integrate a standard signal processing pipeline to simulate end-to-end ultrasound image formation. We showcase the efficacy of the proposed pipeline by modeling synthetic scenes featuring highly reflective objects, such as bones. In doing so, our proposed approach, UltraRay, not only enhances the overall visual quality but also improves the realism of the simulated images by accurately capturing secondary reflections and reducing unnatural artifacts. By building on top of a differentiable framework, the proposed pipeline lays the groundwork for a fast and differentiable ultrasound simulation tool necessary for gradient-based optimization, enabling advanced ultrasound beamforming strategies, neural network integration, and accurate inverse scene reconstruction.
|
[
{
"version": "v1",
"created": "Fri, 10 Jan 2025 10:07:41 GMT"
},
{
"version": "v2",
"created": "Wed, 13 Aug 2025 09:43:19 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 15:03:30 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Duelmer",
"Felix",
""
],
[
"Azampour",
"Mohammad Farid",
""
],
[
"Wysocki",
"Magdalena",
""
],
[
"Navab",
"Nassir",
""
]
] |
2501.06398
|
Lingjiong Zhu
|
Dan Pirjol, Lingjiong Zhu
|
VIX options in the SABR model
|
16 pages, 1 figure, 1 table
|
Operations Research Letters 2025, Volume 63, 107347
| null | null |
q-fin.PR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study the pricing of VIX options in the SABR model $dS_t = \sigma_t S_t^\beta dB_t, d\sigma_t = \omega \sigma_t dZ_t$ where $B_t,Z_t$ are standard Brownian motions correlated with correlation $\rho<0$ and $0 \leq \beta < 1$. VIX is expressed as a risk-neutral conditional expectation of an integral over the volatility process $v_t = S_t^{\beta-1} \sigma_t$. We show that $v_t$ is the unique solution to a one-dimensional diffusion process. Using the Feller test, we show that $v_t$ explodes in finite time with non-zero probability. As a consequence, VIX futures and VIX call prices are infinite, and VIX put prices are zero for any maturity. As a remedy, we propose a capped volatility process by capping the drift and diffusion terms in the $v_{t}$ process such that it becomes non-explosive and well-behaved, and study the short-maturity asymptotics for the pricing of VIX options.
|
[
{
"version": "v1",
"created": "Sat, 11 Jan 2025 01:05:48 GMT"
},
{
"version": "v2",
"created": "Thu, 31 Jul 2025 21:26:54 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Pirjol",
"Dan",
""
],
[
"Zhu",
"Lingjiong",
""
]
] |
2501.07626
|
Trey Qingyun Yang
|
Trey Qingyun Yang, Christoph Federrath
|
Protostellar disc structure and dynamics during star formation from cloud-scale initial conditions
|
19 pages, 18 figures, 3 tables. Submitted to MNRAS. See https://treyyang.com/projects/disc2024/ for animations prior to publication. Alfv\'en surface analysis added; velocity study corrected; references added
|
Mon Not R Astron Soc (2025) 1969-1987
|
10.1093/mnras/staf1088
| null |
astro-ph.SR astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
The early evolution of protostellar, star-forming discs, including their density structure, turbulence, magnetic dynamics, and accretion variability, remains poorly understood. We present high-resolution magnetohydrodynamic simulations, using adaptive mesh refinement to capture detailed disc dynamics down to sub-AU scales. Starting from initial conditions derived from a molecular cloud simulation, we model the collapse of a dense core into a protostellar disc over 10,000 yr following sink particle (star) formation, achieving a maximum effective resolution of 0.63 AU. This simulation traces the evolution of the disc density, accretion rates, turbulence, and magnetic field structures. We find that the protostellar disc grows to a diameter of approximately 100 AU, with mass accretion occurring in episodic bursts influenced by the turbulence of the core from which the disc builds up. The disc is highly turbulent with a sonic Mach number of $\sim2$. Episodic accretion events within the disc cause intermittent increases in mass and magnetic energy density, resulting in an equipartition of the thermal and magnetic pressure, i.e., leading to an Alfv\'en Mach number of $\sim2$. Some regions above and below the disc mid-plane show sub-Alfv\'enic conditions with intermittent outflow activity. The disc density profiles steepen over time, following a power law consistent with observed young stellar discs and the minimum mass solar nebula. These results underscore the role of turbulence in early accretion variability and offer new insights into the physical and magnetic structure of young protostellar discs, especially with respect to their turbulent components.
|
[
{
"version": "v1",
"created": "Mon, 13 Jan 2025 19:00:00 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Jun 2025 18:00:03 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Yang",
"Trey Qingyun",
""
],
[
"Federrath",
"Christoph",
""
]
] |
2501.08033
|
Gabriele Di Luzio
|
Gabriele Di Luzio and Giacomo Morelli
|
Robust Semiparametric Graphical Models with Skew-Elliptical Distributions
|
3 images and 4 tables
| null | null | null |
stat.ME
|
http://creativecommons.org/licenses/by/4.0/
|
We propose semiparametric estimators called elliptical skew-(S)KEPTIC for efficiently and robustly estimating non-Gaussian graphical models, relaxing the assumption of semiparametric elliptical distributions to the family of meta skew-elliptical, which accommodates a skewness component. Theoretically, we demonstrate that the elliptical skew-(S)KEPTIC estimators achieve robust convergence rates in both graph recovery and parameters estimation. We conduct numerical simulations that prove the reliable graph recovery performance of the elliptical skew-(S)KEPTIC estimators. Finally, the new method is applied to the daily log-returns of the stocks of the S\&P500 index and shows higher sparsity compared to the Gaussian copula graphical models.
|
[
{
"version": "v1",
"created": "Tue, 14 Jan 2025 11:37:50 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 10:55:29 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Di Luzio",
"Gabriele",
""
],
[
"Morelli",
"Giacomo",
""
]
] |
2501.08887
|
Guillaume O. Berger
|
Guillaume O. Berger, Rapha\"el M. Jungers
|
PAC Learnability of Scenario Decision-Making Algorithms: Necessary Conditions and Sufficient Conditions
| null | null | null | null |
cs.LG math.OC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We investigate the Probably Approximately Correct (PAC) property of scenario decision algorithms, which refers to their ability to produce decisions with an arbitrarily low risk of violating unknown safety constraints, provided a sufficient number of realizations of these constraints are sampled. While several PAC sufficient conditions for such algorithms exist in the literature -- such as the finiteness of the VC dimension of their associated classifiers, or the existence of a compression scheme -- it remains unclear whether these conditions are also necessary. In this work, we demonstrate through counterexamples that these conditions are not necessary in general. These findings stand in contrast to binary classification learning, where analogous conditions are both sufficient and necessary for a family of classifiers to be PAC. Furthermore, we extend our analysis to stable scenario decision algorithms, a broad class that includes practical methods like scenario optimization. Even under this additional assumption, we show that the aforementioned conditions remain unnecessary. Furthermore, we introduce a novel quantity, called the dVC dimension, which serves as an analogue to the VC dimension for scenario decision algorithms. We prove that the finiteness of this dimension is a PAC necessary condition for scenario decision algorithms. This allows to (i) guide algorithm users and designers to recognize algorithms that are not PAC, and (ii) contribute to a comprehensive characterization of PAC scenario decision algorithms.
|
[
{
"version": "v1",
"created": "Wed, 15 Jan 2025 15:57:13 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 10:01:41 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Berger",
"Guillaume O.",
""
],
[
"Jungers",
"Raphaël M.",
""
]
] |
2501.09993
|
Yeonseok Jeong
|
Yeonseok Jeong, Minsoo Kim, Seung-won Hwang, Byung-Hak Kim
|
Agent-as-Judge for Factual Summarization of Long Narratives
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Large Language Models (LLMs) have demonstrated near-human performance in summarization tasks based on traditional metrics such as ROUGE and BERTScore. However, these metrics do not adequately capture critical aspects of summarization quality, such as factual accuracy, particularly for long narratives (>100K tokens). Recent advances, such as LLM-as-a-Judge, address the limitations of metrics based on lexical similarity but still exhibit factual inconsistencies, especially in understanding character relationships and states. In this work, we introduce NarrativeFactScore, a novel "Agent-as-a-Judge" framework for evaluating and refining summaries. By leveraging a Character Knowledge Graph (CKG) extracted from input and generated summaries, NarrativeFactScore assesses the factual consistency and provides actionable guidance for refinement, such as identifying missing or erroneous facts. We demonstrate the effectiveness of NarrativeFactScore through a detailed workflow illustration and extensive validation on widely adopted benchmarks, achieving superior performance compared to competitive methods. Our results highlight the potential of agent-driven evaluation systems to improve the factual reliability of LLM-generated summaries.
|
[
{
"version": "v1",
"created": "Fri, 17 Jan 2025 07:23:06 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 11:18:41 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Jeong",
"Yeonseok",
""
],
[
"Kim",
"Minsoo",
""
],
[
"Hwang",
"Seung-won",
""
],
[
"Kim",
"Byung-Hak",
""
]
] |
2501.10053
|
Chuzhan Hao
|
Wenfeng Feng and Chuzhan Hao and Yuewei Zhang and Guochao Jiang and Jingyi Song and Hao Wang
|
AirRAG: Autonomous Strategic Planning and Reasoning Steer Retrieval Augmented Generation
|
20 pages, EMNLP25 Accepted
| null | null | null |
cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Leveraging the autonomous decision-making capabilities of large language models (LLMs) has demonstrated superior performance in reasoning tasks. However, despite the success of iterative or agentic retrieval-augmented generation (RAG) techniques, these methods are often constrained to a single solution space when confronted with complex problems. In this paper, we propose a novel thinking pattern in RAG that integrates autonomous strategic planning with efficient reasoning actions, significantly activating intrinsic reasoning capabilities and expanding the solution space of specific tasks via Monte Carlo Tree Search (MCTS), which we refer to as AirRAG. Specifically, our approach designs five fundamental reasoning actions, which are expanded to a broad tree-based reasoning space using MCTS. The approach also incorporates self-consistency verification to explore potential reasoning paths and inference scaling law. Additionally, computationally optimal strategies are employed to allocate more inference resources to key actions, thereby enhancing overall performance. Experimental results demonstrate the effectiveness of AirRAG, showing significant performance gains on complex question-answering datasets. Furthermore, AirRAG is flexible and lightweight, making it easy to integrate with other advanced technologies and models.
|
[
{
"version": "v1",
"created": "Fri, 17 Jan 2025 09:16:13 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Feb 2025 15:20:47 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 07:38:28 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Feng",
"Wenfeng",
""
],
[
"Hao",
"Chuzhan",
""
],
[
"Zhang",
"Yuewei",
""
],
[
"Jiang",
"Guochao",
""
],
[
"Song",
"Jingyi",
""
],
[
"Wang",
"Hao",
""
]
] |
2501.10913
|
Junsung Park
|
Junsung Park, Jungbeom Lee, Jongyoon Song, Sangwon Yu, Dahuin Jung, Sungroh Yoon
|
Know "No" Better: A Data-Driven Approach for Enhancing Negation Awareness in CLIP
|
Accepted to ICCV 2025
| null | null | null |
cs.CV cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
While CLIP has significantly advanced multimodal understanding by bridging vision and language, the inability to grasp negation - such as failing to differentiate concepts like "parking" from "no parking" - poses substantial challenges. By analyzing the data used in the public CLIP model's pre-training, we posit this limitation stems from a lack of negation-inclusive data. To address this, we introduce data generation pipelines that employ a large language model (LLM) and a multimodal LLM to produce negation-inclusive captions. Fine-tuning CLIP with data generated from our pipelines, we develop NegationCLIP, which enhances negation awareness while preserving the generality. Moreover, to enable a comprehensive evaluation of negation understanding, we propose NegRefCOCOg-a benchmark tailored to test VLMs' ability to interpret negation across diverse expressions and positions within a sentence. Experiments on various CLIP architectures validate the effectiveness of our data generation pipelines in enhancing CLIP's ability to perceive negation accurately. Additionally, NegationCLIP's enhanced negation awareness has practical applications across various multimodal tasks, demonstrated by performance gains in text-to-image generation and referring image segmentation.
|
[
{
"version": "v1",
"created": "Sun, 19 Jan 2025 01:17:05 GMT"
},
{
"version": "v2",
"created": "Mon, 31 Mar 2025 06:38:48 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 05:29:07 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Park",
"Junsung",
""
],
[
"Lee",
"Jungbeom",
""
],
[
"Song",
"Jongyoon",
""
],
[
"Yu",
"Sangwon",
""
],
[
"Jung",
"Dahuin",
""
],
[
"Yoon",
"Sungroh",
""
]
] |
2501.11110
|
Junjie Wang
|
Yiyao Yu, Yuxiang Zhang, Dongdong Zhang, Xiao Liang, Hengyuan Zhang, Xingxing Zhang, Mahmoud Khademi, Hany Awadalla, Junjie Wang, Yujiu Yang, Furu Wei
|
Chain-of-Reasoning: Towards Unified Mathematical Reasoning in Large Language Models via a Multi-Paradigm Perspective
|
Accepted to ACL 2025 (Main)
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large Language Models (LLMs) have made notable progress in mathematical reasoning, yet often rely on single-paradigm reasoning, limiting their effectiveness across diverse tasks. We introduce Chain-of-Reasoning (CoR), a novel unified framework integrating multiple reasoning paradigms--Natural Language Reasoning (NLR), Algorithmic Reasoning (AR), and Symbolic Reasoning (SR)--to enable synergistic collaboration. CoR generates multiple potential answers via different reasoning paradigms and synthesizes them into a coherent final solution. We propose a Progressive Paradigm Training (PPT) strategy for models to progressively master these paradigms, leading to CoR-Math-7B. Experimental results demonstrate that CoR-Math-7B significantly outperforms current SOTA models, achieving up to a 41.0% absolute improvement over GPT-4o in theorem proving and a 15.0% improvement over RL-based methods on the MATH benchmark in arithmetic tasks. These results show the enhanced mathematical comprehension ability of our model, enabling zero-shot generalization across tasks.
|
[
{
"version": "v1",
"created": "Sun, 19 Jan 2025 16:53:26 GMT"
},
{
"version": "v2",
"created": "Mon, 12 May 2025 13:04:44 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 10:56:53 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Yu",
"Yiyao",
""
],
[
"Zhang",
"Yuxiang",
""
],
[
"Zhang",
"Dongdong",
""
],
[
"Liang",
"Xiao",
""
],
[
"Zhang",
"Hengyuan",
""
],
[
"Zhang",
"Xingxing",
""
],
[
"Khademi",
"Mahmoud",
""
],
[
"Awadalla",
"Hany",
""
],
[
"Wang",
"Junjie",
""
],
[
"Yang",
"Yujiu",
""
],
[
"Wei",
"Furu",
""
]
] |
2501.11138
|
Isabelle Goldstein
|
Isabelle S. Goldstein, Louis E. Strigari
|
Kinematics of the Sagittarius Dwarf Spheroidal core: A 5D Analysis for a 6D Methodology with Gaia DR3
|
13 pages, 10 figures, 1 auxiliary text file
|
Mon Not R Astron Soc (2025) 1016-1027
|
10.1093/mnras/staf1026
| null |
astro-ph.GA astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Using Gaia DR3 data, we examine the kinematics of the central core of the Sagittarius (Sgr) dwarf spheroidal galaxy using data which includes proper motions and line-of-sight velocities for member stars in addition to their projected positions. We extract a sample of bright stars that are high-probability members of Sgr. We model the distances to these stars, which is the only missing phase-space component measurement from our 5D sample, highlighting how their corresponding uncertainties propagate to affect the kinematics. Using line-of-sight velocity data only, which are not affected by the distance uncertainties, and assuming a jeans-based equilibrium analysis, we obtain a velocity anisotropy of $\beta_a = -2.24 \pm 1.99$, which implies a system with tangentially-biased orbits. With the full 5D data, we project that the data will significantly improve upon measurements of the log-slope of the dark matter density profile and the stellar velocity anisotropy. Tests with mock distance data show an improvement of anisotropy errors of approximately an order of magnitude, and log slope at the half light radius of approximately half an order of magnitude.
