id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.08237 | Bailey Miller | Bailey Miller, Rohan Sawhney, Keenan Crane, Ioannis Gkioulekas | Solving partial differential equations in participating media | SIGGRAPH 2025. Project page https://imaging.cs.cmu.edu/volumetric_walk_on_spheres | null | 10.1145/3731152 | null | cs.GR cs.NA math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of solving partial differential equations (PDEs) in domains with complex microparticle geometry that is impractical, or intractable, to model explicitly. Drawing inspiration from volume rendering, we propose tackling this problem by treating the domain as a participating medium that models microparticle geometry stochastically, through aggregate statistical properties (e.g., particle density). We first introduce the problem setting of PDE simulation in participating media. We then specialize to exponential media and describe the properties that make them an attractive model of microparticle geometry for PDE simulation problems. We use these properties to develop two new algorithms, volumetric walk on spheres and volumetric walk on stars, that generalize previous Monte Carlo algorithms to enable efficient and discretization-free simulation of linear elliptic PDEs (e.g., Laplace) in participating media. We demonstrate experimentally that our algorithms can solve Laplace boundary value problems with complex microparticle geometry more accurately and more efficiently than previous approaches, such as ensemble averaging and homogenization. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:12:22 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Miller",
"Bailey",
""
],
[
"Sawhney",
"Rohan",
""
],
[
"Crane",
"Keenan",
""
],
[
"Gkioulekas",
"Ioannis",
""
]
] |
2506.08238 | Elli Anastasiadi | Parosh Abdulla, Elli Anastasiadi, Mohamed Faouzi Atig, Samuel Grahn | Verification of the Release-Acquire Semantics | null | null | null | null | cs.PL | http://creativecommons.org/licenses/by/4.0/ | The Release-Acquire (RA) semantics and its variants are some of the most fundamental models of concurrent semantics for architectures, programming languages, and distributed systems. Several steps have been taken in the direction of testing such semantics, where one is interested in whether a single program execution is consistent with a memory model. The more general verification problem, i.e., checking whether all allowed program runs are consistent with a memory model, has still not been studied as much. The purpose of this work is to bridge this gap. We tackle the verification problem, where, given an implementation described as a register machine, we check if any of its runs violates the RA semantics or its Strong (SRA) and Weak (WRA) variants. We show that verifying WRA in this setup is in O([)n5 ], while verifying the RA and SRA is in both NP- and coNP-hard, and provide a PSPACE upper bound. This both answers some fundamental questions about the complexity of these problems, but also provides insights on the expressive power of register machines as a model. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:15:18 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Abdulla",
"Parosh",
""
],
[
"Anastasiadi",
"Elli",
""
],
[
"Atig",
"Mohamed Faouzi",
""
],
[
"Grahn",
"Samuel",
""
]
] |
2506.08239 | Jiawang Li | Jiawang Li | Low-Cost Wideband Tilted Beam Antenna for Millimeter-wave Vehicle Applications | null | null | null | null | eess.SP | http://creativecommons.org/licenses/by/4.0/ | To facilitate vehicle coverage for millimeter-wave applications, this communication presents a low-cost, wideband tilted-beam antenna. A novel design is proposed in which a slot antenna is both directly excited and electromagnetically coupled to a monopole array. This slot-monopole configuration is inherently robust against substrate losses, enabling low-cost fabrication while maintaining high realized gain and compact size. Furthermore, the slot-fed structure effectively excites multiple resonant modes within the monopole array, resulting in a significantly enhanced bandwidth. Experimental results demonstrate that the antenna achieves a -10-dB impedance bandwidth of over 76.5% (20-44.78 GHz) and a peak realized gain of 6.1 dBi. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:17:22 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Li",
"Jiawang",
""
]
] |
2506.08240 | Dongkyu Cho | Dongkyu Cho, Rumi Chunara | Dealing with the Evil Twins: Improving Random Augmentation by Addressing Catastrophic Forgetting of Diverse Augmentations | 12 pages, 6 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data augmentation is a promising tool for enhancing out-of-distribution generalization, where the key is to produce diverse, challenging variations of the source domain via costly targeted augmentations that maximize its generalization effect. Conversely, random augmentation is inexpensive but is deemed suboptimal due to its limited effect. In this paper, we revisit random augmentation and explore methods to address its shortcomings. We show that the stochastic nature of random augmentation can produce a set of colliding augmentations that distorts the learned features, similar to catastrophic forgetting. We propose a simple solution that improves the generalization effect of random augmentation by addressing forgetting, which displays strong generalization performance across various single source domain generalization (sDG) benchmarks. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:18:13 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Cho",
"Dongkyu",
""
],
[
"Chunara",
"Rumi",
""
]
] |
2506.08241 | Christopher Danner | Christopher A. Danner, Eric L. Sandquist, Gail H. Schaefer, Luigi R. Bedin, Christopher D. Farrington, Cyprien Lanthermann, Stefan Kraus, Robert Klement, Narsireddy Anugu, John D. Monnier, Jerome A. Orosz, Isabelle Codron, Tyler Gardner, Mayra Gutierrez, Benjamin R. Setterholm, Jeremy Jones, Becky Flores | Precise Age For The Binary HD 21278 In The Young Alpha Persei Cluster | 31 pages, 19 figures, Accepted for publication in the Astrophysical Journal | null | null | null | astro-ph.SR astro-ph.GA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present a study of the double-lined spectroscopic binary HD 21278 that contains one of the brightest main sequence stars in the young $\alpha$ Persei open cluster. We analyzed new spectra and reanalyzed archived spectra to measure precise new radial velocity curves for the binary. We also obtained interferometric data using the CHARA Array at Mount Wilson to measure the sky positions of the two stars and the inclination of the $\sim$ 2 milliarcsecond orbit. We determine that the two stars have masses of $5.381 \pm 0.084 M_{\odot}$ and $3.353 \pm 0.064 M_{\odot}$. From isochrone fits, we find the cluster's age to be $49 \pm 7$ Myr (using PARSEC models) or $49.5 \pm 6$ Myr (MIST models). Finally, we revisit the massive white dwarfs that are candidate escapees from the $\alpha$ Persei cluster to try to better characterize the massive end of the white dwarf initial-final mass relation. The implied progenitor masses challenge the idea that Chandrasekhar-mass white dwarfs are made by single stars with masses near $8 \msun$. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:19:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Danner",
"Christopher A.",
""
],
[
"Sandquist",
"Eric L.",
""
],
[
"Schaefer",
"Gail H.",
""
],
[
"Bedin",
"Luigi R.",
""
],
[
"Farrington",
"Christopher D.",
""
],
[
"Lanthermann",
"Cyprien",
""
],
[
"Kraus",
"Stefan",
""
],
[
"Klement",
"Robert",
""
],
[
"Anugu",
"Narsireddy",
""
],
[
"Monnier",
"John D.",
""
],
[
"Orosz",
"Jerome A.",
""
],
[
"Codron",
"Isabelle",
""
],
[
"Gardner",
"Tyler",
""
],
[
"Gutierrez",
"Mayra",
""
],
[
"Setterholm",
"Benjamin R.",
""
],
[
"Jones",
"Jeremy",
""
],
[
"Flores",
"Becky",
""
]
] |
2506.08243 | Zhenjiang Mao | Zhenjiang Mao, Artem Bisliouk, Rohith Reddy Nama and Ivan Ruchkin | Temporalizing Confidence: Evaluation of Chain-of-Thought Reasoning with Signal Temporal Logic | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have shown impressive performance in mathematical reasoning tasks when guided by Chain-of-Thought (CoT) prompting. However, they tend to produce highly confident yet incorrect outputs, which poses significant risks in domains like education, where users may lack the expertise to assess reasoning steps. To address this, we propose a structured framework that models stepwise confidence as a temporal signal and evaluates it using Signal Temporal Logic (STL). In particular, we define formal STL-based constraints to capture desirable temporal properties and compute robustness scores that serve as structured, interpretable confidence estimates. Our approach also introduces a set of uncertainty reshaping strategies to enforce smoothness, monotonicity, and causal consistency across the reasoning trajectory. Experiments show that our approach consistently improves calibration metrics and provides more reliable uncertainty estimates than conventional confidence aggregation and post-hoc calibration. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:21:12 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Mao",
"Zhenjiang",
""
],
[
"Bisliouk",
"Artem",
""
],
[
"Nama",
"Rohith Reddy",
""
],
[
"Ruchkin",
"Ivan",
""
]
] |
2506.08244 | Riccardo Ali | Riccardo Ali, Pietro Li\`o, Jamie Vicary | Parameter-free approximate equivariance for tasks with finite group symmetry | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Equivariant neural networks incorporate symmetries through group actions, embedding them as an inductive bias to improve performance on a wide variety of tasks. However, existing equivariant methods can be computationally intensive, with high parameter counts, and are often tied to a specific architecture. We propose a simple zero-parameter approach that imposes approximate equivariance for a finite group in the latent representation, as an additional term in the loss function. We conduct experiments which allow the network to learn a group representation on the latent space, and show in every case it prefers to learn the regular representation. Fixing this action on the latent space, this yields a simple method to impose approximate equivariance as an additional loss penalty. We benchmark our approach on three datasets and compare it against several existing equivariant methods, showing that in many cases it achieves similar or better performance for a fraction of the parameters. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:23:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ali",
"Riccardo",
""
],
[
"Liò",
"Pietro",
""
],
[
"Vicary",
"Jamie",
""
]
] |
2506.08245 | Jorge Zuniga | Jorge Zuniga | Fast Ramanujan-type Series for Logarithms. Part I | 18 pages. PARI GP program is embedded in accompanying TeX file which must be downloaded, code extracted and saved as a GP script file | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This report introduces new series and variations of some hypergeometric type identities for fast computing of logarithms $\log\,p$ for small positive integers $p$. These series were found using Wilf Zeilberger (WZ) method and/or integer detection algorithms (LLL) providing highly efficient linearly convergent rational approximants for these constants. Some of the new identities are of $_4F_3$ type, but higher ones are found as well and hypergeometric series for log p, with variable p, have been derived. Found identities are proven by I. classical Beta Integral methods, II. some hypergometric closed forms and III. rational certificates from the WZ method. Since they are very fast, these series are particularly suitable to be embodied in mathematical software being implemented in binary splitting form which produces very efficient algorithms. Over 10e12 decimal places have been obtained for some logarithms in reasonable time. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:24:07 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zuniga",
"Jorge",
""
]
] |
2506.08246 | Maru Sarazola | Andrew Gill and Maru Sarazola | Thomason's colimit theorem for the double category of elements | 14 pages + references | null | null | null | math.CT math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that, for any 2-category $C$ and 2-functor $F\colon C \to Cat$, the double category of elements $\iint_C F$ introduced by Grandis and Par\'e satisfies a version of Thomason's colimit theorem; that is, there is a weak homotopy equivalence $B hocolim F\simeq B(\iint_C F)$. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:26:12 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Gill",
"Andrew",
""
],
[
"Sarazola",
"Maru",
""
]
] |
2506.08249 | Ken Gu | Ken Gu, Zhihan Zhang, Kate Lin, Yuwei Zhang, Akshay Paruchuri, Hong Yu, Mehran Kazemi, Kumar Ayush, A. Ali Heydari, Maxwell A. Xu, Girish Narayanswamy, Yun Liu, Ming-Zher Poh, Yuzhe Yang, Mark Malhotra, Shwetak Patel, Hamid Palangi, Xuhai Xu, Daniel McDuff, Tim Althoff, and Xin Liu | RADAR: Benchmarking Language Models on Imperfect Tabular Data | null | null | null | null | cs.DB cs.CL | http://creativecommons.org/licenses/by/4.0/ | Language models (LMs) are increasingly being deployed to perform autonomous data analyses. However, their data awareness -- the ability to recognize, reason over, and appropriately handle data artifacts such as missing values, outliers, and logical inconsistencies -- remains underexplored. These artifacts are especially common in real-world tabular data and, if mishandled, can significantly compromise the validity of analytical conclusions. To address this gap, we present RADAR, a benchmark for systematically evaluating data-aware reasoning on tabular data. We develop a framework to simulate data artifacts via programmatic perturbations to enable targeted evaluation of model behavior. RADAR comprises 2980 table query pairs, grounded in real-world data spanning 9 domains and 5 data artifact types. In addition to evaluating artifact handling, RADAR systematically varies table size to study how reasoning performance holds when increasing table size. Our evaluation reveals that, despite decent performance on tables without data artifacts, frontier models degrade significantly when data artifacts are introduced, exposing critical gaps in their capacity for robust, data-aware analysis. Designed to be flexible and extensible, RADAR supports diverse perturbation types and controllable table sizes, offering a valuable resource for advancing tabular reasoning. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:32:47 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Gu",
"Ken",
""
],
[
"Zhang",
"Zhihan",
""
],
[
"Lin",
"Kate",
""
],
[
"Zhang",
"Yuwei",
""
],
[
"Paruchuri",
"Akshay",
""
],
[
"Yu",
"Hong",
""
],
[
"Kazemi",
"Mehran",
""
],
[
"Ayush",
"Kumar",
""
],
[
"Heydari",
"A. Ali",
""
],
[
"Xu",
"Maxwell A.",
""
],
[
"Narayanswamy",
"Girish",
""
],
[
"Liu",
"Yun",
""
],
[
"Poh",
"Ming-Zher",
""
],
[
"Yang",
"Yuzhe",
""
],
[
"Malhotra",
"Mark",
""
],
[
"Patel",
"Shwetak",
""
],
[
"Palangi",
"Hamid",
""
],
[
"Xu",
"Xuhai",
""
],
[
"McDuff",
"Daniel",
""
],
[
"Althoff",
"Tim",
""
],
[
"Liu",
"Xin",
""
]
] |
2506.08250 | Erica Hammerstein | Erica Hammerstein, S. Bradley Cenko, Igor Andreoni, Panos Charalampopoulos, Ryan Chornock, Raffaella Margutti, Brendan O'Connor, Steve Schulze, Jesper Sollerman, Sudhanshu Barway, Varun Bhalerao, Anupama G.C., Harsh Kumar, Ester Marini. Diego Paris, Daniel A. Perley, Andrea Rossi, Yuhan Yao | The Jetted Tidal Disruption Event AT2022cmc: Investigating Connections to the Optical Tidal Disruption Event Population and Spectral Subclasses Through Late-Time Follow-up | 20 pages, 10 figures, 6 tables; submitted to ApJ | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | AT2022cmc is the first on-axis jetted tidal disruption event (TDE) to be discovered at optical wavelengths. The optically bright nature of AT2022cmc presents an unprecedented opportunity to place this jetted TDE in the context of the larger optically selected thermal TDE population and explore potential connections to optical TDE subclasses, particularly the class of luminous TDEs that lack optical spectral features. In this work we present late-time optical observations of AT2022cmc, both imaging and spectroscopy, that extend the optical dataset to $\sim 160$ days from the first detection in the observed frame. The light curve clearly evolves from red to blue, which we interpret as a transition from a non-thermally dominated spectral energy distribution (SED) to thermally dominated SED. By accounting for the non-thermal emission evident in the optical SED at early times, we extract the properties of the thermal emission and compare to a sample of optically selected thermal TDEs. We find that the properties of AT2022cmc are consistent with previous correlations found for the evolution and properties of thermal TDEs, with the thermal properties of AT2022cmc aligning with the class of featureless and luminous TDEs. The confirmation of this similarity motivates the importance of prompt and multi-wavelength follow-up of featureless and luminous TDEs in order to further explore the connection they have with jetted TDEs. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:33:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Hammerstein",
"Erica",
""
],
[
"Cenko",
"S. Bradley",
""
],
[
"Andreoni",
"Igor",
""
],
[
"Charalampopoulos",
"Panos",
""
],
[
"Chornock",
"Ryan",
""
],
[
"Margutti",
"Raffaella",
""
],
[
"O'Connor",
"Brendan",
""
],
[
"Schulze",
"Steve",
""
],
[
"Sollerman",
"Jesper",
""
],
[
"Barway",
"Sudhanshu",
""
],
[
"Bhalerao",
"Varun",
""
],
[
"C.",
"Anupama G.",
""
],
[
"Kumar",
"Harsh",
""
],
[
"Paris",
"Ester Marini. Diego",
""
],
[
"Perley",
"Daniel A.",
""
],
[
"Rossi",
"Andrea",
""
],
[
"Yao",
"Yuhan",
""
]
] |
2506.08251 | Maicon Ribeiro Correa | Abimael F. D. Loula and Maicon R. Correa and Jo\~ao N. C. Guerreiro and Elson M. Toledo | On Finite Element Methods for Heterogeneous Elliptic Problems | arXiv admin note: text overlap with arXiv:2505.13924 | International Journal of Solids and Structures, 45:1 (2008), pp 6436-6450 | 10.1016/j.ijsolstr.2008.08.005 | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dealing with variational formulations of second order elliptic problems with discontinuous coefficients, we recall a single field minimization problem of an extended functional presented by Bevilacqua et al (1974), which we associate with the basic idea supporting discontinuous Galerkin finite element methods. We review residual based stabilized mixed methods applied to Darcy flow in homogeneous porous media and extend them to heterogeneous media with an interface of discontinuity. For smooth interfaces, the proposed formulations preserve the continuity of the flux and exactly imposes the constraint between the tangent components of Darcy velocity on the interface. Convergence studies for a heterogeneous and anisotropic porous medium confirm the same rates of convergence predicted for homogeneous problem with smooth solutions. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:39:48 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Loula",
"Abimael F. D.",
""
],
[
"Correa",
"Maicon R.",
""
],
[
"Guerreiro",
"João N. C.",
""
],
[
"Toledo",
"Elson M.",
""
]
] |
2506.08252 | Amisha Srivastava | Amisha Srivastava, Samit S. Miftah, Hyunmin Kim, Debjit Pal, Kanad Basu | PoSyn: Secure Power Side-Channel Aware Synthesis | null | null | null | null | cs.CR cs.AR | http://creativecommons.org/licenses/by/4.0/ | Power Side-Channel (PSC) attacks exploit power consumption patterns to extract sensitive information, posing risks to cryptographic operations crucial for secure systems. Traditional countermeasures, such as masking, face challenges including complex integration during synthesis, substantial area overhead, and susceptibility to optimization removal during logic synthesis. To address these issues, we introduce PoSyn, a novel logic synthesis framework designed to enhance cryptographic hardware resistance against PSC attacks. Our method centers on optimal bipartite mapping of vulnerable RTL components to standard cells from the technology library, aiming to minimize PSC leakage. By utilizing a cost function integrating critical characteristics from both the RTL design and the standard cell library, we strategically modify mapping criteria during RTL-to-netlist conversion without altering design functionality. Furthermore, we theoretically establish that PoSyn minimizes mutual information leakage, strengthening its security against PSC vulnerabilities. We evaluate PoSyn across various cryptographic hardware implementations, including AES, RSA, PRESENT, and post-quantum cryptographic algorithms such as Saber and CRYSTALS-Kyber, at technology nodes of 65nm, 45nm, and 15nm. Experimental results demonstrate a substantial reduction in success rates for Differential Power Analysis (DPA) and Correlation Power Analysis (CPA) attacks, achieving lows of 3% and 6%, respectively. TVLA analysis further confirms that synthesized netlists exhibit negligible leakage. Additionally, compared to conventional countermeasures like masking and shuffling, PoSyn significantly lowers attack success rates, achieving reductions of up to 72%, while simultaneously enhancing area efficiency by as much as 3.79 times. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:41:47 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Srivastava",
"Amisha",
""
],
[
"Miftah",
"Samit S.",
""
],
[
"Kim",
"Hyunmin",
""
],
[
"Pal",
"Debjit",
""
],
[
"Basu",
"Kanad",
""
]
] |
2506.08253 | Roberto Le\'on-Montiel | Jorge-Alberto Peralta-\'Angeles, Mingyuan Hong, Mario A. Quiroz-Ju\'arez, Omar S. Maga\~na-Loaiza, Roberto de J. Le\'on-Montiel | High-density three-dimensional holography using rapid modulation of light | 7 pages, 3 figures | null | null | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the most common methods for reconstructing three-dimensional (3D) images of real objects is digital holography. This technique relies on the use of electro-optical devices that modify the phase or amplitude of light fields in a controlled manner, the so-called spatial light modulators. However, given that holography typically requires coherent light sources, a common problem with three-dimensional projection is the crosstalk between the layers that make up the 3D object. This limits full-depth control and directly affects image quality. Interestingly, in the past few years, several methods have proven to be effective in breaking layer crosstalk by erasing the spatial coherence of light. A drawback of such solutions is that, in many cases, additional optical resources are required to achieve such a task. In this work, we present a method for high-density reconstruction of three-dimensional objects using rapid modulation of light fields by means of digital micromirror devices (DMDs). The 3D reconstruction is performed by discretizing the object into multiplane light-point contours, where the resolution of the contours is controlled by the density of the light points. This allows us to achieve point separations, in the transverse plane, as small as 100 $\mu$m. The high refresh rate of the DMD ($\sim 10$ kHz) allows for a reconstruction where each point of the 3D image is spatially and temporally controlled by independent amplitude holograms, thus effectively eliminating coherence-induced multiplane crosstalk without the need for additional optical elements. Because of its simplicity and versatility, we believe that our method provides a practical route toward compact, high-resolution 3D holographic projectors. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:42:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Peralta-Ángeles",
"Jorge-Alberto",
""
],
[
"Hong",
"Mingyuan",
""
],
[
"Quiroz-Juárez",
"Mario A.",
""
],
[
"Magaña-Loaiza",
"Omar S.",
""
],
[
"León-Montiel",
"Roberto de J.",
""
]
] |
2506.08254 | Abdel Nasser Tawfik | A. Tawfik (Egyptian Ctr. Theor. Phys., Cairo and Future U. in Egypt), S. G. Elgendi (Islamic U. in Madinah), M. Hanafy (Benha U.) | Vacuum Homogeneous and Nonhomogeneous Metrics with Conventional and Quantized Metric Tensor: Singular or Nonsingular Solution | 35 pages, 7 figures, accepted for publication in Annals of Physics | null | null | ECTP-2024-12, WLCAPP-2024-12, FUE-2024-12 | gr-qc | http://creativecommons.org/licenses/by/4.0/ | To investigate whether the Universe underwent a singularity or maintained a nonsingular state, we carry out analytical and numerical analyses of the evolution of the Raychaudhuri equations in vacuum, alongside homogeneous and nonhomogeneous cosmic backgrounds. The results obtained from the Schwarzschild, Friedmann--Lemaitre--Robertson--Walker (FLRW), and Einstein--Gilbert--Straus (EGS) metrics are systematically compared. Analyzing the results from both, conventional and quantized metric tensor, it revealed insights into the nature of initial and spatial singularities. Results associated with the Schwarzschild metric demonstrate a positive evolution that corresponds with a reduction in radial distance (nonsingularity). In contrast, the proposed quantization reverses this trend, leading to a negative evolution (singularity). The situation is similar for the FLRW metric, where the suggested quantization results in a positive evolution as cosmic time decreases, in contrast to the classical and conventional metrics, which are associated with negative evolution. The analysis of the EGS metric reveals that classical evolution remains positively oriented, particularly with a reduction in radial distance. Moreover, the introduction of quantized and conventional metric tensors fully retrains the cosmic time dependence. The results obtained are a rightful recognition of the substantial efforts dedicated to the establishment of the Swiss-cheese model, demonstrating that the EGS metric indeed facilitates the temporal and spatial development of our Universe. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:42:51 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tawfik",
"A.",
"",
"Egyptian Ctr. Theor. Phys., Cairo and Future U. in Egypt"
],
[
"Elgendi",
"S. G.",
"",
"Islamic U. in Madinah"
],
[
"Hanafy",
"M.",
"",
"Benha U."
