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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.14038 | Nabil Omi | Nabil Omi, Siddhartha Sen, Ali Farhadi | Load Balancing Mixture of Experts with Similarity Preserving Routers | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Sparse Mixture of Experts (MoE) models offer a scalable and efficient architecture for training large neural networks by activating only a subset of parameters ("experts") for each input. A learned router computes a distribution over these experts, and assigns input tokens to a small subset. However, without auxiliary balancing mechanisms, routers often converge to using only a few experts, severely limiting model capacity and degrading performance. Most current load balancing mechanisms encourage a distribution over experts that resembles a roughly uniform distribution of experts per token. During training, this can result in inconsistent routing behavior, resulting in the model spending its capacity to learn redundant knowledge. We address this by introducing a novel load balancing loss that preserves token-wise relational structure, encouraging consistent expert choices for similar inputs during training. Our experimental results show that applying our loss to the router results in 36% faster convergence and lower redundancy compared to a popular load balancing loss. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:22:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Omi",
"Nabil",
""
],
[
"Sen",
"Siddhartha",
""
],
[
"Farhadi",
"Ali",
""
]
] |
2506.14039 | Ioannis Mandralis | Ioannis Mandralis, Richard M. Murray, Morteza Gharib | Quadrotor Morpho-Transition: Learning vs Model-Based Control Strategies | null | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quadrotor Morpho-Transition, or the act of transitioning from air to ground through mid-air transformation, involves complex aerodynamic interactions and a need to operate near actuator saturation, complicating controller design. In recent work, morpho-transition has been studied from a model-based control perspective, but these approaches remain limited due to unmodeled dynamics and the requirement for planning through contacts. Here, we train an end-to-end Reinforcement Learning (RL) controller to learn a morpho-transition policy and demonstrate successful transfer to hardware. We find that the RL control policy achieves agile landing, but only transfers to hardware if motor dynamics and observation delays are taken into account. On the other hand, a baseline MPC controller transfers out-of-the-box without knowledge of the actuator dynamics and delays, at the cost of reduced recovery from disturbances in the event of unknown actuator failures. Our work opens the way for more robust control of agile in-flight quadrotor maneuvers that require mid-air transformation. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:23:28 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mandralis",
"Ioannis",
""
],
[
"Murray",
"Richard M.",
""
],
[
"Gharib",
"Morteza",
""
]
] |
2506.14040 | Md. Nazmus Sakib | Md Nazmus Sakib | An Interdisciplinary Review of Commonsense Reasoning and Intent Detection | null | null | null | null | cs.CL cs.HC | http://creativecommons.org/licenses/by/4.0/ | This review explores recent advances in commonsense reasoning and intent detection, two key challenges in natural language understanding. We analyze 28 papers from ACL, EMNLP, and CHI (2020-2025), organizing them by methodology and application. Commonsense reasoning is reviewed across zero-shot learning, cultural adaptation, structured evaluation, and interactive contexts. Intent detection is examined through open-set models, generative formulations, clustering, and human-centered systems. By bridging insights from NLP and HCI, we highlight emerging trends toward more adaptive, multilingual, and context-aware models, and identify key gaps in grounding, generalization, and benchmark design. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:26:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sakib",
"Md Nazmus",
""
]
] |
2506.14041 | David Kaiser | Alexandra P. Klipfel, Peter Fisher, and David I. Kaiser | Hawking Radiation Signatures from Primordial Black Holes Transiting the Inner Solar System: Prospects for Detection | 23pp, 17 figures | null | null | Preprint MIT-CTP/5878 | astro-ph.CO astro-ph.HE hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Primordial black holes (PBHs) arise from the collapse of density perturbations in the early universe and serve as a dark matter (DM) candidate and a probe of fundamental physics. There remains an unconstrained ``asteroid-mass'' window where PBHs of masses $10^{17} {\rm g} \lesssim M \lesssim 10^{23} {\rm g}$ could comprise up to $100\%$ of the dark matter. Current $e^{\pm}$ Hawking radiation constraints on the DM fraction of PBHs are set by comparing observed spatial- and time-integrated cosmic ray flux measurements with predicted Hawking emission fluxes from the galactic DM halo. These constraints depend on cosmic ray production and propagation models, the galactic DM density distribution, and the PBH mass function. We propose to mitigate these model dependencies by developing a new local, time-dependent Hawking radiation signature to detect low-mass PBHs transiting through the inner Solar System. We calculate transit rates for PBHs that form with initial masses $M \lesssim 5\times10^{17}\text{g}$. We then simulate time-dependent positron signals from individual PBH flybys as measured by the Alpha Magnetic Spectrometer (AMS) experiment in low-Earth orbit. We find that AMS is sensitive to PBHs with masses $M\lesssim 2\times10^{14} \, {\rm g}$ due to its lower energy threshold of $500 \, {\rm MeV}$. We demonstrate that a dataset of daily positron fluxes over the energy range $5-500 \, {\rm MeV}$, with similar levels of precision to the existing AMS data, would enable detection of PBHs drawn from present-day distributions that peak within the asteroid-mass window. Our simulations yield ${\cal O} (1)$ detectable PBH transits per year across wide regions of parameter space, which may be used to constrain PBH mass functions. This technique could be extended to detect $\gamma$-ray and X-ray Hawking emission to probe further into the asteroid-mass window. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:31:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Klipfel",
"Alexandra P.",
""
],
[
"Fisher",
"Peter",
""
],
[
"Kaiser",
"David I.",
""
]
] |
2506.14042 | Bernardo Anibal Subercaseaux Roa | Bernardo Subercaseaux | Asymptotically Smaller Encodings for Graph Problems and Scheduling | null | null | null | null | cs.LO cs.AI cs.DS | http://creativecommons.org/licenses/by/4.0/ | We show how several graph problems (e.g., vertex-cover, independent-set, $k$-coloring) can be encoded into CNF using only $O(|V|^2 / \lg |V|)$ many clauses, as opposed to the $\Omega(|V|^2)$ constraints used by standard encodings. This somewhat surprising result is a simple consequence of a result of Erd\H{o}s, Chung, and Spencer (1983) about biclique coverings of graphs, and opens theoretical avenues to understand the success of "Bounded Variable Addition'' (Manthey, Heule, and Biere, 2012) as a preprocessing tool. Finally, we show a novel encoding for independent sets in some dense interval graphs using only $O(|V| \lg |V|)$ clauses (the direct encoding uses $\Omega(|V|^2)$), which we have successfully applied to a string-compression encoding posed by Bannai et al. (2022). As a direct byproduct, we obtain a reduction in the encoding size of a scheduling problem posed by Mayank and Modal (2020) from $O(NMT^2)$ to $O(NMT + M T^2 \lg T)$, where $N$ is the number of tasks, $T$ the total timespan, and $M$ the number of machines. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:31:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Subercaseaux",
"Bernardo",
""
]
] |
2506.14043 | Nathaniel Hermann | Nathaniel G. Hermann and M. Shane Hutson | Anomalous diffusion for mass transport phenomena I: Analytic solutions to time fractional diffusion | 23 pages, 11 figures | null | null | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | Mass transport problems are ubiquitous in diverse fields of physics and engineering. With the development of fractional calculus, many have taken to studying problems of fractional mass transport either through numerical simulations or through complex mathematical structures (e.g. Fox-H functions). Here, we present a set of analytic solutions to common time fractional diffusion problems, written in terms of Mittag-Leffler and M-Wright functions, as well as generalized fractional error and complementary error functions derived within. We additionally show how time fractional diffusion is a generalization of a two-parameter stretched-time fractional diffusion process. Finally we present a procedure to take canonical solutions to mass transport problems with Fickian diffusion and extend these to systems with anomalous diffusion. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:33:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hermann",
"Nathaniel G.",
""
],
[
"Hutson",
"M. Shane",
""
]
] |
2506.14044 | Matheus Rolim Sales | Matheus Rolim Sales, Leonardo Costa de Souza, Daniel Borin, Michele Mugnaine, Jos\'e Danilo Szezech Jr., Ricardo Luiz Viana, Iber\^e Luiz Caldas, Edson Denis Leonel, Chris G. Antonopoulos | pynamicalsys: A Python toolkit for the analysis of dynamical systems | null | null | null | null | nlin.CD physics.comp-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Since Lorenz's seminal work on a simplified weather model, the numerical analysis of nonlinear dynamical systems has become one of the main subjects of research in physics. Despite of that, there remains a need for accessible, efficient, and easy-to-use computational tools to study such systems. In this paper, we introduce pynamicalsys, a simple yet powerful open-source Python module for the analysis of nonlinear dynamical systems. In particular, pynamicalsys implements tools for trajectory simulation, bifurcation diagrams, Lyapunov exponents and several others chaotic indicators, period orbit detection and their manifolds, as well as escape and basins analysis. It also includes many built-in models and the use of custom models is straighforward. We demonstrate the capabilities of pynamicalsys through a series of examples that reproduces well-known results in the literature while developing the mathematical analysis at the same time. We also provide the Jupyter notebook containing all the code used in this paper, including performance benchmarks. pynamicalsys is freely available via the Python Package Index (PyPI) and is indented to support both research and teaching in nonlinear dynamics. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:35:48 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sales",
"Matheus Rolim",
""
],
[
"de Souza",
"Leonardo Costa",
""
],
[
"Borin",
"Daniel",
""
],
[
"Mugnaine",
"Michele",
""
],
[
"Szezech",
"José Danilo",
"Jr."