|
[
{
"version": "v1",
"created": "Sun, 19 Jan 2025 18:29:30 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Jul 2025 18:44:05 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Goldstein",
"Isabelle S.",
""
],
[
"Strigari",
"Louis E.",
""
]
] |
2501.11418
|
Jan Nowosielski
|
Jan Nowosielski, Mateusz Mazelanik, Wojciech Wasilewski and Micha{\l} Parniak
|
Superheterodyne Rydberg S-band receiver with a multi-tone local oscillator based on an atomic transition loop
|
10 pages, 6 figures. This article draws and extends the idea from arXiv:2405.02901
| null | null | null |
physics.optics physics.atom-ph quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Atomic-vapor sensors based on Rydberg atoms now face a transition towards practical applications, with several outstanding challenges. To achieve the best sensitivities, a superheterodyne mode of operation is desired, which requires the presence of a local oscillator in the vapor cell. This local oscillator hinders several advantages of the sensor, such as stealthy and all-optical operation. We propose and realize a detection scheme, which avoids some of those problems by using multi-tone mixing, where direct usage of the local oscillator at the same frequency is not required. Our scheme is further elaborated on using efficient theoretical methods to predict the performance of the sensor. Our sensor operates at the S-band frequency, known for its usage in IEEE 802.11 (Wi-Fi) networks, without interfering with the signal itself.
|
[
{
"version": "v1",
"created": "Mon, 20 Jan 2025 11:34:56 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Jul 2025 11:31:30 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 08:25:52 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Nowosielski",
"Jan",
""
],
[
"Mazelanik",
"Mateusz",
""
],
[
"Wasilewski",
"Wojciech",
""
],
[
"Parniak",
"Michał",
""
]
] |
2501.12116
|
Pedro Taranc\'on-\'Alvarez
|
Pedro Taranc\'on-\'Alvarez, Pablo Tejerina-P\'erez, Raul Jimenez, Pavlos Protopapas
|
Efficient PINNs via Multi-Head Unimodular Regularization of the Solutions Space
| null |
Commun Phys 8, 335 (2025)
|
10.1038/s42005-025-02248-1
| null |
cs.LG cs.AI hep-th math.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Non-linear differential equations are a fundamental tool to describe different phenomena in nature. However, we still lack a well-established method to tackle stiff differential equations. Here we present a machine learning framework to facilitate the solution of nonlinear multiscale differential equations and, especially, inverse problems using Physics-Informed Neural Networks (PINNs). This framework is based on what is called \textit{multi-head} (MH) training, which involves training the network to learn a general space of all solutions for a given set of equations with certain variability, rather than learning a specific solution of the system. This setup is used with a second novel technique that we call Unimodular Regularization (UR) of the latent space of solutions. We show that the multi-head approach, combined with Unimodular Regularization, significantly improves the efficiency of PINNs by facilitating the transfer learning process thereby enabling the finding of solutions for nonlinear, coupled, and multiscale differential equations.
|
[
{
"version": "v1",
"created": "Tue, 21 Jan 2025 13:25:56 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 09:33:34 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Tarancón-Álvarez",
"Pedro",
""
],
[
"Tejerina-Pérez",
"Pablo",
""
],
[
"Jimenez",
"Raul",
""
],
[
"Protopapas",
"Pavlos",
""
]
] |
2501.12618
|
Ao Li
|
Ao Li, Byeongjee Kang, Vasudev Vikram, Isabella Laybourn, Samvid Dharanikota, Shrey Tiwari, and Rohan Padhye
|
Fray: An Efficient General-Purpose Concurrency Testing Platform for the JVM (Extended Version)
| null | null | null | null |
cs.PL cs.DC cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Concurrency bugs are hard to discover and reproduce. Prior work has developed sophisticated algorithms to search for concurrency bugs, such as partial order sampling (POS); however, fundamental limitations with existing platforms for concurrency control hinder effective testing of real-world software. We observe that the design space for concurrency control on managed code involves complex trade-offs between expressibility, applicability, and maintainability on the one hand, and bug-finding efficiency on the other hand.
This paper presents Fray, a platform for performing push-button concurrency testing of data-race-free JVM programs. The key insight behind Fray is that effective controlled concurrency testing requires orchestrating thread interleavings without replacing existing concurrency primitives, while encoding their semantics for faithfully expressing the set of all possible program behaviors. Fray incorporates a novel concurrency control mechanism called shadow locking, designed to make controlled concurrency testing practical and efficient for JVM programs. In an empirical evaluation on 53 benchmark programs with known bugs (SCTBench and JaConTeBe), Fray with random search finds 70% more bugs than JPF and 77% more bugs than RR's chaos mode. We also demonstrate Fray's push-button applicability on 2,655 tests from Apache Kafka, Lucene, and Google Guava. In these mature projects, Fray successfully discovered 18 real-world concurrency bugs that can cause 363 tests to fail reproducibly.
|
[
{
"version": "v1",
"created": "Wed, 22 Jan 2025 03:56:55 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 13:11:55 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Li",
"Ao",
""
],
[
"Kang",
"Byeongjee",
""
],
[
"Vikram",
"Vasudev",
""
],
[
"Laybourn",
"Isabella",
""
],
[
"Dharanikota",
"Samvid",
""
],
[
"Tiwari",
"Shrey",
""
],
[
"Padhye",
"Rohan",
""
]
] |
2501.12986
|
James Sullivan
|
James Sullivan, Zoltan Haiman, Mihir Kulkarni, Eli Visbal
|
Can supermassive stars form in protogalaxies due to internal Lyman-Werner feedback?
|
19 pages, 21 figures, and 1 table
|
Mon Not R Astron Soc (2025) 822-838
|
10.1093/mnras/staf1269
| null |
astro-ph.GA astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Population III stars are possible precursors to early massive and supermassive black holes (BHs). The presence of soft UV Lyman Werner (LW) background radiation can suppress Population III star formation in minihalos and allow them to form in pristine atomic cooling halos. In the absence of molecular hydrogen ($\rm H_2$) cooling, atomic-cooling halos enable rapid collapse with suppressed fragmentation. High background LW fluxes from preceding star-formation have been proposed to dissociate $\rm H_2$. This flux can be supplemented by LW radiation from one or more Population III star(s) in the same halo, reducing the necessary background level. Here we consider atomic-cooling halos in which multiple protostellar cores form close to one another nearly simultaneously. We assess whether the first star's LW radiation can dissociate nearby $\rm H_2$, enabling the prompt formation of a second, supermassive star (SMS) from warm, atomically-cooled gas. We use a set of hydrodynamical simulations with the code ENZO, with identical LW backgrounds centered on a halo with two adjacent collapsing gas clumps. When an additional large local LW flux is introduced, we observe immediate reductions in both the accretion rates and the stellar masses that form within these clumps. While the LW flux reduces the $\text{H}_2$ fraction and increases the gas temperature, the halo core's potential well is too shallow to promptly heat the gas to $\gtrsim$ 1000 K and increase the accretion rate onto the second protostar. We conclude that internal LW feedback inside atomic-cooling halos is unlikely to facilitate the formation of SMSs or massive BH seeds.
|
[
{
"version": "v1",
"created": "Wed, 22 Jan 2025 16:18:45 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Sullivan",
"James",
""
],
[
"Haiman",
"Zoltan",
""
],
[
"Kulkarni",
"Mihir",
""
],
[
"Visbal",
"Eli",
""
]
] |
2501.13097
|
Xiaoyu Liu
|
Xiaoyu Liu, Benjamin F. Schiffer, Jordi Tura
|
Preparing low-variance states using a distributed quantum algorithm
|
12+22 pages, 5+7 figures
|
Quantum 9, 1838 (2025)
|
10.22331/q-2025-08-28-1838
| null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Quantum computers are a highly promising tool for efficiently simulating quantum many-body systems. The preparation of their eigenstates is of particular interest and can be addressed, e.g., by quantum phase estimation algorithms. The routine then acts as an effective filtering operation, reducing the energy variance of the initial state. In this work, we present a distributed quantum algorithm inspired by iterative phase estimation to prepare low-variance states. Our method uses a single auxiliary qubit per quantum device, which controls its dynamics, and a postselection strategy for a joint quantum measurement on such auxiliary qubits. In the multi-device case, the result of this measurement heralds the successful runs of the protocol. This allows us to demonstrate that our distributed algorithm reduces the energy variance faster compared to single-device implementations, thereby highlighting the potential of distributed algorithms for near-term and early fault-tolerant devices.
|
[
{
"version": "v1",
"created": "Wed, 22 Jan 2025 18:55:51 GMT"
},
{
"version": "v2",
"created": "Tue, 19 Aug 2025 09:33:16 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Liu",
"Xiaoyu",
""
],
[
"Schiffer",
"Benjamin F.",
""
],
[
"Tura",
"Jordi",
""
]
] |
2501.15000
|
Zhongpu Chen
|
Zhongpu Chen, Yinfeng Liu, Long Shi, Xingyan Chen, Yu Zhao, Fuji Ren
|
MDEval: Evaluating and Enhancing Markdown Awareness in Large Language Models
|
WWW 2025
| null | null | null |
cs.CL cs.IR
|
http://creativecommons.org/licenses/by/4.0/
|
Large language models (LLMs) are expected to offer structured Markdown responses for the sake of readability in web chatbots (e.g., ChatGPT). Although there are a myriad of metrics to evaluate LLMs, they fail to evaluate the readability from the view of output content structure. To this end, we focus on an overlooked yet important metric -- Markdown Awareness, which directly impacts the readability and structure of the content generated by these language models. In this paper, we introduce MDEval, a comprehensive benchmark to assess Markdown Awareness for LLMs, by constructing a dataset with 20K instances covering 10 subjects in English and Chinese. Unlike traditional model-based evaluations, MDEval provides excellent interpretability by combining model-based generation tasks and statistical methods. Our results demonstrate that MDEval achieves a Spearman correlation of 0.791 and an accuracy of 84.1% with human, outperforming existing methods by a large margin. Extensive experimental results also show that through fine-tuning over our proposed dataset, less performant open-source models are able to achieve comparable performance to GPT-4o in terms of Markdown Awareness. To ensure reproducibility and transparency, MDEval is open sourced at https://github.com/SWUFE-DB-Group/MDEval-Benchmark.
|
[
{
"version": "v1",
"created": "Sat, 25 Jan 2025 00:26:01 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 02:36:57 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Chen",
"Zhongpu",
""
],
[
"Liu",
"Yinfeng",
""
],
[
"Shi",
"Long",
""
],
[
"Chen",
"Xingyan",
""
],
[
"Zhao",
"Yu",
""
],
[
"Ren",
"Fuji",
""
]
] |
2501.15752
|
Mojdeh Saleh
|
Mojdeh Saleh, Abhishek Kumar, Dmitrii L. Maslov, Saurabh Maiti
|
Resonant Edelstein and inverse-Edelstein effects, charge-to-spin conversion, and spin pumping from chiral-spin modes
| null | null | null | null |
cond-mat.mes-hall cond-mat.str-el
|
http://creativecommons.org/licenses/by/4.0/
|
Spin-orbit coupling in systems with broken inversion symmetry gives rise to the Edelstein effect, which is the spin polarization induced by an electric field or current, and the inverse-Edelstein effect (also known as the spin-galvanic effect), which is the electric current induced by an oscillatory magnetic field or spin polarization. At the same time, an interplay between spin-orbit coupling and electron-electron interaction leads to a special type of collective excitations -- chiral-spin modes -- which are oscillations of spin polarization in the absence of a magnetic field. As a result, both Edelstein and inverse-Edelstein effects exhibit resonances at the frequencies of chiral-spin collective modes. Here, we present a detailed study of the effect of electron correlation on the resonances in Edelstein and inverse-Edelstein effects in a single-valley two-dimensional electron gas and in a multi-valley Dirac system with proximity-induced spin-orbit coupling. While the chiral-spin modes involve both in-plane and out-of-plane oscillations of spins, we show that only the in-plane modes are responsible for the above resonances. In the multi-valley system, electron correlation splits the in-plane modes into two. We study the spectral weight distribution between the two modes over a large parameter space of intra- and inter-valley interactions. Finally, we demonstrate that using the chiral-spin modes one can get a resonant enhancement of charge-to-spin conversion and gain a directional control of the injected spins in the spin-pumping process, both of which are relevant to spintronics.
|
[
{
"version": "v1",
"created": "Mon, 27 Jan 2025 03:45:14 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 22:15:36 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Saleh",
"Mojdeh",
""
],
[
"Kumar",
"Abhishek",
""
],
[
"Maslov",
"Dmitrii L.",
""
],
[
"Maiti",
"Saurabh",
""
]
] |
2501.16301
|
Bram Vancraeynest-De Cuiper
|
Bram Vancraeynest-De Cuiper and Clement Delcamp
|
Twisted gauging and topological sectors in (2+1)d abelian lattice gauge theories
| null |
SciPost Phys. 19, 054 (2025)
|
10.21468/SciPostPhys.19.2.054
| null |
cond-mat.str-el hep-th quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Given a two-dimensional quantum lattice model with an abelian gauge theory interpretation, we investigate a duality operation that amounts to gauging its invertible 1-form symmetry, followed by gauging the resulting 0-form symmetry in a twisted way via a choice of discrete torsion. Using tensor networks, we introduce explicit lattice realisations of the so-called condensation defects, which are obtained by gauging the 1-form symmetry along submanifolds of spacetime, and employ the same calculus to realise the duality operators. By leveraging these tensor network operators, we compute the non-trivial interplay between symmetry-twisted boundary conditions and charge sectors under the duality operation, enabling us to construct isometries relating the dual Hamiltonians. Whenever a lattice gauge theory is left invariant under the duality operation, we explore the possibility of promoting the self-duality to an internal symmetry. We argue that this results in a symmetry structure that encodes the 2-representations of a 2-group.
|
[
{
"version": "v1",
"created": "Mon, 27 Jan 2025 18:39:10 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 12:43:28 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Cuiper",
"Bram Vancraeynest-De",
""
],
[
"Delcamp",
"Clement",
""
]
] |
2501.17468
|
Philip Schniter
|
Matt C. Bendel, Saurav K. Shastri, Rizwan Ahmad, and Philip Schniter
|
Solving Inverse Problems using Diffusion with Iterative Colored Renoising
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Imaging inverse problems can be solved in an unsupervised manner using pre-trained diffusion models, but doing so requires approximating the gradient of the measurement-conditional score function in the diffusion reverse process. We show that the approximations produced by existing methods are relatively poor, especially early in the reverse process, and so we propose a new approach that iteratively reestimates and "renoises" the estimate several times per diffusion step. This iterative approach, which we call Fast Iterative REnoising (FIRE), injects colored noise that is shaped to ensure that the pre-trained diffusion model always sees white noise, in accordance with how it was trained. We then embed FIRE into the DDIM reverse process and show that the resulting "DDfire" offers state-of-the-art accuracy and runtime on several linear inverse problems, as well as phase retrieval. Our implementation is at https://github.com/matt-bendel/DDfire
|
[
{
"version": "v1",
"created": "Wed, 29 Jan 2025 08:20:05 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Apr 2025 14:24:44 GMT"
},
{
"version": "v3",
"created": "Thu, 10 Jul 2025 17:09:37 GMT"
},
{
"version": "v4",
"created": "Wed, 27 Aug 2025 14:14:40 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bendel",
"Matt C.",
""
],
[
"Shastri",
"Saurav K.",
""
],
[
"Ahmad",
"Rizwan",
""
],
[
"Schniter",
"Philip",
""
]
] |
2501.18015
|
Jonas M. K\"ubler
|
Jonas M K\"ubler, Yu-Xiang Wang, Shoham Sabach, Navid Ansari, Matth\"aus Kleindessner, Kailash Budhathoki, Volkan Cevher, George Karypis
|
A Proximal Operator for Inducing 2:4-Sparsity
| null |
Transactions on Machine Learning Research, 2835-8856, 2025 (https://openreview.net/forum?id=AsFbXRIe4q)
| null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Recent hardware advancements in AI Accelerators and GPUs allow to efficiently compute sparse matrix multiplications, especially when 2 out of 4 consecutive weights are set to zero. However, this so-called 2:4 sparsity usually comes at a decreased accuracy of the model. We derive a regularizer that exploits the local correlation of features to find better sparsity masks in trained models. We minimize the regularizer jointly with a local squared loss by deriving the proximal operator for which we show that it has an efficient solution in the 2:4-sparse case. After optimizing the mask, we use maskedgradient updates to further minimize the local squared loss. We illustrate our method on toy problems and apply it to pruning entire large language models up to 70B parameters. On models up to 13B we improve over previous state of the art algorithms, whilst on 70B models we match their performance.