]
] |
2506.08255 | Patryk Krukowski | Patryk Krukowski,{\L}ukasz Gorczyca, Piotr Helm, Kamil Ksi\k{a}\.zek, Przemys{\l}aw Spurek | SHIELD: Secure Hypernetworks for Incremental Expansion Learning Defense | null | null | null | null | cs.LG cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | Traditional deep neural networks suffer from several limitations, including catastrophic forgetting. When models are adapted to new datasets, they tend to quickly forget previously learned knowledge. Another significant issue is the lack of robustness to even small perturbations in the input data. In practice, we can often easily perform adversarial attacks and change the network's predictions, adding minimal noise to the input. Dedicated architectures and training procedures can solve each of the above problems separately. Unfortunately, currently, no model can simultaneously address both catastrophic forgetting and vulnerability to adversarial attacks. We introduce SHIELD (Secure Hypernetworks for Incremental Expansion and Learning Defense), a novel approach that integrates a hypernetwork-based continual learning approach with interval arithmetic. SHIELD use the hypernetwork to transfer trainable task embedding vectors into the weights of a target model dedicated to specific data. This paradigm allows for the dynamic generation of separate networks for each subtask, while the hypernetwork aggregates and analyzes information across all tasks. The target model takes in the input a data sample with a defined interval range, and by creating a hypercube, produces a prediction for the given range. Therefore, such target models provide strict guarantees against all possible attacks for data samples within the interval range. Our approach enhances security without sacrificing network adaptability, addressing the overlooked challenge of safety in continual learning. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:43:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Krukowski",
"Patryk",
""
],
[
"Gorczyca",
"Łukasz",
""
],
[
"Helm",
"Piotr",
""
],
[
"Książek",
"Kamil",
""
],
[
"Spurek",
"Przemysław",
""
]
] |
2506.08256 | Victor Pambuccian | Hala King, Victor Pambuccian | Generalized Schatunowsky theorem in a weak arithmetic | null | null | null | null | math.LO | http://creativecommons.org/licenses/by/4.0/ | Schatunowsky's 1893 theorem, that 30 is the largest number all of whose totatives are primes, has been recently generalized by Kaneko and Nakai. In its generalized form, it states the finiteness of the set of all positive numbers $n$, which, for a fixed prime $p$, have the property that all of $n$'s totatives that are not divisible by any prime less than or equal to $p$ are prime numbers. It is this generalized form that we show holds in a weak arithmetic | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:44:09 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"King",
"Hala",
""
],
[
"Pambuccian",
"Victor",
""
]
] |
2506.08257 | Lukas Lao Beyer | L. Lao Beyer, T. Li, X. Chen, S. Karaman, K. He | Highly Compressed Tokenizer Can Generate Without Training | Main manuscript: 9 pages, 7 figures. Appendix: 8 pages, 9 figures. To appear in the Proceedings of the 42nd International Conference on Machine Learning | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Commonly used image tokenizers produce a 2D grid of spatially arranged tokens. In contrast, so-called 1D image tokenizers represent images as highly compressed one-dimensional sequences of as few as 32 discrete tokens. We find that the high degree of compression achieved by a 1D tokenizer with vector quantization enables image editing and generative capabilities through heuristic manipulation of tokens, demonstrating that even very crude manipulations -- such as copying and replacing tokens between latent representations of images -- enable fine-grained image editing by transferring appearance and semantic attributes. Motivated by the expressivity of the 1D tokenizer's latent space, we construct an image generation pipeline leveraging gradient-based test-time optimization of tokens with plug-and-play loss functions such as reconstruction or CLIP similarity. Our approach is demonstrated for inpainting and text-guided image editing use cases, and can generate diverse and realistic samples without requiring training of any generative model. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:45:03 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Beyer",
"L. Lao",
""
],
[
"Li",
"T.",
""
],
[
"Chen",
"X.",
""
],
[
"Karaman",
"S.",
""
],
[
"He",
"K.",
""
]
] |
2506.08258 | Lorenzo Arboit | Lorenzo Arboit, Dennis N. Schneider, Toby Collins, Daniel A. Hashimoto, Silvana Perretta, Bernard Dallemagne, Jacques Marescaux, EAES Working Group, Nicolas Padoy, Pietro Mascagni | Surgeons Awareness, Expectations, and Involvement with Artificial Intelligence: a Survey Pre and Post the GPT Era | 11 pages, 3 figures | null | null | null | cs.CY cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Artificial Intelligence (AI) is transforming medicine, with generative AI models like ChatGPT reshaping perceptions of its potential. This study examines surgeons' awareness, expectations, and involvement with AI in surgery through comparative surveys conducted in 2021 and 2024. Two cross-sectional surveys were distributed globally in 2021 and 2024, the first before an IRCAD webinar and the second during the annual EAES meeting. The surveys assessed demographics, AI awareness, expectations, involvement, and ethics (2024 only). The surveys collected a total of 671 responses from 98 countries, 522 in 2021 and 149 in 2024. Awareness of AI courses rose from 14.5% in 2021 to 44.6% in 2024, while course attendance increased from 12.9% to 23%. Despite this, familiarity with foundational AI concepts remained limited. Expectations for AI's role shifted in 2024, with hospital management gaining relevance. Ethical concerns gained prominence, with 87.2% of 2024 participants emphasizing accountability and transparency. Infrastructure limitations remained the primary obstacle to implementation. Interdisciplinary collaboration and structured training were identified as critical for successful AI adoption. Optimism about AI's transformative potential remained high, with 79.9% of respondents believing AI would positively impact surgery and 96.6% willing to integrate AI into their clinical practice. Surgeons' perceptions of AI are evolving, driven by the rise of generative AI and advancements in surgical data science. While enthusiasm for integration is strong, knowledge gaps and infrastructural challenges persist. Addressing these through education, ethical frameworks, and infrastructure development is essential. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:46:08 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Arboit",
"Lorenzo",
""
],
[
"Schneider",
"Dennis N.",
""
],
[
"Collins",
"Toby",
""
],
[
"Hashimoto",
"Daniel A.",
""
],
[
"Perretta",
"Silvana",
""
],
[
"Dallemagne",
"Bernard",
""
],
[
"Marescaux",
"Jacques",
""
],
[
"Group",
"EAES Working",
""
],
[
"Padoy",
"Nicolas",
""
],
[
"Mascagni",
"Pietro",
""
]
] |
2506.08259 | Andrew McCormack | Andrew McCormack | On the Existence of Unbiased Hypothesis Tests: An Algebraic Approach | 43 pages, 27 pages in the main text | null | null | null | math.ST stat.ME stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In hypothesis testing problems the property of strict unbiasedness describes whether a test is able to discriminate, in the sense of a difference in power, between any distribution in the null hypothesis space and any distribution in the alternative hypothesis space. In this work we examine conditions under which unbiased tests exist for discrete statistical models. It is shown that the existence of an unbiased test can be reduced to an algebraic criterion; an unbiased test exists if and only if there exists a polynomial that separates the null and alternative hypothesis sets. This places a strong, semialgebraic restriction on the classes of null hypotheses that have unbiased tests. The minimum degree of a separating polynomial coincides with the minimum sample size that is needed for an unbiased test to exist, termed the unbiasedness threshold. It is demonstrated that Gr\"obner basis techniques can be used to provide upper bounds for, and in many cases exactly find, the unbiasedness threshold. Existence questions for uniformly most powerful unbiased tests are also addressed, where it is shown that whether such a test exists can depend subtly on the specified level of the test and the sample size. Numerous examples, concerning tests in contingency tables, linear, log-linear, and mixture models are provided. All of the machinery developed in this work is constructive in the sense that when a test with a certain property is shown to exist it is possible to explicitly construct this test. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:46:11 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"McCormack",
"Andrew",
""
]
] |
2506.08260 | Wanjing Anya Ma | Wanjing Anya Ma, Michael Flor, Zuowei Wang | Automatic Generation of Inference Making Questions for Reading Comprehension Assessments | Accepted to the 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025), co-located with the ACL 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Inference making is an essential but complex skill in reading comprehension (RC). Some inferences require resolving references across sentences, and some rely on using prior knowledge to fill in the detail that is not explicitly written in the text. Diagnostic RC questions can help educators provide more effective and targeted reading instruction and interventions for school-age students. We introduce a taxonomy of inference types for RC and use it to analyze the distribution of items within a diagnostic RC item bank. Next, we present experiments using GPT-4o to generate bridging-inference RC items for given reading passages via few-shot prompting, comparing conditions with and without chain-of-thought prompts. Generated items were evaluated on three aspects: overall item quality, appropriate inference type, and LLM reasoning, achieving high inter-rater agreements above 0.90. Our results show that GPT-4o produced 93.8% good-quality questions suitable for operational use in grade 3-12 contexts; however, only 42.6% of the generated questions accurately matched the targeted inference type. We conclude that combining automatic item generation with human judgment offers a promising path toward scalable, high-quality diagnostic RC assessments. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:50:12 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ma",
"Wanjing Anya",
""
],
[
"Flor",
"Michael",
""
],
[
"Wang",
"Zuowei",
""
]
] |
2506.08261 | Sandeep Sen | Sandeep Sen | Towards universally optimal sorting algorithms | 8 pages | null | null | null | cs.DS cs.CC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We formalize a new paradigm for optimality of algorithms, that generalizes worst-case optimality based only on input-size to problem-dependent parameters including implicit ones. We re-visit some existing sorting algorithms from this perspective, and also present a novel measure of sortedness that leads to an optimal algorithm based on partition sort. This paradigm of measuring efficiency of algorithms looks promising for further interesting applications beyond the existing ones. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:55:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Sen",
"Sandeep",
""
]
] |
2506.08262 | Pavlo Mozharovskyi | Leonardo Leone, Pavlo Mozharovskyi, David Bounie | Massive parallelization of projection-based depths | null | null | null | null | stat.CO math.OC | http://creativecommons.org/licenses/by-sa/4.0/ | This article introduces a novel methodology for the massive parallelization of projection-based depths, addressing the computational challenges of data depth in high-dimensional spaces. We propose an algorithmic framework based on Refined Random Search (RRS) and demonstrate significant speedup (up to 7,000 times faster) on GPUs. Empirical results on synthetic data show improved precision and reduced runtime, making the method suitable for large-scale applications. The RRS algorithm (and other depth functions) are available in the Python-library data-depth (https://data-depth.github.io/) with ready-to-use tools to implement and to build upon this work. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:56:33 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Leone",
"Leonardo",
""
],
[
"Mozharovskyi",
"Pavlo",
""
],
[
"Bounie",
"David",
""
]
] |
2506.08263 | Pouya Agheli | Pouya Agheli, Tugce Kobal, Fran\c{c}ois Durand, Matthew Andrews | Learning-Based Multiuser Scheduling in MIMO-OFDM Systems with Hybrid Beamforming | To appear in the proceedings of the European Conference on Networks and Communications (EuCNC) & 6G Summit, 2025 | null | null | null | cs.IT cs.LG eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the multiuser scheduling problem in multiple-input multiple-output (MIMO) systems using orthogonal frequency division multiplexing (OFDM) and hybrid beamforming in which a base station (BS) communicates with multiple users over millimeter wave (mmWave) channels in the downlink. Improved scheduling is critical for enhancing spectral efficiency and the long-term performance of the system from the perspective of proportional fairness (PF) metric in hybrid beamforming systems due to its limited multiplexing gain. Our objective is to maximize PF by properly designing the analog and digital precoders within the hybrid beamforming and selecting the users subject to the number of radio frequency (RF) chains. Leveraging the characteristics of mmWave channels, we apply a two-timescale protocol. On a long timescale, we assign an analog beam to each user. Scheduling the users and designing the digital precoder are done accordingly on a short timescale. To conduct scheduling, we propose combinatorial solutions, such as greedy and sorting algorithms, followed by a machine learning (ML) approach. Our numerical results highlight the trade-off between the performance and complexity of the proposed approaches. Consequently, we show that the choice of approach depends on the specific criteria within a given scenario. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 21:59:05 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Agheli",
"Pouya",
""
],
[
"Kobal",
"Tugce",
""
],
[
"Durand",
"François",
""
],
[
"Andrews",
"Matthew",
""
]
] |
2506.08265 | Claudia Rella | Veronica Fantini and Claudia Rella | Modular resurgence, $q$-Pochhammer symbols, and quantum operators from mirror curves | 49 pages | null | null | null | hep-th math-ph math.AG math.CV math.MP math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Building on the results of [1,2], we study the resurgence of $q$-Pochhammer symbols and determine their summability and quantum modularity properties. We construct a new, infinite family of pairs of modular resurgent series from the asymptotic expansions of sums of $q$-Pochhammer symbols weighted by suitable Dirichlet characters. These weighted sums fit into the modular resurgence paradigm and provide further evidence supporting our conjectures in [1]. In the context of the topological string/spectral theory correspondence for toric Calabi-Yau threefolds, Kashaev and Mari\~no proved that the spectral traces of canonical quantum operators associated with local weighted projective planes can be expressed as sums of $q$-Pochhammer symbols. Exploiting this relation, we show that an exact strong-weak resurgent symmetry, first observed by the second author in [3] and fully formalized in [2] for local $\mathbb{P}^2$, applies to all local $\mathbb{P}^{m,n}$, albeit stripped of some of the underlying number-theoretic properties. Under some assumptions, these properties are restored when considering linear combinations of the spectral traces that reproduce the weighted sums above. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:00:13 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Fantini",
"Veronica",
""
],
[
"Rella",
"Claudia",
""
]
] |
2506.08266 | Yaswanth Chittepu | Yaswanth Chittepu, Blossom Metevier, Will Schwarzer, Austin Hoag, Scott Niekum, Philip S. Thomas | Reinforcement Learning from Human Feedback with High-Confidence Safety Constraints | 20 pages, 6 figures, 4 tables, Second Reinforcement Learning Conference (RLC 2025) | null | null | null | cs.LG cs.AI cs.CL stat.AP | http://creativecommons.org/licenses/by/4.0/ | Existing approaches to language model alignment often treat safety as a tradeoff against helpfulness, which can lead to unacceptable responses in sensitive domains. To ensure reliable performance in such settings, we propose High-Confidence Safe Reinforcement Learning from Human Feedback (HC-RLHF), a method that provides high-confidence safety guarantees while maximizing helpfulness. Similar to previous methods, HC-RLHF explicitly decouples human preferences into helpfulness and harmlessness (safety), which are learned by training a reward model and a cost model, respectively. It then employs a two-step process to find safe solutions. In the first step, it optimizes the reward function under an intentionally pessimistic version of the cost constraint. In the second step, the trained model undergoes a safety test to verify whether its performance stays within an upper-confidence bound of the actual cost constraint. We provide a theoretical analysis of HC-RLHF, including proof that it will not return an unsafe solution with a probability greater than a user-specified threshold. For our empirical analysis, we apply HC-RLHF to align three different language models (Qwen2-1.5B, Qwen2.5-3B, and LLaMa3.2-3B) with human preferences. Our results demonstrate that HC-RLHF produces safe models with high probability and can improve harmlessness and helpfulness compared to previous methods. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:03:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Chittepu",
"Yaswanth",
""
],
[
"Metevier",
"Blossom",
""
],
[
"Schwarzer",
"Will",
""
],
[
"Hoag",
"Austin",
""
],
[
"Niekum",
"Scott",
""
],
[
"Thomas",
"Philip S.",
""
]
] |
2506.08267 | Mansooreh Montazerin | Mansooreh Montazerin, Majd Al Aawar, Antonio Ortega, Ajitesh Srivastava | Sparse Interpretable Deep Learning with LIES Networks for Symbolic Regression | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Symbolic regression (SR) aims to discover closed-form mathematical expressions that accurately describe data, offering interpretability and analytical insight beyond standard black-box models. Existing SR methods often rely on population-based search or autoregressive modeling, which struggle with scalability and symbolic consistency. We introduce LIES (Logarithm, Identity, Exponential, Sine), a fixed neural network architecture with interpretable primitive activations that are optimized to model symbolic expressions. We develop a framework to extract compact formulae from LIES networks by training with an appropriate oversampling strategy and a tailored loss function to promote sparsity and to prevent gradient instability. After training, it applies additional pruning strategies to further simplify the learned expressions into compact formulae. Our experiments on SR benchmarks show that the LIES framework consistently produces sparse and accurate symbolic formulae outperforming all baselines. We also demonstrate the importance of each design component through ablation studies. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:05:53 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Montazerin",
"Mansooreh",
""
],
[
"Aawar",
"Majd Al",
""
],
[
"Ortega",
"Antonio",
""
],
[
"Srivastava",
"Ajitesh",
""
]
] |