],
[
"Viana",
"Ricardo Luiz",
""
],
[
"Caldas",
"Iberê Luiz",
""
],
[
"Leonel",
"Edson Denis",
""
],
[
"Antonopoulos",
"Chris G.",
""
]
] |
2506.14045 | Martin Klissarov | Martin Klissarov, Akhil Bagaria, Ziyan Luo, George Konidaris, Doina Precup, Marlos C. Machado | Discovering Temporal Structure: An Overview of Hierarchical Reinforcement Learning | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Developing agents capable of exploring, planning and learning in complex open-ended environments is a grand challenge in artificial intelligence (AI). Hierarchical reinforcement learning (HRL) offers a promising solution to this challenge by discovering and exploiting the temporal structure within a stream of experience. The strong appeal of the HRL framework has led to a rich and diverse body of literature attempting to discover a useful structure. However, it is still not clear how one might define what constitutes good structure in the first place, or the kind of problems in which identifying it may be helpful. This work aims to identify the benefits of HRL from the perspective of the fundamental challenges in decision-making, as well as highlight its impact on the performance trade-offs of AI agents. Through these benefits, we then cover the families of methods that discover temporal structure in HRL, ranging from learning directly from online experience to offline datasets, to leveraging large language models (LLMs). Finally, we highlight the challenges of temporal structure discovery and the domains that are particularly well-suited for such endeavours. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:36:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Klissarov",
"Martin",
""
],
[
"Bagaria",
"Akhil",
""
],
[
"Luo",
"Ziyan",
""
],
[
"Konidaris",
"George",
""
],
[
"Precup",
"Doina",
""
],
[
"Machado",
"Marlos C.",
""
]
] |
2506.14046 | David Kogan | David Kogan, Max Schumacher, Sam Nguyen, Masanori Suzuki, Melissa Smith, Chloe Sophia Bellows, Jared Bernstein | Ace-CEFR -- A Dataset for Automated Evaluation of the Linguistic Difficulty of Conversational Texts for LLM Applications | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | There is an unmet need to evaluate the language difficulty of short, conversational passages of text, particularly for training and filtering Large Language Models (LLMs). We introduce Ace-CEFR, a dataset of English conversational text passages expert-annotated with their corresponding level of text difficulty. We experiment with several models on Ace-CEFR, including Transformer-based models and LLMs. We show that models trained on Ace-CEFR can measure text difficulty more accurately than human experts and have latency appropriate to production environments. Finally, we release the Ace-CEFR dataset to the public for research and development. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:40:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kogan",
"David",
""
],
[
"Schumacher",
"Max",
""
],
[
"Nguyen",
"Sam",
""
],
[
"Suzuki",
"Masanori",
""
],
[
"Smith",
"Melissa",
""
],
[
"Bellows",
"Chloe Sophia",
""
],
[
"Bernstein",
"Jared",
""
]
] |
2506.14047 | Igor Dolinka | Igor Dolinka, Ganna Kudryavtseva | On Universal Inverse Monoids with the $F$-Inverse Property | 20 pages, 1 figure | null | null | null | math.GR | http://creativecommons.org/licenses/by/4.0/ | An inverse monoid $M$ is called $F$-inverse if each $\sigma$-class of $M$, where $\sigma$ is the minimum group congruence of $M$, has a maximum element with respect to the natural order of $M$. Upon Kinyon's observation made in 2018 that $F$-inverse monoids form a variety in the enriched signature involving an additional unary operation mapping each element of $M$ to the maximum element of its $\sigma$-class, a separate theory of $F$-inverse monoids is developing in recent years, greatly aided by a recent solution to the long-standing problem of Henckell and Rhodes on finite $F$-inverse covers of finite inverse monoids. In particular, a description of the universal object $F(G,X)$ in the class of $X$-generated $F$-inverse monoids with a prescribed maximum group image was provided recently. In this paper we introduce yet another object $M_F(G,X)$, the one that is universal for inverse monoids (so, not in the enriched signature) that are $X$-generated with maximum group image a quotient of $G$ and which happen to be $F$-inverse. We obtain presentations and geometric models based on Cayley graphs of groups for such inverse monoids. As an application, we are able to characterise all one-relator special inverse monoids $\mathrm{Inv}\langle{X}\,|\,{w=1}\rangle$ with a cyclically reduced relator word $w$ enjoying the $F$-inverse property. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 22:49:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Dolinka",
"Igor",
""
],
[
"Kudryavtseva",
"Ganna",
""
]
] |
2506.14049 | Mitja Fridman | Saurya Das, Mitja Fridman, Gaetano Lambiase | Testing the Quantum Equivalence Principle with Gravitational Waves | 16 pages, no figures | JHEAp \textbf{48}, 100413 (2025) | 10.1016/j.jheap.2025.100413 | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study modifications of gravitational wave observables, such as the wave amplitude and frequency, which follow from the quantum equivalence principle, and are expressed in terms of the inertial, gravitational and rest masses of the LIGO/Virgo mirrors. We provide bounds on the violations of the quantum equivalence principle by comparing the results with the most resolved gravitational wave events observed by the LIGO/Virgo collaboration. The formalism is equally applicable to other future ground and space-based gravitational wave detectors. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:04:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Das",
"Saurya",
""
],
[
"Fridman",
"Mitja",
""
],
[
"Lambiase",
"Gaetano",
""
]
] |
2506.14050 | Konstantinos Varvarezos | Hyunki Min, Konstantinos Varvarezos | Bordered contact invariants and half Giroux torsion | 17 pages; 6 figures | null | null | null | math.GT math.SG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that there exist infinitely many closed contact 3-manifolds containing half Giroux torsion along a separating torus whose contact invariants do not vanish. This provides counterexamples to Ghiggini's conjecture and suggests that separating half Giroux torsion may not obstruct symplectic fillability. The main tools are the bordered contact invariants recently developed by the authors and the innermost contact structures on knot complements. We also show that there exists a closed contact 3-manifold containing convex torsion along a separating torus with non-vanishing contact invariant, which implies that $2\pi$ is the minimal amount of twisting necessary to ensure vanishing of the contact invariant. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:16:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Min",
"Hyunki",
""
],
[
"Varvarezos",
"Konstantinos",
""
]
] |
2506.14051 | Jiyuan Tan | Jiyuan Tan, Jose Blanchet, Vasilis Syrgkanis | Estimation of Treatment Effects in Extreme and Unobserved Data | null | null | null | null | stat.ML cs.LG math.ST stat.ME stat.TH | http://creativecommons.org/licenses/by/4.0/ | Causal effect estimation seeks to determine the impact of an intervention from observational data. However, the existing causal inference literature primarily addresses treatment effects on frequently occurring events. But what if we are interested in estimating the effects of a policy intervention whose benefits, while potentially important, can only be observed and measured in rare yet impactful events, such as extreme climate events? The standard causal inference methodology is not designed for this type of inference since the events of interest may be scarce in the observed data and some degree of extrapolation is necessary. Extreme Value Theory (EVT) provides methodologies for analyzing statistical phenomena in such extreme regimes. We introduce a novel framework for assessing treatment effects in extreme data to capture the causal effect at the occurrence of rare events of interest. In particular, we employ the theory of multivariate regular variation to model extremities. We develop a consistent estimator for extreme treatment effects and present a rigorous non-asymptotic analysis of its performance. We illustrate the performance of our estimator using both synthetic and semi-synthetic data. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:17:44 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tan",
"Jiyuan",
""
],
[
"Blanchet",
"Jose",
""
],
[
"Syrgkanis",
"Vasilis",
""
]
] |
2506.14052 | Ananya Joshi | Ananya Joshi, George Khoury, Christodoulas Floudas | Inhibiting Alzheimer's Disease by Targeting Aggregation of Beta-Amyloid | null | null | null | null | q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Alzheimer's disease is characterized by dangerous amyloid plaques formed by deposits of the protein Beta-Amyloid aggregates in the brain. The specific amino acid sequence that is responsible for the aggregates of Beta-Amyloid is lys-leu-val-phe-phe (KLVFF). KLVFF aggregation inhibitors, which we design in this paper, prevent KLVFF from binding with itself to form oligomers or fibrils (and eventually plaques) that cause neuronal death. Our binder-blocker peptides are designed such that, on one side, they bind strongly to KLVFF, and on the other side, they disrupt critical interactions, thus preventing aggregation. Our methods use optimization techniques and molecular simulations and identify 10 candidate sequences for trial of the 3.2 million possible sequences. This approach for inhibitor identification can be generalized to other diseases characterized by protein aggregation, such as Parkinson's, Huntington's, and prion diseases. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:20:33 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Joshi",
"Ananya",
""
],
[
"Khoury",
"George",
""
],
[
"Floudas",
"Christodoulas",
""
]
] |
2506.14053 | Kaio Felipe Ramos Nascimento | Kaio F. R. Nascimento, Daniel M. Castro, Gustavo G. Cambrainha and Mauro Copelli | The phenomenological renormalization group in neuronal models near criticality | 8 pages, 4 figures | null | null | null | cond-mat.dis-nn cond-mat.stat-mech nlin.AO physics.data-an | http://creativecommons.org/licenses/by/4.0/ | The phenomenological renormalization group (PRG) has been applied to the study of scaleinvariant phenomena in neuronal data, providing evidence for critical phenomena in the brain. However, it remains unclear how reliably these observed signatures indicate genuine critical behavior, as it is not well established how close to criticality a system must be for them to emerge. Here, we rely on neuronal models with known critical points to investigate under which conditions the PRG procedure yields consistent results. We discuss how the time-binning step of data preprocessing can crucially affect the final results, and propose a data-driven method to adapt the time bin in order to circumvent this issue. Under these conditions, the PRG method only detects scaling behavior in neuronal models within a very narrow range of the critical point, lending credence to the conclusions drawn from PRG results in experimental data. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:20:47 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Nascimento",
"Kaio F. R.",
""
],
[
"Castro",
"Daniel M.",
""
],
[
"Cambrainha",
"Gustavo G.",
""
],
[
"Copelli",
"Mauro",
""
]
] |
2506.14054 | Joshua Fan | Joshua Fan, Haodi Xu, Feng Tao, Md Nasim, Marc Grimson, Yiqi Luo, Carla P. Gomes | Scientifically-Interpretable Reasoning Network (ScIReN): Uncovering the Black-Box of Nature | 28 pages, 9 figures, submitted to NeurIPS 2025 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural networks are a powerful tool for learning patterns from data. However, they do not respect known scientific laws, nor can they reveal novel scientific insights due to their black-box nature. In contrast, scientific reasoning distills biological or physical principles from observations and controlled experiments, and quantitatively interprets them with process-based models made of mathematical equations. Yet, process-based models rely on numerous free parameters that must be set in an ad-hoc manner, and thus often fit observations poorly in cross-scale predictions. While prior work has embedded process-based models in conventional neural networks, discovering interpretable relationships between parameters in process-based models and input features is still a grand challenge for scientific discovery. We thus propose Scientifically-Interpretable Reasoning Network (ScIReN), a fully-transparent framework that combines interpretable neural and process-based reasoning. An interpretable encoder predicts scientifically-meaningful latent parameters, which are then passed through a differentiable process-based decoder to predict labeled output variables. ScIReN also uses a novel hard-sigmoid constraint layer to restrict latent parameters to meaningful ranges defined by scientific prior knowledge, further enhancing its interpretability. While the embedded process-based model enforces established scientific knowledge, the encoder reveals new scientific mechanisms and relationships hidden in conventional black-box models. We apply ScIReN on two tasks: simulating the flow of organic carbon through soils, and modeling ecosystem respiration from plants. In both tasks, ScIReN outperforms black-box networks in predictive accuracy while providing substantial scientific interpretability -- it can infer latent scientific mechanisms and their relationships with input features. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:21:37 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Fan",
"Joshua",
""
],
[
"Xu",
"Haodi",
""
],
[
"Tao",
"Feng",
""
],
[
"Nasim",
"Md",
""
],
[
"Grimson",
"Marc",
""
],
[
"Luo",
"Yiqi",
""
],
[
"Gomes",
"Carla P.",
""
]
] |
2506.14055 | Yutian Tang | Yutian Tang, Hongchen Cao, Yuxi Chen, David Lo | Characterising Bugs in Jupyter Platform | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | As a representative literate programming platform, Jupyter is widely adopted by developers, data analysts, and researchers for replication, data sharing, documentation, interactive data visualization, and more. Understanding the bugs in the Jupyter platform is essential for ensuring its correctness, security, and robustness. Previous studies focused on code reuse, restoration, and repair execution environment for Jupyter notebooks. However, the bugs in Jupyter notebooks' hosting platform Jupyter are not investigated. In this paper, we investigate 387 bugs in the Jupyter platform. These Jupyter bugs are classified into 11 root causes and 11 bug symptoms. We identify 14 major findings for developers. More importantly, our study opens new directions in building tools for detecting and fixing bugs in the Jupyter platform. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:24:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tang",