|
[
{
"version": "v1",
"created": "Wed, 29 Jan 2025 22:05:17 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Kübler",
"Jonas M",
""
],
[
"Wang",
"Yu-Xiang",
""
],
[
"Sabach",
"Shoham",
""
],
[
"Ansari",
"Navid",
""
],
[
"Kleindessner",
"Matthäus",
""
],
[
"Budhathoki",
"Kailash",
""
],
[
"Cevher",
"Volkan",
""
],
[
"Karypis",
"George",
""
]
] |
2501.18507
|
Teresa Krick
|
Teresa Krick and Agnes Szanto
|
A note on the multivariate symmetric Hermite Interpolant
|
17 pages. To Appear in J.Symb.Comput
| null | null | null |
math.CA math.AC
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In this note we explicit the notion of Hermite interpolant of a multivariate symmetric polynomial, generalizing the notion of Lagrange interpolant to the case when there are roots coalescence, an extension of the results on the symmetric Hermite interpolation basis by M.-F. Roy and A. Szpirglas.
|
[
{
"version": "v1",
"created": "Thu, 30 Jan 2025 17:16:12 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 20:09:34 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Krick",
"Teresa",
""
],
[
"Szanto",
"Agnes",
""
]
] |
2501.18561
|
Antonio Agresti
|
Antonio Agresti and Mark Veraar
|
Nonlinear SPDEs and Maximal Regularity: An Extended Survey
|
Typos corrected. To appear in Nonlinear Differential Equations and Applications
| null | null | null |
math.PR math.AP math.CA math.FA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this survey, we provide an in-depth exposition of our recent results on the well-posedness theory for stochastic evolution equations, employing maximal regularity techniques. The core of our approach is an abstract notion of critical spaces, which, when applied to nonlinear SPDEs, coincides with the concept of scaling-invariant spaces. This framework leads to several sharp blow-up criteria and enables one to obtain instantaneous regularization results. Additionally, we refine and unify our previous results, while also presenting several new contributions.
In the second part of the survey, we apply the abstract results to several concrete SPDEs. In particular, we give applications to stochastic perturbations of quasi-geostrophic equations, Navier-Stokes equations, and reaction-diffusion systems (including Allen--Cahn, Cahn--Hilliard and Lotka--Volterra models). Moreover, for the Navier--Stokes equations, we establish new Serrin-type blow-up criteria. While some applications are addressed using $L^2$-theory, many require a more general $L^p(L^q)$-framework. In the final section, we outline several open problems, covering both abstract aspects of stochastic evolution equations, and concrete questions in the study of linear and nonlinear SPDEs.
|
[
{
"version": "v1",
"created": "Thu, 30 Jan 2025 18:36:28 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Feb 2025 12:33:49 GMT"
},
{
"version": "v3",
"created": "Wed, 14 May 2025 11:57:11 GMT"
},
{
"version": "v4",
"created": "Thu, 17 Jul 2025 17:58:27 GMT"
},
{
"version": "v5",
"created": "Wed, 27 Aug 2025 08:29:47 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Agresti",
"Antonio",
""
],
[
"Veraar",
"Mark",
""
]
] |
2502.00521
|
Ali \"Ovg\"un Assoc.Prof.Dr.
|
Vitalii Vertogradov, Ali \"Ovg\"un, Daniil Shatov
|
Formation of regular black hole from baryonic matter
|
12 pages. To be published in Chinese Physics C
|
Chin. Phys. C 49, no.11,115103 (2025)
|
10.1088/1674-1137/ade95c
| null |
gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
We present a family of exact, singularity-free solutions describing the collapse of baryonic matter characterized by a barotropic equation of state whose coefficient $\alpha(r,v)$ varies in both radius and time. By matching these interior solutions to the Husain exterior metric, we obtain a self-consistent, dynamical spacetime representing a regular black hole. Although the pressure profile of our models grows with radius and eventually violates the dominant energy condition beyond a critical surface-necessitating an external junction to ensure a globally well-defined spacetime-the interior solution remains non-singular throughout the collapse. We further analyze the optical properties of these regular black holes and find that both the photon sphere radius and the corresponding shadow radius increase monotonically as the local equation of state parameter $\alpha$ is raised. Moreover, the matching interface between the interior and exterior metrics naturally suggests a phase transition in the collapsing fluid, which can postpone the formation of an apparent horizon. Taken together, our results not only highlight novel physical features of horizon formation in regular collapse models but also identify characteristic shadow signatures that could be tested by future observations.
|
[
{
"version": "v1",
"created": "Sat, 01 Feb 2025 18:49:33 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Jun 2025 05:58:38 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Vertogradov",
"Vitalii",
""
],
[
"Övgün",
"Ali",
""
],
[
"Shatov",
"Daniil",
""
]
] |
2502.00995
|
Paolo Bertozzini -
|
Paolo Bertozzini, Roberto Conti, Wicharn Lewkeeratiyutkul, Kasemsun Rutamorn
|
Spectral Theory for Non-full Commutative C*-categories
|
AMS-LaTeX-2e, 34 pages, extended introduction, typos corrected, no change in the main text and in the results
|
Advances in Operator Theory 10(4):80 (2025)
| null | null |
math.OA math.CT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We extend the spectral theory of commutative C*-categories to the non full-case, introducing a suitable notion of spectral spaceoid provinding a duality between a category of "non-trivial" *-functors of non-full commutative C*-categories and a category of Takahashi morphisms of "non-full spaceoids" (here defined). As a byproduct we obtain a spectral theorem for a non-full generalization of imprimitivity Hilbert C*-bimodules over commutative unital C*-algebras via continuous sections vanishing at infinity of a Hilbert C*-line-bundle over the graph of a homeomorphism between open subsets of the corresponding Gel'fand spectra of the C*-algebras.
|
[
{
"version": "v1",
"created": "Mon, 03 Feb 2025 02:25:13 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Feb 2025 02:19:32 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bertozzini",
"Paolo",
""
],
[
"Conti",
"Roberto",
""
],
[
"Lewkeeratiyutkul",
"Wicharn",
""
],
[
"Rutamorn",
"Kasemsun",
""
]
] |
2502.01700
|
Mohammad Amin Hasanpour
|
Mohammad Amin Hasanpour, Mikkel Kirkegaard, Xenofon Fafoutis
|
EdgeMark: An Automation and Benchmarking System for Embedded Artificial Intelligence Tools
| null |
Journal of Systems Architecture (2025)
|
10.1016/j.sysarc.2025.103488
| null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
The integration of artificial intelligence (AI) into embedded devices, a paradigm known as embedded artificial intelligence (eAI) or tiny machine learning (TinyML), is transforming industries by enabling intelligent data processing at the edge. However, the many tools available in this domain leave researchers and developers wondering which one is best suited to their needs. This paper provides a review of existing eAI tools, highlighting their features, trade-offs, and limitations. Additionally, we introduce EdgeMark, an open-source automation system designed to streamline the workflow for deploying and benchmarking machine learning (ML) models on embedded platforms. EdgeMark simplifies model generation, optimization, conversion, and deployment while promoting modularity, reproducibility, and scalability. Experimental benchmarking results showcase the performance of widely used eAI tools, including TensorFlow Lite Micro (TFLM), Edge Impulse, Ekkono, and Renesas eAI Translator, across a wide range of models, revealing insights into their relative strengths and weaknesses. The findings provide guidance for researchers and developers in selecting the most suitable tools for specific application requirements, while EdgeMark lowers the barriers to adoption of eAI technologies.
|
[
{
"version": "v1",
"created": "Mon, 03 Feb 2025 08:28:01 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hasanpour",
"Mohammad Amin",
""
],
[
"Kirkegaard",
"Mikkel",
""
],
[
"Fafoutis",
"Xenofon",
""
]
] |
2502.01793
|
Junyeong Ahn
|
Junyeong Ahn and Ashvin Vishwanath
|
Circular-polarization-selective perfect reflection from chiral superconductors
|
11 pages, 5 figures
|
Nat. Commun. 16, 6493 (2025)
|
10.1038/s41467-025-61658-5
| null |
cond-mat.mes-hall cond-mat.supr-con
|
http://creativecommons.org/licenses/by/4.0/
|
Integrating mirrors with magnetic components is crucial for constructing chiral optical cavities, which provide tunable platforms for time-reversal-asymmetric light-matter interactions. Here, we introduce single-crystal circular-polarization-selective mirrors based on chiral superconductors, which break time-reversal symmetry themselves, eliminating the need for additional components. We show that a circular-polarization-selective perfect reflection (CSPR) occurs for strong-coupling superconductors in the BCS-BEC crossover regime or beyond if the optical Hall conductivity is significant in the unit of conductivity quantum per unit layer, $e^2/ha_z$, where $a_z$ is the lattice constant along the surface normal. While the optical Hall conductivity in chiral superconductors is typically tiny, we classify three routes to obtain a large value. We demonstrate the significant optical Hall conductivity and the resulting CSPR with two examples: (1) superconductivity in doped quantum Hall insulators and (2) chiral pairing that preserves the Bogoliubov Fermi surfaces in the weak-pairing limit. We also discuss the application of our theory to the recently discovered chiral superconducting phase in rhombohedral graphene. Our theory reveals the potential of these classes of chiral superconductors as promising elements for building high-quality-factor terahertz chiral cavities.
|
[
{
"version": "v1",
"created": "Mon, 03 Feb 2025 20:14:47 GMT"
},
{
"version": "v2",
"created": "Wed, 18 Jun 2025 02:20:31 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Ahn",
"Junyeong",
""
],
[
"Vishwanath",
"Ashvin",
""
]
] |
2502.02171
|
Mohamed Youssef
|
Mohamed Youssef, Jian Peng, Oliver Bimber
|
DeepForest: Sensing Into Self-Occluding Volumes of Vegetation With Aerial Imaging
| null | null | null | null |
cs.CV eess.IV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Access to below-canopy volumetric vegetation data is crucial for understanding ecosystem dynamics. We address the long-standing limitation of remote sensing to penetrate deep into dense canopy layers. LiDAR and radar are currently considered the primary options for measuring 3D vegetation structures, while cameras can only extract the reflectance and depth of top layers. Using conventional, high-resolution aerial images, our approach allows sensing deep into self-occluding vegetation volumes, such as forests. It is similar in spirit to the imaging process of wide-field microscopy, but can handle much larger scales and strong occlusion. We scan focal stacks by synthetic-aperture imaging with drones and reduce out-of-focus signal contributions using pre-trained 3D convolutional neural networks with mean squared error (MSE) as the loss function. The resulting volumetric reflectance stacks contain low-frequency representations of the vegetation volume. Combining multiple reflectance stacks from various spectral channels provides insights into plant health, growth, and environmental conditions throughout the entire vegetation volume. Compared with simulated ground truth, our correction leads to ~x7 average improvements (min: ~x2, max: ~x12) for forest densities of 220 trees/ha - 1680 trees/ha. In our field experiment, we achieved an MSE of 0.05 when comparing with the top-vegetation layer that was measured with classical multispectral aerial imaging.
|
[
{
"version": "v1",
"created": "Tue, 04 Feb 2025 09:45:49 GMT"
},
{
"version": "v2",
"created": "Tue, 20 May 2025 12:05:04 GMT"
},
{
"version": "v3",
"created": "Thu, 31 Jul 2025 10:02:07 GMT"
},
{
"version": "v4",
"created": "Wed, 27 Aug 2025 11:09:24 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Youssef",
"Mohamed",
""
],
[
"Peng",
"Jian",
""
],
[
"Bimber",
"Oliver",
""
]
] |
2502.02306
|
Ryan Kozlowski
|
Ryan Kozlowski, Luis A. Pugnaloni
|
Flow rate from a vertical silo with a tilted orifice
|
25 pages total: 2 pages with graphical abstract and highlights for preprint, 18 pages for main text, and 5 pages for supplemental materials. 16 figures in main text, 4 figures in supplemental materials. (Videos will be provided with Supplemental Materials upon peer-reviewed publication.)
| null |
10.1016/j.powtec.2025.121189
| null |
cond-mat.soft
|
http://creativecommons.org/licenses/by/4.0/
|
The flow of dry granular materials from silos is of great practical interest in industry and of theoretical import for understanding multiphase dynamics. Recent studies have demonstrated that one way to control the rate of flow from a silo is to tilt it. However, this may not be practical in many industrial applications. Here, we demonstrate in experiments of quasi-2D silo discharge of monodisperse grains that the flow rate can be modulated by rotating the orifice - through elevating and shifting one side of the base - instead of tilting the entire silo. We use high-speed image analysis to track the average motion of grains in the silo. We first show that the flow rate decreases with orifice angle, but that this decrease is not as strong as when a silo is tilted or when a lateral orifice is used. However, with the addition of a grain-sized ridge on each side of the orifice, the flow rate collapses with prior tilted-silo results. We then characterize the flow velocity of grains exiting the orifice and highlight key features of the stagnant zones and slip zones on each orifice side. Finally, we model our results based on these measurements, demonstrating the importance of horizontal creep along slip zones next to the orifice and the narrowest opening cross-section through which the material flows. These findings reveal a simple method for controlling both flow rate and direction, and highlight the importance of both dynamics within and geometry of the stagnant zones near the orifice.
|
[
{
"version": "v1",
"created": "Tue, 04 Feb 2025 13:20:04 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Kozlowski",
"Ryan",
""
],
[
"Pugnaloni",
"Luis A.",
""
]
] |
2502.02836
|
Michael Reitz
|
Michael Reitz, Stephan van den Wildenberg, Arghadip Koner, George C. Schatz, Joel Yuen-Zhou
|
Quantum theory of surface lattice resonances
|
9 pages main text (6 figures), 9 pages Supplement (2 figures). [v2] includes a section on nonlinear pump-probe spectroscopy of SLRs
| null | null | null |
quant-ph cond-mat.mes-hall physics.optics
|
http://creativecommons.org/licenses/by/4.0/
|
The collective interactions of nanoparticles arranged in periodic structures give rise to high-$Q$ in-plane diffractive modes known as surface lattice resonances (SLRs). While these resonances and their broader implications have been extensively studied within the framework of classical electrodynamics and linear response theory, a quantum optical theory capable of describing the dynamics of these structures, especially in the presence of material nonlinearities, beyond the use of \textit{ad hoc} few-mode approximations is largely missing. To this end, we consider here a lattice of metallic nanoparticles coupled to the background electromagnetic field and derive the quantum input-output relations within the electric dipole approximation. As a first application, we discuss coupling of the nanoparticle array to an external array of emitters and illustrate how the formalism can be extended to molecular optomechanics. Specifically, we show how the high $Q$-factors provided by SLRs can facilitate coupling between the nanoparticle array and collective vibrational modes. Finally, we examine a scenario in which the metallic nanoparticles themselves are replaced with saturable excitonic emitters, illustrating how the nonlinearity of the emitters can be used to switch the SLR condition between different electronic transitions. We further show, using a perturbative approach that explicitly accounts for population dynamics, how the resulting switching dynamics can be observed and characterized in pump-probe experiments. Our work provides a seamless framework to rigorously model SLR phenomena with quantum emitters without resorting to phenomenological treatments of either the electromagnetic environment or the material degrees of freedom.