2506.08268 | Devesh Nandal | Devesh Nandal, Ga\"el Buldgen, Daniel J. Whalen, John Regan, Tyrone E. Woods, Jonathan C. Tan | The Evolution Of Rotating Supermassive Pop III Stars On The Main Sequence | 10 pages, 5 figures | null | null | null | astro-ph.SR astro-ph.GA astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The detection of billion-solar-mass supermassive black holes (SMBHs) within the first billion years of cosmic history challenges conventional theories of black hole formation and growth. Simultaneously, recent JWST observations revealing exceptionally high nitrogen-to-oxygen abundance ratios in galaxies at high redshifts raise critical questions about rapid chemical enrichment mechanisms operating in the early universe. Supermassive stars (SMSs) with masses of 1000 to 10000 M$_{\odot}$ are promising candidates to explain these phenomena, but existing models have so far neglected the pivotal role of stellar rotation. Here, we present the first comprehensive evolutionary models of rotating Pop III SMSs computed using the GENEC stellar evolution code, including detailed treatments of rotation-induced chemical mixing, angular momentum transport, and mass loss driven by the $\Omega\Gamma$ limit. We demonstrate that rotation significantly enlarges the convective core and extends stellar lifetimes by up to 20%, with moderate enhancement of mass-loss rates as stars approach critical rotation thresholds. Our results further indicate that the cores of SMSs rotate relatively slowly (below $\sim 200$ km s$^{-1}$), resulting in dimensionless spin parameters $a* < 0.1$ for intermediate-mass black hole (IMBH) remnants that are notably lower than theoretical maximum spins. These findings highlight rotation as a key factor in determining the structural evolution, chemical yields, and black hole spin properties of SMSs, providing critical insights to interpret observational signatures from the high-redshift universe. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:08:11 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nandal",
"Devesh",
""
],
[
"Buldgen",
"Gaël",
""
],
[
"Whalen",
"Daniel J.",
""
],
[
"Regan",
"John",
""
],
[
"Woods",
"Tyrone E.",
""
],
[
"Tan",
"Jonathan C.",
""
]
] |
2506.08269 | Takeshi Fukuyama | Takeshi Fukuyama | Infrared Divergence in QED and the Fluctuation of Electromagnetic Fields | 6pages, 3 figures | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We reconsider the well known infrared divergence (IRD) in QED from the various graphs and relate with the electromagnetic fluctuation etc. based on the Fock space formalism. It may be useful for discriminating it from the alternative approach based on the dressed asymptotic space. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:17:40 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Fukuyama",
"Takeshi",
""
]
] |
2506.08271 | Joseph McDonald | Chun-Yu Lin, Joseph McDonald | Functional monadic ortholattices and locally finite $\sigma$-free polyadic ortholattices | null | null | null | null | math.LO | http://creativecommons.org/licenses/by/4.0/ | In this paper, we show that every monadic ortholattice is isomorphic to a functional one, thereby resolving a recent question posed by Harding. We then study certain substitution-free reducts of the polyadic ortholattices, which we call locally finite $\sigma$-free polyadic ortholattices, and provide an analogous functional representation result. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:24:05 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Lin",
"Chun-Yu",
""
],
[
"McDonald",
"Joseph",
""
]
] |
2506.08272 | Tarushri N. S. | Tarushri N. S. | Universal Differential Equations for Scientific Machine Learning of Node-Wise Battery Dynamics in Smart Grids | null | null | null | null | cs.LG eess.SP | http://creativecommons.org/licenses/by/4.0/ | Universal Differential Equations (UDEs), which blend neural networks with physical differential equations, have emerged as a powerful framework for scientific machine learning (SciML), enabling data-efficient, interpretable, and physically consistent modeling. In the context of smart grid systems, modeling node-wise battery dynamics remains a challenge due to the stochasticity of solar input and variability in household load profiles. Traditional approaches often struggle with generalization and fail to capture unmodeled residual dynamics. This work proposes a UDE-based approach to learn node-specific battery evolution by embedding a neural residual into a physically inspired battery ODE. Synthetic yet realistic solar generation and load demand data are used to simulate battery dynamics over time. The neural component learns to model unobserved or stochastic corrections arising from heterogeneity in node demand and environmental conditions. Comprehensive experiments reveal that the trained UDE aligns closely with ground truth battery trajectories, exhibits smooth convergence behavior, and maintains stability in long-term forecasts. These findings affirm the viability of UDE-based SciML approaches for battery modeling in decentralized energy networks and suggest broader implications for real-time control and optimization in renewable-integrated smart grids. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:24:55 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"S.",
"Tarushri N.",
""
]
] |
2506.08273 | Bart{\l}omiej Dyda | Bart{\l}omiej Dyda | Fractional and non-fractional Hardy inequality on~a~lattice~$\Z^d$ | 15 pages | null | null | null | math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present simple proofs of a~discrete fractional and non-fractional Hardy inequality, Our constants are explicit, but not optimal. In the class of power weights, we get a~complete picture of when the non-fractional Hardy inequality holds, for any dimension $d$ of the lattice $\Z^d$ and exponent $0<p<\infty$. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:28:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Dyda",
"Bartłomiej",
""
]
] |
2506.08275 | Olfa Draouil | Olfa Draouil, Rahma Yasmina Moulay Hachemi and Bernt {\O}ksendal | The stochastic heat inclusion with fractional time driven by time-space Brownian and L\'evy white noise | null | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We study a time-fractional stochastic heat inclusion driven by additive time-space Brownian and L\'evy white noise. The fractional time derivative is interpreted as the Caputo derivative of order $\alpha \in (0,2).$ We show the following: \\ a) If a solution exists, then it is a fixed point of a specific set-valued map.\\ b) Conversely, any fixed point of this map is a solution of the heat inclusion.\\ c) Finally, we show that there is at least one fixed point of this map, thereby proving that there is at least one solution of the time-fractional stochastic heat inclusion.
A solution $Y(t,x)$ is called \emph{mild} if $\E[Y^2(t,x)] < \infty$ for all $t,x$. We show that the solution is mild if\\ $\alpha=1$ \& $d=1,$ \ or \ $\alpha \geq 1$ \& $d\in \{1,2\}$.
On the other hand, if $\alpha < 1$ we show that the solution is not mild for any space dimension $d$. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:34:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Draouil",
"Olfa",
""
],
[
"Hachemi",
"Rahma Yasmina Moulay",
""
],
[
"Øksendal",
"Bernt",
""
]
] |
2506.08276 | Yichuan Wang | Yichuan Wang, Shu Liu, Zhifei Li, Yongji Wu, Ziming Mao, Yilong Zhao, Xiao Yan, Zhiying Xu, Yang Zhou, Ion Stoica, Sewon Min, Matei Zaharia, Joseph E. Gonzalez | LEANN: A Low-Storage Vector Index | null | null | null | null | cs.DB cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Embedding-based search is widely used in applications such as recommendation and retrieval-augmented generation (RAG). Recently, there is a growing demand to support these capabilities over personal data stored locally on devices. However, maintaining the necessary data structure associated with the embedding-based search is often infeasible due to its high storage overhead. For example, indexing 100 GB of raw data requires 150 to 700 GB of storage, making local deployment impractical. Reducing this overhead while maintaining search quality and latency becomes a critical challenge. In this paper, we present LEANN, a storage-efficient approximate nearest neighbor (ANN) search index optimized for resource-constrained personal devices. LEANN combines a compact graph-based structure with an efficient on-the-fly recomputation strategy to enable fast and accurate retrieval with minimal storage overhead. Our evaluation shows that LEANN reduces index size to under 5% of the original raw data, achieving up to 50 times smaller storage than standard indexes, while maintaining 90% top-3 recall in under 2 seconds on real-world question answering benchmarks. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:43:30 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Yichuan",
""
],
[
"Liu",
"Shu",
""
],
[
"Li",
"Zhifei",
""
],
[
"Wu",
"Yongji",
""
],
[
"Mao",
"Ziming",
""
],
[
"Zhao",
"Yilong",
""
],
[
"Yan",
"Xiao",
""
],
[
"Xu",
"Zhiying",
""
],
[
"Zhou",
"Yang",
""
],
[
"Stoica",
"Ion",
""
],
[
"Min",
"Sewon",
""
],
[
"Zaharia",
"Matei",
""
],
[
"Gonzalez",
"Joseph E.",
""
]
] |
2506.08277 | Subba Reddy Oota | Subba Reddy Oota, Khushbu Pahwa, Prachi Jindal, Satya Sai Srinath Namburi, Maneesh Singh, Tanmoy Chakraborty, Bapi S. Raju, Manish Gupta | Instruction-Tuned Video-Audio Models Elucidate Functional Specialization in the Brain | 39 pages, 22 figures | null | null | null | q-bio.NC cs.AI cs.CL cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recent voxel-wise multimodal brain encoding studies have shown that multimodal large language models (MLLMs) exhibit a higher degree of brain alignment compared to unimodal models in both unimodal and multimodal stimulus settings. More recently, instruction-tuned multimodal models have shown to generate task-specific representations that align strongly with brain activity. However, prior work evaluating the brain alignment of MLLMs has primarily focused on unimodal settings or relied on non-instruction-tuned multimodal models for multimodal stimuli. To address this gap, we investigated brain alignment, that is, measuring the degree of predictivity of neural activity recorded while participants were watching naturalistic movies (video along with audio) with representations derived from MLLMs. We utilized instruction-specific embeddings from six video and two audio instruction-tuned MLLMs. Experiments with 13 video task-specific instructions show that instruction-tuned video MLLMs significantly outperform non-instruction-tuned multimodal (by 15%) and unimodal models (by 20%). Our evaluation of MLLMs for both video and audio tasks using language-guided instructions shows clear disentanglement in task-specific representations from MLLMs, leading to precise differentiation of multimodal functional processing in the brain. We also find that MLLM layers align hierarchically with the brain, with early sensory areas showing strong alignment with early layers, while higher-level visual and language regions align more with middle to late layers. These findings provide clear evidence for the role of task-specific instructions in improving the alignment between brain activity and MLLMs, and open new avenues for mapping joint information processing in both the systems. We make the code publicly available [https://github.com/subbareddy248/mllm_videos]. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:48:36 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Oota",
"Subba Reddy",
""
],
[
"Pahwa",
"Khushbu",
""
],
[
"Jindal",
"Prachi",
""
],
[
"Namburi",
"Satya Sai Srinath",
""
],
[
"Singh",
"Maneesh",
""
],
[
"Chakraborty",
"Tanmoy",
""
],
[
"Raju",
"Bapi S.",
""
],
[
"Gupta",
"Manish",
""
]
] |
2506.08278 | Kyung-Eun Choi | Kyung-Eun Choi, Oleksiy V. Agapitov, Dae-Young Lee, Forrest Mozer, Jia Huang, Lucas Colomban, Jaye L. Verniero, Nour Raouafi | Switchbacks near Boundaries of Small-scale Magnetic Flux Ropes in the Young Solar Wind from Parker Solar Probe Observations | null | null | null | null | astro-ph.SR physics.space-ph | http://creativecommons.org/licenses/by/4.0/ | The Parker Solar Probe (PSP) mission has revealed frequent occurrences of switchbacks (SBs) and small-scale magnetic flux ropes (SMFRs) as prominent structures within the solar wind. These mesoscale features are observed across all heliocentric distances, with heightened activity in the young solar wind, such as successive SMFRs, blobs, and SBs using PSP in situ observations. One study, in particular, focuses on SMFRs observed during the intervals of PSP co-rotating with the Sun, which suggests a similar source of the observed solar wind. In this letter, we identified SBs at the boundaries of SMFRs as a regularly observed phenomenon and found instances where SBs and SMFRs co-occur, with the significance level $\alpha<0.05$. The SMFR-related SBs - observed at the leading and trailing edges of an SMFR - exhibit well-organized axial co-orientations, with their polarity flipping, meaning the radial direction remains constrained while the transversal field reverses. Furthermore, the axial field directions of SMFRs-related SBs appear to be more closely connected than to another SB that is spatially closer and are linked to the SMFR orientation. Our analysis of their relative geometry, which examines the alignment between SBs and the SMFR axis, reveals a distinct tendency emphasizing their correlation, further supporting the idea that the axes of SMFR-related SBs are presumably determined by the SMFR orientation. Observations suggest that a fraction of SBs is spatially and temporally associated with SMFRs, implying that processes related to SMFR boundaries may contribute to SB formation, or that SBs tend to develop in magnetic environments shaped by SMFRs. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:48:41 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Choi",
"Kyung-Eun",
""
],
[
"Agapitov",
"Oleksiy V.",
""
],
[
"Lee",
"Dae-Young",
""
],
[
"Mozer",
"Forrest",
""
],
[
"Huang",
"Jia",
""
],
[
"Colomban",
"Lucas",
""
],
[
"Verniero",
"Jaye L.",
""
],
[
"Raouafi",
"Nour",
""
]
] |
2506.08279 | Andrew Jaegle | Aditi Sundararaman and Amogh Adishesha and Andrew Jaegle and Dan Bigioi and Hyoung-Kyu Song and Jon Kyl and Justin Mao and Kevin Lan and Mojtaba Komeili and ShahRukh Athar and Sheila Babayan and Stanislau Beliasau and William Buchwalter | Seeing Voices: Generating A-Roll Video from Audio with Mirage | Technical report website: mirage.app/research/seeing-voices, product website: mirage.app | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | From professional filmmaking to user-generated content, creators and consumers have long recognized that the power of video depends on the harmonious integration of what we hear (the video's audio track) with what we see (the video's image sequence). Current approaches to video generation either ignore sound to focus on general-purpose but silent image sequence generation or address both visual and audio elements but focus on restricted application domains such as re-dubbing. We introduce Mirage, an audio-to-video foundation model that excels at generating realistic, expressive output imagery from scratch given an audio input. When integrated with existing methods for speech synthesis (text-to-speech, or TTS), Mirage results in compelling multimodal video. When trained on audio-video footage of people talking (A-roll) and conditioned on audio containing speech, Mirage generates video of people delivering a believable interpretation of the performance implicit in input audio. Our central technical contribution is a unified method for training self-attention-based audio-to-video generation models, either from scratch or given existing weights. This methodology allows Mirage to retain generality as an approach to audio-to-video generation while producing outputs of superior subjective quality to methods that incorporate audio-specific architectures or loss components specific to people, speech, or details of how images or audio are captured. We encourage readers to watch and listen to the results of Mirage for themselves (see paper and comments for links). | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:56:02 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Sundararaman",
"Aditi",
""
],
[
"Adishesha",
"Amogh",
""
],
[
"Jaegle",
"Andrew",
""
],
[
"Bigioi",
"Dan",
""
],
[
"Song",
"Hyoung-Kyu",
""
],
[
"Kyl",
"Jon",
""
],
[
"Mao",
"Justin",
""
],
[
"Lan",
"Kevin",
""
],
[
"Komeili",
"Mojtaba",
""
],
[
"Athar",
"ShahRukh",
""
],
[
"Babayan",
"Sheila",
""
],
[
"Beliasau",
"Stanislau",
""
],
[
"Buchwalter",
"William",
""
]
] |
2506.08280 | Daniel H. Pak | Daniel H. Pak, Shubh Thaker, Kyle Baylous, Xiaoran Zhang, Danny Bluestein, James S. Duncan | Snap-and-tune: combining deep learning and test-time optimization for high-fidelity cardiovascular volumetric meshing | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-quality volumetric meshing from medical images is a key bottleneck for physics-based simulations in personalized medicine. For volumetric meshing of complex medical structures, recent studies have often utilized deep learning (DL)-based template deformation approaches to enable fast test-time generation with high spatial accuracy. However, these approaches still exhibit limitations, such as limited flexibility at high-curvature areas and unrealistic inter-part distances. In this study, we introduce a simple yet effective snap-and-tune strategy that sequentially applies DL and test-time optimization, which combines fast initial shape fitting with more detailed sample-specific mesh corrections. Our method provides significant improvements in both spatial accuracy and mesh quality, while being fully automated and requiring no additional training labels. Finally, we demonstrate the versatility and usefulness of our newly generated meshes via solid mechanics simulations in two different software platforms. Our code is available at https://github.com/danpak94/Deep-Cardiac-Volumetric-Mesh. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 22:58:11 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Pak",
"Daniel H.",
""
],
[
"Thaker",
"Shubh",
""
],
[
"Baylous",
"Kyle",
""
],
[
"Zhang",
"Xiaoran",
""
],
[
"Bluestein",
"Danny",
""
],
[
"Duncan",
"James S.",
""
]
] |
2506.08281 | Seick Kim | Jongkeun Choi, Hongjie Dong, Seick Kim | Refined regularity at critical points for linear elliptic equations | 31 pages | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the regularity of solutions to linear elliptic equations in both divergence and non-divergence forms, particularly when the principal coefficients have Dini mean oscillation. We show that if a solution $u$ to a divergence-form equation satisfies $Du(x^o)=0$ at a point, then the second derivative $D^2u(x^o)$ exists and satisfies sharp continuity estimates. As a consequence, we obtain ``$C^{2,\alpha}$ regularity'' at critical points when the coefficients of $L$ are $C^\alpha$. This result refines a theorem of Teixeira (Math. Ann. 358 (2014), no. 1--2, 241--256) in the linear setting, where both linear and nonlinear equations were considered. We also establish an analogous result for equations in non-divergence form. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:02:52 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Choi",
"Jongkeun",
""
],
[
"Dong",
"Hongjie",
""
],
[
"Kim",
"Seick",
""
]
] |
2506.08282 | Monte Fischer | Monte Fischer, Peter W. Glynn | The Law of Large Numbers and CLT for Non-stationary Markov Jump Processes Exhibiting Time-of-Day Effects | null | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we develop a general law of large numbers and central limit theorem for finite state Markov jump processes with non-stationary transition rates. Such models commonly arise in service operations and manufacturing applications in which time-of-day, day-of-week, and secular effects are of first order importance in predicting system behavior. Our theorems allow for non-stationary reward environments that continuously accumulate reward, while including contributions from non-stationary lump-sum rewards of random size that are collected at jump times of the underlying process, jump times of a Poisson process modulated by the underlying process, or scheduled deterministic times. As part of our development, we also obtain a new central limit theorem for the special case in which the jump process and reward structure is periodic (as may occur over a weekly time interval), as well as for jump process models with resetting. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:09:28 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Fischer",
"Monte",
""
],
[
"Glynn",
"Peter W.",
""
]
] |