"Yutian",
""
],
[
"Cao",
"Hongchen",
""
],
[
"Chen",
"Yuxi",
""
],
[
"Lo",
"David",
""
]
] |
2506.14056 | Fan Lei | Fan Lei, David A. Sampson, Jiayi Hong, Yuxin Ma, Giuseppe Mascaro, Dave White, Rimjhim Agarwal, and Ross Maciejewski | FEWSim: A Visual Analytic Framework for Exploring the Nexus of Food-Energy-Water Simulations | Accepted by IEEE Computer Graphics and Applications (CG&A) | null | 10.1109/MCG.2025.3581004 | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | The interdependencies of food, energy, and water (FEW) systems create a nexus opportunity to explore the strengths and vulnerabilities of individual and cross-sector interactions within FEW systems. However, the variables quantifying nexus interactions are hard to observe, which hinders the cross-sector analysis. To overcome such challenges, we present FEWSim, a visual analytics framework designed to support domain experts in exploring and interpreting simulation results from a coupled FEW model. FEWSim employs a three-layer asynchronous architecture: the model layer integrates food, energy, and water models to simulate the FEW nexus; the middleware layer manages scenario configuration and execution; and the visualization layer provides interactive visual exploration of simulated time-series results across FEW sectors. The visualization layer further facilitates the exploration across multiple scenarios and evaluates scenario differences in performance using sustainability indices of the FEW nexus. We demonstrate the utility of FEWSim through a case study for the Phoenix Active Management Area (AMA) in Arizona. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:29:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lei",
"Fan",
""
],
[
"Sampson",
"David A.",
""
],
[
"Hong",
"Jiayi",
""
],
[
"Ma",
"Yuxin",
""
],
[
"Mascaro",
"Giuseppe",
""
],
[
"White",
"Dave",
""
],
[
"Agarwal",
"Rimjhim",
""
],
[
"Maciejewski",
"Ross",
""
]
] |
2506.14057 | Yash Vekaria | Yash Vekaria (1), Yohan Beugin (2), Shaoor Munir (1), Gunes Acar (3), Nataliia Bielova (4), Steven Englehardt (5), Umar Iqbal (6), Alexandros Kapravelos (7), Pierre Laperdrix (8), Nick Nikiforakis (9), Jason Polakis (10), Franziska Roesner (11), Zubair Shafiq (1), and Sebastian Zimmeck (12) ((1) University of California Davis, USA, (2) University of Wisconsin-Madison, USA, (3) Radboud University, Netherlands, (4) Inria Centre at Universite Cote d'Azur, France, (5) Independent Researcher, USA, (6) Washington University in St. Louis, USA, (7) North Carolina State University, USA, (8) Centre National de la Recherche Scientifique, France, (9) Stony Brook University, USA, (10) University of Illinois Chicago, USA, (11) University of Washington, USA, (12) Wesleyan University, USA) | SoK: Advances and Open Problems in Web Tracking | Extended Version is available at: https://github.com/privacysandstorm/sok-advances-open-problems-web-tracking | null | null | null | cs.CR cs.CY cs.NI | http://creativecommons.org/licenses/by/4.0/ | Web tracking is a pervasive and opaque practice that enables personalized advertising, retargeting, and conversion tracking. Over time, it has evolved into a sophisticated and invasive ecosystem, employing increasingly complex techniques to monitor and profile users across the web. The research community has a long track record of analyzing new web tracking techniques, designing and evaluating the effectiveness of countermeasures, and assessing compliance with privacy regulations. Despite a substantial body of work on web tracking, the literature remains fragmented across distinctly scoped studies, making it difficult to identify overarching trends, connect new but related techniques, and identify research gaps in the field. Today, web tracking is undergoing a once-in-a-generation transformation, driven by fundamental shifts in the advertising industry, the adoption of anti-tracking countermeasures by browsers, and the growing enforcement of emerging privacy regulations. This Systematization of Knowledge (SoK) aims to consolidate and synthesize this wide-ranging research, offering a comprehensive overview of the technical mechanisms, countermeasures, and regulations that shape the modern and rapidly evolving web tracking landscape. This SoK also highlights open challenges and outlines directions for future research, aiming to serve as a unified reference and introductory material for researchers, practitioners, and policymakers alike. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:30:54 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Vekaria",
"Yash",
"",
"University of California Davis, USA"
],
[
"Beugin",
"Yohan",
"",
"University of Wisconsin-Madison, USA"
],
[
"Munir",
"Shaoor",
"",
"University of California Davis, USA"
],
[
"Acar",
"Gunes",
"",
"Radboud University, Netherlands"
],
[
"Bielova",
"Nataliia",
"",
"Inria Centre at Universite Cote d'Azur, France"
],
[
"Englehardt",
"Steven",
"",
"Independent Researcher, USA"
],
[
"Iqbal",
"Umar",
"",
"Washington University in St. Louis, USA"
],
[
"Kapravelos",
"Alexandros",
"",
"North Carolina State University, USA"
],
[
"Laperdrix",
"Pierre",
"",
"Centre National de la Recherche Scientifique, France"
],
[
"Nikiforakis",
"Nick",
"",
"Stony Brook University, USA"
],
[
"Polakis",
"Jason",
"",
"University of Illinois Chicago, USA"
],
[
"Roesner",
"Franziska",
"",
"University of Washington, USA"
],
[
"Shafiq",
"Zubair",
"",
"University of California Davis, USA"
],
[
"Zimmeck",
"Sebastian",
"",
"Wesleyan University, USA"
]
] |
2506.14058 | Ali Baheri | Ali Baheri | Implicit Constraint-Aware Off-Policy Correction for Offline Reinforcement Learning | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Offline reinforcement learning promises policy improvement from logged interaction data alone, yet state-of-the-art algorithms remain vulnerable to value over-estimation and to violations of domain knowledge such as monotonicity or smoothness. We introduce implicit constraint-aware off-policy correction, a framework that embeds structural priors directly inside every Bellman update. The key idea is to compose the optimal Bellman operator with a proximal projection on a convex constraint set, which produces a new operator that (i) remains a $\gamma$-contraction, (ii) possesses a unique fixed point, and (iii) enforces the prescribed structure exactly. A differentiable optimization layer solves the projection; implicit differentiation supplies gradients for deep function approximators at a cost comparable to implicit Q-learning. On a synthetic Bid-Click auction -- where the true value is provably monotone in the bid -- our method eliminates all monotonicity violations and outperforms conservative Q-learning and implicit Q-learning in return, regret, and sample efficiency. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:32:26 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Baheri",
"Ali",
""
]
] |
2506.14059 | Shakib Mustavee | Shakib Mustavee, Shaurya Agarwal, and Arvind Singh | A Stochastic Differential Equation Framework for Modeling Queue Length Dynamics Inspired by Self-Similarity | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This article develops a stochastic differential equation (SDE) for modeling the temporal evolution of queue length dynamics at signalized intersections. Inspired by the observed quasiperiodic and self-similar characteristics of the queue length dynamics, the proposed model incorporates three properties into the SDE: (i) mean reversion with periodic mean, (ii) multiplicative noise, and (iii) fractional Brownian motion. It replicates key statistical features observed in real data, including the probability distribution function (PDF) and PSD of queue lengths. To our knowledge, this is the first equation-based model for queue dynamics. The proposed approach offers a transparent, data-consistent framework that may help inform and enhance the design of black-box learning algorithms with underlying traffic physics. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:40:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Mustavee",
"Shakib",
""
],
[
"Agarwal",
"Shaurya",
""
],
[
"Singh",
"Arvind",
""
]
] |
2506.14060 | Taras Banakh | Taras Banakh | Linear Geometry and Algebra | 45 chapters, 792 pages | null | null | null | math.AG math.AC math.GR math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Linear Geometry studies geometric properties which can be expressed via the notion of a line. All information about lines is encoded in a ternary relation called a line relation. A set endowed with a line relation is called a liner. So, Linear Geometry studies liners. Imposing some additional axioms on a liner, we obtain some special classes of liners: regular, projective, affine, proaffine, etc. Linear Geometry includes Affine and Projective Geometries and is a part of Incidence Geometry. The aim of this book is to present a self-contained logical development of Linear Geometry, starting with some intuitive acceptable geometric axioms and ending with algebraic structures that necessarily arise from studying the structure of geometric objects that satisfy those simple and intuitive geometric axioms. We shall meet many quite exotic algebraic structures that arise this way: magmas, loops, ternary-ring, quasi-fields, alternative rings, procorps, profields, etc. We strongly prefer (synthetic) geometric proofs and use tools of analytic geometry only when no purely geometric proof is available. Liner Geometry has been developed by many great mathematicians since times of Antiquity (Thales, Euclides, Proclus, Pappus), through Renaissance (Descartes, Desargues), Early Modernity (Playfair, Gauss, Lobachevski, Bolyai, Poncelet, Steiner, M\"obius), Late Modernity Times (Steinitz, Klein, Hilbert, Moufang, Hessenberg, Jordan, Beltrami, Fano, Gallucci, Veblen, Wedderburn, Lenz, Barlotti) till our contempories (Hartshorne, Hall, Buekenhout, Gleason, Kantor, Doyen, Hubault, Dembowski, Klingenberg, Grundh\"ofer). | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:44:20 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Banakh",
"Taras",
""
]
] |
2506.14062 | Adriano Meligrana | Lilith Orion Hafner, Adriano Meligrana | An Exact and Efficient Sampler for Dynamic Discrete Distributions | 8 pages, 5 figures | null | null | null | cs.DS math.PR stat.CO | http://creativecommons.org/licenses/by/4.0/ | Sampling from a dynamic discrete distribution involves sampling from a dynamic set of weighted elements, where elements can be added or removed at any stage of the sampling process. Although efficient for static sets, the Alias method becomes impractical in dynamic settings due to the need to reconstruct the sampler after each update, which incurs a computational cost proportional to the size of the distribution, making it unsuitable for applications requiring frequent insertions, deletions, or weight adjustments. To address this limitation, different approaches have been studied, such as the Forest of Trees method and the BUcket Sampling (BUS) method. However, all previous methods suffered from numerical issues which can bias the sampling process. In this paper, we describe EBUS (Exact BUcket Sampling), the first exact algorithm with $O(1)$ sampling and update cost. The sampler can be updated by base-$b$ numbers with bounded precision and exponent, and sample the distribution of its elements exactly and efficiently. We provide also a state of the art implementation of the method using IEEE 64-bit floating point numbers which we empirically show to be more efficient than several implementations of previous inexact methods. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:46:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hafner",
"Lilith Orion",
""
],
[
"Meligrana",
"Adriano",
""
]
] |
2506.14063 | Wenjin Zhao | Wenjin Zhao, Zui Tao, Yichi Zhang, Bowen Shen, Zhongdong Han, Patrick Kn\"uppel, Yihang Zeng, Zhengchao Xia, Kenji Watanabe, Takashi Taniguchi, Jie Shan, Kin Fai Mak | Emergence of Chern metal in a moir\'e Kondo lattice | null | null | null | null | cond-mat.str-el cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | A Chern metal is a two-dimensional metallic state of matter carrying chiral edge states. It can emerge as a doped Chern insulator, but theoretical studies have also predicted its emergence near a Kondo breakdown separating a metallic chiral spin liquid and a heavy Fermi liquid in a frustrated lattice. To date, the latter exotic scenario has not been realized. Here, we report the observation of a Chern metal at the onset of the magnetic Kondo breakdown in a frustrated moir\'e Kondo lattice--angle-aligned MoTe2/WSe2 bilayers. The state is compressible and is manifested by a nearly quantized Hall resistance but a finite longitudinal resistance that arises from a bad metallic bulk. The state also separates an itinerant and a heavy Fermi liquid and appears far away from the band inversion critical point of the material, thus ruling out its origin from simply doping a Chern insulator. We demonstrate the presence of a chiral edge state by nonlocal transport measurements and current-induced quantum anomalous Hall breakdown. Magnetic circular dichroism measurements further reveal a magnetization plateau for the Chern metal before a metamagnetic transition at the Kondo breakdown. Our results open an opportunity for moir\'e engineering of exotic quantum phases of matter through the close interplay between band topology and Kondo interactions. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:47:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhao",
"Wenjin",
""
],
[
"Tao",
"Zui",
""
],
[
"Zhang",
"Yichi",
""
],
[
"Shen",
"Bowen",
""
],
[
"Han",
"Zhongdong",
""
],
[
"Knüppel",
"Patrick",
""
],
[
"Zeng",
"Yihang",
""
],
[
"Xia",
"Zhengchao",
""
],
[
"Watanabe",
"Kenji",
""
],
[
"Taniguchi",
"Takashi",
""
],
[
"Shan",
"Jie",
""
],
[
"Mak",
"Kin Fai",
""
]
] |