|
[
{
"version": "v1",
"created": "Wed, 05 Feb 2025 02:47:07 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 19:29:34 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Reitz",
"Michael",
""
],
[
"Wildenberg",
"Stephan van den",
""
],
[
"Koner",
"Arghadip",
""
],
[
"Schatz",
"George C.",
""
],
[
"Yuen-Zhou",
"Joel",
""
]
] |
2502.03030
|
Arthur Hughes
|
Arthur Hughes, Layla Parast, Rodolphe Thi\'ebaut, Boris P. Hejblum
|
RISE: Two-Stage Rank-Based Identification of High-Dimensional Surrogate Markers Applied to Vaccinology
| null | null | null | null |
stat.ME stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
In vaccine trials with long-term participant follow-up, it is of great importance to identify surrogate markers that accurately infer long-term immune responses. These markers offer practical advantages such as providing early, indirect evidence of vaccine efficacy, and can accelerate vaccine development while identifying potential biomarkers. High-throughput technologies like RNA-sequencing have emerged as promising tools for understanding complex biological systems and informing new treatment strategies. However, these data are high-dimensional, presenting unique statistical challenges for existing surrogate marker identification methods. We introduce Rank-based Identification of high-dimensional SurrogatE Markers (RISE), a novel approach designed for small sample, high-dimensional settings typical in modern vaccine experiments. RISE employs a non-parametric univariate test to screen variables for promising candidates, followed by surrogate evaluation on independent data. Our simulation studies demonstrate RISE's desirable properties, including type one error rate control and empirical power under various conditions. Applying RISE to a clinical trial for inactivated influenza vaccination, we sought to identify genes whose expression could serve as a surrogate for the induced immune response. This analysis revealed a signature of genes appearing to function as a reasonable surrogate for the neutralising antibody response. Pathways related to innate antiviral signalling and interferon stimulation were strongly represented in this derived surrogate, providing a clear immunological interpretation.
|
[
{
"version": "v1",
"created": "Wed, 05 Feb 2025 09:33:42 GMT"
},
{
"version": "v2",
"created": "Tue, 20 May 2025 17:14:08 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 13:51:51 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hughes",
"Arthur",
""
],
[
"Parast",
"Layla",
""
],
[
"Thiébaut",
"Rodolphe",
""
],
[
"Hejblum",
"Boris P.",
""
]
] |
2502.04197
|
William Meier
|
William R. Meier, David E. Graf, Brenden R. Ortiz, Shirin Mozaffari and David Mandrus
|
Pressure suppresses the density wave order in kagome metal LuNb$_6$Sn$_6$
|
6 pages, 3 figures
|
Phys. Rev. Materials 9, L082001 (2025)
|
10.1103/hsdc-9j7p
| null |
cond-mat.str-el
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Dancing tins pair up,
But compressing the framework
Thwarts the displacements.
The density waves that develop in kagome metals ScV$_{6}$Sn$_{6}$ and LuNb$_{6}$Sn$_{6}$ at low temperature appear to arise from under-filled atomic columns within a V-Sn or Nb-Sn scaffolding. Compressing this network with applied pressure in ScV$_{6}$Sn$_{6}$ suppressed the structural transition temperature by constraining atomic rattling and inhibiting the shifts that define the structural modulation. We predicted that the density wave transition in LuNb$_{6}$Sn$_{6}$ at 68 K would be suppressed by pressure as well. In this brief study, we examine the pressure dependence of the density wave transition by measuring resistance vs temperature up to 2.26 GPa. We found the transition temperature is smoothly depressed and disappears around 1.9 GPa. This result not only addresses our prediction, but strengthens the rattling chains origin of structural instabilities in the HfFe$_{6}$Ge$_{6}$-type kagome metals.
|
[
{
"version": "v1",
"created": "Thu, 06 Feb 2025 16:36:45 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 15:03:21 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Meier",
"William R.",
""
],
[
"Graf",
"David E.",
""
],
[
"Ortiz",
"Brenden R.",
""
],
[
"Mozaffari",
"Shirin",
""
],
[
"Mandrus",
"David",
""
]
] |
2502.04610
|
Jian Li
|
Dominik Kwietniak, Jian Li, Habibeh Pourmand
|
A note on logarithmic mean equicontinuity
|
17 pages, to appear in J. Dynam. Differential Equations
| null | null | null |
math.DS
|
http://creativecommons.org/licenses/by/4.0/
|
We study the set of harmonic limits of empirical measures in topological dynamical systems. We obtain a characterization of unique ergodicity based of logarithmic (harmonic) mean convergence in place of Ces\`aro convergence. We introduce logarithmic mean equicontinuity and show that a topological dynamical system is logarithmically mean equicontinuous if and only if it is mean equicontinuous.
|
[
{
"version": "v1",
"created": "Fri, 07 Feb 2025 02:06:34 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 03:56:05 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Kwietniak",
"Dominik",
""
],
[
"Li",
"Jian",
""
],
[
"Pourmand",
"Habibeh",
""
]
] |
2502.05380
|
Sarah Lotspeich
|
Sarah C. Lotspeich, Sheetal Kedar, Rabeya Tahir, Aidan D. Keleghan, Amelia Miranda, Stephany N. Duda, Michael P. Bancks, Brian J. Wells, Ashish K. Khanna, Joseph Rigdon
|
Overcoming data challenges through enriched validation and targeted sampling to measure whole-person health in electronic health records
|
18 pages, 1 table, 7 figures, supplementary materials and code on GitHub
| null | null | null |
stat.ME stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
The allostatic load index (ALI) is a 10-component measure of whole-person health. Data from electronic health records (EHR) present a huge opportunity to operationalize the ALI in learning health systems; however, these data are prone to missingness and errors. Validation (e.g., through chart reviews) provides better-quality data, but realistically, only a subset of patients' data can be validated, and most protocols do not recover missing data. Using a representative sample of 1000 patients from the EHR at an extensive learning health system (100 of whom could be validated), we propose methods to design, conduct, and analyze statistically efficient and robust studies of ALI and healthcare utilization. Employing semiparametric maximum likelihood estimation, we robustly incorporate all available patient information into statistical models. Using targeted design strategies, we examine ways to select the most informative patients for validation. Incorporating clinical expertise, we devise a novel validation protocol to promote EHR data quality and completeness. Chart reviews uncovered few errors (99% matched source documents) and recovered some missing data through auxiliary information in patients' charts. On average, validation increased the number of non-missing ALI components per patient from 6 to 7. Through simulations based on preliminary data, residual sampling was identified as the most informative strategy for completing our validation study. Incorporating validation data, statistical models indicated that worse whole-person health (higher ALI) was associated with higher odds of engaging in the healthcare system, adjusting for age.
|
[
{
"version": "v1",
"created": "Fri, 07 Feb 2025 23:25:01 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Feb 2025 13:08:47 GMT"
},
{
"version": "v3",
"created": "Mon, 24 Feb 2025 19:38:24 GMT"
},
{
"version": "v4",
"created": "Sat, 19 Jul 2025 19:36:41 GMT"
},
{
"version": "v5",
"created": "Wed, 27 Aug 2025 14:27:18 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Lotspeich",
"Sarah C.",
""
],
[
"Kedar",
"Sheetal",
""
],
[
"Tahir",
"Rabeya",
""
],
[
"Keleghan",
"Aidan D.",
""
],
[
"Miranda",
"Amelia",
""
],
[
"Duda",
"Stephany N.",
""
],
[
"Bancks",
"Michael P.",
""
],
[
"Wells",
"Brian J.",
""
],
[
"Khanna",
"Ashish K.",
""
],
[
"Rigdon",
"Joseph",
""
]
] |
2502.07208
|
Hyeonuk Nam
|
Hyeonuk Nam, Seong-Hu Kim, Deokki Min, Byeong-Yun Ko, Yong-Hwa Park
|
Towards Understanding of Frequency Dependence on Sound Event Detection
|
Accepted to IEEE/ACM TASLP
| null | null | null |
eess.AS cs.SD
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this work, we conduct an in-depth analysis of two frequency-dependent methods for sound event detection (SED): FilterAugment and frequency dynamic convolution (FDY conv). The goal is to better understand their characteristics and behaviors in the context of SED. While SED has been rapidly advancing through the adoption of various deep learning techniques from other pattern recognition fields, such adopted techniques are often not suitable for SED. To address this issue, two frequency-dependent SED methods were previously proposed: FilterAugment, a data augmentation randomly weighting frequency bands, and FDY conv, an architecture applying frequency adaptive convolution kernels. These methods have demonstrated superior performance in SED, and we aim to further analyze their detailed effectiveness and characteristics in SED. We compare class-wise performance to find out specific pros and cons of FilterAugment and FDY conv. We apply Gradient-weighted Class Activation Mapping (Grad-CAM), which highlights time-frequency region that is more inferred by the model, on SED models with and without frequency masking and two types of FilterAugment to observe their detailed characteristics. We propose simpler frequency dependent convolution methods and compare them with FDY conv to further understand which components of FDY conv affects SED performance. Lastly, we apply PCA to show how FDY conv adapts dynamic kernel across frequency dimensions on different sound event classes. The results and discussions demonstrate that frequency dependency plays a significant role in sound event detection and further confirms the effectiveness of frequency dependent methods on SED.
|
[
{
"version": "v1",
"created": "Tue, 11 Feb 2025 03:07:03 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 08:09:06 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Nam",
"Hyeonuk",
""
],
[
"Kim",
"Seong-Hu",
""
],
[
"Min",
"Deokki",
""
],
[
"Ko",
"Byeong-Yun",
""
],
[
"Park",
"Yong-Hwa",
""
]
] |
2502.09242
|
Lukas Buess
|
Lukas Buess, Matthias Keicher, Nassir Navab, Andreas Maier, Soroosh Tayebi Arasteh
|
From large language models to multimodal AI: A scoping review on the potential of generative AI in medicine
| null |
Biomed. Eng. Lett. 15 (2025)
|
10.1007/s13534-025-00497-1
| null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Generative artificial intelligence (AI) models, such as diffusion models and OpenAI's ChatGPT, are transforming medicine by enhancing diagnostic accuracy and automating clinical workflows. The field has advanced rapidly, evolving from text-only large language models for tasks such as clinical documentation and decision support to multimodal AI systems capable of integrating diverse data modalities, including imaging, text, and structured data, within a single model. The diverse landscape of these technologies, along with rising interest, highlights the need for a comprehensive review of their applications and potential. This scoping review explores the evolution of multimodal AI, highlighting its methods, applications, datasets, and evaluation in clinical settings. Adhering to PRISMA-ScR guidelines, we systematically queried PubMed, IEEE Xplore, and Web of Science, prioritizing recent studies published up to the end of 2024. After rigorous screening, 144 papers were included, revealing key trends and challenges in this dynamic field. Our findings underscore a shift from unimodal to multimodal approaches, driving innovations in diagnostic support, medical report generation, drug discovery, and conversational AI. However, critical challenges remain, including the integration of heterogeneous data types, improving model interpretability, addressing ethical concerns, and validating AI systems in real-world clinical settings. This review summarizes the current state of the art, identifies critical gaps, and provides insights to guide the development of scalable, trustworthy, and clinically impactful multimodal AI solutions in healthcare.
|
[
{
"version": "v1",
"created": "Thu, 13 Feb 2025 11:57:51 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Buess",
"Lukas",
""
],
[
"Keicher",
"Matthias",
""
],
[
"Navab",
"Nassir",
""
],
[
"Maier",
"Andreas",
""
],
[
"Arasteh",
"Soroosh Tayebi",
""
]
] |
2502.10761
|
Bolin Li
|
Bolin Li, Wentao Zhang, Xuecong Huang, Lijun Zhu, and Han Ding
|
A Whole-Body Disturbance Rejection Control Framework for Dynamic Motions in Legged Robots
|
have been accepted for IEEE RA-L
| null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This letter presents a control framework for legged robots that enables self-perception and resistance to external disturbances and model uncertainties. First, a novel disturbance estimator is proposed, integrating adaptive control and extended state observers (ESO) to estimate external disturbances and model uncertainties. This estimator is embedded within the whole-body control framework to compensate for disturbances in the legged system. Second, a comprehensive whole-body disturbance rejection control framework (WB-DRC) is introduced, accounting for the robot's full-body dynamics. Compared to previous whole-body control frameworks, WB-DRC effectively handles external disturbances and model uncertainties, with the potential to adapt to complex terrain. Third, simulations of both biped and quadruped robots are conducted in the Gazebo simulator to demonstrate the effectiveness and versatility of WB-DRC. Finally, extensive experimental trials on the quadruped robot validate the robustness and stability of the robot system using WB-DRC under various disturbance conditions.
|
[
{
"version": "v1",
"created": "Sat, 15 Feb 2025 10:57:27 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Feb 2025 01:44:51 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 03:00:17 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Li",
"Bolin",
""
],
[
"Zhang",
"Wentao",
""
],
[
"Huang",
"Xuecong",
""
],
[
"Zhu",
"Lijun",
""
],
[
"Ding",
"Han",
""
]
] |
2502.11504
|
Janak M. Patel
|
Janak M. Patel, Milad Ramezankhani, Anirudh Deodhar, Dagnachew Birru
|
Accelerated Gradient-based Design Optimization Via Differentiable Physics-Informed Neural Operator: A Composites Autoclave Processing Case Study
|
15 pages, 7 figures
|
j.compositesb. 1359-8368 (2025) 112935
|
10.1016/j.compositesb.2025.112935
| null |
cs.LG cs.AI cs.NA math.NA
|
http://creativecommons.org/licenses/by/4.0/
|
Simulation and optimization are crucial for advancing the engineering design of complex systems and processes. Traditional optimization methods require substantial computational time and effort due to their reliance on resource-intensive simulations, such as finite element analysis, and the complexity of rigorous optimization algorithms. Data-agnostic AI-based surrogate models, such as Physics-Informed Neural Operators (PINOs), offer a promising alternative to these conventional simulations, providing drastically reduced inference time, unparalleled data efficiency, and zero-shot super-resolution capability. However, the predictive accuracy of these models is often constrained to small, low-dimensional design spaces or systems with relatively simple dynamics. To address this, we introduce a novel Physics-Informed DeepONet (PIDON) architecture, which extends the capabilities of conventional neural operators to effectively model the nonlinear behavior of complex engineering systems across high-dimensional design spaces and a wide range of dynamic design configurations. This new architecture outperforms existing SOTA models, enabling better predictions across broader design spaces. Leveraging PIDON's differentiability, we integrate a gradient-based optimization approach using the Adam optimizer to efficiently determine optimal design variables. This forms an end-to-end gradient-based optimization framework that accelerates the design process while enhancing scalability and efficiency. We demonstrate the effectiveness of this framework in the optimization of aerospace-grade composites curing processes achieving a 3x speedup in obtaining optimal design variables compared to gradient-free methods. Beyond composites processing, the proposed model has the potential to be used as a scalable and efficient optimization tool for broader applications in advanced engineering and digital twin systems.