2506.08283 | Jianling Wang | Haoting Wang, Jianling Wang, Hao Li, Fangjun Yi, Mengyu Fu, Youwei Zhang, Yifan Liu, Liang Liu, Minmin Chen, Ed H. Chi, Lichan Hong and Haokai Lu | Serendipitous Recommendation with Multimodal LLM | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Conventional recommendation systems succeed in identifying relevant content but often fail to provide users with surprising or novel items. Multimodal Large Language Models (MLLMs) possess the world knowledge and multimodal understanding needed for serendipity, but their integration into billion-item-scale platforms presents significant challenges. In this paper, we propose a novel hierarchical framework where fine-tuned MLLMs provide high-level guidance to conventional recommendation models, steering them towards more serendipitous suggestions. This approach leverages MLLM strengths in understanding multimodal content and user interests while retaining the efficiency of traditional models for item-level recommendation. This mitigates the complexity of applying MLLMs directly to vast action spaces. We also demonstrate a chain-of-thought strategy enabling MLLMs to discover novel user interests by first understanding video content and then identifying relevant yet unexplored interest clusters. Through live experiments within a commercial short-form video platform serving billions of users, we show that our MLLM-powered approach significantly improves both recommendation serendipity and user satisfaction. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:13:22 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Haoting",
""
],
[
"Wang",
"Jianling",
""
],
[
"Li",
"Hao",
""
],
[
"Yi",
"Fangjun",
""
],
[
"Fu",
"Mengyu",
""
],
[
"Zhang",
"Youwei",
""
],
[
"Liu",
"Yifan",
""
],
[
"Liu",
"Liang",
""
],
[
"Chen",
"Minmin",
""
],
[
"Chi",
"Ed H.",
""
],
[
"Hong",
"Lichan",
""
],
[
"Lu",
"Haokai",
""
]
] |
2506.08284 | Raymond Tuminaro Dr. | Raymond Tuminaro and Christian Glusa | A structure preserving H-curl algebraic multigrid method for the eddy current equations | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new algebraic multigrid method (AMG) is presented for solving the linear systems associated with the eddy current approximation to the Maxwell equations. This AMG method extends an idea proposed by Reitzinger and Schoberl. The main feature of the Reitzinger and Schoberl algorithm (RSAMG) is that it maintains null-space properties of the Curl-Curl operator throughout all levels of the AMG hierarchy. It does this by enforcing a commuting relationship involving grid transfers and the discrete gradient operator. This null-space preservation property is critical to the algorithm's success, however enforcing this commuting relationship is non-trivial except in the special case where one leverages a piece-wise constant nodal interpolation operator. For this reason, mesh independent convergence rates are generally not observed for RSAMG due to its reliance on sub-optimal piece-wise constant interpolation. We present a new AMG algorithm that enforces the same commuting relationship. The main advance is that the new structure preserving H-curl algorithm (SpHcurlAMG) does not rely on piece-wise constant interpolation and can leverage fairly general and more sophisticated nodal interpolation operators. The key idea is to employ energy minimization AMG (EAMG) to construct edge interpolation grid transfers and to enforce the commuting relationship by embedding it as constraints within an EAMG procedure. While it might appear that solving such a constrained energy minimization is costly, we illustrate how this is not the case in our context. Numerical results are then given demonstrating mesh independent convergence over a range of test problems. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:18:12 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tuminaro",
"Raymond",
""
],
[
"Glusa",
"Christian",
""
]
] |
2506.08285 | Akshay Subramaniam | Akshay Subramaniam, Dale Durran, David Pruitt, Nathaniel Cresswell-Clay, William Yik | Imposing the Fundamental Dynamical Constraint of Hydrostatic Balance to Improve Global ML Weather Prediction | 13 pages, 7 figures | null | null | null | physics.ao-ph | http://creativecommons.org/licenses/by-sa/4.0/ | Forecasting weather accurately and efficiently is a critical capability in our ability to adapt to climate change. Data driven approaches to this problem have enjoyed much success recently providing forecasts with accuracy comparable to physics based numerical prediction models but at significantly reduced computational expense. However, these models typically do not incorporate any physics priors. In this work, we demonstrate improved skill of data driven weather prediction approaches by incorporating physical constraints, specifically in the context of the DLWP model (Karlbauer et. al. 2024). Near hydrostatic balance, between the vertical pressure gradient and gravity, is one of the most fundamental and well satisfied constraints on atmospheric motions. We impose this balance through both hard and soft constraints, and demonstrate that the soft constraint improves the RMSE of many forecast fields, particularly at lead times beyond 7-10 days. The positive influence of hydrostatic balance is also clearly evident in improving the physicality and strength of a 10-day forecast for hurricane Irma. These results show that adding appropriate physical constraints can improve the skill and fidelity of data driven weather models in a way that does not impose any significant additional memory capacity or scalability challenges. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:18:50 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Subramaniam",
"Akshay",
""
],
[
"Durran",
"Dale",
""
],
[
"Pruitt",
"David",
""
],
[
"Cresswell-Clay",
"Nathaniel",
""
],
[
"Yik",
"William",
""
]
] |
2506.08286 | Gregor Schiwietz | G. Schiwietz and P.L. Grande | A non-local exchange potential for electronic structure calculations | 24 pages, 7 figures | null | null | null | cond-mat.mtrl-sci quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we describe a model for the exchange interaction of electrons, as it follows from the Pauli exclusion principle. Starting from Hartree-Fock theory and making use of the free electron-gas model we propose a simple scheme to calculate the exchange-potential for atoms as well as simple molecules and solids. This method assures the correct asymptotic long-range behavior of the potential, contrary to local-density approximations that rely on a strongly simplified generalization of the well-known Kohn-Sham or Slater exchange interaction. Furthermore, our results approach the Kohn-Sham results in the interstitial space of solids. As a benchmark test, total energies and eigenenergies for atoms from He to Xe computed within our model are compared to other calculations as well as to experimental data. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:27:27 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Schiwietz",
"G.",
""
],
[
"Grande",
"P. L.",
""
]
] |
2506.08287 | Farmer Schlutzenberg | Farmer Schlutzenberg | Full normalization for $\kappa^+$-supercompactness | 24 pages | null | null | null | math.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend the normalization results of the author's paper "Full normalization for transfinite stacks" [5] to mice at the level of $\kappa^+$-supercompactness: given a normal iteration strategy $\Sigma$ for such a mouse $M$, with both $M$ and $\Sigma$ satisfying certain condensation properties, we extend $\Sigma$ to a strategy $\Sigma^*$ for stacks of normal trees, such that every iterate via $\Sigma^*$ is in fact a normal iterate via $\Sigma$. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:28:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Schlutzenberg",
"Farmer",
""
]
] |
2506.08288 | Ulisses Ribeiro | U. Ribeiro, F. Avila, C. R. Bom, C. Franco, A. Cortesi, C. Mendes de Oliveira, A. Kanaan, T. Ribeiro, W. Schoenell, E. Telles and A. Bernui | A cosmology weakly dependent measurement of 2D Baryon Acoustic Oscillations scale from the Southern Photometric Local Universe Survey | 12 pages, 10 figures | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | Baryon Acoustic Oscillations (BAO) provide a robust standard ruler for observational cosmology, enabling precise constraints on the expansion history of the Universe. We present a weakly model-dependent measurement of the BAO angular scale in the low-redshift Universe using the blue galaxies from the Southern Photometric Local Universe Survey (S-PLUS). Our analysis is based on the 2-point angular correlation function applied to a selected photometric sample of $5977$ galaxies with redshifts $0.03 \leq z \leq 0.1$. To account for photometric redshift uncertainties, we implement a resampling technique using the probability distribution function of each galaxy. Angular correlations are computed using the Landy-Szalay estimator; the uncertainties are quantified using a set of $1000$ log-normal mock catalogues. Our 2-point angular correlation analyses reveal a prominent BAO signal that after a shift correction, due to the projection effect caused by the finite thickness of the redshift bin, provides the transversal BAO measurement: $\theta_{BAO} = 21.81^{\circ} \pm 0.85^{\circ}$, at $z_{eff} = 0.075$, detected with a statistical significance of $3.22 \sigma$. In addition, we performed consistency tests that support the robustness of our result. Our measurement constitutes the first robust detection of the transversal BAO scale: at the lowest-redshift in the Universe and using multi-band (narrow+wide) photometry data from the S-PLUS. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:37:36 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ribeiro",
"U.",
""
],
[
"Avila",
"F.",
""
],
[
"Bom",
"C. R.",
""
],
[
"Franco",
"C.",
""
],
[
"Cortesi",
"A.",
""
],
[
"de Oliveira",
"C. Mendes",
""
],
[
"Kanaan",
"A.",
""
],
[
"Ribeiro",
"T.",
""
],
[
"Schoenell",
"W.",
""
],
[
"Telles",
"E.",
""
],
[
"Bernui",
"A.",
""
]
] |
2506.08289 | Weiller Felipe Chaves Barboza | W.F.C. Barboza, T.F. Cruz, R.B. Leal | Stereographic Projections on Some Quadric Surfaces | null | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | In this work, we present an adaptation of the classical stereographic projection, originally formulated for the sphere, now considering the context of the ellipsoid and the elliptic paraboloid. We begin by constructing the stereographic projections for both quadric surfaces separately, analyzing the geometric particularities of each surface and the challenges arising from their variable curvatures and, in the case of the paraboloid, its non-compactness. In the final part of the work, we establish results concerning the eccentricities, curvatures, arc length, and areas of the ellipses formed by the intersection of the quadrics with horizontal sections and their corresponding projections onto the plane-xy. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:38:51 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Barboza",
"W. F. C.",
""
],
[
"Cruz",
"T. F.",
""
],
[
"Leal",
"R. B.",
""
]
] |
2506.08290 | Hyunmi Song | Sangeun Yu, Hyunmi Song, Kwang-il Seon | Ly$\alpha$ radiative transfer modeling for 163 MUSE Ly$\alpha$-emitting galaxies at $z=$3--6 | Accepted for publication in ApJ; 18 pages, 14 figures | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We utilized Ly$\alpha$ radiative transfer calculations from \citet{Song2020} to investigate the properties of extended Ly$\alpha$ halos around star-forming galaxies in the \textit{Hubble} Ultra Deep Field, observed by the Multi-Unit Spectroscopic Explorer. Expanding on the work of \citet{Song2020}, which was limited to eight galaxies, we derived best-fit models for a significantly larger sample of 163 galaxies, which successfully reproduced both their Ly$\alpha$ spectra and surface brightness profiles (SBPs). These best-fit models suggest a broad medium distribution surrounding each galaxy, with low expanding velocities at large radii. This conclusion could not have been drawn from modeling either the spectrum or SBP alone, but only through simultaneous modeling of both. Our correlation analysis between observables and model parameters reveals that the spatial extent of Ly$\alpha$ halos is primarily determined by the extents of the medium and the source, while the spectral peak shift and full width at half maximum are governed mainly by optical depth, with the velocity structure of the medium playing a secondary yet non-negligible role. The fact that various correlations derived from the full set of models and those from the best-fit subset can differ significantly highlights the complex and interdependent nature of Ly$\alpha$ radiative transfer. All model parameters interact to shape the observed Ly$\alpha$ features in a non-trivial way. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:39:54 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yu",
"Sangeun",
""
],
[
"Song",
"Hyunmi",
""
],
[
"Seon",
"Kwang-il",
""
]
] |
2506.08291 | Won Kyung Do | Won Kyung Do, Matthew Strong, Aiden Swann, Boshu Lei, and Monroe Kennedy III | TensorTouch: Calibration of Tactile Sensors for High Resolution Stress Tensor and Deformation for Dexterous Manipulation | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Advanced dexterous manipulation involving multiple simultaneous contacts across different surfaces, like pinching coins from ground or manipulating intertwined objects, remains challenging for robotic systems. Such tasks exceed the capabilities of vision and proprioception alone, requiring high-resolution tactile sensing with calibrated physical metrics. Raw optical tactile sensor images, while information-rich, lack interpretability and cross-sensor transferability, limiting their real-world utility. TensorTouch addresses this challenge by integrating finite element analysis with deep learning to extract comprehensive contact information from optical tactile sensors, including stress tensors, deformation fields, and force distributions at pixel-level resolution. The TensorTouch framework achieves sub-millimeter position accuracy and precise force estimation while supporting large sensor deformations crucial for manipulating soft objects. Experimental validation demonstrates 90% success in selectively grasping one of two strings based on detected motion, enabling new contact-rich manipulation capabilities previously inaccessible to robotic systems. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:49:03 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Do",
"Won Kyung",
""
],
[
"Strong",
"Matthew",
""
],
[
"Swann",
"Aiden",
""
],
[
"Lei",
"Boshu",
""
],
[
"Kennedy",
"Monroe",
"III"
]
] |
2506.08292 | Zhanke Zhou | Xie Yi, Zhanke Zhou, Chentao Cao, Qiyu Niu, Tongliang Liu, Bo Han | From Debate to Equilibrium: Belief-Driven Multi-Agent LLM Reasoning via Bayesian Nash Equilibrium | Accepted by ICML 2025 | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-agent frameworks can substantially boost the reasoning power of large language models (LLMs), but they typically incur heavy computational costs and lack convergence guarantees. To overcome these challenges, we recast multi-LLM coordination as an incomplete-information game and seek a Bayesian Nash equilibrium (BNE), in which each agent optimally responds to its probabilistic beliefs about the strategies of others. We introduce Efficient Coordination via Nash Equilibrium (ECON), a hierarchical reinforcement-learning paradigm that marries distributed reasoning with centralized final output. Under ECON, each LLM independently selects responses that maximize its expected reward, conditioned on its beliefs about co-agents, without requiring costly inter-agent exchanges. We mathematically prove that ECON attains a markedly tighter regret bound than non-equilibrium multi-agent schemes. Empirically, ECON outperforms existing multi-LLM approaches by 11.2% on average across six benchmarks spanning complex reasoning and planning tasks. Further experiments demonstrate ECON's ability to flexibly incorporate additional models, confirming its scalability and paving the way toward larger, more powerful multi-LLM ensembles. The code is publicly available at: https://github.com/tmlr-group/ECON. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:49:14 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yi",
"Xie",
""
],
[
"Zhou",
"Zhanke",
""
],
[
"Cao",
"Chentao",
""
],
[
"Niu",
"Qiyu",
""
],
[
"Liu",
"Tongliang",
""
],
[
"Han",
"Bo",
""
]
] |
2506.08293 | Jason Gleghorn | Logan Hallee, Nikolaos Rafailidis, David B. Bichara, Jason P. Gleghorn | Diffusion Sequence Models for Enhanced Protein Representation and Generation | 20 pages, 15 figures | null | null | null | q-bio.BM | http://creativecommons.org/licenses/by-sa/4.0/ | Proteins are fundamental to biology, executing diverse functions through complex physicochemical interactions, and they hold transformative potential across medicine, materials science, and environmental applications. Protein Language Models (pLMs) aim to unlock insights from the vast space of unlabeled protein sequences by learning rich, semantic representations from primary sequences via masked language modeling. However, these models typically exhibit limited generative capacity. In this work, we introduce the Diffusion Sequence Model (DSM), a novel pLM trained with masked diffusion to enable both high-quality representation learning and generative protein design. DSM builds upon the ESM2 architecture by incorporating a masked forward diffusion process inspired by the LLaDA framework. After training, DSM is capable of generating diverse, biomimetic sequences that align with expected amino acid compositions, secondary structures, and predicted functions, even with 90\% token corruption. Furthermore, DSM's learned representations match or exceed those of similarly sized pLMs on downstream tasks. We also introduce DSM(ppi), a variant fine-tuned to generate protein binders by attending to target sequences. We demonstrate DSM(ppi)'s effectiveness on the challenging Bench-tested Binder Benchmark (BenchBB), where both DSM and DSM(ppi) produce candidates with superior predicted binding affinity compared to known binders. Our results establish masked diffusion as a powerful paradigm for unifying protein representation and generation in a single framework. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:50:11 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Hallee",
"Logan",
""
],
[
"Rafailidis",
"Nikolaos",
""
],
[
"Bichara",
"David B.",
""
],
[
"Gleghorn",
"Jason P.",
""
]
] |
2506.08294 | Ruanqianqian (Lisa) Huang | Ruanqianqian Huang, Ayana Monroe, Peli de Halleux, Sorin Lerner, Nikolaj Bj{\o}rner | Z3Guide: A Scalable, Student-Centered, and Extensible Educational Environment for Logic Modeling | null | null | null | null | cs.HC cs.LO | http://creativecommons.org/licenses/by/4.0/ | Constraint-satisfaction problems (CSPs) are ubiquitous, ranging from budgeting for grocery shopping to verifying software behavior. Logic modeling helps solve CSPs programmatically using SMT solvers. Despite its importance in many Computer Science disciplines, resources for teaching and learning logic modeling are scarce and scattered, and challenges remain in designing educational environments for logic modeling that are accessible and meet the needs of teachers and students. This paper explores how to design such an environment and probes the impact of the design on the learning experience. From a need-finding interview study and a design iteration with teachers of logic modeling, we curated 10 design guidelines spanning three main requirements: providing easy access, supporting various educational modalities, and allowing extensions for customized pedagogical needs. We implemented nine guidelines in Z3Guide, an open-source browser-based tool. Using Z3Guide in a logic modeling learning workshop with more than 100 students, we gathered positive feedback on its support for learning and identified opportunities for future improvements. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:56:15 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Huang",
"Ruanqianqian",
""
],
[
"Monroe",
"Ayana",
""
],
[
"de Halleux",
"Peli",
""
],
[
"Lerner",
"Sorin",
""
],
[
"Bjørner",
"Nikolaj",
""
]
] |