2506.14064 | Nicolas Navarre | Iona Carslaw, Sivan Milton, Nicolas Navarre, Ciyang Qing and Wataru Uegaki | Automatic Extraction of Clausal Embedding Based on Large-Scale English Text Data | Accepted in the Society for Computation in Linguistics | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | For linguists, embedded clauses have been of special interest because of their intricate distribution of syntactic and semantic features. Yet, current research relies on schematically created language examples to investigate these constructions, missing out on statistical information and naturally-occurring examples that can be gained from large language corpora. Thus, we present a methodological approach for detecting and annotating naturally-occurring examples of English embedded clauses in large-scale text data using constituency parsing and a set of parsing heuristics. Our tool has been evaluated on our dataset Golden Embedded Clause Set (GECS), which includes hand-annotated examples of naturally-occurring English embedded clause sentences. Finally, we present a large-scale dataset of naturally-occurring English embedded clauses which we have extracted from the open-source corpus Dolma using our extraction tool. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:48:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Carslaw",
"Iona",
""
],
[
"Milton",
"Sivan",
""
],
[
"Navarre",
"Nicolas",
""
],
[
"Qing",
"Ciyang",
""
],
[
"Uegaki",
"Wataru",
""
]
] |
2506.14065 | Ethan Chien | Ethan Chien and Jan Steckel | Dynamically Tunable Helical Antenna | 7 pages, 5 figures | null | null | null | eess.SP | http://creativecommons.org/licenses/by/4.0/ | Unmanned aerial FPV systems demand ultra-low latency, high-reliability communication links. At high speeds and in cluttered environments, Doppler shifts and rapid multipath changes can dramatically raise packet error rates. This paper investigates these phenomena in the context of ExpressLRS (ELRS) long-range FPV control links and demonstrates a novel solution: real-time geometry tuning of a circularly polarized helical antenna array. This study integrates Maxwell-equation-based full-wave simulations (via Ansys HFSS) with controlled, blind field trials to validate performance. A new analysis framework incorporates Doppler-induced frequency offset into the antenna's radiation pattern and the system's error model. Compared to a conventional fixed antenna, the adaptive helical array shows a 20-30% PER reduction when drones exceed 150 mph. The adaptive system automatically adjusts coil pitch and diameter to retune the antenna as flight parameters (velocity, attitude) change. Measured VSWR stays near unity, preventing transmitter reflection spikes. RSSI variation is reduced by half, indicating stronger link stability in urban multi-path. A regression analysis confirms that the reduction in PER due to tuning is highly statistically significant. Calibration data and error analyses are provided to validate our methodology. These findings advance the understanding of high-mobility UAV communication channels and demonstrate that reconfigurable hardware-here, mechanically tunable helices-can effectively counter Doppler and multi-path impairments. The findings inform new design principles for UAV antenna arrays and suggest a path toward AI-integrated adaptive RF systems for drone swarms and racing platforms. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:48:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chien",
"Ethan",
""
],
[
"Steckel",
"Jan",
""
]
] |
2506.14066 | Malak Mansour | Ali Abouzeid, Malak Mansour, Chengsong Hu, Dezhen Song | A Point Cloud Completion Approach for the Grasping of Partially Occluded Objects and Its Applications in Robotic Strawberry Harvesting | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | In robotic fruit picking applications, managing object occlusion in unstructured settings poses a substantial challenge for designing grasping algorithms. Using strawberry harvesting as a case study, we present an end-to-end framework for effective object detection, segmentation, and grasp planning to tackle this issue caused by partially occluded objects. Our strategy begins with point cloud denoising and segmentation to accurately locate fruits. To compensate for incomplete scans due to occlusion, we apply a point cloud completion model to create a dense 3D reconstruction of the strawberries. The target selection focuses on ripe strawberries while categorizing others as obstacles, followed by converting the refined point cloud into an occupancy map for collision-aware motion planning. Our experimental results demonstrate high shape reconstruction accuracy, with the lowest Chamfer Distance compared to state-of-the-art methods with 1.10 mm, and significantly improved grasp success rates of 79.17%, yielding an overall success-to-attempt ratio of 89.58\% in real-world strawberry harvesting. Additionally, our method reduces the obstacle hit rate from 43.33% to 13.95%, highlighting its effectiveness in improving both grasp quality and safety compared to prior approaches. This pipeline substantially improves autonomous strawberry harvesting, advancing more efficient and reliable robotic fruit picking systems. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:49:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Abouzeid",
"Ali",
""
],
[
"Mansour",
"Malak",
""
],
[
"Hu",
"Chengsong",
""
],
[
"Song",
"Dezhen",
""
]
] |
2506.14067 | Minjae Lee | Minjae Lee, Yoonjae Jung and Sangdon Park | A Regret Perspective on Online Selective Generation | 10 pages | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language generative models increasingly interact with humans, while their falsified responses raise concerns. To address this hallucination effect, selectively abstaining from answering, called selective generation, provides an effective way for generators to control the hallucination when it is unsure of their answers. However, as selective generators are interacting under non-stochastic environments and having partial feedback from users on selective generation (e.g., thumbs up or down on the selected answer), learning methods for selective generation under such practical setups are crucial but currently missing. To address these limitations, we propose an online learning algorithm for selective generation under partial feedback. In particular, as learning under partial feedback is well-studied by multi-armed bandit problems, we reduce selective generation to bandits and provide a novel conversion lemma from bandits back to selective generation to leverage any known bandit algorithms and theoretical properties. This mainly connects regret guarantees of bandits to false discovery rate (FDR) guarantees of selective generation for controlling hallucination. However, naively exploiting known bandit algorithms and their regret bounds suffers from slow convergence speed in practice due the nature of partial feedback. To overcome this, we exploit a unique structure of arms in selective generation for feedback unlocking, i.e., unlocking unknown feedback from observed feedback. We theoretically and empirically evaluate the efficacy of the proposed online selective generation algorithm under partial feedback over diverse data environment setups, resulting in controlling a desired FDR, while maintaining reasonable selection efficiency, i.e., the ratio of non-abstaining answers, compared to baselines. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:51:30 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Lee",
"Minjae",
""
],
[
"Jung",
"Yoonjae",
""
],
[
"Park",
"Sangdon",
""
]
] |
2506.14068 | Thomas Hahn | Noam Avidan, Thomas A. Hahn, Joseph M. Renes, Rotem Arnon | Fully Quantum Computational Entropies | 47 pages, 3 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum information theory, particularly its entropic formulations, has made remarkable strides in characterizing quantum systems and tasks. However, a critical dimension remains underexplored: computational efficiency. While classical computational entropies integrate complexity and feasibility into information measures, analogous concepts have yet to be rigorously developed in the quantum setting. In this work, we lay the basis for a new quantum computational information theory. Such a theory will allow studying efficient -- thus relevant in practice -- manipulation of quantum information. We introduce two innovative entropies: quantum computational min- and max-entropies (along with their smooth variants). Our quantum computational min-entropy is both the fully quantum counterpart of the classical unpredictability entropy, as well as the computational parallel to the quantum min-entropy. We establish a series of essential properties for this new entropy, including data processing and a chain rule. The quantum computational max-entropy is defined via a duality relation and gains operational meaning through an alternative formulation that we derive. Notably, it captures the efficiency of entanglement distillation with the environment, restricted to local quantum circuits of bounded size. With the introduction of our computational entropies and their study, this work marks a critical step toward a quantum information theory that incorporates computational elements. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:56:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Avidan",
"Noam",
""
],
[
"Hahn",
"Thomas A.",
""
],
[
"Renes",
"Joseph M.",
""
],
[
"Arnon",
"Rotem",
""
]
] |
2506.14069 | Sonja Farr | Sonja Farr | $\mathbb{E}_2$-algebra structures on the derived center of an algebraic scheme | 55 pages. Comments welcome | null | null | null | math.AT math.AG math.QA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper provides an explicit interface between J. Lurie's work on higher centers, and the Hochschild cohomology of an algebraic $\mathbb{k}$-scheme within the framework of deformation quantization. We first recover a canonical solution to Deligne's conjecture on Hochschild cochains in the affine and global cases, even for singular schemes, by exhibiting the Hochschild complex as an $\infty$-operadic center. We then prove that this universal $\mathbb{E}_2$-algebra structure precisely agrees with the classical Gerstenhaber bracket and cup product on cohomology in the affine and smooth cases. This last statement follows from our main technical result which allows us to extract the Gerstenhaber bracket of any $\mathbb{E}_2$-algebra obtained from a 2-algebra via Lurie's Dunn Additivity Theorem. | [
{
"version": "v1",
"created": "Mon, 16 Jun 2025 23:59:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Farr",
"Sonja",
""
]
] |
2506.14070 | Xinglei Wang | Xinglei Wang, Tao Cheng, Stephen Law, Zichao Zeng, Ilya Ilyankou, Junyuan Liu, Lu Yin, Weiming Huang, Natchapon Jongwiriyanurak | Into the Unknown: Applying Inductive Spatial-Semantic Location Embeddings for Predicting Individuals' Mobility Beyond Visited Places | 10 pages, 5 figures | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Predicting individuals' next locations is a core task in human mobility modelling, with wide-ranging implications for urban planning, transportation, public policy and personalised mobility services. Traditional approaches largely depend on location embeddings learned from historical mobility patterns, limiting their ability to encode explicit spatial information, integrate rich urban semantic context, and accommodate previously unseen locations. To address these challenges, we explore the application of CaLLiPer -- a multimodal representation learning framework that fuses spatial coordinates and semantic features of points of interest through contrastive learning -- for location embedding in individual mobility prediction. CaLLiPer's embeddings are spatially explicit, semantically enriched, and inductive by design, enabling robust prediction performance even in scenarios involving emerging locations. Through extensive experiments on four public mobility datasets under both conventional and inductive settings, we demonstrate that CaLLiPer consistently outperforms strong baselines, particularly excelling in inductive scenarios. Our findings highlight the potential of multimodal, inductive location embeddings to advance the capabilities of human mobility prediction systems. We also release the code and data (https://github.com/xlwang233/Into-the-Unknown) to foster reproducibility and future research. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:00:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Xinglei",
""
],
[
"Cheng",
"Tao",
""
],
[
"Law",
"Stephen",
""
],
[
"Zeng",
"Zichao",
""
],
[
"Ilyankou",
"Ilya",
""
],
[
"Liu",
"Junyuan",
""
],
[
"Yin",
"Lu",
""
],
[
"Huang",
"Weiming",
""
],
[
"Jongwiriyanurak",
"Natchapon",
""
]
] |
2506.14071 | Amir A. Khodahami | Amir A. Khodahami and Azizollah Azizi | A revision to the QES prescription | 11 pages, 2 figures | null | null | null | hep-th | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we propose a revision to the Quantum Extremal Surface (QES) prescription, which plays a crucial role in describing the entanglement entropy of black holes. While derivations exist for the original QES prescription using the replica trick, they rely on a saddle point approximation that faces challenges when the number of replicas approaches unity. We highlight these challenges and derive a refined formula for the entanglement entropy that incorporates a weighted summation over multiple surfaces, contrasting the QES prescription's reliance on a single extremal surface. We validate our formulation by demonstrating consistency with expected results for both the early and late stages of black hole evaporation, confirming its alignment with the unitary evolution depicted by the Page curve. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:03:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Khodahami",
"Amir A.",
""
],
[
"Azizi",
"Azizollah",
""
]
] |
2506.14072 | John Campbell | John M. Campbell | A generalization of Deterministic Finite Automata related to discharging | Submitted for publication | null | null | null | cs.FL math.CO | http://creativecommons.org/licenses/by/4.0/ | Deterministic Finite Automata (DFAs) are of central importance in automata theory. In view of how state diagrams for DFAs are defined using directed graphs, this leads us to introduce a generalization of DFAs related to a method widely used in graph theory referred to as the discharging method. Given a DFA $(Q, \Sigma, \delta, q_{0}, F)$, the transition function $\delta\colon Q \times \Sigma \to Q$ determines a directed path in the corresponding state diagram based on an input string $a_{1} a_{2} \cdots a_{n}$ consisting of characters in $\Sigma$, and our generalization can be thought of as being based on how each vertex in $D$ ''discharges'' rational values to adjacent vertices (by analogy with the discharging method) depending on the string $a_{1} a_{2} \cdots a_{n}$ and according to a fixed set of rules. We formalize this notion and pursue an exploration of the notion of a Discharging Deterministic Finite Automaton (DDFA) introduced in this paper. Our DDFA construction gives rise to a ring structure consisting of sequences that we refer to as being quasi-$k$-regular, and this ring generalizes the ring of $k$-regular sequences introduced by Allouche and Shallit. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:03:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Campbell",