|
[
{
"version": "v1",
"created": "Mon, 17 Feb 2025 07:11:46 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Patel",
"Janak M.",
""
],
[
"Ramezankhani",
"Milad",
""
],
[
"Deodhar",
"Anirudh",
""
],
[
"Birru",
"Dagnachew",
""
]
] |
2502.11779
|
Xuan Ren
|
Xuan Ren, Qi Chen, Lingqiao Liu
|
Efficient Response Generation Strategy Selection for Fine-Tuning Large Language Models Through Self-Aligned Perplexity
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Fine-tuning large language models (LLMs) typically relies on producing large sets of input-output pairs. Yet for a given question, there can be many valid outputs. In practice, these outputs are often derived by distilling knowledge from teacher models, and they can vary depending on the specific teacher model or prompting strategy employed. Recent findings show that how these training outputs are generated can significantly affect the performance of the fine-tuned model, raising an important question: how do we pick the best data generation method from among numerous possibilities? Rather than exhaustively training and evaluating on each candidate, this paper proposes a scalable approximate method that assesses a small subset of generated data to estimate its suitability for a specific target LLM. Our central idea is that effective outputs should be familiar to the target LLM. While previous work measures familiarity with perplexity, we find that perplexity might be suboptimal in characterizing familiarity through empirical analyses and practical observations. To address this, we introduce self-aligned perplexity, a novel metric capturing how closely candidate outputs adhere to the target LLM's own style and reasoning patterns. In this way, we can identify the most effective generation strategy on a small sample, then apply it to produce the complete training set. We demonstrate that training on data generated by the chosen method yields significant improvements across diverse reasoning-focused benchmarks, particularly in cases where different candidate methods lead to highly divergent training outcomes. Our implementation is publicly available at https://github.com/XuanRen4470/SPPL.
|
[
{
"version": "v1",
"created": "Mon, 17 Feb 2025 13:14:11 GMT"
},
{
"version": "v2",
"created": "Tue, 08 Apr 2025 13:56:38 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 04:45:16 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Ren",
"Xuan",
""
],
[
"Chen",
"Qi",
""
],
[
"Liu",
"Lingqiao",
""
]
] |
2502.11833
|
Ulrich S. Schwarz
|
Nils O. Winkler, Oliver M. Drozdowski, Falko Ziebert and Ulrich S. Schwarz (Heidelberg University)
|
Active gel theory for cell migration with two myosin isoforms
|
Revtex, 14 pages, 9 figures, 63 references
|
Phys. Rev. E. 112, 024413 (Aug 26 2025)
|
10.1103/8h9z-1qfs
| null |
q-bio.CB cond-mat.soft physics.bio-ph
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Myosin II molecular motors slide actin filaments relatively to each other and are essential for force generation, motility and mechanosensing in animal cells. For non-muscle cells, evolution has resulted in three different isoforms, which have different properties concerning the motor cycle and also occur in different abundances in the cells, but their respective biological and physical roles are not fully understood. Here we use active gel theory to demonstrate the complementary roles of isoforms A and B for cell migration. We first show that our model can be derived both from coarse-graining kinetic equations and from nonequilibrium thermodynamics as the macroscopic limit of a two-component Tonks gas. We then parametrize the model and show that motile solutions exist, in which the more abundant and more dynamic isoform A is localized to the front and the stronger isoform B to the rear, in agreement with experiments. Exploring parameter space beyond the isoform parameters typical for animal cells, we also find cell oscillations in length and velocity, which might be realized for genetically engineered systems. We also describe an analytical solution for the stiff limit, which then is used to calculate a state diagram, and the effect of actin polymerization at the boundaries, that leads to an imperfect pitchfork bifurcation. Our findings highlight the importance of including isoform-specific molecular details to describe whole cell behavior.
|
[
{
"version": "v1",
"created": "Mon, 17 Feb 2025 14:28:21 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Jul 2025 20:49:00 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Winkler",
"Nils O.",
"",
"Heidelberg University"
],
[
"Drozdowski",
"Oliver M.",
"",
"Heidelberg University"
],
[
"Ziebert",
"Falko",
"",
"Heidelberg University"
],
[
"Schwarz",
"Ulrich S.",
"",
"Heidelberg University"
]
] |
2502.12496
|
Frances Kuo
|
Alexander Keller, Frances Y. Kuo, Dirk Nuyens, and Ian H. Sloan
|
Regularity and tailored regularization of Deep Neural Networks, with application to parametric PDEs in uncertainty quantification
| null | null | null | null |
math.NA cs.NA
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In this paper we consider Deep Neural Networks (DNNs) with a smooth activation function as surrogates for high-dimensional functions that are somewhat smooth but costly to evaluate. We consider the standard (non-periodic) DNNs as well as propose a new model of periodic DNNs which are especially suited for a class of periodic target functions when Quasi-Monte Carlo lattice points are used as training points. We study the regularity of DNNs by obtaining explicit bounds on all mixed derivatives with respect to the input parameters. The bounds depend on the neural network parameters as well as the choice of activation function. By imposing restrictions on the network parameters to match the regularity features of the target functions, we prove that DNNs with $N$ tailor-constructed lattice training points can achieve the generalization error (or $L_2$ approximation error) bound ${\tt tol} + \mathcal{O}(N^{-r/2})$, where ${\tt tol}\in (0,1)$ is the tolerance achieved by the training error in practice, and $r = 1/p^*$, with $p^*$ being the ``summability exponent'' of a sequence that characterises the decay of the input variables in the target functions, and with the implied constant independent of the dimensionality of the input data. We apply our analysis to popular models of parametric elliptic PDEs in uncertainty quantification. In our numerical experiments, we restrict the network parameters during training by adding tailored regularization terms, and we show that for an algebraic equation mimicking the parametric PDE problems the DNNs trained with tailored regularization perform significantly better.
|
[
{
"version": "v1",
"created": "Tue, 18 Feb 2025 03:21:53 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 15:28:20 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Keller",
"Alexander",
""
],
[
"Kuo",
"Frances Y.",
""
],
[
"Nuyens",
"Dirk",
""
],
[
"Sloan",
"Ian H.",
""
]
] |
2502.13212
|
Zheng-Yang Zhuang
|
Zheng-Yang Zhuang, Di Zhu, Zhigang Wu, and Zhongbo Yan
|
Cartesian Nodal Lines and Magnetic Kramers Weyl Nodes in Spin-Split Antiferromagnets
|
5 pages, 4 figures for main text; 9 pages, 2 figures for supplemental materials
| null | null | null |
cond-mat.mes-hall cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
When band degeneracy occurs in a spin-split band structure, it gives rise to divergent Berry curvature and distinctive topological boundary states, resulting in a variety of fascinating effects. We show that three-dimensional spin-split antiferromagnets, characterized by symmetry-constrained momentum-dependent spin splitting and zero net magnetization, can host two unique forms of symmetry-protected band degeneracy: Cartesian nodal lines in the absence of spin-orbit coupling, and magnetic Kramers Weyl nodes when spin-orbit coupling is present. Remarkably, these band degeneracies not only produce unique patterns of Berry-curvature distributions but also give rise to topological boundary states with unconventional spin textures. Furthermore, we find that these band degeneracies can lead to strong or even quantized anomalous Hall effects and quantized circular photogalvanic effects under appropriate conditions. Our study suggests that spin-split antiferromagnets provide a fertile ground for exploring unconventional topological phases.
|
[
{
"version": "v1",
"created": "Tue, 18 Feb 2025 19:00:04 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 11:18:38 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Zhuang",
"Zheng-Yang",
""
],
[
"Zhu",
"Di",
""
],
[
"Wu",
"Zhigang",
""
],
[
"Yan",
"Zhongbo",
""
]
] |
2502.13295
|
Alexander Bondarenko
|
Alexander Bondarenko, Denis Volk, Dmitrii Volkov, Jeffrey Ladish
|
Demonstrating specification gaming in reasoning models
|
Updated with o3 results, fixed fonts
| null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We demonstrate LLM agent specification gaming by instructing models to win against a chess engine. We find reasoning models like OpenAI o3 and DeepSeek R1 will often hack the benchmark by default, while language models like GPT-4o and Claude 3.5 Sonnet need to be told that normal play won't work to hack.
We improve upon prior work like (Hubinger et al., 2024; Meinke et al., 2024; Weij et al., 2024) by using realistic task prompts and avoiding excess nudging. Our results suggest reasoning models may resort to hacking to solve difficult problems, as observed in OpenAI (2024)'s o1 Docker escape during cyber capabilities testing.
|
[
{
"version": "v1",
"created": "Tue, 18 Feb 2025 21:32:24 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2025 13:42:18 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 11:15:11 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bondarenko",
"Alexander",
""
],
[
"Volk",
"Denis",
""
],
[
"Volkov",
"Dmitrii",
""
],
[
"Ladish",
"Jeffrey",
""
]
] |
2502.13788
|
Carlos Carneiro
|
Carlos R. Melo-Carneiro, Thomas E. Collett, Lindsay J. Oldham, Wolfgang J. R. Enzi, Cristina Furlanetto, Ana L. Chies-Santos, Tian Li
|
Unveiling a 36 Billion Solar Mass Black Hole at the Centre of the Cosmic Horseshoe Gravitational Lens
|
20 pages, 12 figures, 8 tables. Accepted for publication at MNRAS
|
Mon Not R Astron Soc (2025) 2853-2871
|
10.1093/mnras/staf1036
| null |
astro-ph.GA astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
Supermassive black holes (SMBHs) are found at the centre of every massive galaxy, with their masses tightly connected to their host galaxies through a co-evolution over cosmic time. For massive ellipticals, the SMBH mass ($M_\text{BH}$) strongly correlates with the central stellar velocity dispersion ($\sigma_e$), via the $M_\text{BH}-\sigma_e$ relation. However, SMBH mass measurements have traditionally relied on central stellar dynamics in nearby galaxies ($z < 0.1$), limiting our ability to explore the SMBHs across cosmic time. In this work, we present a self-consistent analysis combining 2D stellar dynamics and lens modelling of the Cosmic Horseshoe gravitational lens system ($z = 0.44$), one of the most massive galaxies ever observed. Using integral-field spectroscopic data from MUSE and high-resolution imaging from HST, we model the radial arc and stellar kinematics, constraining the galaxy's central mass distribution and SMBH mass. Bayesian model comparison yields a $5\sigma$ detection of an ultramassive black hole (UMBH) with $\log_{10}(M_\text{BH}/M_{\odot}) = 10.56^{+0.07}_{-0.08} \pm (0.12)^\text{sys}$, consistent across various systematic tests. Our findings place the Cosmic Horseshoe $\sim$$1.5\sigma$ above the $M_\text{BH}-\sigma_e$ relation, supporting an emerging trend observed in BGCs and other massive galaxies. This suggests a steeper $M_\text{BH}-\sigma_e$ relationship at the highest masses, potentially driven by a different co-evolution of SMBHs and their host galaxies. Future surveys will uncover more radial arcs, enabling the detection of SMBHs over a broader redshift and mass range. These discoveries will further refine our understanding of the $M_\text{BH}-\sigma_e$ relation and its evolution across cosmic time.
|
[
{
"version": "v1",
"created": "Wed, 19 Feb 2025 14:57:39 GMT"
},
{
"version": "v2",
"created": "Wed, 26 Feb 2025 19:06:21 GMT"
},
{
"version": "v3",
"created": "Fri, 11 Jul 2025 19:50:16 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Melo-Carneiro",
"Carlos R.",
""
],
[
"Collett",
"Thomas E.",
""
],
[
"Oldham",
"Lindsay J.",
""
],
[
"Enzi",
"Wolfgang J. R.",
""
],
[
"Furlanetto",
"Cristina",
""
],
[
"Chies-Santos",
"Ana L.",
""
],
[
"Li",
"Tian",
""
]
] |
2502.14029
|
Peter Sims
|
Peter H. Sims, Steven G. Murray, Judd D. Bowman, John P. Barrett, Rigel C. Cappallo, Colin J. Lonsdale, Nivedita Mahesh, Raul A. Monsalve, Alan E. E. Rogers, Titu Samson, and Akshatha K. Vydula
|
A general Bayesian model-validation framework based on null-test evidence ratios, with an example application to global 21-cm cosmology
|
20 pages, 5 figures; accepted for publication in MNRAS
|
Mon Not R Astron Soc (2025) 2262-2281
|
10.1093/mnras/staf1109
| null |
astro-ph.CO astro-ph.IM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Comparing composite models for multi-component observational data is a prevalent scientific challenge. When fitting composite models, there exists the potential for systematics from a poor fit of one model component to be absorbed by another, resulting in the composite model providing an accurate fit to the data in aggregate but yielding biased a posteriori estimates for individual components. We begin by defining a classification scheme for composite model comparison scenarios, identifying two categories: category I, where models with accurate and predictive components are separable through Bayesian comparison of the unvalidated composite models, and category II, where models with accurate and predictive components may not be separable due to interactions between components, leading to spurious detections or biased signal estimation. To address the limitations of category II model comparisons, we introduce the Bayesian Null Test Evidence Ratio-based (BaNTER) validation framework. Applying this classification scheme and BaNTER to a composite model comparison problem in 21-cm cosmology, where minor systematics from imperfect foreground modelling can bias global 21-cm signal recovery, we validate six composite models using mock data. We show that incorporating BaNTER alongside Bayes-factor-based comparison reliably ensures unbiased inferences of the signal of interest across both categories, positioning BaNTER as a valuable addition to Bayesian inference workflows with potential applications across diverse fields.
|
[
{
"version": "v1",
"created": "Wed, 19 Feb 2025 19:00:00 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Sims",
"Peter H.",
""
],
[
"Murray",
"Steven G.",
""
],
[
"Bowman",
"Judd D.",
""
],
[
"Barrett",
"John P.",
""
],
[
"Cappallo",
"Rigel C.",
""
],
[
"Lonsdale",
"Colin J.",
""
],
[
"Mahesh",
"Nivedita",
""
],
[
"Monsalve",
"Raul A.",
""
],
[
"Rogers",
"Alan E. E.",
""
],
[
"Samson",
"Titu",
""
],
[
"Vydula",
"Akshatha K.",
""
]
] |
2502.14131
|
Enoch Hyunwook Kang
|
Enoch H. Kang, Hema Yoganarasimhan, Lalit Jain
|
An Empirical Risk Minimization Approach for Offline Inverse RL and Dynamic Discrete Choice Model
| null | null | null | null |
cs.LG cs.AI econ.EM
|
http://creativecommons.org/licenses/by/4.0/
|
We study the problem of estimating Dynamic Discrete Choice (DDC) models, also known as offline Maximum Entropy-Regularized Inverse Reinforcement Learning (offline MaxEnt-IRL) in machine learning. The objective is to recover reward or $Q^*$ functions that govern agent behavior from offline behavior data. In this paper, we propose a globally convergent gradient-based method for solving these problems without the restrictive assumption of linearly parameterized rewards. The novelty of our approach lies in introducing the Empirical Risk Minimization (ERM) based IRL/DDC framework, which circumvents the need for explicit state transition probability estimation in the Bellman equation. Furthermore, our method is compatible with non-parametric estimation techniques such as neural networks. Therefore, the proposed method has the potential to be scaled to high-dimensional, infinite state spaces. A key theoretical insight underlying our approach is that the Bellman residual satisfies the Polyak-Lojasiewicz (PL) condition -- a property that, while weaker than strong convexity, is sufficient to ensure fast global convergence guarantees. Through a series of synthetic experiments, we demonstrate that our approach consistently outperforms benchmark methods and state-of-the-art alternatives.