2506.08295 | Zhanke Zhou | Zhanke Zhou, Xiao Feng, Zhaocheng Zhu, Jiangchao Yao, Sanmi Koyejo, Bo Han | From Passive to Active Reasoning: Can Large Language Models Ask the Right Questions under Incomplete Information? | Accepted by ICML 2025 | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While existing benchmarks probe the reasoning abilities of large language models (LLMs) across diverse domains, they predominantly assess passive reasoning, providing models with all the information needed to reach a solution. By contrast, active reasoning-where an LLM must interact with external systems to acquire missing evidence or data-has received little systematic attention. To address this shortfall, we present AR-Bench, a novel benchmark designed explicitly to evaluate an LLM's active reasoning skills. AR-Bench comprises three task families-detective cases, situation puzzles, and guessing numbers-that together simulate real-world, agentic scenarios and measure performance across commonsense, logical, and symbolic reasoning challenges. Empirical evaluation on AR-Bench demonstrates that contemporary LLMs exhibit pronounced difficulties with active reasoning: they frequently fail to acquire or leverage the information needed to solve tasks. This gap highlights a stark divergence between their passive and active reasoning abilities. Moreover, ablation studies indicate that even advanced strategies, such as tree-based searching or post-training approaches, yield only modest gains and fall short of the levels required for real-world deployment. Collectively, these findings highlight the critical need to advance methodology for active reasoning, e.g., incorporating interactive learning, real-time feedback loops, and environment-aware objectives for training. The benchmark is publicly available at: https://github.com/tmlr-group/AR-Bench. | [
{
"version": "v1",
"created": "Mon, 09 Jun 2025 23:56:41 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zhou",
"Zhanke",
""
],
[
"Feng",
"Xiao",
""
],
[
"Zhu",
"Zhaocheng",
""
],
[
"Yao",
"Jiangchao",
""
],
[
"Koyejo",
"Sanmi",
""
],
[
"Han",
"Bo",
""
]
] |
2506.08297 | Nhat Thanh Tran | Nhat Thanh Tran, Fanghui Xue, Shuai Zhang, Jiancheng Lyu, Yunling Zheng, Yingyong Qi, Jack Xin | SEMA: a Scalable and Efficient Mamba like Attention via Token Localization and Averaging | 15 pages, figures 3 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Attention is the critical component of a transformer. Yet the quadratic computational complexity of vanilla full attention in the input size and the inability of its linear attention variant to focus have been challenges for computer vision tasks. We provide a mathematical definition of generalized attention and formulate both vanilla softmax attention and linear attention within the general framework. We prove that generalized attention disperses, that is, as the number of keys tends to infinity, the query assigns equal weights to all keys. Motivated by the dispersion property and recent development of Mamba form of attention, we design Scalable and Efficient Mamba like Attention (SEMA) which utilizes token localization to avoid dispersion and maintain focusing, complemented by theoretically consistent arithmetic averaging to capture global aspect of attention. We support our approach on Imagenet-1k where classification results show that SEMA is a scalable and effective alternative beyond linear attention, outperforming recent vision Mamba models on increasingly larger scales of images at similar model parameter sizes. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:03:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tran",
"Nhat Thanh",
""
],
[
"Xue",
"Fanghui",
""
],
[
"Zhang",
"Shuai",
""
],
[
"Lyu",
"Jiancheng",
""
],
[
"Zheng",
"Yunling",
""
],
[
"Qi",
"Yingyong",
""
],
[
"Xin",
"Jack",
""
]
] |
2506.08298 | Trung-Kien Nguyen | Trung-Kien Nguyen, Heng Ping, Shixuan Li, Peiyu Zhang, Nikos Kanakaris, Nicholas Kotov, Paul Bogdan | H$^2$GFM: Towards unifying Homogeneity and Heterogeneity on Text-Attributed Graphs | null | null | null | null | cs.LG cs.SI | http://creativecommons.org/licenses/by/4.0/ | The growing interests and applications of graph learning in diverse domains have propelled the development of a unified model generalizing well across different graphs and tasks, known as the Graph Foundation Model (GFM). Existing research has leveraged text-attributed graphs (TAGs) to tackle the heterogeneity in node features among graphs. However, they primarily focus on homogeneous TAGs (HoTAGs), leaving heterogeneous TAGs (HeTAGs), where multiple types of nodes/edges reside, underexplored. To enhance the capabilities and applications of GFM, we introduce H$^2$GFM, a novel framework designed to generalize across both HoTAGs and HeTAGs. Our model projects diverse meta-relations among graphs under a unified textual space, and employs a context encoding to capture spatial and higher-order semantic relationships. To achieve robust node representations, we propose a novel context-adaptive graph transformer (CGT), effectively capturing information from both context neighbors and their relationships. Furthermore, we employ a mixture of CGT experts to capture the heterogeneity in structural patterns among graph types. Comprehensive experiments on a wide range of HoTAGs and HeTAGs as well as learning scenarios demonstrate the effectiveness of our model. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:03:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nguyen",
"Trung-Kien",
""
],
[
"Ping",
"Heng",
""
],
[
"Li",
"Shixuan",
""
],
[
"Zhang",
"Peiyu",
""
],
[
"Kanakaris",
"Nikos",
""
],
[
"Kotov",
"Nicholas",
""
],
[
"Bogdan",
"Paul",
""
]
] |
2506.08299 | Kangning Yang | Kangning Yang, Ling Ouyang, Huiming Sun, Jie Cai, Lan Fu, Jiaming Ding, Chiu Man Ho, Zibo Meng | OpenRR-1k: A Scalable Dataset for Real-World Reflection Removal | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Reflection removal technology plays a crucial role in photography and computer vision applications. However, existing techniques are hindered by the lack of high-quality in-the-wild datasets. In this paper, we propose a novel paradigm for collecting reflection datasets from a fresh perspective. Our approach is convenient, cost-effective, and scalable, while ensuring that the collected data pairs are of high quality, perfectly aligned, and represent natural and diverse scenarios. Following this paradigm, we collect a Real-world, Diverse, and Pixel-aligned dataset (named OpenRR-1k dataset), which contains 1,000 high-quality transmission-reflection image pairs collected in the wild. Through the analysis of several reflection removal methods and benchmark evaluation experiments on our dataset, we demonstrate its effectiveness in improving robustness in challenging real-world environments. Our dataset is available at https://github.com/caijie0620/OpenRR-1k. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:04:47 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yang",
"Kangning",
""
],
[
"Ouyang",
"Ling",
""
],
[
"Sun",
"Huiming",
""
],
[
"Cai",
"Jie",
""
],
[
"Fu",
"Lan",
""
],
[
"Ding",
"Jiaming",
""
],
[
"Ho",
"Chiu Man",
""
],
[
"Meng",
"Zibo",
""
]
] |
2506.08300 | Matteo Cargnelutti | Matteo Cargnelutti, Catherine Brobston, John Hess, Jack Cushman, Kristi Mukk, Aristana Scourtas, Kyle Courtney, Greg Leppert, Amanda Watson, Martha Whitehead, Jonathan Zittrain | Institutional Books 1.0: A 242B token dataset from Harvard Library's collections, refined for accuracy and usability | null | null | null | null | cs.CL cs.DL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) use data to learn about the world in order to produce meaningful correlations and predictions. As such, the nature, scale, quality, and diversity of the datasets used to train these models, or to support their work at inference time, have a direct impact on their quality. The rapid development and adoption of LLMs of varying quality has brought into focus the scarcity of publicly available, high-quality training data and revealed an urgent need to ground the stewardship of these datasets in sustainable practices with clear provenance chains. To that end, this technical report introduces Institutional Books 1.0, a large collection of public domain books originally digitized through Harvard Library's participation in the Google Books project, beginning in 2006. Working with Harvard Library, we extracted, analyzed, and processed these volumes into an extensively-documented dataset of historic texts. This analysis covers the entirety of Harvard Library's collection scanned as part of that project, originally spanning 1,075,899 volumes written in over 250 different languages for a total of approximately 250 billion tokens. As part of this initial release, the OCR-extracted text (original and post-processed) as well as the metadata (bibliographic, source, and generated) of the 983,004 volumes, or 242B tokens, identified as being in the public domain have been made available. This report describes this project's goals and methods as well as the results of the analyses we performed, all in service of making this historical collection more accessible and easier for humans and machines alike to filter, read and use. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:11:30 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Cargnelutti",
"Matteo",
""
],
[
"Brobston",
"Catherine",
""
],
[
"Hess",
"John",
""
],
[
"Cushman",
"Jack",
""
],
[
"Mukk",
"Kristi",
""
],
[
"Scourtas",
"Aristana",
""
],
[
"Courtney",
"Kyle",
""
],
[
"Leppert",
"Greg",
""
],
[
"Watson",
"Amanda",
""
],
[
"Whitehead",
"Martha",
""
],
[
"Zittrain",
"Jonathan",
""
]
] |
2506.08301 | Yuji Yamaguchi | Y. Yamaguchi, M. Niikura, R. Mizuno, M. Tampo, M. Harada, N. Kawamura, I. Umegaki, S. Takeshita, K. Haga | Measurement of radionuclide production probabilities in negative muon nuclear capture and validation of Monte Carlo simulation model | null | null | null | null | nucl-ex | http://creativecommons.org/publicdomain/zero/1.0/ | As part of the development of a sample radioactivity calculation program, we have measured radionuclide production probabilities in negative muon nuclear capture to update experimental data and to validate a calculation dataset obtained by a Monte Carlo simulation code. The probabilities have been obtained by an activation experiment on $^{27}$Al, $^\mathrm{nat}$Si, $^{59}$Co, and $^\mathrm{nat}$Ta targets. The obtained probabilities expand the validation scope to the radionuclide production processes outside of the existing data coverage. By comparing the resultant probabilities with the calculated dataset, it has been revealed that the dataset is generally on the {\color{black}safe} side in radioactivity estimation and needs to be corrected in the following three cases: (i) isomer production; (ii) radionuclide production by the multiple neutron emission; (iii) radionuclide production by particle emissions involving a proton. The present probabilities and the new findings on the correction provide valuable clues to improvements of the simulation models. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:15:09 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yamaguchi",
"Y.",
""
],
[
"Niikura",
"M.",
""
],
[
"Mizuno",
"R.",
""
],
[
"Tampo",
"M.",
""
],
[
"Harada",
"M.",
""
],
[
"Kawamura",
"N.",
""
],
[
"Umegaki",
"I.",
""
],
[
"Takeshita",
"S.",
""
],
[
"Haga",
"K.",
""
]
] |
2506.08302 | Georg Engelhardt | Georg Engelhardt, Konstantin Dorfman, Zhedong Zhang | Photon-resolved Floquet theory approach to spectroscopic quantum sensing | 33 pages including appendices; 6 pictures; comments are welcome | null | null | null | quant-ph physics.optics | http://creativecommons.org/licenses/by/4.0/ | Spectroscopic methods play a vital role in quantum sensing, which uses the quantized nature of atoms or molecules to reach astonishing precision for sensing of, e.g., electric or magnetic fields. In the theoretical treatment, one typically invokes semiclassical methods to describe the light-matter interaction between quantum emitters, e.g., atoms or molecules, and a strong coherent laser field. However, these semiclassical approaches struggle to predict the stochastic measurement fluctuations beyond the mean value, necessary to predict the sensitivity of spectroscopic quantum sensing protocols. Here, we develop a theoretical framework based on the recently developed Photon-resolved Floquet theory (PRFT) which is capable to predict the measurement statistics describing higher order statistics of coherent quantum states of light. The PRFT constructs flow equations for the cumulants of the photonic measurement statistics utilizing only the semiclassical dynamics of the matter system. We apply the PRFT to spectroscopic quantum sensing using dissipative two-level and four-level systems (describing electric field sensing with Rydberg atoms), and demonstrate how to calculate the Fisher information of the measurement statistics with respect to various system parameters. In doing so, we demonstrate that the PRFT is a flexible tool allowing to improve the sensitivity of spectroscopic quantum sensing devices by several orders of magnitudes. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:18:02 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Engelhardt",
"Georg",
""
],
[
"Dorfman",
"Konstantin",
""
],
[
"Zhang",
"Zhedong",
""
]
] |
2506.08303 | Thomas M. Kwok | Thomas M. Kwok, Hilary HY Cheng, Wai Tuck Chow | EMG-Driven Stiffness-Modulating Palpation for Telerehabilitation | Accepted by the Workshop on Human-Robot Contact and Manipulation (HRCM 2025) at RSS Conference 2025 | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | In this work, we introduce HJ-Pal, a lightweight wearable haptic device that leverages EMG-driven honeycomb jamming to render muscle activation as kinesthetic feedback, enabling remote palpation for small muscle assessment in telerehabilitation. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:26:35 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Kwok",
"Thomas M.",
""
],
[
"Cheng",
"Hilary HY",
""
],
[
"Chow",
"Wai Tuck",
""
]
] |
2506.08304 | Davi Arrais Nobre | Davi Arrais Nobre, Karen C. Abbott, Jonathan Machta, Alan Hastings | Long-range dispersal promotes spatial synchrony but reduces the length and time scales of synchronous fluctuations | null | null | null | null | q-bio.PE cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Synchronous oscillations of spatially disjunct populations are widely observed in ecology. Even in the absence of spatially synchronized exogenous forces, metapopulations may synchronize via dispersal. For many species, most dispersal is local, but rare long-distance dispersal events also occur. While even small amounts of long-range dispersal are known to be important for processes like invasion and spatial spread rates, their potential influence on population synchrony is often overlooked, since local dispersal on its own can be strongly synchronizing. In this work, we investigate the effect of random, rare, long-range dispersal on the spatial synchrony of a metapopulation and find profound effects not only on synchrony but also on properties of the resulting spatial patterns. While controlling for the overall amount of emigration from each local subpopulation, we vary the fraction of dispersal that occurs locally (to nearest neighbors) versus globally (to random locations, irrespective of distance). Using a metric that measures the instantaneous level of global synchrony, we show that this form of long-range dispersal significantly favors the spatially synchronous state and homogenizes the population by decreasing the size of clusters of subpopulations that are out of phase with the rest of the metapopulation. Moreover, the addition of non-local dispersal significantly decreases the equilibration time of the metapopulation. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:27:39 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Nobre",
"Davi Arrais",
""
],
[
"Abbott",
"Karen C.",
""
],
[
"Machta",
"Jonathan",
""
],
[
"Hastings",
"Alan",
""
]
] |
2506.08305 | Ashish Srivastava | Kulumani M. Rangaswamy, Ashish K Srivastava | Graded Naimark's Problem for Leavitt Path Algebras | null | null | null | null | math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we study the graded version of Naimark's problem for Leavitt path algebras considering them as $\mathbb{Z}$-graded algebras. Several characterizations are obtained of a Leavitt path algebra $L$ of an arbitrary graph $E$ over a field $\mathbb{K}$ over which any two graded-simple modules are graded isomorphic. Such a Leavitt path algebra $L$ is shown to be graded isomorphic to the algebra of graded infinite matrices having at most finitely many non-zero entries from the ring $R$ where $R=\mathbb{K}$ or $R=\mathbb{K}[x,x^{-1}]$. Equivalently, $L$ is a graded-simple ring which is graded-semisimple, that is, $L$ is a graded direct sum of graded-isomorphic graded-simple left $L$-modules. Graphically, the graph $E$ is shown to be row-finite, downward directed and the vertex set $E^{0}$ is the hereditary saturated closure of a single vertex $v$ which is either a line point or lies on a cycle without exits. We also characterize Leavitt path algebras possessing at most countably many isomorphism classes of graded-simple left modules. Examples are constructed illustrating these results. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:31:48 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Rangaswamy",
"Kulumani M.",
""
],
[
"Srivastava",
"Ashish K",
""
]
] |
2506.08306 | Rithwik Sudharsan | Tuan Truong, Rithwik Sudharsan, Yibo Yang, Peter Xiangyuan Ma, Ruihan Yang, Stephan Mandt, Joshua S. Bloom | AstroCompress: A benchmark dataset for multi-purpose compression of astronomical data | ICLR 2025 conference paper. See reviews at https://openreview.net/forum?id=kQCHCkNk7s | null | null | null | cs.AI astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The site conditions that make astronomical observatories in space and on the ground so desirable -- cold and dark -- demand a physical remoteness that leads to limited data transmission capabilities. Such transmission limitations directly bottleneck the amount of data acquired and in an era of costly modern observatories, any improvements in lossless data compression has the potential scale to billions of dollars worth of additional science that can be accomplished on the same instrument. Traditional lossless methods for compressing astrophysical data are manually designed. Neural data compression, on the other hand, holds the promise of learning compression algorithms end-to-end from data and outperforming classical techniques by leveraging the unique spatial, temporal, and wavelength structures of astronomical images. This paper introduces AstroCompress: a neural compression challenge for astrophysics data, featuring four new datasets (and one legacy dataset) with 16-bit unsigned integer imaging data in various modes: space-based, ground-based, multi-wavelength, and time-series imaging. We provide code to easily access the data and benchmark seven lossless compression methods (three neural and four non-neural, including all practical state-of-the-art algorithms). Our results on lossless compression indicate that lossless neural compression techniques can enhance data collection at observatories, and provide guidance on the adoption of neural compression in scientific applications. Though the scope of this paper is restricted to lossless compression, we also comment on the potential exploration of lossy compression methods in future studies. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:32:30 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Truong",
"Tuan",
""
],
[
"Sudharsan",
"Rithwik",
""
],
[
"Yang",
"Yibo",
""
],
[
"Ma",
"Peter Xiangyuan",
""
],
[
"Yang",
"Ruihan",
""
],
[
"Mandt",
"Stephan",
""
],
[
"Bloom",
"Joshua S.",
""
]
] |
2506.08307 | Zhenghua Xu | Zhenghua Xu, Chao Ding, Haiyan Wang | Monogenic functions over real alternative *-algebras: the several hypercomplex variables case | 20 pages | null | null | null | math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The notion of monogenic (or regular) functions, which is a correspondence of holomorphic functions, has been studied extensively in hypercomplex analysis, including quaternionic, octonionic, and Clifford analysis. Recently, the concept of monogenic functions over real alternative $\ast$-algebras has been introduced to unify several classical monogenic functions theories. In this paper, we initiate the study of monogenic functions of several hypercomplex variables over real alternative $\ast$-algebras, which naturally extends the theory of several complex variables to a very general setting. In this new setting, we develop some fundamental properties, such as Bochner-Martinelli formula, Plemelj-Sokhotski formula, and Hartogs extension theorem. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:33:46 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Xu",