"John M.",
""
]
] |
2506.14073 | Han Wu | Timo Sprekeler, Han Wu, Zhiwen Zhang | Numerical approximation of effective diffusivities in homogenization of nondivergence-form equations with large drift by a Lagrangian method | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | In this paper, we study numerical methods for the homogenization of linear second-order elliptic equations in nondivergence-form with periodic diffusion coefficients and large drift terms. Upon noting that the effective diffusion matrix can be characterized through the long-time variance of an associated diffusion process, we construct a Lagrangian numerical scheme based on a direct simulation of the underlying stochastic differential equation and utilizing the framework of modified equations, thereby avoiding the need to solve the Fokker--Planck--Kolmogorov equation. Through modified equation analysis, we derive higher-order weak convergence rates for our method. Finally, we conduct numerical experiments to demonstrate the accuracy of the proposed method. The results show that the method efficiently computes effective diffusivities, even in high dimensions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:08:12 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sprekeler",
"Timo",
""
],
[
"Wu",
"Han",
""
],
[
"Zhang",
"Zhiwen",
""
]
] |
2506.14074 | Nathaniel Pinckney | Nathaniel Pinckney, Chenhui Deng, Chia-Tung Ho, Yun-Da Tsai, Mingjie Liu, Wenfei Zhou, Brucek Khailany, Haoxing Ren | Comprehensive Verilog Design Problems: A Next-Generation Benchmark Dataset for Evaluating Large Language Models and Agents on RTL Design and Verification | 16 pages with appendix | null | null | null | cs.LG cs.AR | http://creativecommons.org/licenses/by/4.0/ | We present the Comprehensive Verilog Design Problems (CVDP) benchmark, a new dataset and infrastructure to advance LLM and agent research in hardware design and verification. CVDP includes 783 problems across 13 task categories, covering RTL generation, verification, debugging, specification alignment, and technical Q&A authored by experienced hardware engineers. Problems are offered in both non-agentic and agentic formats. The benchmark introduces more realistic and challenging contexts than prior work, with state-of-the-art models achieving no more than 34% pass@1 on code generation. Agentic tasks$\unicode{x2013}$especially those involving RTL reuse and verification$\unicode{x2013}$are particularly difficult. Evaluation uses open-source tools and model scoring infrastructure, with comprehension tasks assessed via BLEU and LLM-based judging. CVDP reveals substantial gaps in current model capabilities, underscoring the need for continued research toward robust, real-world hardware design automation. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:11:13 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pinckney",
"Nathaniel",
""
],
[
"Deng",
"Chenhui",
""
],
[
"Ho",
"Chia-Tung",
""
],
[
"Tsai",
"Yun-Da",
""
],
[
"Liu",
"Mingjie",
""
],
[
"Zhou",
"Wenfei",
""
],
[
"Khailany",
"Brucek",
""
],
[
"Ren",
"Haoxing",
""
]
] |
2506.14075 | Hillol Biswas | Hillol Biswas | Comparing a Few Qubit Systems for Superconducting Hardware Compatibility and Circuit Design Sensitivity in Qiskit | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | In the current quantum computing ecosystem, building complex and integrated circuits for addressing real-world problems often involves using basic historical components like Bell states to take advantage of superposition, entanglement, and coherence. The availability of the simulator and the IBM quantum computing endeavor through Qiskit further broadens the scope of application based on domain use. In the NISQ era, however, error mitigation is a criterion that is applied when comparing a QPU run result with the simulator. As the real-world scope of application for quantum computing is being broadened, using simulators as a baseline offers confidence; however, concerns about computation resources and availability arise in the high-speed computational regime. As complex problems entail using a larger number of superconducting physical qubits, what should be the basis of framing a quantum circuit for building quantum algorithms in a real-world scenario, given hardware compatibility that should ideally provide confidence baselining with the simulator outcome? This work implements three base circuits for different qubit systems in the simulator and corresponding 127-qubit IBM Sherbrooke superconducting quantum processing units (QPU) to explore the tradeoff between generalizability, sensitivity of circuit design to parameters, noise resilience, resource planning, and efficient qubit usage insights. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:18:46 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Biswas",
"Hillol",
""
]
] |
2506.14076 | Juan Carlos Trelles Arjona | Juan Carlos Trelles Arjona | Quiet Sun Magnetism from ViSP Data Multiline Inversions Around 630.1 nm | null | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Quiet Sun magnetism plays an important role in the global energy balance of the solar atmosphere. The study of the magnetism of the quiet Sun has been a major effort in the last decades, and, as a result, very important advances in our knowledge have been achieved. The recent commissioning of the Daniel K. Inouye Solar Telescope, the largest ground-based solar telescope in the world today, provides us with data of high spectropolarimetric quality and high spatial resolution. The combination of data acquired with the ViSP instrument on the DKIST telescope around 630.1 nm and multiline inversions enables the achievement of exceptionally high spatial resolution in the solar atmosphere, both at the surface and in depth. In this paper we determine the log(gf) values of the spectral lines around 630.1 nm observed with the ViSP instrument and search for the best multi-inversion strategy, by means of MHD simulations, to analyze quiet Sun magnetism data within such spectral range. We present results of quiet Sun magnetism with special emphasis on its evolution with optical depth. In the internetwork, we find the decrease with height of the averaged magnetic field strength and the zone from which the fields become more horizontal. Additionally, we explore the depth-dependent evolution of the networks canopy, uncovering intriguing insights into its thermodynamic and magnetic properties. Notably, we detect a temperature enhancement in very localized areas of the vicinity of the network and distinguish mass motions with opposing velocities inside the flux tubes. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:20:15 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Arjona",
"Juan Carlos Trelles",
""
]
] |
2506.14077 | Jie Wu | J.Y. Shen, G.F. Chen, Y.C. Zhang, G.Y. Xi, J.Y. He, X.B. Cheng, J. Wu | Electronic nematic normal and superconducting state in electron-doped copper-oxide superconductors | null | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The similarities and differences between hole- and electron-doped cuprates are central to studies of high-temperature superconductivity. While electronic nematicity is found to be pervasive in hole-doped cuprates, iron-based superconductors, and other unconventional superconductors, evidence for electronic nematicity in electron-doped cuprates remains elusive. Here, we discover that the normal state of electron-doped Sr0.9La0.1CuO2 (SLCO) is nematic by the angle-resolved resistivity (ARR) method and the uncovered ground state at zero temperature is also nematic when superconductivity is suppressed by an applied magnetic field. As we deliberately change the substrate from tetragonal KTaO3(001) (KTO) to orthorhombic GdScO3(110) (GSO), the nematic director of SLCO is pinned by the epitaxial strain but the nematic amplitude remains roughly the same, implying that the nematicity originates from electron-electron correlations. The nematicity is significantly enhanced by the presence of superconducting fluctuations and its amplitude increases appreciably as the effective doping level of SLCO is lowered from optimal to underdoped. Thus, electronic nematicity is intrinsic to high-temperature superconductors regardless of differences in the structural and electronic configurations corresponding to hole or electron doping. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:20:24 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Shen",
"J. Y.",
""
],
[
"Chen",
"G. F.",
""
],
[
"Zhang",
"Y. C.",
""
],
[
"Xi",
"G. Y.",
""
],
[
"He",
"J. Y.",
""
],
[
"Cheng",
"X. B.",
""
],
[
"Wu",
"J.",
""
]
] |
2506.14078 | Yonggeun Jung | Yonggeun Jung | Machine Learning-Based Estimation of Monthly GDP | This version: June 16, 2025. Replication package and data are available at https://github.com/Yonggeun-Jung/monthly_gdp | null | null | null | econ.EM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a scalable framework to estimate monthly GDP using machine learning methods. We apply Multi-Layer Perceptron (MLP), Long Short-Term Memory networks (LSTM), Extreme Gradient Boosting (XGBoost), and Elastic Net regression to map monthly indicators to quarterly GDP growth, and reconcile the outputs with actual aggregates. Using data from China, Germany, the UK, and the US, our method delivers robust performance across varied data environments. Benchmark comparisons with prior US studies and UK official statistics validate its accuracy. The approach offers a flexible and data-driven tool for high-frequency macroeconomic monitoring and policy analysis. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:25:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jung",
"Yonggeun",
""
]
] |
2506.14079 | Matthew Toles | Matthew Toles, Rattandeep Singh, Isaac Song Zhou Yu | FormGym: Doing Paperwork with Agents | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Completing paperwork is a challenging and time-consuming problem. Form filling is especially challenging in the pure-image domain without access to OCR, typeset PDF text, or a DOM. For computer agents, it requires multiple abilities, including multi-modal understanding, information retrieval, and tool-use. We present a novel form-filling benchmark consisting of 432 fields spread across 55 documents and 3 tasks, requiring knowledge of 236 features per user. We find that baseline VLAs achieve less than 1% accuracy in most cases, primarily due to poor localization ability. GUI agents also struggle, scoring between 10.6-68.0% despite high cost and latency. Therefore, we also contribute FieldFinder, a tool to assist LLMs in identifying where to place text on a form. With FieldFinder, all models achieve equal or better performance in all six study conditions, with a maximum increase from 2% to 56%. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:32:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Toles",
"Matthew",
""
],
[
"Singh",
"Rattandeep",
""
],
[
"Yu",
"Isaac Song Zhou",
""
]
] |
2506.14080 | Sonika Johri | Michael Kubal, Sonika Johri | A Quantum Platform for Multiomics Data | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The complexity of biological systems, governed by molecular interactions across hierarchical scales, presents a challenge for computational modeling. While advances in multiomic profiling have enabled precise measurements of biological components, classical computational approaches remain limited in capturing emergent dynamics critical for understanding disease mechanisms and therapeutic interventions. Quantum computing offers a new paradigm for addressing classically intractable problems, yet its integration into biological research remains nascent due to scalability barriers and accessibility gaps. Here, we introduce a hybrid quantum-classical machine learning platform designed to bridge this gap, with an encode-search-build approach which allows for efficiently extracting the most relevant information from biological data to \underline{encode} into a quantum state, provably efficient training algorithms to \underline{search} for optimal parameters, and a stacking strategy that allows one to systematically \underline{build} more complex models as more quantum resources become available. We propose to demonstrate the platform's utility through two initial use cases: quantum-enhanced classification of phenotypic states from molecular variables and prediction of temporal evolution in biological systems. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:33:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kubal",
"Michael",
""
],
[
"Johri",
"Sonika",
""
]
] |
2506.14081 | Philip Wellnitz | Marco Bressan, Julian Brinkmann, Holger Dell, Marc Roth and Philip Wellnitz | The Complexity of Counting Small Sub-Hypergraphs | null | null | null | null | cs.CC cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Subgraph counting is a fundamental and well-studied problem whose computational complexity is well understood. Quite surprisingly, the hypergraph version of subgraph counting has been almost ignored. In this work, we address this gap by investigating the most basic sub-hypergraph counting problem: given a (small) hypergraph $H$ and a (large) hypergraph $G$, compute the number of sub-hypergraphs of $G$ isomorphic to $H$. Formally, for a family $\mathcal{H}$ of hypergraphs, let #Sub($\mathcal{H}$) be the restriction of the problem to $H \in \mathcal{H}$; the induced variant #IndSub($\mathcal{H}$) is defined analogously. Our main contribution is a complete classification of the complexity of these problems. Assuming the Exponential Time Hypothesis, we prove that #Sub($\mathcal{H}$) is fixed-parameter tractable if and only if $\mathcal{H}$ has bounded fractional co-independent edge-cover number, a novel graph parameter we introduce. Moreover, #IndSub($\mathcal{H}$) is fixed-parameter tractable if and only if $\mathcal{H}$ has bounded fractional edge-cover number. Both results subsume pre-existing results for graphs as special cases. We also show that the fixed-parameter tractable cases of #Sub($\mathcal{H}$) and #IndSub($\mathcal{H}$) are unlikely to be in polynomial time, unless respectively #P = P and Graph Isomorphism $\in$ P. This shows a separation with the special case of graphs, where the fixed-parameter tractable cases are known to actually be in polynomial time. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:33:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Bressan",
"Marco",
""
],
[
"Brinkmann",
"Julian",
""
],
[
"Dell",
"Holger",
""
],
[
"Roth",
"Marc",
""
],
[
"Wellnitz",
"Philip",
""
]
] |