|
[
{
"version": "v1",
"created": "Wed, 19 Feb 2025 22:22:20 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 21:42:24 GMT"
},
{
"version": "v3",
"created": "Tue, 06 May 2025 17:12:37 GMT"
},
{
"version": "v4",
"created": "Fri, 18 Jul 2025 01:39:40 GMT"
},
{
"version": "v5",
"created": "Wed, 27 Aug 2025 07:01:51 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Kang",
"Enoch H.",
""
],
[
"Yoganarasimhan",
"Hema",
""
],
[
"Jain",
"Lalit",
""
]
] |
2502.14822
|
Zhichao Xu
|
Zhichao Xu, Fengran Mo, Zhiqi Huang, Crystina Zhang, Puxuan Yu, Bei Wang, Jimmy Lin, Vivek Srikumar
|
A Survey of Model Architectures in Information Retrieval
| null | null | null | null |
cs.IR
|
http://creativecommons.org/licenses/by/4.0/
|
The period from 2019 to the present has represented one of the biggest paradigm shifts in information retrieval (IR) and natural language processing (NLP), culminating in the emergence of powerful large language models (LLMs) from 2022 onward. Methods leveraging pretrained encoder-only models (e.g., BERT) and LLMs have outperformed many previous approaches, particularly excelling in zero-shot scenarios and complex reasoning tasks. This work surveys the evolution of model architectures in IR, focusing on two key aspects: backbone models for feature extraction and end-to-end system architectures for relevance estimation. The review intentionally separates architectural considerations from training methodologies to provide a focused analysis of structural innovations in IR systems. We trace the development from traditional term-based methods to modern neural approaches, particularly highlighting the impact of transformer-based models and subsequent large language models (LLMs). We conclude with a forward-looking discussion of emerging challenges and future directions, including architectural optimizations for performance and scalability, handling of multimodal, multilingual data, and adaptation to novel application domains such as autonomous search agents that is beyond traditional search paradigms.
|
[
{
"version": "v1",
"created": "Thu, 20 Feb 2025 18:42:58 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 23:10:28 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Xu",
"Zhichao",
""
],
[
"Mo",
"Fengran",
""
],
[
"Huang",
"Zhiqi",
""
],
[
"Zhang",
"Crystina",
""
],
[
"Yu",
"Puxuan",
""
],
[
"Wang",
"Bei",
""
],
[
"Lin",
"Jimmy",
""
],
[
"Srikumar",
"Vivek",
""
]
] |
2502.14851
|
Sophie Hoyland
|
Sophie Hoyland, Hans A. Winther, Daniela Saadeh, Kazuya Koyama, Albert Izard
|
Fast Generation of Weak Lensing Maps in Modified Gravity with COLA
|
18 pages, 12 figures; Accepted for publication in MNRAS
|
Mon Not R Astron Soc (2025) 3167-3183
|
10.1093/mnras/staf1071
| null |
astro-ph.CO gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
Accurate predictions of weak lensing observables are essential for understanding the large-scale structure of the Universe and probing the nature of gravity. In this work, we present a lightcone implementation to generate maps of the weak lensing convergence field using the COmoving Lagrangian Acceleration (COLA) method. The lightcone is constructed in spherical shells from the source to the observer following an onion representation of the Universe. We validate the COLA-generated convergence maps in General Relativity by comparing five statistics to those of maps obtained with publically available high-resolution $N$-body simulations: the power spectrum, bispectrum, probability distribution function, peak counts and Minkowski functionals. The convergence power spectrum is accurate to within $5\%$ up to $\ell\sim500$ and to within $10\%$ up to $\ell\sim750$, confirming the accuracy of this method on both linear and non-linear scales. For the probability distribution function, peak counts and Minkowski functionals, we determine the map pixel resolution required for COLA to capture the statistical features of the $N$-body convergence maps. Our validation tests provide a baseline for the convergence map specifications at which we can trust COLA for each statistic considered. Using these map specifications, we extend our analyses to two representative theories of Modified Gravity, and demonstrate their imprints on the five convergence statistics considered. This work represents a step towards precise weak lensing predictions under both General Relativity and Modified Gravity with reduced computational cost, providing a robust framework to explore the nature of gravity using field-level inference.
|
[
{
"version": "v1",
"created": "Thu, 20 Feb 2025 18:57:06 GMT"
},
{
"version": "v2",
"created": "Tue, 01 Jul 2025 10:06:22 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hoyland",
"Sophie",
""
],
[
"Winther",
"Hans A.",
""
],
[
"Saadeh",
"Daniela",
""
],
[
"Koyama",
"Kazuya",
""
],
[
"Izard",
"Albert",
""
]
] |
2502.14959
|
Henry Whitehead
|
Henry Whitehead, Connar Rowan, Bence Kocsis
|
3D Adiabatic Simulations of Binary Black Hole Formation in AGN
|
23 pages, 18 figures, submitted to MNRAS
|
Mon Not R Astron Soc (2025) 1033-1055
|
10.1093/mnras/staf1271
| null |
astro-ph.HE astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
We investigate close encounters between initially unbound black holes (BHs) in the gaseous discs of active galactic nuclei (AGN), performing the first 3D non-isothermal hydrodynamical simulations of gas-assisted binary BH formation. We discuss a suite of 135 simulations, considering 9 AGN disc environments and 15 BH impact parameters. We find that the gas distribution within the Hill sphere about an isolated embedded BH is akin to a spherically symmetric star with a low-mass convective envelope and a BH core, with large convective currents driving strong outflows away from the midplane. We find that Coriolis force acting on the outflow results in winds, analogous to cyclones, that counter-rotate with respect to the midplane flow within the Hill sphere. We confirm the existence of strong thermal blasts due to minidisc collisions during BH close encounters, as predicted in our previous 2D studies. We document binary formation across a wide range of environments, finding formation likelihood is increased when the gas mass in the Hill sphere is large, allowing for easier binary formation in the outer AGN disc. We provide a comprehensive overview of the SMBH's role in binary formation, investigating how binary formation in intermediate density environments is biased towards certain binary orientations. We offer two models for predicting dissipation by gas during close encounters, as a function of the ambient Hill mass alone, or with the periapsis depth. We use these models to motivate a prescription for binary formation likelihood that can be readily applied to Monte-Carlo simulations of AGN evolution.
|
[
{
"version": "v1",
"created": "Thu, 20 Feb 2025 19:00:03 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Whitehead",
"Henry",
""
],
[
"Rowan",
"Connar",
""
],
[
"Kocsis",
"Bence",
""
]
] |
2502.15348
|
Zhong Wang
|
Yi Zhang, Fan Wei, Jingyi Li, Yan Wang, Yanyan Yu, Jianli Chen, Zipo Cai, Xinyu Liu, Wei Wang, Sensen Yao, Peng Wang, Zhong Wang
|
Constructing a Norm for Children's Scientific Drawing: Distribution Features Based on Semantic Similarity of Large Language Models
| null |
biology methods and protocols, 2025, 10(01)
|
10.1093/biomethods/bpaf062
| null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The use of children's drawings to examining their conceptual understanding has been proven to be an effective method, but there are two major problems with previous research: 1. The content of the drawings heavily relies on the task, and the ecological validity of the conclusions is low; 2. The interpretation of drawings relies too much on the subjective feelings of the researchers. To address this issue, this study uses the Large Language Model (LLM) to identify 1420 children's scientific drawings (covering 9 scientific themes/concepts), and uses the word2vec algorithm to calculate their semantic similarity. The study explores whether there are consistent drawing representations for children on the same theme, and attempts to establish a norm for children's scientific drawings, providing a baseline reference for follow-up children's drawing research. The results show that the representation of most drawings has consistency, manifested as most semantic similarity>0.8. At the same time, it was found that the consistency of the representation is independent of the accuracy (of LLM's recognition), indicating the existence of consistency bias. In the subsequent exploration of influencing factors, we used Kendall rank correlation coefficient to investigate the effects of "sample size", "abstract degree", and "focus points" on drawings, and used word frequency statistics to explore whether children represented abstract themes/concepts by reproducing what was taught in class. It was found that accuracy (of LLM's recognition) is the most sensitive indicator, and data such as sample size and semantic similarity are related to it; The consistency between classroom experiments and teaching purpose is also an important factor, many students focus more on the experiments themselves rather than what they explain.
|
[
{
"version": "v1",
"created": "Fri, 21 Feb 2025 10:02:15 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 06:28:48 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Zhang",
"Yi",
""
],
[
"Wei",
"Fan",
""
],
[
"Li",
"Jingyi",
""
],
[
"Wang",
"Yan",
""
],
[
"Yu",
"Yanyan",
""
],
[
"Chen",
"Jianli",
""
],
[
"Cai",
"Zipo",
""
],
[
"Liu",
"Xinyu",
""
],
[
"Wang",
"Wei",
""
],
[
"Yao",
"Sensen",
""
],
[
"Wang",
"Peng",
""
],
[
"Wang",
"Zhong",
""
]
] |
2502.15378
|
Yanyu Chen
|
Yi-Jun Chang, Yanyu Chen, Dipan Dey, Gopinath Mishra, Hung Thuan Nguyen, Bryce Sanchez
|
Optimal Distributed Replacement Paths
|
PODC 2025
| null | null | null |
cs.DS cs.DC
|
http://creativecommons.org/licenses/by/4.0/
|
We study the replacement paths problem in the $\mathsf{CONGEST}$ model of distributed computing. Given an $s$-$t$ shortest path $P$, the goal is to compute, for every edge $e$ in $P$, the shortest-path distance from $s$ to $t$ avoiding $e$. For unweighted directed graphs, we establish the tight randomized round complexity bound for this problem as $\widetilde{\Theta}(n^{2/3} + D)$ by showing matching upper and lower bounds. Our upper bound extends to $(1+\epsilon)$-approximation for weighted directed graphs. Our lower bound applies even to the second simple shortest path problem, which asks only for the smallest replacement path length. These results improve upon the very recent work of Manoharan and Ramachandran (SIROCCO 2024), who showed a lower bound of $\widetilde{\Omega}(n^{1/2} + D)$ and an upper bound of $\widetilde{O}(n^{2/3} + \sqrt{n h_{st}} + D)$, where $h_{st}$ is the number of hops in the given $s$-$t$ shortest path $P$.
|
[
{
"version": "v1",
"created": "Fri, 21 Feb 2025 11:03:29 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Feb 2025 17:30:35 GMT"
},
{
"version": "v3",
"created": "Tue, 06 May 2025 02:45:53 GMT"
},
{
"version": "v4",
"created": "Wed, 27 Aug 2025 11:11:50 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Chang",
"Yi-Jun",
""
],
[
"Chen",
"Yanyu",
""
],
[
"Dey",
"Dipan",
""
],
[
"Mishra",
"Gopinath",
""
],
[
"Nguyen",
"Hung Thuan",
""
],
[
"Sanchez",
"Bryce",
""
]
] |
2502.15525
|
Xiaoli Wang
|
Xiaoli Wang, Sipu Ruan, Xin Meng, Gregory Chirikjian
|
Enhanced Probabilistic Collision Detection for Motion Planning Under Sensing Uncertainty
| null | null | null | null |
cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Probabilistic collision detection (PCD) is essential in motion planning for robots operating in unstructured environments, where considering sensing uncertainty helps prevent damage. Existing PCD methods mainly used simplified geometric models and addressed only position estimation errors. This paper presents an enhanced PCD method with two key advancements: (a) using superquadrics for more accurate shape approximation and (b) accounting for both position and orientation estimation errors to improve robustness under sensing uncertainty. Our method first computes an enlarged surface for each object that encapsulates its observed rotated copies, thereby addressing the orientation estimation errors. Then, the collision probability under the position estimation errors is formulated as a chance-constraint problem that is solved with a tight upper bound. Both the two steps leverage the recently developed normal parameterization of superquadric surfaces. Results show that our PCD method is twice as close to the Monte-Carlo sampled baseline as the best existing PCD method and reduces path length by 30% and planning time by 37%, respectively. A Real2Sim2Real pipeline further validates the importance of considering orientation estimation errors, showing that the collision probability of executing the planned path in simulation is only 2%, compared to 9% and 29% when considering only position estimation errors or no errors at all.
|
[
{
"version": "v1",
"created": "Fri, 21 Feb 2025 15:27:29 GMT"
},
{
"version": "v2",
"created": "Mon, 19 May 2025 20:46:35 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 07:16:49 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Wang",
"Xiaoli",
""
],
[
"Ruan",
"Sipu",
""
],
[
"Meng",
"Xin",
""
],
[
"Chirikjian",
"Gregory",
""
]
] |
2502.17359
|
Alexander Hammerl
|
Alexander Hammerl, Ravi Seshadri, Thomas Kj{\ae}r Rasmussen, Otto Anker Nielsen
|
Travel Time Reliability in Stochastic Kinematic Flow Models
| null | null | null | null |
physics.soc-ph math-ph math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
This paper analyzes the time-dependent relationship between the mean and variance of travel time on a single corridor under rush hour like congestion patterns. To model this phenomenon, we apply the LWR ((Lighthill & Whitham, 1955), (Richards, 1956)) theory on a homogenous freeway with a discontinuous bottleneck at its downstream end, assuming a uni-modal demand profile with a stochastic peak. We establish conditions for typical counterclockwise hysteresis loops under these assumptions. It is demonstrated that shapes of the fundamental diagram which always produce a counterclockwise loop can be interpreted as an indication of aggressive driving behavior, while deviations may occur under defensive driving. This classification enables a detailed explanation of the qualitative physical mechanisms behind this pattern, as well as an analysis of the causes for quantitatively limited deviations. Some of the mathematical properties of the LWR model identified in our analysis have not yet been addressed in the literature and we critically examine the extent to which these reflect actual traffic flow behavior. Our considerations are supported by numerical experiments. The obtained results aim to improve the fundamental understanding of the physical causes of this hysteresis pattern and to facilitate its better estimation in traffic planning and control.
|
[
{
"version": "v1",
"created": "Mon, 24 Feb 2025 17:37:15 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 12:17:53 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hammerl",
"Alexander",
""
],
[
"Seshadri",
"Ravi",
""
],
[
"Rasmussen",
"Thomas Kjær",
""
],
[
"Nielsen",
"Otto Anker",
""
]
] |
2502.17401
|
Giulia Cusin
|
Giulia Cusin, Cyril Pitrou, Martin Pijnenburg and Alberto Sesana
|
Measuring anisotropies in the PTA band with cross-correlations
|
12 pages, 3 figures
|
Mon Not R Astron Soc (2025) 2884-2896
|
10.1093/mnras/staf1074
| null |
gr-qc astro-ph.CO astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
The astrophysical gravitational wave background in the nanohertz (nHz) band is expected to be primarily composed of the superposition of signals from binaries of supermassive black holes. The spatial discreteness of these sources introduces shot noise, which, in certain regimes, would overwhelm efforts to measure the anisotropy of the gravitational wave background. Moreover, the fact the time-residual map has a finite angular resolution and the presence of pulsar noise, affects our ability to construct the angular power spectrum of the anisotropy from a time-residual map (finite resolution noise). In this work, we explicitly demonstrate, starting from first principles, that cross-correlating a gravitational wave background map with a sufficiently dense galaxy survey can mitigate this issue. This approach could potentially reveal underlying properties of the gravitational wave background that would otherwise remain obscured. We quantify both the shot noise and the finite resolution noise level and show that cross-correlating the gravitational wave background with a galaxy catalog improves by more than one order of magnitude the prospects for a first detection of the background anisotropy by a gravitational wave observatory operating in the nHz frequency range. In particular, we find that with a futuristic scenario with an effective number of frequencies equal to $N_f=10$, the detection of the spectral amplitude can be achieved combining the first $20$ multipoles, with a threshold to resolve single events SKA-like. Increasing observation time, pulsar number or reducing the pulsar white noise considerably improves the detection significance.