"Zhenghua",
""
],
[
"Ding",
"Chao",
""
],
[
"Wang",
"Haiyan",
""
]
] |
2506.08308 | Yong Zhang | Yu Li, Zhixuan Li, Manting Xie, Yong Zhang | An efficient Fourier spectral algorithm for the Bogoliubov-de Gennes excitation eigenvalue problem | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose an efficient Fourier spectral algorithm for an eigenvalue problem, that is, the Bogoliubov-de Gennes (BdG) equation arsing from spin-1 Bose-Einstein condensates (BEC) to describe the elementary/collective excitations around the mean-field ground state. The BdG equation is essentially a constrained eigenvalue/eigenfunction system. Firstly, we investigate its analytical properties, including exact eigenpairs, generalized nullspace, and bi-orthogonality of eigenspaces. Secondly, by combining the standard Fourier spectral method for spatial discretization and a stable Gram-Schmidt bi-orthogonal algorithm, we develop a subspace iterative solver for such a large-scale dense eigenvalue problem, and it proves to be numerically stable, efficient, and accurate. Our solver is matrix-free and the operator-function evaluation is accelerated by discrete Fast Fourier Transform (FFT) with almost optimal efficiency. Therefore, it is memory-friendly and efficient for large-scale problems. Furthermore, we give a rigorous and detailed numerical analysis on the stability and spectral convergence. Finally, we present extensive numerical results to illustrate the spectral accuracy and efficiency, and investigate the excitation spectrum and Bogoliubov amplitudes around the ground state in 1-3 spatial dimensions. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:35:34 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Li",
"Yu",
""
],
[
"Li",
"Zhixuan",
""
],
[
"Xie",
"Manting",
""
],
[
"Zhang",
"Yong",
""
]
] |
2506.08310 | Yingjie Luo | Debesh Bhattacharjee (1), Eduard P. Kontar (1) and Yingjie Luo (1) ((1) School of Physics & Astronomy, University of Glasgow, Glasgow, UK) | Total Power and Low-energy Cut-off Time Evolution of Solar Flare Accelerated Electrons Using X-Ray Observations and Warm-Target Model | Acceptted by APJ,12 Figures | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | A primary characteristic of solar flares is the efficient acceleration of electrons to nonthermal deka-keV energies. While hard X-Ray (HXR) observation of bremsstrahlung emission serves as the key diagnostic of these electrons. In this study, we investigate the time evolution of flare-accelerated electrons using the warm-target model. This model, unlike the commonly used cold-target model, can determine the low-energy cut-off in the nonthermal electron distribution, so that the energetics of nonthermal electrons can be deduced more accurately. Here, we examine the time-evolution of nonthermal electrons in flares well-observed by the RHESSI and the Solar Orbiter (SolO, using the STIX instrument) spacecrafts. Using spectroscopic and imaging HXR observations, the time evolution of the low-energy cut-off of the accelerated electron distribution, the total power of nonthermal electrons, total rate of nonthermal electrons, and excess thermal emission measure from the nonthermal electrons, are investigated. We find that the time profile of the low-energy cut-off of the accelerated electron distribution shows a high-low-high trend around the HXR bursts of flares, while the time evolution of the total rate of injected electrons shows a low-high-low behavior. Although the total power of nonthermal electrons is sensitive to the cut-off energy, the temporal variation of the flare power follows the temporal variation of the acceleration rate. We further find that the highest contribution of the excess thermal emission measure coming from thermalization of injected electrons takes place around the hard X-ray peak. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:40:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bhattacharjee",
"Debesh",
"",
"School of Physics & Astronomy, University of Glasgow, Glasgow, UK"
],
[
"Kontar",
"Eduard P.",
"",
"School of Physics & Astronomy, University of Glasgow, Glasgow, UK"
],
[
"Luo",
"Yingjie",
"",
"School of Physics & Astronomy, University of Glasgow, Glasgow, UK"
]
] |
2506.08311 | Ira Ceka | Ira Ceka, Saurabh Pujar, Shyam Ramji, Luca Buratti, Gail Kaiser, and Baishakhi Ray | Understanding Software Engineering Agents Through the Lens of Traceability: An Empirical Study | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by/4.0/ | With the advent of large language models (LLMs), software engineering agents (SWE agents) have emerged as a powerful paradigm for automating a range of software tasks -- from code generation and repair to test case synthesis. These agents operate autonomously by interpreting user input and responding to environmental feedback. While various agent architectures have demonstrated strong empirical performance, the internal decision-making worfklows that drive their behavior remain poorly understood. Deeper insight into these workflows hold promise for improving both agent reliability and efficiency. In this work, we present the first systematic study of SWE agent behavior through the lens of execution traces. Our contributions are as follows: (1) we propose the first taxonomy of decision-making pathways across five representative agents; (2) using this taxonomy, we identify three core components essential to agent success -- bug localization, patch generation, and reproduction test generation -- and study each in depth; (3) we study the impact of test generation on successful patch production; and analyze strategies that can lead to successful test generation; (4) we further conduct the first large-scale code clone analysis comparing agent-generated and developer-written patches and provide a qualitative study revealing structural and stylistic differences in patch content. Together, these findings offer novel insights into agent design and open avenues for building agents that are both more effective and more aligned with human development practices. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:41:54 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ceka",
"Ira",
""
],
[
"Pujar",
"Saurabh",
""
],
[
"Ramji",
"Shyam",
""
],
[
"Buratti",
"Luca",
""
],
[
"Kaiser",
"Gail",
""
],
[
"Ray",
"Baishakhi",
""
]
] |
2506.08312 | Tom\'as Gonz\'alez | Tom\'as Gonz\'alez, Giulia Fanti, Aaditya Ramdas | Private Evolution Converges | null | null | null | null | cs.LG cs.CR cs.DS math.PR math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Private Evolution (PE) is a promising training-free method for differentially private (DP) synthetic data generation. While it achieves strong performance in some domains (e.g., images and text), its behavior in others (e.g., tabular data) is less consistent. To date, the only theoretical analysis of the convergence of PE depends on unrealistic assumptions about both the algorithm's behavior and the structure of the sensitive dataset. In this work, we develop a new theoretical framework to explain PE's practical behavior and identify sufficient conditions for its convergence. For $d$-dimensional sensitive datasets with $n$ data points from a bounded domain, we prove that PE produces an $(\epsilon, \delta)$-DP synthetic dataset with expected 1-Wasserstein distance of order $\tilde{O}(d(n\epsilon)^{-1/d})$ from the original, establishing worst-case convergence of the algorithm as $n \to \infty$. Our analysis extends to general Banach spaces as well. We also connect PE to the Private Signed Measure Mechanism, a method for DP synthetic data generation that has thus far not seen much practical adoption. We demonstrate the practical relevance of our theoretical findings in simulations. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:47:09 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"González",
"Tomás",
""
],
[
"Fanti",
"Giulia",
""
],
[
"Ramdas",
"Aaditya",
""
]
] |
2506.08313 | Akinwale Famotire | Oriyomi Ahmad Hassan, Aisha Tunrayo Maradesa, Abdulazeez Toyosi Alabi, Oyejide Surajudeen Salam, Ajani Busari, Akinwale Victor Famotire, Habeeb Abiodun Afolabi, Solomon Adeleke, and Abayomi Ayodele Akomolafe | A New Lifetime Distribution: Exponentiated Exponential-Pareto-HalfNormal Mixture Model for Biomedical Applications | 22 pages, 9 figures | null | null | null | stat.AP stat.ME | http://creativecommons.org/licenses/by/4.0/ | This study introduces the Exponentiated-Exponential-Pareto-Half Normal Mixture Distribution (EEPHND), a novel hybrid model developed to overcome the limitations of classical distributions in modeling complex real-world data. By compounding the Exponentiated-Exponential-Pareto (EEP) and Half-Normal distributions through a mixture mechanism, EEPHND effectively captures both early-time symmetry and long-tail behavior, features which are commonly observed in survival and reliability data. The model offers closed-form expressions for its probability density, cumulative distribution, survival and hazard functions, moments, and reliability metrics, ensuring analytical traceability and interpretability in the presence of censoring and heterogeneous risk dynamics. When applied to a real-world lung cancer dataset, EEPHND outperformed competing models in both goodness-of-fit and predictive accuracy, achieving a Concordance Index (CI) of 0.9997. These results highlight its potential as a flexible and powerful tool for survival analysis and biomedical engineering. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:48:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Hassan",
"Oriyomi Ahmad",
""
],
[
"Maradesa",
"Aisha Tunrayo",
""
],
[
"Alabi",
"Abdulazeez Toyosi",
""
],
[
"Salam",
"Oyejide Surajudeen",
""
],
[
"Busari",
"Ajani",
""
],
[
"Famotire",
"Akinwale Victor",
""
],
[
"Afolabi",
"Habeeb Abiodun",
""
],
[
"Adeleke",
"Solomon",
""
],
[
"Akomolafe",
"Abayomi Ayodele",
""
]
] |
2506.08314 | Sibo Zhao | Sibo Zhao, Michael Bewong, Selasi Kwashie, Junwei Hu, and Zaiwen Feng | Rule-Assisted Attribute Embedding | ECML-PKDD2025 | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recommendation systems often overlook the rich attribute information embedded in property graphs, limiting their effectiveness. Existing graph convolutional network (GCN) models either ignore attributes or rely on simplistic <user, item, attribute> triples, failing to capture deeper semantic structures. We propose RAE (Rule- Assisted Approach for Attribute Embedding), a novel method that improves recommendations by mining semantic rules from property graphs to guide attribute embedding. RAE performs rule-based random walks to generate enriched attribute representations, which are integrated into GCNs. Experiments on real-world datasets (BlogCatalog, Flickr) show that RAE outperforms state-of-the-art baselines by 10.6% on average in Recall@20 and NDCG@20. RAE also demonstrates greater robustness to sparse data and missing attributes, highlighting the value of leveraging structured attribute information in recommendation tasks. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:51:03 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zhao",
"Sibo",
""
],
[
"Bewong",
"Michael",
""
],
[
"Kwashie",
"Selasi",
""
],
[
"Hu",
"Junwei",
""
],
[
"Feng",
"Zaiwen",
""
]
] |
2506.08315 | Christopher Jungkind | Christopher M. Jungkind, Brian C. Seymour, Andrew Laeuger, Yanbei Chen | Prospects for High-Frequency Gravitational-Wave Detection with GEO600 | 19 pages, 8 figures, submitted to PRD | null | null | null | gr-qc astro-ph.IM | http://creativecommons.org/licenses/by/4.0/ | Current ground-based interferometers are optimized for sensitivity from a few tens of Hz to about 1 kHz. While they are not currently utilized for GW detection, interferometric detectors also feature narrow bands of strong sensitivity at higher frequencies where the sideband fields created by a GW are resonantly amplified in the optical system. For certain interferometer configurations, small changes to system parameters allow the narrow band of high sensitivity to be scanned over a much larger range of frequencies, potentially enabling broadband detection at high frequencies. In this paper, we investigate whether simply modifying the detuning angle of the signal-recycling mirror of the GEO600 interferometer can make this experiment sensitive to GWs in the kilohertz frequency range. We compute the strain sensitivity for GEO600 across a frequency range from several kHz to tens of kHz for various detuning angles. We also show that LIGO cannot attain the same effect assuming that the optical components are not changed due to the narrow band response of the Fabry-Perot cavities. We then calculate the sensitivity of GEO600 to various proposed high-frequency GW sources and compare it to the sensitivity of other ground-based detectors. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:51:26 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Jungkind",
"Christopher M.",
""
],
[
"Seymour",
"Brian C.",
""
],
[
"Laeuger",
"Andrew",
""
],
[
"Chen",
"Yanbei",
""
]
] |
2506.08316 | Alan Amin | Alan N. Amin, Nate Gruver, Andrew Gordon Wilson | Why Masking Diffusion Works: Condition on the Jump Schedule for Improved Discrete Diffusion | Code available at: https://github.com/AlanNawzadAmin/SCUD | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Discrete diffusion models, like continuous diffusion models, generate high-quality samples by gradually undoing noise applied to datapoints with a Markov process. Gradual generation in theory comes with many conceptual benefits; for example, inductive biases can be incorporated into the noising Markov process, and access to improved sampling algorithms. In practice, however, the consistently best performing discrete diffusion model is, surprisingly, masking diffusion, which does not denoise gradually. Here we explain the superior performance of masking diffusion by noting that it makes use of a fundamental difference between continuous and discrete Markov processes: discrete Markov processes evolve by discontinuous jumps at a fixed rate and, unlike other discrete diffusion models, masking diffusion builds in the known distribution of jump times and only learns where to jump to. We show that we can similarly bake in the known distribution of jump times into any discrete diffusion model. The resulting models - schedule-conditioned discrete diffusion (SCUD) - generalize classical discrete diffusion and masking diffusion. By applying SCUD to models with noising processes that incorporate inductive biases on images, text, and protein data, we build models that outperform masking. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 00:58:25 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Amin",
"Alan N.",
""
],
[
"Gruver",
"Nate",
""
],
[
"Wilson",
"Andrew Gordon",
""
]
] |
2506.08317 | Jiewon Park | Christine Breiner, Jiewon Park | Quantitative Rigidity Using Colding's Monotonicity Formulas for Ricci Curvature | null | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | In \cite{Colding}, Colding proved monotonicity formulas for the Green function on manifolds with nonnegative Ricci curvature. Inspired by the sharp estimates relating the pinching of monotone quantities to the splitting function in \cite{cjn}, in this paper we investigate quantitative control obtained from pinching of Colding's monotone functionals. From the Green functions with poles at $(k+1)$-many independent points, $k$-splitting functions are constructed with regularity quantitatively controlled by the pinching. Moreover, the pinching at these independent points controls the distance to the nearest cone of the form $\mathbb{R}^k \times C(X)$. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:01:45 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Breiner",
"Christine",
""
],
[
"Park",
"Jiewon",
""
]
] |
2506.08318 | Jean Dolbeault | Jean Dolbeault and Rupert L. Frank and Jonte Weixler | Symmetry and symmetry breaking in interpolation inequalities for two-dimensional spinors -- Preliminary results | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | On the two-dimensional Euclidean space, we study a spinorial analogue of the Caffarelli-Kohn-Nirenberg inequality involving weighted gradient norms. This (SCKN) inequality is equivalent to a spinorial Gagliardo-Nirenberg type interpolation inequality on a cylinder as well as to an interpolation inequality involving Aharonov-Bohm magnetic fields, which was analyzed in a paper of 2020. We examine the symmetry properties of optimal functions by linearizing the associated functional around radial minimizers. We prove that the stability of the linearized problem is equivalent to the positivity of a $2\times2$ matrix-valued differential operator. We study the positivity issue via a combination of analytical arguments and numerical computations. In particular, our results provide numerical evidence that the region of symmetry breaking extends beyond what was previously known, while the threshold of the known symmetry region is linearly stable. Altogether, we obtain refined estimates of the phase transition between symmetry and symmetry breaking. Our results also put in evidence striking differences with the three-dimensional (SCKN) inequality that was recently investigated. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:04:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Dolbeault",
"Jean",
""
],
[
"Frank",
"Rupert L.",
""
],
[
"Weixler",
"Jonte",
""
]
] |
2506.08319 | Ao Jin | Ao Jin, Qinyi Wang, Sijie Wen, Ya Liu, Ganghui Shen, Panfeng Huang, Fan Zhang | DEKC: Data-Enable Control for Tethered Space Robot Deployment in the Presence of Uncertainty via Koopman Operator Theory | 12 pages | null | null | null | eess.SY cs.RO cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work focuses the deployment of tethered space robot in the presence of unknown uncertainty. A data-enable framework called DEKC which contains offline training part and online execution part is proposed to deploy tethered space robot in the presence of uncertainty. The main idea of this work is modeling the unknown uncertainty as a dynamical system, which enables high accuracy and convergence of capturing uncertainty. The core part of proposed framework is a proxy model of uncertainty, which is derived from data-driven Koopman theory and is separated with controller design. In the offline stage, the lifting functions associated with Koopman operator are parameterized with deep neural networks. Then by solving an optimization problem, the lifting functions are learned from sampling data. In the online execution stage, the proxy model cooperates the learned lifting functions obtained in the offline phase to capture the unknown uncertainty. Then the output of proxy model is compensated to the baseline controller such that the effect of uncertainty can be attenuated or even eliminated. Furthermore, considering some scenarios in which the performance of proxy model may weaken, a receding-horizon scheme is proposed to update the proxy model online. Finally, the extensive numerical simulations demonstrate the effectiveness of our proposed framework. The implementation of proposed DEKC framework is publicly available at https://github.com/NPU-RCIR/DEKC.git. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:04:29 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Jin",
"Ao",
""
],
[
"Wang",
"Qinyi",
""
],
[
"Wen",
"Sijie",
""
],
[
"Liu",
"Ya",
""
],
[
"Shen",
"Ganghui",
""
],
[
"Huang",
"Panfeng",
""
],
[
"Zhang",
"Fan",
""
]
] |
2506.08320 | Aditya Patwardhan | Vivek Vaidya, Aditya Patwardhan, Ashish Kundu | How Good LLM-Generated Password Policies Are? | 11 pages, 2 Tables, 9 figures, 3 Algorithms | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Generative AI technologies, particularly Large Language Models (LLMs), are rapidly being adopted across industry, academia, and government sectors, owing to their remarkable capabilities in natural language processing. However, despite their strengths, the inconsistency and unpredictability of LLM outputs present substantial challenges, especially in security-critical domains such as access control. One critical issue that emerges prominently is the consistency of LLM-generated responses, which is paramount for ensuring secure and reliable operations.