2506.14082 | Daisuke Sato | Dye SK Sato, Yuji Yagi, Ryo Okuwaki and Yukitoshi Fukahata | Smooth surface reconstruction of earthquake faults from distributed potency beachballs | 35 pages, 13 figures | null | null | null | physics.geo-ph stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Earthquake faults approximate smooth surfaces of displacement discontinuity in a coarse-grained linear continuum, and the beachball nodal planes of source inelastic strain (potency) indicate the normal vector (n-vector) of the crack face. This theoretical relation has been used to interpret estimated potency as geometrical information of earthquake faults. We formulate an inverse analysis for reconstructing a smooth, three-dimensional fault surface from a given potency field. The analysis is grounded on the definition of the n-vector that is orthogonality to a surface. It is shown from this definition that, whereas the estimated n-vector field is always explained by a unique curve in two dimensions, the surface reconstruction from the estimated n-vector field is an overdetermined problem with no solution in three dimensions. We resolve this overdetermination by adopting the original definition of a fault surface as a map that approximates a thin sheet of potency onto a displacement discontinuity across a literal face. An analytical expression for this fault reconstruction is derived and confirmed to reproduce the original three-dimensional smooth surface from a synthesized noisy n-vector field. By incorporating this result into surface reconstruction based on potency density tensor inversion, we estimate the three-dimensional source fault geometry of the 2013 Balochistan earthquake. The estimated geometry is more consistent with the fault trace than the quasi-two-dimensional shape reconstruction as anticipated by our numerical experiments. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:43:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sato",
"Dye SK",
""
],
[
"Yagi",
"Yuji",
""
],
[
"Okuwaki",
"Ryo",
""
],
[
"Fukahata",
"Yukitoshi",
""
]
] |
2506.14083 | Zhicheng Zhang | Zhicheng Zhang and Yoshihiko Susuki and Atsushi Okazaki | Extracting transient Koopman modes from short-term weather simulations with sparsity-promoting dynamic mode decomposition | 38 pages, 21 figures, | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Convective features-here represented as warm bubble-like patterns-reveal essential, high-level information about how short-term weather dynamics evolve within a high-dimensional state space. We introduce a data-driven framework that uncovers transient dynamics captured by Koopman modes responsible for these structures and traces their emergence, growth, and decay. Our approach incorporates the sparsity-promoting dynamic mode decomposition into the framework of Koopman mode decomposition, yielding a few number of selected modes whose sparse amplitudes highlight dominant transient structures. By tuning the sparsity weight, we balance reconstruction accuracy and model complexity. We illustrate the methodology on weather simulations, using the magnitude of velocity and vorticity fields as distinct observable datasets. The resulting sparse dominant Koopman modes capture the transient evolution of bubble-like pattern and can reduce the dimensionality of the weather system model, offering an efficient surrogate for diagnostic and forecasting tasks. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:48:25 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Zhicheng",
""
],
[
"Susuki",
"Yoshihiko",
""
],
[
"Okazaki",
"Atsushi",
""
]
] |
2506.14084 | Taehee Jeong | Taehee Jeong | Lightweight Relevance Grader in RAG | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Retrieval-Augmented Generation (RAG) addresses limitations of large language models (LLMs) by leveraging a vector database to provide more accurate and up-to-date information. When a user submits a query, RAG executes a vector search to find relevant documents, which are then used to generate a response. However, ensuring the relevance of retrieved documents with a query would be a big challenge. To address this, a secondary model, known as a relevant grader, can be served to verify its relevance. To reduce computational requirements of a relevant grader, a lightweight small language model is preferred. In this work, we finetuned llama-3.2-1b as a relevant grader and achieved a significant increase in precision from 0.1301 to 0.7750. Its precision is comparable to that of llama-3.1-70b. Our code is available at https://github.com/taeheej/Lightweight-Relevance-Grader-in-RAG. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 00:52:21 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Jeong",
"Taehee",
""
]
] |
2506.14085 | Toru Takahashi | Toru Takahashi, Tatsuya Tokito, Yi Cui, Toshiro Matsumoto | A shape optimisation of mutual inductances among coils | null | null | 10.1016/j.apm.2025.116257 | null | math-ph math.MP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper introduces a shape optimisation framework for achieving desired mutual inductances (MIs) among coils in 3D space. Utilising a wire modelling approach, the coils are discretised using B-spline curves, with control points (CPs) serving as design variables. The key contribution is the derivation of the shape derivative of the objective function in terms of MIs, enabling the use of gradient-based quasi-Newton optimisation methods. A coil length constraint is also incorporated. The study demonstrates the effectiveness of the framework through numerical examples, validating the theoretical and numerical developments. This approach addresses the largely unexplored area of magnetostatic MI optimisation within the wire modelling framework, offering a computationally efficient alternative to finite element methods etc. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:01:57 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Takahashi",
"Toru",
""
],
[
"Tokito",
"Tatsuya",
""
],
[
"Cui",
"Yi",
""
],
[
"Matsumoto",
"Toshiro",
""
]
] |
2506.14086 | Kaustubh Dhole | Rahul Seetharaman, Kaustubh D. Dhole, Aman Bansal | InsertRank: LLMs can reason over BM25 scores to Improve Listwise Reranking | null | null | null | null | cs.IR cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have demonstrated significant strides across various information retrieval tasks, particularly as rerankers, owing to their strong generalization and knowledge-transfer capabilities acquired from extensive pretraining. In parallel, the rise of LLM-based chat interfaces has raised user expectations, encouraging users to pose more complex queries that necessitate retrieval by ``reasoning'' over documents rather than through simple keyword matching or semantic similarity. While some recent efforts have exploited reasoning abilities of LLMs for reranking such queries, considerable potential for improvement remains. In that regards, we introduce InsertRank, an LLM-based reranker that leverages lexical signals like BM25 scores during reranking to further improve retrieval performance. InsertRank demonstrates improved retrieval effectiveness on -- BRIGHT, a reasoning benchmark spanning 12 diverse domains, and R2MED, a specialized medical reasoning retrieval benchmark spanning 8 different tasks. We conduct an exhaustive evaluation and several ablation studies and demonstrate that InsertRank consistently improves retrieval effectiveness across multiple families of LLMs, including GPT, Gemini, and Deepseek models. %In addition, we also conduct ablation studies on normalization by varying the scale of the BM25 scores, and positional bias by shuffling the order of the documents. With Deepseek-R1, InsertRank achieves a score of 37.5 on the BRIGHT benchmark. and 51.1 on the R2MED benchmark, surpassing previous methods. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:04:45 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Seetharaman",
"Rahul",
""
],
[
"Dhole",
"Kaustubh D.",
""
],
[
"Bansal",
"Aman",
""
]
] |
2506.14087 | Zhongzheng Qiao | Zhongzheng Qiao, Chenghao Liu, Yiming Zhang, Ming Jin, Quang Pham, Qingsong Wen, P.N. Suganthan, Xudong Jiang, Savitha Ramasamy | Multi-Scale Finetuning for Encoder-based Time Series Foundation Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Time series foundation models (TSFMs) demonstrate impressive zero-shot performance for time series forecasting. However, an important yet underexplored challenge is how to effectively finetune TSFMs on specific downstream tasks. While naive finetuning can yield performance gains, we argue that it falls short of fully leveraging TSFMs' capabilities, often resulting in overfitting and suboptimal performance. Given the diverse temporal patterns across sampling scales and the inherent multi-scale forecasting capabilities of TSFMs, we adopt a causal perspective to analyze finetuning process, through which we highlight the critical importance of explicitly modeling multiple scales and reveal the shortcomings of naive approaches. Focusing on \textit{encoder-based} TSFMs, we propose \textbf{M}ulti\textbf{\textsc{s}}cale \textbf{\textsc{f}}ine\textbf{\textsc{t}}uning (\textbf{MSFT}), a simple yet general framework that explicitly integrates multi-scale modeling into the finetuning process. Experimental results on three different backbones (\moirai, \moment\ and \units) demonstrate that TSFMs finetuned with MSFT not only outperform naive and typical parameter efficient finetuning methods but also surpass state-of-the-art deep learning methods. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:06:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Qiao",
"Zhongzheng",
""
],
[
"Liu",
"Chenghao",
""
],
[
"Zhang",
"Yiming",
""
],
[
"Jin",
"Ming",
""
],
[
"Pham",
"Quang",
""
],
[
"Wen",
"Qingsong",
""
],
[
"Suganthan",
"P. N.",
""
],
[
"Jiang",
"Xudong",
""
],
[
"Ramasamy",
"Savitha",
""
]
] |
2506.14088 | Kolja Knauer | Kolja Knauer, Alvaro Soto Gomez | What is and is not inside a Cayley graph? | 9 pages, 2 figures | null | null | null | math.CO cs.DM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this note we show that there is a cubic graph of girth $5$ that is not a subgraph of any minimal Cayley graph. On the other hand, we show that any Generalized Petersen Graph $G(n,k)$ with $\gcd(n,k)=1$ is an induced subgraph of a minimal Cayley graph. These results give insights into two comments of L\'aszl\'o Babai in [L. Babai, \emph{Automorphism groups, isomorphism, reconstruction}. Graham, R. L. (ed.) et al., Handbook of combinatorics. Vol. 1-2, 1994]. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:06:54 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Knauer",
"Kolja",
""
],
[
"Gomez",
"Alvaro Soto",
""
]
] |
2506.14089 | Lao Tse L\'opez PhD | A. Carrillo-Monteverde and L. L\'opez-Lozano | Dark Matter signals in solar neutrinos fluxes as probe of non-linear symmetry breaking | 15 page, 15 figures | null | null | null | hep-ph astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Dark matter (DM) particles gravitationally captured by the Sun can accumulate in its core and subsequently annihilate, producing neutrino fluxes that may be detectable on Earth. The intensity of these fluxes is highly sensitive to the properties of the underlying DM model, especially when the DM candidate is a scalar particle originating from spontaneous or non-linear symmetry breaking mechanisms. In this work, we explore the potential of solar neutrino fluxes to distinguish between the Standard Model extended by a scalar singlet and the non-linear Higgs portal scenarios in the context of a future DM discovery. We compute the expected neutrino fluxes within the regions of parameter space consistent with both relic density and current direct detection limits. Our results show that the non-linear model predicts neutrino fluxes that are systematically larger than those of the linear case, typically by at least one order of magnitude, and up to six orders of magnitude for DM masses around 1 TeV. These findings suggest that solar neutrino observations could provide a valuable probe to discriminate between these competing dark matter frameworks. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:11:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Carrillo-Monteverde",
"A.",
""
],
[
"López-Lozano",
"L.",
""
]
] |
2506.14090 | Xichao Zhang | X. Zhang, C. Reichhardt, C.J.O. Reichhardt, Y. Zhou, Y. Xu, M. Mochizuki | Asymmetric Diffusion of Chiral Skyrmions | 6 pages, 3 figures | null | null | null | cond-mat.mes-hall cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Brownian dynamics of chiral matter in chiral environment may give rise to emergent phenomena that could not be observed in achiral environment. Here, we report the asymmetric diffusion of chiral skyrmions in two chambers separated by a chiral gate. The topology-determined Brownian gyromotion of skyrmions could lead to effective interactions between skyrmions and chamber walls with a sense of clockwise or counter-clockwise rotation. By fabricating a chiral gate separating two chambers, skyrmions could demonstrate asymmetric diffusion through the gate to approach a nonequilibrium state before reaching the thermal equilibrium. We focus on the asymmetric diffusion of skyrmions that depends on the gate chirality and opening width. Although the simulated diffusion of skyrmions is affected by the skyrmion density, which varies with time due to the diffusion and annihilation, the simulated outcomes are generally in line with simple theories assuming time-independent diffusion rates. Our results uncover asymmetric diffusive behaviors of chiral skyrmions interacting with chiral nanostructures, which will appeal to the wide audience interested in hard condensed matter systems, magnetism, active matter, and statistical physics. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:13:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"X.",
""
],
[
"Reichhardt",
"C.",
""
],
[
"Reichhardt",
"C. J. O.",
""
],
[
"Zhou",
"Y.",
""
],
[
"Xu",
"Y.",
""
],
[
"Mochizuki",
"M.",
""
]
] |
2506.14091 | Renhao Tian | Jianfeng Huang, Renhao Tian, Yulin Zhao | A Chebyshev criterion for at most two non-zero limit cycles in Abel equations | null | null | null | null | math.CA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper investigates the Abel equation $\dot{x}=A(t)x^{3}+B(t)x^{2}$ on an interval $[0,T]$. The Smale-Pugh problem asks whether the maximum number of limit cycles of the equation is bounded in terms of a given class of coefficients. We establish for the first time a Chebyshev criterion, providing a positive answer to the problem when this class spanned by an extended Chebyshev system (ET-system) $\mathcal{F}=\{f_{0},f_{1},f_{2}\}$ on $[0,T)$ with $f_{0}\not=0$.