|
[
{
"version": "v1",
"created": "Mon, 24 Feb 2025 18:29:34 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 15:16:16 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Cusin",
"Giulia",
""
],
[
"Pitrou",
"Cyril",
""
],
[
"Pijnenburg",
"Martin",
""
],
[
"Sesana",
"Alberto",
""
]
] |
2502.18079
|
Tomasz Bazylewicz
|
T. Bazylewicz, M. Szczepanik, J. K{\l}os, and J. K. Korbicz
|
Decoherence from the light bending interaction
| null |
Phys. Rev. D 112, 046016 2025
|
10.1103/qxdn-hw9y
| null |
quant-ph gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We analyse a decoherence effect, caused by the gravitational interaction between a massive body and the electromagnetic field. Assuming a quantum version of the light bending interaction, we show that it leads to decoherence of the mass if the light is not observed. Using the extreme weakness of the gravitational coupling, we derive explicitly the decoherence lengthscales for general states of the central mass and for both thermal and coherent light. Predictably, the effect is very faint for anything but hugely energetic light, however from the fundamental point of view of co-existence of both gravitation and quantum theories, it is there. Since effectively the studied system is a quantum optomechanical system, we hope our results, properly rescaled, will be also useful in optomechanics.
|
[
{
"version": "v1",
"created": "Tue, 25 Feb 2025 10:47:06 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 23:46:35 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bazylewicz",
"T.",
""
],
[
"Szczepanik",
"M.",
""
],
[
"Kłos",
"J.",
""
],
[
"Korbicz",
"J. K.",
""
]
] |
2502.19596
|
Nayoung Choi
|
Nayoung Choi, Grace Byun, Andrew Chung, Ellie S. Paek, Shinsun Lee, Jinho D. Choi
|
Reference-Aligned Retrieval-Augmented Question Answering over Heterogeneous Proprietary Documents
|
Accepted to CIKM 2025 Applied Research Track
| null | null | null |
cs.AI cs.IR
|
http://creativecommons.org/licenses/by/4.0/
|
Proprietary corporate documents contain rich domain-specific knowledge, but their overwhelming volume and disorganized structure make it difficult even for employees to access the right information when needed. For example, in the automotive industry, vehicle crash-collision tests, each costing hundreds of thousands of dollars, produce highly detailed documentation. However, retrieving relevant content during decision-making remains time-consuming due to the scale and complexity of the material. While Retrieval-Augmented Generation (RAG)-based Question Answering (QA) systems offer a promising solution, building an internal RAG-QA system poses several challenges: (1) handling heterogeneous multi-modal data sources, (2) preserving data confidentiality, and (3) enabling traceability between each piece of information in the generated answer and its original source document. To address these, we propose a RAG-QA framework for internal enterprise use, consisting of: (1) a data pipeline that converts raw multi-modal documents into a structured corpus and QA pairs, (2) a fully on-premise, privacy-preserving architecture, and (3) a lightweight reference matcher that links answer segments to supporting content. Applied to the automotive domain, our system improves factual correctness (+1.79, +1.94), informativeness (+1.33, +1.16), and helpfulness (+1.08, +1.67) over a non-RAG baseline, based on 1-5 scale ratings from both human and LLM judge.
|
[
{
"version": "v1",
"created": "Wed, 26 Feb 2025 22:20:08 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 20:00:15 GMT"
},
{
"version": "v3",
"created": "Mon, 16 Jun 2025 14:27:30 GMT"
},
{
"version": "v4",
"created": "Tue, 19 Aug 2025 19:27:41 GMT"
},
{
"version": "v5",
"created": "Wed, 27 Aug 2025 03:09:19 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Choi",
"Nayoung",
""
],
[
"Byun",
"Grace",
""
],
[
"Chung",
"Andrew",
""
],
[
"Paek",
"Ellie S.",
""
],
[
"Lee",
"Shinsun",
""
],
[
"Choi",
"Jinho D.",
""
]
] |
2502.20034
|
Hongseok Oh
|
Hongseok Oh, Wonseok Hwang
|
Do Vision Encoders Truly Explain Object Hallucination?: Mitigating Object Hallucination via Simple Fine-Grained CLIPScore
| null | null | null | null |
cs.CV cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Recently, Large Vision-Language Models (LVLMs) show remarkable performance across various domains. However, these models suffer from object hallucination. This study revisits the previous claim that the cause of such hallucinations lies in the limited representational capacity of the vision encoder. Our analysis implies that the capacity of the vision encoder is not necessarily a major limiting factor in detecting object hallucination. Based on this insight, we propose Fine-grained CLIPScore (F-CLIPScore), a simple yet effective evaluation metric that enhances object-level granularity by incorporating text embeddings at the noun level. Evaluations on the OHD-Caps benchmark show that F-CLIPScore significantly outperforms conventional CLIPScore in accuracy by a large margin of \textbf{39.6\%} without additional training. We further demonstrate that F-CLIPScore-based data filtering reduces object hallucination in LVLM (4.9\% in POPE).
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 12:20:02 GMT"
},
{
"version": "v2",
"created": "Sun, 18 May 2025 11:29:25 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 10:19:08 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Oh",
"Hongseok",
""
],
[
"Hwang",
"Wonseok",
""
]
] |
2502.20251
|
Hyungrok Kim
|
Alex S. Arvanitakis, Leron Borsten, Dimitri Kanakaris and Hyungrok Kim
|
Homotopy Manin Theories: Generalising Third-Way, Yang-Mills and Integrable Sigma Models
|
24 pages
|
J.Phys.A 58 (2025) 355401
|
10.1088/1751-8121/adfc17
| null |
hep-th math-ph math.MP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Manin theories are a class of non-topological deformations of Chern-Simons theories that naturally realise the third-way mechanism and furthermore admit localisation despite not being supersymmetric in the usual sense. In this paper, we extend this construction to higher dimensions, thereby producing a large class of examples of third-way-type theories. Furthermore, the construction naturally yields Yang-Baxter integrable deformations of the principal chiral model as well as gravitational models various dimensions.
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 16:36:12 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Arvanitakis",
"Alex S.",
""
],
[
"Borsten",
"Leron",
""
],
[
"Kanakaris",
"Dimitri",
""
],
[
"Kim",
"Hyungrok",
""
]
] |
2502.20336
|
Nikolaos Rekatsinas
|
Lewin Ernst, Nikolaos Rekatsinas, and Karsten Urban
|
A posteriori certification for neural networks approximations to PDEs
| null | null | null | null |
math.NA cs.NA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose rigorous lower and upper bounds for neural network (NN) approximations to PDEs by efficiently computing the Riesz representations of suitable extension and restrictions of the NN residual towards geometrically simpler domains, which are either embedded or enveloping the original domain. Error bounds are proven and detailed for elliptic as well as parabolic problems. Numerical experiments show the good quantitative behaviour of the derived upper and lower error bounds.
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 18:05:23 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 17:41:12 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Ernst",
"Lewin",
""
],
[
"Rekatsinas",
"Nikolaos",
""
],
[
"Urban",
"Karsten",
""
]
] |
2502.20478
|
Jun Hou
|
Jun Hou, Lucy Lu Wang
|
Explainable AI for Clinical Outcome Prediction: A Survey of Clinician Perceptions and Preferences
| null |
2025 AMIA Informatics Summit
| null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Explainable AI (XAI) techniques are necessary to help clinicians make sense of AI predictions and integrate predictions into their decision-making workflow. In this work, we conduct a survey study to understand clinician preference among different XAI techniques when they are used to interpret model predictions over text-based EHR data. We implement four XAI techniques (LIME, Attention-based span highlights, exemplar patient retrieval, and free-text rationales generated by LLMs) on an outcome prediction model that uses ICU admission notes to predict a patient's likelihood of experiencing in-hospital mortality. Using these XAI implementations, we design and conduct a survey study of 32 practicing clinicians, collecting their feedback and preferences on the four techniques. We synthesize our findings into a set of recommendations describing when each of the XAI techniques may be more appropriate, their potential limitations, as well as recommendations for improvement.
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 19:30:20 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hou",
"Jun",
""
],
[
"Wang",
"Lucy Lu",
""
]
] |
2502.20523
|
Arsen Herasymchuk
|
A. A. Herasymchuk, E. V. Gorbar, P. O. Sukhachov
|
Viscoelastic tensor and hydrodynamics of altermagnets
|
23 pages, 8 figures; published version
|
Phys. Rev. B 112, 075150 (2025)
|
10.1103/3vw7-vrkj
| null |
cond-mat.str-el cond-mat.mes-hall
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We calculate the viscoelasticity tensor for altermagnets and formulate the corresponding hydrodynamic equations. The anisotropy of altermagnetic Fermi surfaces allows for additional terms in the viscoelasticity tensor and is manifested in transport properties, including electron and spin flows in a channel and nonlocal responses. In the channel geometry, the altermagnetic spin splitting leads to nontrivial spin density and spin current. Like the electric current, the spin current acquires a Poiseuille profile for no-slip boundary conditions. In nonlocal responses, the altermagnetic anisotropy affects current streamlines and electric potential distributions in the viscous regime. Our results provide signatures of the hydrodynamic transport regime in altermagnets, potentially facilitating its experimental studies and discovery.
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 21:18:12 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Aug 2025 17:34:00 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 17:11:13 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Herasymchuk",
"A. A.",
""
],
[
"Gorbar",
"E. V.",
""
],
[
"Sukhachov",
"P. O.",
""
]
] |
2502.20586
|
Albert Tseng
|
Albert Tseng and Tao Yu and Youngsuk Park
|
Training LLMs with MXFP4
|
AISTATS 2025, block-scaled FP4 (MXFP4, NVFP4, etc.) training
| null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Low precision (LP) datatypes such as MXFP4 can accelerate matrix multiplications (GEMMs) and reduce training costs. However, directly using MXFP4 instead of BF16 during training significantly degrades model quality. In this work, we present the first near-lossless training recipe that uses MXFP4 GEMMs, which are $2\times$ faster than FP8 on supported hardware. Our key insight is to compute unbiased gradient estimates with stochastic rounding (SR), resulting in more accurate model updates. However, directly applying SR to MXFP4 can result in high variance from block-level outliers, harming convergence. To overcome this, we use the random Hadamard tranform to theoretically bound the variance of SR. We train GPT models up to 6.7B parameters and find that our method induces minimal degradation over mixed-precision BF16 training. Our recipe computes $>1/2$ the training FLOPs in MXFP4, enabling an estimated speedup of $>1.3\times$ over FP8 and $>1.7\times$ over BF16 during backpropagation.
|
[
{
"version": "v1",
"created": "Thu, 27 Feb 2025 23:01:31 GMT"
},
{
"version": "v2",
"created": "Tue, 04 Mar 2025 04:58:12 GMT"
},
{
"version": "v3",
"created": "Tue, 26 Aug 2025 23:08:09 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Tseng",
"Albert",
""
],
[
"Yu",
"Tao",
""
],
[
"Park",
"Youngsuk",
""
]
] |
2502.21119
|
Katherine Ormerod
|
Katherine Ormerod, Joris Witstok, Renske Smit, Anna de Graaff, Jakob M. Helton, Michael V. Maseda, Irene Shivaei, Andrew J. Bunker, Stefano Carniani, Francesco D'Eugenio, Rachana Bhatawdekar, Jacopo Chevallard, Marijn Franx, Nimisha Kumari, Roberto Maiolino, Pierluigi Rinaldi, Brant Robertson, Sandro Tacchella
|
Detection of the 2175{\AA} UV Bump at z>7: Evidence for Rapid Dust Evolution in a Merging Reionisation-Era Galaxy
|
Accepted for publication in MNRAS
|
Mon Not R Astron Soc (2025) 1136-1154
|
10.1093/mnras/staf1228
| null |
astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
Dust is a fundamental component of the interstellar medium (ISM) within galaxies, as dust grains are highly efficient absorbers of UV and optical photons. Accurately quantifying this obscuration is crucial for interpreting galaxy spectral energy distributions (SEDs). The extinction curves in the Milky Way (MW) and Large Magellanic Cloud (LMC) exhibit a strong feature known as the 2175A UV bump, most often attributed to small carbonaceous dust grains. This feature was recently detected in faint galaxies out to z=7.55 suggesting rapid formation channels. Here we report the detection of a strong UV bump in a luminous Lyman-break galaxy at z = 7.11235, GNWY-7379420231, through observations taken as part of the NIRSpec Wide GTO survey. We fit a dust attenuation curve that is consistent with the MW extinction curve within 1{\sigma}, in a galaxy just ~700 Myr after the Big Bang. From the integrated spectrum, we infer a young mass-weighted age (t* ~ 22-59 Myr) for this galaxy, however spatially resolved SED fitting unveils the presence of an older stellar population (t* ~ 252 Myr). Furthermore, morphological analysis provides evidence for a potential merger. The underlying older stellar population suggests the merging system could be pre-enriched, with the dust illuminated by a merger-induced starburst. Moreover, turbulence driven by stellar feedback in this bursty region may be driving PAH formation through top-down shattering. The presence of a UV bump in GNWY-7379420231 solidifies growing evidence for the rapid evolution of dust properties within the first billion years of cosmic time.
|
[
{
"version": "v1",
"created": "Fri, 28 Feb 2025 14:55:25 GMT"
},
{
"version": "v2",
"created": "Fri, 25 Jul 2025 14:41:10 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Ormerod",
"Katherine",
""
],
[
"Witstok",
"Joris",
""
],
[
"Smit",
"Renske",
""
],
[
"de Graaff",
"Anna",
""
],
[
"Helton",
"Jakob M.",
""
],
[
"Maseda",
"Michael V.",
""
],
[
"Shivaei",
"Irene",
""
],
[
"Bunker",
"Andrew J.",
""
],
[
"Carniani",
"Stefano",
""
],
[
"D'Eugenio",
"Francesco",
""
],
[
"Bhatawdekar",
"Rachana",
""
],
[
"Chevallard",
"Jacopo",
""
],
[
"Franx",
"Marijn",
""
],
[
"Kumari",
"Nimisha",
""
],
[
"Maiolino",
"Roberto",
""
],
[
"Rinaldi",
"Pierluigi",
""
],
[
"Robertson",
"Brant",
""
],
[
"Tacchella",
"Sandro",
""
]
] |
2503.00552
|
Zheng Zhao
|
Zheng Zhao, Attaphon Kaewsnod, Kai Xu, Nattapat Tagsinsit, Xuyang Liu, Ayut Limphirat, Yupeng Yan
|
Study of $1^{--}$ P wave charmoniumlike and bottomoniumlike tetraquark spectroscopy
| null | null | null | null |
hep-ph
|
http://creativecommons.org/publicdomain/zero/1.0/
|
The masses of $1^{--}$ P-wave charmonium-like and bottomonium-like tetraquark states are calculated in a constituent quark model (CQM) where the Cornell-like potential and Breit-Fermi interaction are employed. All model parameters were imported from previous work, and predetermined by studying the low-lying conventional S- and P-wave light, charmed, bottom, charmonium, and bottomonium meson mass spectra. The lowest $1^{--}$ tetraquark mass is predicted to be around 4.15 GeV. The decay widths of $1^{--}$ P-wave tetraquark states are calculated for possible two-body strong decay channels within the rearrangement mechanism, including $\omega \chi_{cJ}$, $\eta J/\psi$, and $\rho J/\psi$ for charmonium-like tetraquarks, and $\omega \chi_{bJ}$ for bottomonium-like tetraquarks. The tetraquark theoretical results are compared with the selected exotic states, also known as Y states, and tentative assignments are suggested. This study suggests that $\psi(4230)$, $\psi(4360)$, $\psi(4660)$, and $\Upsilon$(10753) may be P-wave tetraquark states and that multiple states might exist around 4.36 GeV.