In this paper, we study the application of LLMs within the context of Cybersecurity Access Control Systems. Specifically, we investigate the consistency and accuracy of LLM-generated password policies, translating natural language prompts into executable pwquality.conf configuration files. Our experimental methodology adopts two distinct approaches: firstly, we utilize pre-trained LLMs to generate configuration files purely from natural language prompts without additional guidance. Secondly, we provide these models with official pwquality.conf documentation to serve as an informative baseline. We systematically assess the soundness, accuracy, and consistency of these AI-generated configurations. Our findings underscore significant challenges in the current generation of LLMs and contribute valuable insights into refining the deployment of LLMs in Access Control Systems. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:12:31 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Vaidya",
"Vivek",
""
],
[
"Patwardhan",
"Aditya",
""
],
[
"Kundu",
"Ashish",
""
]
] |
2506.08321 | Manooshree Patel | Manooshree Patel, Rayna Bhattacharyya, Thomas Lu, Arnav Mehta, Niels Voss, Narges Norouzi, Gireeja Ranade | LeanTutor: A Formally-Verified AI Tutor for Mathematical Proofs | null | null | null | null | cs.AI cs.HC cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present LeanTutor, a Large Language Model (LLM)-based tutoring system for math proofs. LeanTutor interacts with the student in natural language, formally verifies student-written math proofs in Lean, generates correct next steps, and provides the appropriate instructional guidance. LeanTutor is composed of three modules: (i) an autoformalizer/proof-checker, (ii) a next-step generator, and (iii) a natural language feedback generator. The first module faithfully autoformalizes student proofs into Lean and verifies proof accuracy via successful code compilation. If the proof has an error, the incorrect step is identified. The next-step generator module outputs a valid next Lean tactic for incorrect proofs via LLM-based candidate generation and proof search. The feedback generator module leverages Lean data to produce a pedagogically-motivated natural language hint for the student user. To evaluate our system, we introduce PeanoBench, a human-written dataset derived from the Natural Numbers Game, consisting of 371 Peano Arithmetic proofs, where each natural language proof step is paired with the corresponding logically equivalent tactic in Lean. The Autoformalizer correctly formalizes 57% of tactics in correct proofs and accurately identifies the incorrect step in 30% of incorrect proofs. In generating natural language hints for erroneous proofs, LeanTutor outperforms a simple baseline on accuracy and relevance metrics. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:12:36 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Patel",
"Manooshree",
""
],
[
"Bhattacharyya",
"Rayna",
""
],
[
"Lu",
"Thomas",
""
],
[
"Mehta",
"Arnav",
""
],
[
"Voss",
"Niels",
""
],
[
"Norouzi",
"Narges",
""
],
[
"Ranade",
"Gireeja",
""
]
] |
2506.08322 | Minghuan Zeng | Minghuan Zeng, Junjie Zeng, Ling Qin, Shiping Feng, Donghui Xu, Rui Wang | Spontaneously broken chiral symmetry in the interacting Kane-Mele model | 18 pages, 11 figures | null | null | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The essential properties of the half-filled interacting Kane-Mele model on a hexagon lattice is
studied using the slave rotor approach. It is shown clearly that a long-range charge-order state with
spontaneously broken chiral symmetry emerges in the weak and moderate interaction regimes, as
well as a presumed site-selected topological Mott insulator state in the stronger interaction regime
with U < UMott, where UMott is the critical interaction strength, and in the case of U > UMott,
the system is transited into the usual topological Mott state. This new charge-order state has
lower energy compared to the usual topological band insulator (TBI) state with chiral symmetry,
and thus is named as non-chiral TBI state. More specifically, in this non-chiral TBI state without
any long-range magnetic order, a long-range charge order with different electron occupation on
two sublattices appears in the absence of external sublattice field. The spontaneously broken chiral
symmetry gives rise to a special helical edge state, which has different spin accumulation on opposite
edges of the cylinder with periodic boundary condition in the zigzag direction, and thus leads to
a net spin current across the system. This net spin current would be further strengthened if the
nearest neighbor electron Coulomb interaction is taken into account as well, because it is favorable
for the long-range charge order with different electron occupation on sublattices. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:14:31 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zeng",
"Minghuan",
""
],
[
"Zeng",
"Junjie",
""
],
[
"Qin",
"Ling",
""
],
[
"Feng",
"Shiping",
""
],
[
"Xu",
"Donghui",
""
],
[
"Wang",
"Rui",
""
]
] |
2506.08323 | Hanyang Wang | Hanyang. Wang and Gary W Slater | Diffusive spreading of a polydisperse polymer solution in a channel | 9 figures, 9pages | null | null | null | cond-mat.soft physics.bio-ph | http://creativecommons.org/licenses/by/4.0/ | Long DNA molecules can be mapped by cutting them with restriction enzymes inside a narrow channel. Once cut, the individual fragments thus produced move away from each other due to diffusion and entropic effects. We investigate how long it takes for these fragments to travel distances large enough for an experimental device to distinguish them and (possibly) estimate their size. In essence, this is a single-file diffusion process in which molecules of different sizes and hence different diffusion coefficients spread out from an initially dense configuration. We use Monte Carlo methods to investigate this class of problems and define the time taken to reach the required final state as a first-passage \textit{spreading time}. Our results demonstrate that the stochastic nature of the diffusion process is as significant as the specifics of the molecular size distribution in determining the spreading time. We examine the relationship between the spreading time and the final space occupied by the fragments as a function of the experimental parameters and determine the fundamental length scale governing this process. We introduce a molecular sequence randomness parameter, $Z$, which is linearly correlated with the final spreading time. Finally, we show that the distribution function of spreading times follows a well-known form for first-passage time problems, and that its variance decreases linearly with the number of fragments. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:20:50 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Hanyang.",
""
],
[
"Slater",
"Gary W",
""
]
] |
2506.08325 | Marcos Matabuena | Marcos Matabuena, Rahul Ghosal, Pavlo Mozharovskyi, Oscar Hernan Madrid Padilla, and Jukka-Pekka Onnela | Model-Free Kernel Conformal Depth Measures Algorithm for Uncertainty Quantification in Regression Models in Separable Hilbert Spaces | arXiv admin note: substantial text overlap with arXiv:2405.13970 | null | null | null | stat.ML cs.LG stat.ME | http://creativecommons.org/licenses/by/4.0/ | Depth measures are powerful tools for defining level sets in emerging, non--standard, and complex random objects such as high-dimensional multivariate data, functional data, and random graphs. Despite their favorable theoretical properties, the integration of depth measures into regression modeling to provide prediction regions remains a largely underexplored area of research. To address this gap, we propose a novel, model-free uncertainty quantification algorithm based on conditional depth measures--specifically, conditional kernel mean embeddings and an integrated depth measure. These new algorithms can be used to define prediction and tolerance regions when predictors and responses are defined in separable Hilbert spaces. The use of kernel mean embeddings ensures faster convergence rates in prediction region estimation. To enhance the practical utility of the algorithms with finite samples, we also introduce a conformal prediction variant that provides marginal, non-asymptotic guarantees for the derived prediction regions. Additionally, we establish both conditional and unconditional consistency results, as well as fast convergence rates in certain homoscedastic settings. We evaluate the finite--sample performance of our model in extensive simulation studies involving various types of functional data and traditional Euclidean scenarios. Finally, we demonstrate the practical relevance of our approach through a digital health application related to physical activity, aiming to provide personalized recommendations | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:25:37 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Matabuena",
"Marcos",
""
],
[
"Ghosal",
"Rahul",
""
],
[
"Mozharovskyi",
"Pavlo",
""
],
[
"Padilla",
"Oscar Hernan Madrid",
""
],
[
"Onnela",
"Jukka-Pekka",
""
]
] |
2506.08326 | Xingbo Fu | Xingbo Fu, Zehong Wang, Zihan Chen, Jiazheng Li, Yaochen Zhu, Zhenyu Lei, Cong Shen, Yanfang Ye, Chuxu Zhang, Jundong Li | Graph Prompting for Graph Learning Models: Recent Advances and Future Directions | Accepted by KDD 2025 Tutorial/Survey Track | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph learning models have demonstrated great prowess in learning expressive representations from large-scale graph data in a wide variety of real-world scenarios. As a prevalent strategy for training powerful graph learning models, the "pre-training, adaptation" scheme first pre-trains graph learning models on unlabeled graph data in a self-supervised manner and then adapts them to specific downstream tasks. During the adaptation phase, graph prompting emerges as a promising approach that learns trainable prompts while keeping the pre-trained graph learning models unchanged. In this paper, we present a systematic review of recent advancements in graph prompting. First, we introduce representative graph pre-training methods that serve as the foundation step of graph prompting. Next, we review mainstream techniques in graph prompting and elaborate on how they design learnable prompts for graph prompting. Furthermore, we summarize the real-world applications of graph prompting from different domains. Finally, we discuss several open challenges in existing studies with promising future directions in this field. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:27:19 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Fu",
"Xingbo",
""
],
[
"Wang",
"Zehong",
""
],
[
"Chen",
"Zihan",
""
],
[
"Li",
"Jiazheng",
""
],
[
"Zhu",
"Yaochen",
""
],
[
"Lei",
"Zhenyu",
""
],
[
"Shen",
"Cong",
""
],
[
"Ye",
"Yanfang",
""
],
[
"Zhang",
"Chuxu",
""
],
[
"Li",
"Jundong",
""
]
] |
2506.08327 | Yuto Kase | Yuto Kase and Kai Ishibe and Ryoma Yasuda and Yudai Washida and Sakiko Hashimoto | Locating Tennis Ball Impact on the Racket in Real Time Using an Event Camera | 17 pages, 10 figures, 3 tables | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In racket sports, such as tennis, locating the ball's position at impact is important in clarifying player and equipment characteristics, thereby aiding in personalized equipment design. High-speed cameras are used to measure the impact location; however, their excessive memory consumption limits prolonged scene capture, and manual digitization for position detection is time-consuming and prone to human error. These limitations make it difficult to effectively capture the entire playing scene, hindering the ability to analyze the player's performance. We propose a method for locating the tennis ball impact on the racket in real time using an event camera. Event cameras efficiently measure brightness changes (called `events') with microsecond accuracy under high-speed motion while using lower memory consumption. These cameras enable users to continuously monitor their performance over extended periods. Our method consists of three identification steps: time range of swing, timing at impact, and contours of ball and racket. Conventional computer vision techniques are utilized along with an original event-based processing to detect the timing at impact (PATS: the amount of polarity asymmetry in time symmetry). The results of the experiments were within the permissible range for measuring tennis players' performance. Moreover, the computation time was sufficiently short for real-time applications. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:29:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Kase",
"Yuto",
""
],
[
"Ishibe",
"Kai",
""
],
[
"Yasuda",
"Ryoma",
""
],
[
"Washida",
"Yudai",
""
],
[
"Hashimoto",
"Sakiko",
""
]
] |
2506.08328 | Junoh Heo | Junoh Heo, Romain Boutelet, Chih-Li Sung | Diffusion Non-Additive Model for Multi-Fidelity Simulations with Tunable Precision | 35 pages including references and 20 pages supplementary | null | null | null | stat.ME | http://creativecommons.org/licenses/by/4.0/ | Computer simulations are indispensable for analyzing complex systems, yet high-fidelity models often incur prohibitive computational costs. Multi-fidelity frameworks address this challenge by combining inexpensive low-fidelity simulations with costly high-fidelity simulations to improve both accuracy and efficiency. However, certain scientific problems demand even more accurate results than the highest-fidelity simulations available, particularly when a tuning parameter controlling simulation accuracy is available, but the exact solution corresponding to a zero-valued parameter remains out of reach. In this paper, we introduce the Diffusion Non-Additive (DNA) model, inspired by generative diffusion models, which captures nonlinear dependencies across fidelity levels using Gaussian process priors and extrapolates to the exact solution. The DNA model: (i) accommodates complex, non-additive relationships across fidelity levels; (ii) employs a nonseparable covariance kernel to model interactions between the tuning parameter and input variables, improving both predictive performance and physical interpretability; and (iii) provides closed-form expressions for the posterior predictive mean and variance, allowing efficient inference and uncertainty quantification. The methodology is validated on a suite of numerical studies and real-world case studies. An R package implementing the proposed methodology is available to support practical applications. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:32:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Heo",
"Junoh",
""
],
[
"Boutelet",
"Romain",
""
],
[
"Sung",
"Chih-Li",
""
]
] |
2506.08330 | Kato Mivule | Kato Mivule, Kenneth Hopkinson | Distortion Search, A Web Search Privacy Heuristic | 11 pages, 11 figures, Future Technologies Conference (FTC) 2017 | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Search engines have vast technical capabilities to retain Internet search logs for each user and thus present major privacy vulnerabilities to both individuals and organizations in revealing user intent. Additionally, many of the web search privacy enhancing tools available today require that the user trusts a third party, which make confidentiality of user intent even more challenging. The user is left at the mercy of the third party without the control over his or her own privacy. In this article, we suggest a user-centric heuristic, Distortion Search, a web search query privacy methodology that works by the formation of obfuscated search queries via the permutation of query keyword categories, and by strategically applying k-anonymised web navigational clicks on URLs and Ads to generate a distorted user profile and thus providing specific user intent and query confidentiality. We provide empirical results via the evaluation of distorted web search queries in terms of retrieved search results and the resulting web ads from search engines. Preliminary experimental results indicate that web search query and specific user intent privacy might be achievable from the user side without the involvement of the search engine or other third parties. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:35:16 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Mivule",
"Kato",
""
],
[
"Hopkinson",
"Kenneth",
""
]
] |
2506.08331 | Pan Zhang | Pan Zhang | Correcting a noisy quantum computer using a quantum computer | 13 pages, 3 figures, comments are welcome | null | null | null | quant-ph cond-mat.dis-nn cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | Quantum computers require error correction to achieve universal quantum computing. However, current decoding of quantum error-correcting codes relies on classical computation, which is slower than quantum operations in superconducting qubits. This discrepancy makes the practical implementation of real-time quantum error correction challenging. In this work, we propose a decoding scheme that leverages the operations of the quantum circuit itself. Given a noisy quantum circuit $A$, we train a decoding quantum circuit $B$ using syndrome measurements to identify the logical operators needed to correct errors in circuit $A$. The trained quantum circuit $B$ can be deployed on quantum devices, such as superconducting qubits, to perform real-time decoding and error correction. Our approach is applicable to general quantum codes with multiple logical qubits and operates efficiently under various noise conditions, and the decoding speed matches the speed of the quantum circuits being corrected. We have conducted numerical experiments using surface codes up to distance 7 under circuit-level noise, demonstrating performance on par with the classical minimum-weight perfect matching algorithm. Interestingly, our method reveals that the traditionally classical task of decoding error-correcting codes can be accomplished without classical devices or measurements. This insight paves the way for the development of self-correcting quantum computers. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:35:21 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Zhang",
"Pan",
""
]
] |
2506.08332 | Zhiang Wang | Amur Ghose, Andrew B. Kahng, Sayak Kundu, and Zhiang Wang | ORFS-agent: Tool-Using Agents for Chip Design Optimization | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Machine learning has been widely used to optimize complex engineering workflows across numerous domains. In the context of integrated circuit design, modern flows (e.g., going from a register-transfer level netlist to physical layouts) involve extensive configuration via thousands of parameters, and small changes to these parameters can have large downstream impacts on desired outcomes - namely design performance, power, and area. Recent advances in Large Language Models (LLMs) offer new opportunities for learning and reasoning within such high-dimensional optimization tasks. In this work, we introduce ORFS-agent, an LLM-based iterative optimization agent that automates parameter tuning in an open-source hardware design flow. ORFS-agent adaptively explores parameter configurations, demonstrating clear improvements over standard Bayesian optimization approaches in terms of resource efficiency and final design metrics. Our empirical evaluations on two different technology nodes and a range of circuit benchmarks indicate that ORFS-agent can improve both routed wirelength and effective clock period by over 13%, all while using 40% fewer optimization iterations. Moreover, by following natural language objectives to trade off certain metrics for others, ORFS-agent demonstrates a flexible and interpretable framework for multi-objective optimization. Crucially, RFS-agent is modular and model-agnostic, and can be plugged in to any frontier LLM without any further fine-tuning. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:38:57 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Ghose",
"Amur",
""
],
[
"Kahng",
"Andrew B.",
""
],
[
"Kundu",
"Sayak",
""
],
[
"Wang",
"Zhiang",
""
]
] |
2506.08333 | Souvik Ray | Shankar Bhamidi, Amarjit Budhiraja, Souvik Ray | Large Deviations for Markovian Graphon Processes and Associated Dynamical Systems on Networks | 55 pages, no figures | null | null | null | math.PR math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider temporal models of rapidly changing Markovian networks modulated by time-evolving spatially dependent kernels that define rates for edge formation and dissolution. Alternatively, these can be viewed as Markovian networks with $O(1)$ jump rates viewed over a long time horizon. In the regimes we consider, the window averages of graphon valued processes over suitable time intervals are natural state descriptors for the system. Under appropriate conditions on the jump-rate kernels, we establish laws of large numbers and large deviation principles(LDP) for the graphon processes averaged over a suitable time window, both in the weak topology and with respect to the cut norm in the associated graphon space. Although the problem setting and analysis are more involved than for the well-studied static random network model, the variational problem associated with the rate function admits an explicit solution, yielding an equally tractable, though different, expression for the rate function, similar to the static case. Using these results, we then establish the LDP for node-valent dynamical systems driven by the underlying evolving network. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:41:25 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Bhamidi",