As an application, we prove that the equation has at most three limit cycles (including $x=0$) when the coefficients $A$ and $B$ are both linear trigonometric functions or quadratic polynomials. This reestablishes the result of Yu et al. (J. Differ. Equ., 2024) and improves the work of Bravo et al. (Disc. Cont. Dyn. Syst., 2015 \& J. Differ. Equ., 2024). We also obtain the same maximum number of limit cycles for the equation with trinomial coefficients. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:14:20 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Huang",
"Jianfeng",
""
],
[
"Tian",
"Renhao",
""
],
[
"Zhao",
"Yulin",
""
]
] |
2506.14092 | Haonan Yin | Haonan Yin, Shai Vardi, Vidyanand Choudhary | Fragile Preferences: A Deep Dive Into Order Effects in Large Language Models | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Large language models (LLMs) are increasingly used in decision-support systems across high-stakes domains such as hiring and university admissions, where decisions often involve selecting among competing alternatives. While prior work has noted positional order biases in LLM-driven comparisons, these biases have not been systematically dissected or linked to underlying preference structures. We provide the first comprehensive investigation of positional biases across multiple LLM architectures and domains, uncovering strong and consistent order effects, including a novel centrality bias not previously documented in human or machine decision-making. We also find a quality-dependent shift: when options are high quality, models exhibit primacy bias, but favor latter options when option quality is low. We further identify a previously undocumented bias favoring certain names over others. To distinguish superficial tie-breaking from true distortions of judgment, we introduce a framework that classifies pairwise preferences as robust, fragile, or indifferent. We show that order effects can lead models to select strictly inferior options, and that positional biases are typically stronger than gender biases. These findings suggest that LLMs are not merely inheriting human-like biases, but exhibit distinct failure modes not seen in human decision-making. We propose targeted mitigation strategies, including a novel use of the temperature parameter, to reduce order-driven distortions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:14:22 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yin",
"Haonan",
""
],
[
"Vardi",
"Shai",
""
],
[
"Choudhary",
"Vidyanand",
""
]
] |
2506.14093 | Rui Yang | Rui Yang, Christopher J. Howland, Hao-Ran Liu, Roberto Verzicco, Detlef Lohse | Bistability in radiatively heated melt ponds | null | null | 10.1103/PhysRevLett.131.234002 | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Melting and solidification processes, intertwined with convective flows, play a fundamental role in geophysical contexts. One of these processes is the formation of melt ponds on glaciers, ice shelves, and sea ice. It is driven by solar radiation and is of great significance for the Earth's heat balance, as it significantly lowers the albedo. Through direct numerical simulations and theoretical analysis, we unveil a bistability phenomenon in the melt pond dynamics. As solar radiation intensity and the melt pond's initial depth vary, an abrupt transition occurs: This tipping point transforms the system from a stable fully frozen state to another stable equilibrium state, characterized by a distinct melt pond depth. The physics of this transition can be understood within a heat flux balance model, which exhibits excellent agreement with our numerical results. Together with the Grossmann-Lohse theory for internally heated convection, the model correctly predicts the bulk temperature and the flow strength within the melt ponds, offering insight into the coupling of phase transitions with adjacent turbulent flows and the interplay between convective melting and radiation-driven processes. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:17:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yang",
"Rui",
""
],
[
"Howland",
"Christopher J.",
""
],
[
"Liu",
"Hao-Ran",
""
],
[
"Verzicco",
"Roberto",
""
],
[
"Lohse",
"Detlef",
""
]
] |
2506.14094 | Thomas Hancock | Thomas O. Hancock and John Buckell | A generalised discrete mixture model to better capture preference heterogeneity in discrete choice data | HESG conference paper | null | null | null | stat.ME | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Arguably the key issue in modelling discrete choice data is capturing preference heterogeneity. This can be through observed characteristics, and/or using techniques for capturing random heterogeneity across respondents. On the latter, in health economics, the two main approaches are the mixed multinomial logit (MMNL) and the latent class (LC) model. In this paper, we revisit the discrete mixture (DM) model as a third alternative to these. The DM model is similar to LC but allows for any combination of preferences across attributes, rather than grouping preferences as is the case in LC. We next develop a generalised discrete mixture (GDM) model. Additional boosting parameters in the class allocation component allow the model to collapse to a standard DM or LC structure as best fits the data at hand. This means that the model, by definition, performs at least as well as the best of a standard DM and a LC model; or better than both. Additional benefits include that it (a) allows the data to tell us the underlying correlations of preferences, (b) does not rely on distributions as is the case for mixed logit models, meaning estimation times are reduced and it does not require assumptions on the distribution of preferences. Exercises on simulated data show the unlikely conditions under which a LC model would be preferred to a DM. The convention of labelling latent classes, we believe, is questionable in many cases. The GDM is suitable in all cases. We show in empirical data that the GDM substantially outperforms LC models, granting a more detailed depiction of respondents' preferences. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:18:29 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hancock",
"Thomas O.",
""
],
[
"Buckell",
"John",
""
]
] |
2506.14096 | Ibne Farabi Shihab | Sanjeda Akter, Ibne Farabi Shihab, Anuj Sharma | Image Segmentation with Large Language Models: A Survey with Perspectives for Intelligent Transportation Systems | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The integration of Large Language Models (LLMs) with computer vision is profoundly transforming perception tasks like image segmentation. For intelligent transportation systems (ITS), where accurate scene understanding is critical for safety and efficiency, this new paradigm offers unprecedented capabilities. This survey systematically reviews the emerging field of LLM-augmented image segmentation, focusing on its applications, challenges, and future directions within ITS. We provide a taxonomy of current approaches based on their prompting mechanisms and core architectures, and we highlight how these innovations can enhance road scene understanding for autonomous driving, traffic monitoring, and infrastructure maintenance. Finally, we identify key challenges, including real-time performance and safety-critical reliability, and outline a perspective centered on explainable, human-centric AI as a prerequisite for the successful deployment of this technology in next-generation transportation systems. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:20:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Akter",
"Sanjeda",
""
],
[
"Shihab",
"Ibne Farabi",
""
],
[
"Sharma",
"Anuj",
""
]
] |
2506.14097 | Bryce Palmer | Bryce Palmer, Hasan Metin Aktulga, Tong Gao | ReLCP: Scalable Complementarity-Based Collision Resolution for Smooth Rigid Bodies | null | null | null | null | cs.RO cond-mat.soft physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a complementarity-based collision resolution algorithm for smooth, non-spherical, rigid bodies. Unlike discrete surface representation approaches, which approximate surfaces using discrete elements (e.g., tessellations or sub-spheres) with constraints between nearby faces, edges, nodes, or sub-objects, our algorithm solves a recursively generated linear complementarity problem (ReLCP) to adaptively identify potential collision locations during the collision resolution procedure. Despite adaptively and in contrast to Newton-esque schemes, we prove conditions under which the resulting solution exists and the center of mass translational and rotational dynamics are unique. Our ReLCP also converges to classical LCP-based collision resolution for sufficiently small timesteps. Because increasing the surface resolution in discrete representation methods necessitates subdividing geometry into finer elements -- leading to a super-linear increase in the number of collision constraints -- these approaches scale poorly with increased surface resolution. In contrast, our adaptive ReLCP framework begins with a single constraint per pair of nearby bodies and introduces new constraints only when unconstrained motion would lead to overlap, circumventing the oversampling required by discrete methods. By requiring one to two orders of magnitude fewer collision constraints to achieve the same surface resolution, we observe 10-100x speedup in densely packed applications. We validate our ReLCP method against multisphere and single-constraint methods, comparing convergence in a two-ellipsoid collision test, scalability and performance in a compacting ellipsoid suspension and growing bacterial colony, and stability in a taut chainmail network, highlighting our ability to achieve high-fidelity surface representations without suffering from poor scalability or artificial surface roughness. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:25:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Palmer",
"Bryce",
""
],
[
"Aktulga",
"Hasan Metin",
""
],
[
"Gao",
"Tong",
""
]
] |
2506.14098 | Ziyuan Tang | Ziyuan Tang, Jie Chen | Toward a Graph Foundation Model: Pre-Training Transformers With Random Walks | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | A foundation model like GPT elicits many emergent abilities, owing to the pre-training with broad inclusion of data and the use of the powerful Transformer architecture. While foundation models in natural languages are prevalent, can we build similar models for graphs? This paper describes an approach toward a graph foundation model that is pre-trained with diverse graph datasets by adapting the Transformer backbone. A central challenge toward this end is how a sequence model encodes graphs of varying sizes and from different domains. We propose representing a node as multiple random walks, such that the Transformer can extract node representations from sequences, which in turn form edge and graph representations. We develop a novel context prediction loss for these random walks and theoretically analyze their expressive power in distinguishing neighborhoods and graphs. We also demonstrate the pre-training of our model and its adaptation to downstream tasks, showcasing its potential as a foundation for processing and reasoning with graph-structured data. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:28:34 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Tang",
"Ziyuan",
""
],
[
"Chen",
"Jie",
""
]
] |
2506.14099 | Thomas Hancock | John Buckell and Alice Wreford and Matthew Quaife and Thomas O. Hancock | A break from the norm? Parametric representations of preference heterogeneity for discrete choice models in health | null | null | null | null | econ.EM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Background: Any sample of individuals has its own, unique distribution of preferences for choices that they make. Discrete choice models try to capture these distributions. Mixed logits are by far the most commonly used choice model in health. A raft of parametric model specifications for these models are available. We test a range of alternatives assumptions, and model averaging, to test if or how model outputs are impacted. Design: Scoping review of current modelling practices. Seven alternative distributions, and model averaging over all distributional assumptions, were compared on four datasets: two were stated preference, one was revealed preference, and one was simulated. Analyses examined model fit, preference distributions, willingness-to-pay, and forecasting. Results: Almost universally, using normal distributions is the standard practice in health. Alternative distributional assumptions outperformed standard practice. Preference distributions and the mean willingness-to-pay varied significantly across specifications, and were seldom comparable to those derived from normal distributions. Model averaging offered distributions allowed for greater flexibility, further gains in fit, reproduced underlying distributions in simulations, and mitigated against analyst bias arising from distribution selection. There was no evidence that distributional assumptions impacted predictions from models. Limitations: Our focus was on mixed logit models since these models are the most common in health, though latent class models are also used. Conclusions: The standard practice of using all normal distributions appears to be an inferior approach for capturing random preference heterogeneity. Implications: Researchers should test alternative assumptions to normal distributions in their models. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:28:38 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Buckell",
"John",
""
],
[
"Wreford",
"Alice",
""
],
[
"Quaife",
"Matthew",
""
],
[
"Hancock",
"Thomas O.",
""
]
] |
2506.14100 | Yupeng Zhou | Yupeng Zhou, Can Cui, Juntong Peng, Zichong Yang, Juanwu Lu, Jitesh H Panchal, Bin Yao, Ziran Wang | A Hierarchical Test Platform for Vision Language Model (VLM)-Integrated Real-World Autonomous Driving | null | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision-Language Models (VLMs) have demonstrated notable promise in autonomous driving by offering the potential for multimodal reasoning through pretraining on extensive image-text pairs. However, adapting these models from broad web-scale data to the safety-critical context of driving presents a significant challenge, commonly referred to as domain shift. Existing simulation-based and dataset-driven evaluation methods, although valuable, often fail to capture the full complexity of real-world scenarios and cannot easily accommodate repeatable closed-loop testing with flexible scenario manipulation. In this paper, we introduce a hierarchical real-world test platform specifically designed to evaluate VLM-integrated autonomous driving systems. Our approach includes a modular, low-latency on-vehicle middleware that allows seamless incorporation of various VLMs, a clearly separated perception-planning-control architecture that can accommodate both VLM-based and conventional modules, and a configurable suite of real-world testing scenarios on a closed track that facilitates controlled yet authentic evaluations. We demonstrate the effectiveness of the proposed platform`s testing and evaluation ability with a case study involving a VLM-enabled autonomous vehicle, highlighting how our test framework supports robust experimentation under diverse conditions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:32:28 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhou",
"Yupeng",
""
],
[
"Cui",
"Can",
""
],
[
"Peng",
"Juntong",
""
],
[
"Yang",
"Zichong",
""
],
[
"Lu",
"Juanwu",
""
],
[
"Panchal",
"Jitesh H",
""
],
[
"Yao",
"Bin",
""
],
[
"Wang",
"Ziran",
""
]
] |