|
[
{
"version": "v1",
"created": "Sat, 01 Mar 2025 16:38:04 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Mar 2025 15:21:21 GMT"
},
{
"version": "v3",
"created": "Wed, 27 Aug 2025 04:17:06 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Zhao",
"Zheng",
""
],
[
"Kaewsnod",
"Attaphon",
""
],
[
"Xu",
"Kai",
""
],
[
"Tagsinsit",
"Nattapat",
""
],
[
"Liu",
"Xuyang",
""
],
[
"Limphirat",
"Ayut",
""
],
[
"Yan",
"Yupeng",
""
]
] |
2503.01487
|
Weijia Wang
|
Simone Naldi (CNRS, XLIM), Mohab Safey El Din (PolSys), Adrien Taylor (SIERRA), Weijia Wang (PolSys)
|
Solving generic parametric linear matrix inequalities
| null |
ISSAC '25: International Symposium on Symbolic and Algebraic Computation, Jul 2025, Guanajuato, Mexico
| null | null |
cs.SC math.AG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We consider linear matrix inequalities (LMIs) $A = A_0 + x_1 A_1 + ... + x_n A_n \succeq 0$ with the $A_i$'s being $m \times m$ symmetric matrices, with entries in a ring $\mathcal{R}$. When $\mathcal{R} = \mathbb{R}$, the feasibility problem consists in deciding whether the $x_i$'s can be instantiated to obtain a positive semidefinite matrix. When $\mathcal{R} = \mathbb{Q}[y_1, ... , y_t]$, the problem asks for a formula on the parameters $y_1, ..., y_t$, which describes the values of the parameters for which the specialized LMI is feasible. This problem can be solved using general quantifier elimination algorithms, with a complexity that is exponential in n. In this work, we leverage the LMI structure of the problem to design an algorithm that computes a formula $\Phi$ describing a dense subset of the feasible region of parameters, under genericity assumptions. The complexity of this algorithm is exponential in n, m and t but becomes polynomial in $n$ when $m$ and $t$ are fixed. We apply the algorithm to a parametric sum-of-squares problem and to the convergence analyses of certain first-order optimization methods, which are both known to be equivalent to the feasibility of certain parametric LMIs, hence demonstrating its practical interest.
|
[
{
"version": "v1",
"created": "Mon, 03 Mar 2025 12:53:35 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 09:34:13 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Naldi",
"Simone",
"",
"CNRS, XLIM"
],
[
"Din",
"Mohab Safey El",
"",
"PolSys"
],
[
"Taylor",
"Adrien",
"",
"SIERRA"
],
[
"Wang",
"Weijia",
"",
"PolSys"
]
] |
2503.01510
|
Alexander Baranov
|
Alexander Baranov, Anna Palatkina, Yulia Makovka, Pavel Braslavski
|
KoWit-24: A Richly Annotated Dataset of Wordplay in News Headlines
|
Accepted to RANLP 2025
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present KoWit-24, a dataset with fine-grained annotation of wordplay in 2,700 Russian news headlines. KoWit-24 annotations include the presence of wordplay, its type, wordplay anchors, and words/phrases the wordplay refers to. Unlike the majority of existing humor collections of canned jokes, KoWit-24 provides wordplay contexts -- each headline is accompanied by the news lead and summary. The most common type of wordplay in the dataset is the transformation of collocations, idioms, and named entities -- the mechanism that has been underrepresented in previous humor datasets. Our experiments with five LLMs show that there is ample room for improvement in wordplay detection and interpretation tasks. The dataset and evaluation scripts are available at https://github.com/Humor-Research/KoWit-24
|
[
{
"version": "v1",
"created": "Mon, 03 Mar 2025 13:24:25 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 20:08:46 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Baranov",
"Alexander",
""
],
[
"Palatkina",
"Anna",
""
],
[
"Makovka",
"Yulia",
""
],
[
"Braslavski",
"Pavel",
""
]
] |
2503.02253
|
Abigale Snortland
|
Abigale Snortland, Katherine Van Ness, Jennifer A. Franck, Ari Athair, Owen Williams, Brian Polagye
|
Experimental identification of blade-level forces, torque, and pitching moment for cross-flow turbines
| null | null | null | null |
physics.flu-dyn
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Cross-flow turbine power is a net sum of power generation from rotating blades and power loss from rotating support structures. While the aggregate forces and torques at the turbine level are important for end use, these can inhibit a deeper understanding of fluid-structure interactions. Identification of blade-level forces and torques allows for specific investigations into how the fluid forcing on the blade drives rotation and can aid blade structural design. Here, we present a physics-based methodology for extracting blade-level forces and torques from experimental measurements at the axis of rotation of a cross-flow turbine, and demonstrate strong agreement with equivalent blade-only simulations. In doing so, we highlight the often-overlooked pitching moment, which offsets continuous increases in power generation from the tangential force and leads to net-zero power generation at freewheel.
|
[
{
"version": "v1",
"created": "Tue, 04 Mar 2025 03:58:19 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 16:47:49 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Snortland",
"Abigale",
""
],
[
"Van Ness",
"Katherine",
""
],
[
"Franck",
"Jennifer A.",
""
],
[
"Athair",
"Ari",
""
],
[
"Williams",
"Owen",
""
],
[
"Polagye",
"Brian",
""
]
] |
2503.02556
|
Bernardo Hummes Flores
|
Stephan Felber, Bernardo Hummes Flores, Hugo Rincon Galeana
|
A Sheaf-Theoretic Characterization of Tasks in Distributed Systems
| null | null | null | null |
cs.DC math.CT
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce a sheaf-theoretic characterization of task solvability in general distributed computing models, unifying distinct approaches to message-passing models. We establish cellular sheaves as a natural mathematical framework for analyzing the global consistency requirements of local computations. Our main contribution is a task sheaf construction that explicitly relates both the distributed system and task, in which terminating solutions are precisely its global sections. We prove that a task can only be solved by a system when such sections exist in a task sheaf obtained from an execution cut, the frontier in which processes have enough information to decide. Our characterization is model-independent, working under varying synchronicity, failures and message adversaries, as long as the model produces runs composed of global states of the system. Furthermore, we show that the cohomology of the task sheaf provides a linear algebraic description of the decision space of the processes, and encodes the obstructions to find solutions. This opens way to a computational approach to protocol synthesis, which we illustrated by deriving a protocol for approximate agreement. This work bridges distributed computing and sheaf theory, providing both theoretical foundations for analyzing task solvability and tools for protocol design leveraging computational topology.
|
[
{
"version": "v1",
"created": "Tue, 04 Mar 2025 12:30:49 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 12:46:30 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Felber",
"Stephan",
""
],
[
"Flores",
"Bernardo Hummes",
""
],
[
"Galeana",
"Hugo Rincon",
""
]
] |
2503.02564
|
Isabella Gierz-Pehla
|
Niklas Hofmann, Johannes Gradl, Leonard Weigl, Stiven Forti, Camilla Coletti, Isabella Gierz
|
Influence of excitation energy on microscopic quantum pathways for ultrafast charge transfer in van der Waals heterostructures
|
16 pages, 5 figures, 2 tables
| null | null | null |
cond-mat.mes-hall cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Efficient charge separation in van der Waals (vdW) heterostructures is crucial for optimizing light-harvesting and detection applications. However, precise control over the microscopic pathways governing ultrafast charge transfer remains an open challenge. These pathways are intrinsically linked to charge transfer states with strongly delocalized wave functions that appear at various momenta in the Brillouin zone. Here, we use time- and angle-resolved photoemission spectroscopy (trARPES) to investigate the possibility of steering carriers through specific charge transfer states in a prototypical WS2-graphene heterostructure. By selectively exciting electron-hole pairs at the K-point and close to the Q-point of WS2 with different pump photon energies, we find that charge separation is faster at higher excitation energies. We attribute this to distinct tunneling mechanisms dictated by the momentum where the initial excitation takes place. Our findings reveal that the microscopic pathways for ultrafast charge transfer in vdW heterostructures are highly sensitive to the excitation energy, an aspect that should be carefully considered when optimizing the performance of optoelectronic devices.
|
[
{
"version": "v1",
"created": "Tue, 04 Mar 2025 12:41:02 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 08:52:52 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Hofmann",
"Niklas",
""
],
[
"Gradl",
"Johannes",
""
],
[
"Weigl",
"Leonard",
""
],
[
"Forti",
"Stiven",
""
],
[
"Coletti",
"Camilla",
""
],
[
"Gierz",
"Isabella",
""
]
] |
2503.03007
|
Evan Scope Crafts
|
Evan Scope Crafts and Umberto Villa
|
Benchmarking Diffusion Annealing-Based Bayesian Inverse Problem Solvers
| null |
IEEE Open J. Signal Process. 6 (2025) 975-991
|
10.1109/OJSP.2025.3597867
| null |
math.OC cs.LG math.PR stat.ML
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In recent years, the ascendance of diffusion modeling as a state-of-the-art generative modeling approach has spurred significant interest in their use as priors in Bayesian inverse problems. However, it is unclear how to optimally integrate a diffusion model trained on the prior distribution with a given likelihood function to obtain posterior samples. While algorithms developed for this purpose can produce high-quality, diverse point estimates of the unknown parameters of interest, they are often tested on problems where the prior distribution is analytically unknown, making it difficult to assess their performance in providing rigorous uncertainty quantification. Motivated by this challenge, this work introduces three benchmark problems for evaluating the performance of diffusion model based samplers. The benchmark problems, which are inspired by problems in image inpainting, x-ray tomography, and phase retrieval, have a posterior density that is analytically known. In this setting, approximate ground-truth posterior samples can be obtained, enabling principled evaluation of the performance of posterior sampling algorithms. This work also introduces a general framework for diffusion model based posterior sampling, Bayesian Inverse Problem Solvers through Diffusion Annealing (BIPSDA). This framework unifies several recently proposed diffusion-model-based posterior sampling algorithms and contains novel algorithms that can be realized through flexible combinations of design choices. We tested the performance of a set of BIPSDA algorithms, including previously proposed state-of-the-art approaches, on the proposed benchmark problems. The results provide insight into the strengths and limitations of existing diffusion-model based posterior samplers, while the benchmark problems provide a testing ground for future algorithmic developments.
|
[
{
"version": "v1",
"created": "Tue, 04 Mar 2025 21:07:15 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 21:47:04 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Crafts",
"Evan Scope",
""
],
[
"Villa",
"Umberto",
""
]
] |
2503.03453
|
Patryk Rygiel
|
Patryk Rygiel, Julian Suk, Kak Khee Yeung, Christoph Brune, Jelmer M. Wolterink
|
Active Learning for Deep Learning-Based Hemodynamic Parameter Estimation
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Hemodynamic parameters such as pressure and wall shear stress play an important role in diagnosis, prognosis, and treatment planning in cardiovascular diseases. These parameters can be accurately computed using computational fluid dynamics (CFD), but CFD is computationally intensive. Hence, deep learning methods have been adopted as a surrogate to rapidly estimate CFD outcomes. A drawback of such data-driven models is the need for time-consuming reference CFD simulations for training. In this work, we introduce an active learning framework to reduce the number of CFD simulations required for the training of surrogate models, lowering the barriers to their deployment in new applications. We propose three distinct querying strategies to determine for which unlabeled samples CFD simulations should be obtained. These querying strategies are based on geometrical variance, ensemble uncertainty, and adherence to the physics governing fluid dynamics. We benchmark these methods on velocity field estimation in synthetic coronary artery bifurcations and find that they allow for substantial reductions in annotation cost. Notably, we find that our strategies reduce the number of samples required by up to 50% and make the trained models more robust to difficult cases. Our results show that active learning is a feasible strategy to increase the potential of deep learning-based CFD surrogates.
|
[
{
"version": "v1",
"created": "Wed, 05 Mar 2025 12:35:54 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 07:44:22 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Rygiel",
"Patryk",
""
],
[
"Suk",
"Julian",
""
],
[
"Yeung",
"Kak Khee",
""
],
[
"Brune",
"Christoph",
""
],
[
"Wolterink",
"Jelmer M.",
""
]
] |
2503.03949
|
Yi-Heng Tsai
|
Yi-Heng Tsai
|
Volumes in Calabi-Yau Complete Intersection of Products of Projective Space
| null | null | null | null |
math.AG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We prove that the birational automorphism group of a general Calabi-yau complete intersection $X$ given by ample divisors in $\mathbb{P}^{n_1}\times\cdots\times\mathbb{P}^{n_l}$ is always Lorentzain. Applying the Kawamata-Morrison cone theorem on such $X$, we compute $\operatorname{vol}_X(D+sA)$ for any divisor $D\in \partial\overline{\operatorname{Eff}}(X)$ and ample divisor $A$ when $s$ is small. We also provide examples of volumes of certain Cartier divisors that involve the digamma function.
|
[
{
"version": "v1",
"created": "Wed, 05 Mar 2025 22:46:05 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 19:36:54 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Tsai",
"Yi-Heng",
""
]
] |
2503.04126
|
Joshua Bird
|
Joshua Bird, Jan Blumenkamp, and Amanda Prorok
|
DVM-SLAM: Decentralized Visual Monocular Simultaneous Localization and Mapping for Multi-Agent Systems
|
Accepted to 2025 IEEE International Conference on Robotics and Automation, pp. 15814-15820
| null | null | null |
cs.RO cs.CV cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
Cooperative Simultaneous Localization and Mapping (C-SLAM) enables multiple agents to work together in mapping unknown environments while simultaneously estimating their own positions. This approach enhances robustness, scalability, and accuracy by sharing information between agents, reducing drift, and enabling collective exploration of larger areas. In this paper, we present Decentralized Visual Monocular SLAM (DVM-SLAM), the first open-source decentralized monocular C-SLAM system. By only utilizing low-cost and light-weight monocular vision sensors, our system is well suited for small robots and micro aerial vehicles (MAVs). DVM-SLAM's real-world applicability is validated on physical robots with a custom collision avoidance framework, showcasing its potential in real-time multi-agent autonomous navigation scenarios. We also demonstrate comparable accuracy to state-of-the-art centralized monocular C-SLAM systems. We open-source our code and provide supplementary material online.
|
[
{
"version": "v1",
"created": "Thu, 06 Mar 2025 06:10:21 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 18:42:54 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Bird",
"Joshua",
""
],
[
"Blumenkamp",
"Jan",
""
],
[
"Prorok",
"Amanda",
""
]
] |
2503.04836
|
Yanfei Li
|
Yanfei Li, Teng Yin, Wenyi Shang, Jingyu Liu, Xi Wang, Kaiyang Zhao
|
PGAD: Prototype-Guided Adaptive Distillation for Multi-Modal Learning in AD Diagnosis
| null | null | null | null |
eess.IV cs.AI cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Missing modalities pose a major issue in Alzheimer's Disease (AD) diagnosis, as many subjects lack full imaging data due to cost and clinical constraints. While multi-modal learning leverages complementary information, most existing methods train only on complete data, ignoring the large proportion of incomplete samples in real-world datasets like ADNI. This reduces the effective training set and limits the full use of valuable medical data. While some methods incorporate incomplete samples, they fail to effectively address inter-modal feature alignment and knowledge transfer challenges under high missing rates. To address this, we propose a Prototype-Guided Adaptive Distillation (PGAD) framework that directly incorporates incomplete multi-modal data into training. PGAD enhances missing modality representations through prototype matching and balances learning with a dynamic sampling strategy. We validate PGAD on the ADNI dataset with varying missing rates (20%, 50%, and 70%) and demonstrate that it significantly outperforms state-of-the-art approaches. Ablation studies confirm the effectiveness of prototype matching and adaptive sampling, highlighting the potential of our framework for robust and scalable AD diagnosis in real-world clinical settings.
|
[
{
"version": "v1",
"created": "Wed, 05 Mar 2025 14:39:31 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 02:10:39 GMT"
}
] | 2025-08-28T00:00:00 |
[
[
"Li",
"Yanfei",
""
],
[
"Yin",
"Teng",
""
],
[
"Shang",
"Wenyi",
""
],
[
"Liu",
"Jingyu",
""
],
[
"Wang",
"Xi",
""
],
[
"Zhao",
"Kaiyang",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.