"Shankar",
""
],
[
"Budhiraja",
"Amarjit",
""
],
[
"Ray",
"Souvik",
""
]
] |
2506.08334 | Weikun Peng | Weikun Peng and Jun Lv and Cewu Lu and Manolis Savva | Generalizable Articulated Object Reconstruction from Casually Captured RGBD Videos | Project website can be found at https://3dlg-hcvc.github.io/video2articulation/ | null | null | null | cs.GR cs.CV | http://creativecommons.org/licenses/by/4.0/ | Articulated objects are prevalent in daily life. Understanding their kinematic structure and reconstructing them have numerous applications in embodied AI and robotics. However, current methods require carefully captured data for training or inference, preventing practical, scalable, and generalizable reconstruction of articulated objects. We focus on reconstruction of an articulated object from a casually captured RGBD video shot with a hand-held camera. A casually captured video of an interaction with an articulated object is easy to acquire at scale using smartphones. However, this setting is quite challenging, as the object and camera move simultaneously and there are significant occlusions as the person interacts with the object. To tackle these challenges, we introduce a coarse-to-fine framework that infers joint parameters and segments movable parts of the object from a dynamic RGBD video. To evaluate our method under this new setting, we build a 20$\times$ larger synthetic dataset of 784 videos containing 284 objects across 11 categories. We compare our approach with existing methods that also take video as input. Experiments show that our method can reconstruct synthetic and real articulated objects across different categories from dynamic RGBD videos, outperforming existing methods significantly. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:41:46 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Peng",
"Weikun",
""
],
[
"Lv",
"Jun",
""
],
[
"Lu",
"Cewu",
""
],
[
"Savva",
"Manolis",
""
]
] |
2506.08335 | Xuan Luo | Xuan Luo, Hua-Xing Chen, Er-Liang Cui, Hui-Min Yang, Dan Zhou, Zhi-Yong Zhou | Radiative decays of $P$-wave charmed baryons in the $SU(3)$ flavor $\bf6_F$ representation | null | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We perform a comprehensive investigation of the radiative decays of $P$-wave charmed baryons in the $SU(3)$ flavor $\mathbf{6}_F$ representation, employing the light-cone QCD sum rule approach within the framework of heavy quark effective theory. We analyze their electromagnetic transitions into ground-state charmed baryons via photon emission. When combined with the mass spectra and strong decay properties previously studied in Ref.~\cite{Yang:2021lce}, our results constitute a systematic and complete QCD sum rule analysis of the $P$-wave singly charmed baryons within the framework of heavy quark effective theory. As summarized in Table~\ref{tab:result}, several excited charmed baryons are found to possess suppressed strong decay widths, thereby rendering their radiative decay channels particularly significant for experimental identification and theoretical understanding. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:43:38 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Luo",
"Xuan",
""
],
[
"Chen",
"Hua-Xing",
""
],
[
"Cui",
"Er-Liang",
""
],
[
"Yang",
"Hui-Min",
""
],
[
"Zhou",
"Dan",
""
],
[
"Zhou",
"Zhi-Yong",
""
]
] |
2506.08337 | Juhyeok Choi | Juhyeok Choi, Chenglin Fan | A Simple Analysis of Discretization Error in Diffusion Models | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Diffusion models, formulated as discretizations of stochastic differential equations (SDEs), achieve state-of-the-art generative performance. However, existing analyses of their discretization error often rely on complex probabilistic tools. In this work, we present a simplified theoretical framework for analyzing the Euler--Maruyama discretization of variance-preserving SDEs (VP-SDEs) in Denoising Diffusion Probabilistic Models (DDPMs), where $ T $ denotes the number of denoising steps in the diffusion process. Our approach leverages Gr\"onwall's inequality to derive a convergence rate of $ \mathcal{O}(1/T^{1/2}) $ under Lipschitz assumptions, significantly streamlining prior proofs. Furthermore, we demonstrate that the Gaussian noise in the discretization can be replaced by a discrete random variable (e.g., Rademacher or uniform noise) without sacrificing convergence guarantees-an insight with practical implications for efficient sampling. Experiments validate our theory, showing that (1) the error scales as predicted, (2) discrete noise achieves comparable sample quality to Gaussian noise, and (3) incorrect noise scaling degrades performance. By unifying simplified analysis and discrete noise substitution, our work bridges theoretical rigor with practical efficiency in diffusion-based generative modeling. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:46:42 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Choi",
"Juhyeok",
""
],
[
"Fan",
"Chenglin",
""
]
] |
2506.08338 | Hirokazu Iwasawa | Ryoichi Asashiba, Reiji Kozuma, and Hirokazu Iwasawa | midr: Learning from Black-Box Models by Maximum Interpretation Decomposition | 20 pages, 10 figures | null | null | null | stat.ME cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The use of appropriate methods of Interpretable Machine Learning (IML) and eXplainable Artificial Intelligence (XAI) is essential for adopting black-box predictive models in fields where model and prediction explainability is required. As a novel tool for interpreting black-box models, we introduce the R package midr, which implements Maximum Interpretation Decomposition (MID). MID is a functional decomposition approach that derives a low-order additive representation of a black-box model by minimizing the squared error between the model's prediction function and this additive representation. midr enables learning from black-box models by constructing a global surrogate model with advanced analytical capabilities. After reviewing related work and the theoretical foundation of MID, we demonstrate the package's usage and discuss some of its key features. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:46:49 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Asashiba",
"Ryoichi",
""
],
[
"Kozuma",
"Reiji",
""
],
[
"Iwasawa",
"Hirokazu",
""
]
] |
2506.08339 | Ling Tang | Ling Tang, Weiyi Xia, Gayatri Viswanathan, Ernesto Soto, Kirill Kovnir, and Cai-Zhuang Wang | Developing a Neural Network Machine Learning Interatomic Potential for Molecular Dynamics Simulations of La-Si-P Systems | null | null | null | null | cond-mat.mtrl-sci cond-mat.dis-nn physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | While molecular dynamics (MD) is a very useful computational method for atomistic simulations, modeling the interatomic interactions for reliable MD simulations of real materials has been a long-standing challenge. In 2007, Behler and Perrinello first proposed and demonstrated an artificial neural network machine learning (ANN-ML) scheme, opening a new paradigm for developing accurate and efficient interatomic potentials for reliable MD simulation studies of the thermodynamics and kinetics of materials. In this paper, we show that an accurate and transferable ANN-ML interatomic potential can be developed for MD simulations of La-Si-P system. The crucial role of training data in the ML potential development is discussed. The developed ANN-ML potential accurately describes not only the energy versus volume curves for all the known elemental, binary, and ternary crystalline structures in La-Si-P system, but also the structures of La-Si-P liquids with various compositions. Using the developed ANN-ML potential, the melting temperatures of several crystalline phases in La-Si-P system are predicted by the coexistence of solid-liquid phases from MD simulations. While the ANN-ML model systematically underestimates the melting temperatures of these phases, the overall trend agrees with experiment. The developed ANN-ML potential is also applied to study the nucleation and growth of LaP as a function of different relative concentrations of Si and P in the La-Si-P liquid, and the obtained results are consistent with experimental observations. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:46:56 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Tang",
"Ling",
""
],
[
"Xia",
"Weiyi",
""
],
[
"Viswanathan",
"Gayatri",
""
],
[
"Soto",
"Ernesto",
""
],
[
"Kovnir",
"Kirill",
""
],
[
"Wang",
"Cai-Zhuang",
""
]
] |
2506.08340 | Emo Todorov | Emo Todorov | Dynamical System Optimization | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop an optimization framework centered around a core idea: once a (parametric) policy is specified, control authority is transferred to the policy, resulting in an autonomous dynamical system. Thus we should be able to optimize policy parameters without further reference to controls or actions, and without directly using the machinery of approximate Dynamic Programming and Reinforcement Learning. Here we derive simpler algorithms at the autonomous system level, and show that they compute the same quantities as policy gradients and Hessians, natural gradients, proximal methods. Analogs to approximate policy iteration and off-policy learning are also available. Since policy parameters and other system parameters are treated uniformly, the same algorithms apply to behavioral cloning, mechanism design, system identification, learning of state estimators. Tuning of generative AI models is not only possible, but is conceptually closer to the present framework than to Reinforcement Learning. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:50:38 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Todorov",
"Emo",
""
]
] |
2506.08341 | Yongqiang Liu | Yong-qiang Liu | A discussion on the symmetry of relativistic Vlasov gas and its accretion in Kerr-Newman black hole | null | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the kinetic properties of collisionless Vlasov gas in Kerr-Newman spacetime, focusing on how spacetime symmetries constrain the distribution functions. By establishing the general relationship between distribution functions and constants of motion (mass $m$, energy $E$, angular momentum $L_z$, Carter constant $L$) along with $L$'s conjugate configuration variable $Q^{3}$, we derive the analytical framework for particle dynamics in stationary axisymmetric spacetimes. Within the Locally Non-Rotating Frame , we compute key observables -- particle number density, energy density, principal pressures, and accretion rates -- resolving coordinate divergences near the event horizon through geometric rescaling. For a plasma obeying the J\"uttner distribution, asymptotic expressions at spatial infinity are explicitly calculated. Numerical analyses reveal: (1) mass and energy accretion rates exhibit nearly identical parametric dependencies on black hole parameters; (2) extremal Kerr-Newman black holes show monotonic enhancement in mass, energy, and angular momentum accretion rates with increasing $a$, contrasting with Kerr black hole behavior. These results provide new insights into accretion mechanisms in extreme spacetime geometries. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:53:12 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Liu",
"Yong-qiang",
""
]
] |
2506.08342 | Tong Li | Tong Li, Michael A. Schmidt, Man Yuan | Reexamining the search for light ALPs at flavor and forward accelerator experiments | 20 pages, 3 tables, 3 figures | null | null | CPPC-2025-05 | hep-ph | http://creativecommons.org/licenses/by/4.0/ | The axion-like particle (ALP) is a well-motivated extension of the Standard Model. In this work, we revisit the sensitivity of forward accelerator experiments to light long-lived ALPs, and analyze flavor constraints. Our analysis incorporates recent measurements of the rare decays $B\to K + X$ and $K\to \pi +X$, which place stringent bounds on quark flavor violation of a light ALP. We consider the complete list of hadronic modes in the calculation of the ALP decay rate and exclusive production channels based on recent improvements. The analysis includes the discussion of tree-level quark flavor-violating couplings in addition to a universal flavor-conserving ALP coupling to fermions and the electroweak ALP couplings. Our results demonstrate the complementarity of heavy meson decays and forward accelerator facilities in probing light ALPs. The interplay between two ALP couplings is also investigated. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:53:45 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Li",
"Tong",
""
],
[
"Schmidt",
"Michael A.",
""
],
[
"Yuan",
"Man",
""
]
] |
2506.08343 | Chenlong Wang | Chenlong Wang, Yuanning Feng, Dongping Chen, Zhaoyang Chu, Ranjay Krishna, Tianyi Zhou | Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiency | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Recent advances in large reasoning models have enabled complex, step-by-step reasoning but often introduce significant overthinking, resulting in verbose and redundant outputs that hinder efficiency. In this study, we examine whether explicit self-reflection, signaled by tokens such as "Wait" and "Hmm", is necessary for advanced reasoning. We propose NoWait, a simple yet effective approach that disables explicit self-reflection by suppressing these tokens during inference. Extensive experiments on ten benchmarks across textual, visual, and video reasoning tasks show that NoWait reduces chain-of-thought trajectory length by up to 27%-51% in five R1-style model series, without compromising model utility. NoWait thus offers a plug-and-play solution for efficient and utility-preserving multimodal reasoning. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:54:04 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Wang",
"Chenlong",
""
],
[
"Feng",
"Yuanning",
""
],
[
"Chen",
"Dongping",
""
],
[
"Chu",
"Zhaoyang",
""
],
[
"Krishna",
"Ranjay",
""
],
[
"Zhou",
"Tianyi",
""
]
] |
2506.08344 | Neset Unver Akmandor | Ne\c{s}et \"Unver Akmandor, Sarvesh Prajapati, Mark Zolotas, and Ta\c{s}k{\i}n Pad{\i}r | Re4MPC: Reactive Nonlinear MPC for Multi-model Motion Planning via Deep Reinforcement Learning | Accepted to the 2025 IEEE International Conference on Automation Science and Engineering (CASE) | null | null | null | cs.RO cs.AI cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional motion planning methods for robots with many degrees-of-freedom, such as mobile manipulators, are often computationally prohibitive for real-world settings. In this paper, we propose a novel multi-model motion planning pipeline, termed Re4MPC, which computes trajectories using Nonlinear Model Predictive Control (NMPC). Re4MPC generates trajectories in a computationally efficient manner by reactively selecting the model, cost, and constraints of the NMPC problem depending on the complexity of the task and robot state. The policy for this reactive decision-making is learned via a Deep Reinforcement Learning (DRL) framework. We introduce a mathematical formulation to integrate NMPC into this DRL framework. To validate our methodology and design choices, we evaluate DRL training and test outcomes in a physics-based simulation involving a mobile manipulator. Experimental results demonstrate that Re4MPC is more computationally efficient and achieves higher success rates in reaching end-effector goals than the NMPC baseline, which computes whole-body trajectories without our learning mechanism. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:58:32 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Akmandor",
"Neşet Ünver",
""
],
[
"Prajapati",
"Sarvesh",
""
],
[
"Zolotas",
"Mark",
""
],
[
"Padır",
"Taşkın",
""
]
] |
2506.08345 | Sa\'ul Blanco | Sa\'ul A. Blanco and Charles Buehrle | On the spectra of prefix-reversal graphs | 18 pages | null | null | null | math.CO cs.DM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we study spectral properties of prefix-reversal graphs. These graphs are obtained by connecting two elements of $C_m\wr S_n$ via prefix reversals. If $m=1,2$, the corresponding prefix-reversal graphs are the classic pancake and burnt pancake graphs. If $m>2$, then one can consider the directed and undirected versions of these graphs. We prove that the spectrum of the undirected prefix-reversal graph $\mathbb{P}_m(n)$ contains all even integers in the interval $[0,2n]\setminus\{2\lfloor n/2\rfloor\}$ and if $m\equiv0\pmod4$, we then show that the spectrum contains all even integers in $[0,2n]$. In the directed case, we show that the spectrum of the directed prefix-reversal graph $P(m,n)$ contains all integers in the interval $[0,n]\setminus\{\lfloor n/2\rfloor\}$. As a consequence, we show that in either case, the prefix-reversal graphs have a small spectral gap. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 01:59:13 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Blanco",
"Saúl A.",
""
],
[
"Buehrle",
"Charles",
""
]
] |
2506.08346 | Wenhan Yao | Wenhan Yao, Fen Xiao, Xiarun Chen, Jia Liu, YongQiang He, Weiping Wen | SPBA: Utilizing Speech Large Language Model for Backdoor Attacks on Speech Classification Models | Accepted by IJCNN 2025 | null | null | null | cs.SD cs.AI cs.CL eess.AS | http://creativecommons.org/publicdomain/zero/1.0/ | Deep speech classification tasks, including keyword spotting and speaker verification, are vital in speech-based human-computer interaction. Recently, the security of these technologies has been revealed to be susceptible to backdoor attacks. Specifically, attackers use noisy disruption triggers and speech element triggers to produce poisoned speech samples that train models to become vulnerable. However, these methods typically create only a limited number of backdoors due to the inherent constraints of the trigger function. In this paper, we propose that speech backdoor attacks can strategically focus on speech elements such as timbre and emotion, leveraging the Speech Large Language Model (SLLM) to generate diverse triggers. Increasing the number of triggers may disproportionately elevate the poisoning rate, resulting in higher attack costs and a lower success rate per trigger. We introduce the Multiple Gradient Descent Algorithm (MGDA) as a mitigation strategy to address this challenge. The proposed attack is called the Speech Prompt Backdoor Attack (SPBA). Building on this foundation, we conducted attack experiments on two speech classification tasks, demonstrating that SPBA shows significant trigger effectiveness and achieves exceptional performance in attack metrics. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 02:01:00 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Yao",
"Wenhan",
""
],
[
"Xiao",
"Fen",
""
],
[
"Chen",
"Xiarun",
""
],
[
"Liu",
"Jia",
""
],
[
"He",
"YongQiang",
""
],
[
"Wen",
"Weiping",
""
]
] |
2506.08347 | Yinan Huang | Yinan Huang, Haoteng Ying, Eli Chien, Rongzhe Wei, Pan Li | Differentially Private Relational Learning with Entity-level Privacy Guarantees | null | null | null | null | cs.LG cs.CR | http://creativecommons.org/licenses/by/4.0/ | Learning with relational and network-structured data is increasingly vital in sensitive domains where protecting the privacy of individual entities is paramount. Differential Privacy (DP) offers a principled approach for quantifying privacy risks, with DP-SGD emerging as a standard mechanism for private model training. However, directly applying DP-SGD to relational learning is challenging due to two key factors: (i) entities often participate in multiple relations, resulting in high and difficult-to-control sensitivity; and (ii) relational learning typically involves multi-stage, potentially coupled (interdependent) sampling procedures that make standard privacy amplification analyses inapplicable. This work presents a principled framework for relational learning with formal entity-level DP guarantees. We provide a rigorous sensitivity analysis and introduce an adaptive gradient clipping scheme that modulates clipping thresholds based on entity occurrence frequency. We also extend the privacy amplification results to a tractable subclass of coupled sampling, where the dependence arises only through sample sizes. These contributions lead to a tailored DP-SGD variant for relational data with provable privacy guarantees. Experiments on fine-tuning text encoders over text-attributed network-structured relational data demonstrate the strong utility-privacy trade-offs of our approach. Our code is available at https://github.com/Graph-COM/Node_DP. | [
{
"version": "v1",
"created": "Tue, 10 Jun 2025 02:03:43 GMT"
}
] | 2025-06-11T00:00:00 | [
[
"Huang",
"Yinan",
""
],
[
"Ying",
"Haoteng",
""
],
[
"Chien",
"Eli",
""
],
[
"Wei",
"Rongzhe",
""
],
[
"Li",
"Pan",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.