2506.14101 | Paul Landes | Paul Landes, Sitara Rao, Aaron Jeremy Chaise, Barbara Di Eugenio | Abstract Meaning Representation for Hospital Discharge Summarization | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The Achilles heel of Large Language Models (LLMs) is hallucination, which has drastic consequences for the clinical domain. This is particularly important with regards to automatically generating discharge summaries (a lengthy medical document that summarizes a hospital in-patient visit). Automatically generating these summaries would free physicians to care for patients and reduce documentation burden. The goal of this work is to discover new methods that combine language-based graphs and deep learning models to address provenance of content and trustworthiness in automatic summarization. Our method shows impressive reliability results on the publicly available Medical Information Mart for Intensive III (MIMIC-III) corpus and clinical notes written by physicians at Anonymous Hospital. rovide our method, generated discharge ary output examples, source code and trained models. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:33:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Landes",
"Paul",
""
],
[
"Rao",
"Sitara",
""
],
[
"Chaise",
"Aaron Jeremy",
""
],
[
"Di Eugenio",
"Barbara",
""
]
] |
2506.14102 | Thomas Hancock | John Buckell and Thomas Hancock | Numerical evaluation of deliberative discussions of the UK food system: stimuli, demographics, and opinion reversion | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | There is increasing acknowledgement - including from the UK government - of the benefit of employing deliberative processes (deliberative fora, citizens' juries, etc.). Evidence suggests that the qualitative reporting of deliberative fora are often unclear or imprecise. If this is the case, their value to policymakers could be diminished. In this study we develop numerical methods of deliberative processes to document people's preferences, as a complement to qualitative analysis. Data are taken from the Food Conversation, a nationwide public consultation on reformations of the food system comprising 345 members of the general public. Each participant attended 5 workshops, each with differing stimuli covering subtopics of the food system. In each workshop, individuals twice reported responsibility, from 0-10, for changing the food system for 5 stakeholders (governments, the food industry, supermarkets, farmers, individuals). Analyses examined individuals' perceptions of food system change responsibility. Governments were most responsible and farmers least so. We assessed variation by workshop content, and by demographics. Reported responsibility changed most for individuals, and changed least for the food industry. We devise a model to document a reversion effect, where shifts in perceptions on responsibility that occurred during workshops waned over time; this was strongest among those who intended to vote (rather than not to). These results can support qualitative analyses and inform food system policy development. These methods are readily adopted for any such deliberative process, allowing for statistical evaluation of whether they can induce opinion change. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:37:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Buckell",
"John",
""
],
[
"Hancock",
"Thomas",
""
]
] |
2506.14103 | Pratheepa Jeganathan | Rajitha Senanayake, Pratheepa Jeganathan | A Robust Nonparametric Framework for Detecting Repeated Spatial Patterns | 39 pages including an Appendix of 17 pages, 39 figures | null | null | null | stat.ME q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Identifying spatially contiguous clusters and repeated spatial patterns (RSP) characterized by similar underlying distributions that are spatially apart is a key challenge in modern spatial statistics. Existing constrained clustering methods enforce spatial contiguity but are limited in their ability to identify RSP. We propose a novel nonparametric framework that addresses this limitation by combining constrained clustering with a post-clustering reassigment step based on the maximum mean discrepancy (MMD) statistic. We employ a block permutation strategy within each cluster that preserves local attribute structure when approximating the null distribution of the MMD. We also show that the MMD$^2$ statistic is asymptotically consistent under second-order stationarity and spatial mixing conditions. This two-stage approach enables the detection of clusters that are both spatially distant and similar in distribution. Through simulation studies that vary spatial dependence, cluster sizes, shapes, and multivariate dimensionality, we demonstrate the robustness of our proposed framework in detecting RSP. We further illustrate its applicability through an analysis of spatial proteomics data from patients with triple-negative breast cancer. Overall, our framework presents a methodological advancement in spatial clustering, offering a flexible and robust solution for spatial datasets that exhibit repeated patterns. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:38:28 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Senanayake",
"Rajitha",
""
],
[
"Jeganathan",
"Pratheepa",
""
]
] |
2506.14104 | Longdi Xian | RuiKun Yang and ZhongLiang Wei and Longdi Xian | Innovating China's Intangible Cultural Heritage with DeepSeek + MidJourney: The Case of Yangliuqing theme Woodblock Prints | null | null | null | null | cs.GR cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Yangliuqing woodblock prints, a cornerstone of China's intangible cultural heritage, are celebrated for their intricate designs and vibrant colors. However, preserving these traditional art forms while fostering innovation presents significant challenges. This study explores the DeepSeek + MidJourney approach to generating creative, themed Yangliuqing woodblock prints focused on the fight against COVID-19 and depicting joyous winners. Using Fr\'echet Inception Distance (FID) scores for evaluation, the method that combined DeepSeek-generated thematic prompts, MidJourney-generated thematic images, original Yangliuqing prints, and DeepSeek-generated key prompts in MidJourney-generated outputs achieved the lowest mean FID score (150.2) with minimal variability ({\sigma} = 4.9). Additionally, feedback from 62 participants, collected via questionnaires, confirmed that this hybrid approach produced the most representative results. Moreover, the questionnaire data revealed that participants demonstrated the highest willingness to promote traditional culture and the strongest interest in consuming the AI-generated images produced through this method. These findings underscore the effectiveness of an innovative approach that seamlessly blends traditional artistic elements with modern AI-driven creativity, ensuring both cultural preservation and contemporary relevance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:47:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yang",
"RuiKun",
""
],
[
"Wei",
"ZhongLiang",
""
],
[
"Xian",
"Longdi",
""
]
] |
2506.14105 | Qipeng Qian | Qipeng Qian, Christos N. Gagatsos | The effect of partial post-selection on quantum discrimination | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The discrimination of quantum states is a central problem in quantum information science and technology. Meanwhile, partial post-selection has emerged as a valuable tool for quantum state engineering. In this work, we bring these two areas together and ask whether partial measurements can, on average, enhance the discrimination performance between two unknown and non-orthogonal pure states. Our framework is general: the two unknown states interact with the same environment-set in a pure state-via an arbitrary unitary transformation. A partial measurement is then performed on one of the output modes, modeled by an arbitrary positive operator-valued measure (POVM). We then allow classical communication to inform the unmeasured mode of the outcome of the partial measurement on the other mode, which is subsequently measured by a POVM that is optimal in the sense that the probability of error is minimized. The two POVMs act locally, but since we allow for classical communication between the two modes, we consider a scheme involving local operations and classical communication (LOCC). Under these considerations, we show that the minimum error probability, averaged over all possible conditional states, cannot be reduced below the minimum error probability of discriminating the original input states. In other words, prior state engineering via partial measurement does not improve the average discrimination performance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:49:58 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Qian",
"Qipeng",
""
],
[
"Gagatsos",
"Christos N.",
""
]
] |
2506.14106 | Yuki Itahashi Dr | Yuki M. Itahashi, Yamato Nohara, Michiya Chazono, Hideki Matsuoka, Koichiro Arioka, Tetsuya Nomoto, Yoshimitsu Kohama, Youichi Yanase, Yoshihiro Iwasa, and Kaya Kobayashi | Misfit layered superconductor (PbSe)1.14(NbSe2)3 with possible layer-selective FFLO state | 23 pages, 5 figures | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Two-dimensional (2D) superconductors are known for their novel emergent phenomena, however, lack of experimental probes beyond resistivity has hindered further exploration of diverse superconducting states. Bulk 2D superconductors, with superconducting layers separated by non-superconducting layers, offer a unique opportunity to break this limit. Here, we synthesized a single crystal of misfit layered compound (PbSe)1.14(NbSe2)3, composed of alternately stacked tri-layer NbSe2 and non-superconducting block layers with incompatible unit cells. Due to its unique structure, 2D Ising superconductivity is maintained even in a bulk form. Resistivity and tunnel diode oscillator measurements reveal two distinct superconducting phases in magnetic field vs. temperature phase diagram. Combined with the theoretical analysis, the high-magnetic-field phase is identified as a layer-selective Fulde-Ferrell-Larkin-Ovchinnikov (FFLO) phase, where Ising and finite-q superconductivity are mixed due to the tri-layer structure. Bulk 2D superconductors with misfit structure offer a novel opportunity for understanding of 2D superconductivity through bulk measurements and interlayer engineering. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:55:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Itahashi",
"Yuki M.",
""
],
[
"Nohara",
"Yamato",
""
],
[
"Chazono",
"Michiya",
""
],
[
"Matsuoka",
"Hideki",
""
],
[
"Arioka",
"Koichiro",
""
],
[
"Nomoto",
"Tetsuya",
""
],
[
"Kohama",
"Yoshimitsu",
""
],
[
"Yanase",
"Youichi",
""
],
[
"Iwasa",
"Yoshihiro",
""
],
[
"Kobayashi",
"Kaya",
""
]
] |
2506.14107 | Jinwoo Hwang | Jinwoo Hwang, Daeun Kim, Sangyeop Lee, Yoonsung Kim, Guseul Heo, Hojoon Kim, Yunseok Jeong, Tadiwos Meaza, Eunhyeok Park, Jeongseob Ahn, Jongse Park | D\'ej\`a Vu: Efficient Video-Language Query Engine with Learning-based Inter-Frame Computation Reuse | Accepted to 2025 VLDB | null | null | null | cs.DC cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recently, Video-Language Models (VideoLMs) have demonstrated remarkable capabilities, offering significant potential for flexible and powerful video query systems. These models typically rely on Vision Transformers (ViTs), which process video frames individually to extract visual embeddings. However, generating embeddings for large-scale videos requires ViT inferencing across numerous frames, posing a major hurdle to real-world deployment and necessitating solutions for integration into scalable video data management systems. This paper introduces D\'ej\`a Vu, a video-language query engine that accelerates ViT-based VideoLMs by reusing computations across consecutive frames. At its core is ReuseViT, a modified ViT model specifically designed for VideoLM tasks, which learns to detect inter-frame reuse opportunities, striking an effective balance between accuracy and reuse. Although ReuseViT significantly reduces computation, these savings do not directly translate into performance gains on GPUs. To overcome this, D\'ej\`a Vu integrates memory-compute joint compaction techniques that convert the FLOP savings into tangible performance gains. Evaluations on three VideoLM tasks show that D\'ej\`a Vu accelerates embedding generation by up to a 2.64x within a 2% error bound, dramatically enhancing the practicality of VideoLMs for large-scale video analytics. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:59:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hwang",
"Jinwoo",
""
],
[
"Kim",
"Daeun",
""
],
[
"Lee",
"Sangyeop",
""
],
[
"Kim",
"Yoonsung",
""
],
[
"Heo",
"Guseul",
""
],
[
"Kim",
"Hojoon",
""
],
[
"Jeong",
"Yunseok",
""
],
[
"Meaza",
"Tadiwos",
""
],
[
"Park",
"Eunhyeok",
""
],
[
"Ahn",
"Jeongseob",
""
],
[
"Park",
"Jongse",
""
]
] |
2506.14108 | Siyi Wang | Siyi Wang, Alexandre Leblanc, Paul D. McNicholas | $\beta$-integrated local depth and corresponding partitioned local depth representation | null | null | null | null | math.ST stat.ME stat.TH | http://creativecommons.org/licenses/by/4.0/ | A novel local depth definition, $\beta$-integrated local depth ($\beta$-ILD), is proposed as a generalization of the local depth introduced by Paindaveine and Van Bever \cite{paindaveine2013depth}, designed to quantify the local centrality of data points. $\beta$-ILD inherits desirable properties from global data depth and remains robust across varying locality levels. A partitioning approach for $\beta$-ILD is introduced, leading to the construction of a matrix that quantifies the contribution of one point to another's local depth, providing a new interpretable measure of local centrality. These concepts are applied to classification and outlier detection tasks, demonstrating significant improvements in the performance of depth-based algorithms. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:59:41 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Siyi",
""
],
[
"Leblanc",
"Alexandre",
""
],
[
"McNicholas",
"Paul D.",
""
]
] |
2506.14109 | Rakesh Singh | Abhilasha Kumari, Gadha Ramesh, Debasish Koner, Rakesh S. Singh, and Mantu Santra | Compositional fluctuations and polymorph selection in crystallization of model soft colloids | null | null | null | null | cond-mat.soft | http://creativecommons.org/licenses/by/4.0/ | Understanding polymorph selection in atomic and molecular systems and its control through thermodynamic conditions and external factors (such as seed characteristics) is fundamental to the design of targeted materials and holds great significance in materials sciences. In this work, using Monte Carlo simulations on the Gaussian Core Model and Hard-Core Yukawa colloidal systems, we investigated the control of polymorph selection and explored the underlying mechanisms by tuning thermodynamic parameters. We demonstrate that by carefully modifying the free energy landscape to render the globally stable face-centered cubic (FCC) phase metastable with respect to the body-centered cubic (BCC) phase, the polymorphic identity of particles transitions from FCC-dominated to BCC-dominated via an intermediate regime where both phases nucleate -- either selectively or competitively -- giving rise to a critical-like composition fluctuation of the growing solid-like cluster during the nucleation process. We further probed the critical solid-like cluster compositions, especially in the vicinity of the triple point where the three phases coexist, and observed an interpenetrating arrangement of FCC- and BCC-like particles rather than a commonly observed non-classical core-shell-like two-step nucleation scenario. In addition, we investigated the polymorph selection signatures encoded in local structural fluctuations of the metastable fluid using a machine learning approach based on structural descriptors derived from persistent homology, a topological data analysis method. We believe that the insights gained from this work have the potential to add to the ongoing efforts to control crystallization pathways to obtain the desired functional material. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 01:59:54 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Kumari",
"Abhilasha",
""
],
[
"Ramesh",
"Gadha",
""
],
[
"Koner",
"Debasish",
""
],
[
"Singh",
"Rakesh S.",
""
],
[
"Santra",
"Mantu",
""
]
] |
2506.14110 | Mingyue Xu | Steve Hanneke and Mingyue Xu | Universal Rates of ERM for Agnostic Learning | Accepted for presentation at the Conference on Learning Theory (COLT) 2025 | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | The universal learning framework has been developed to obtain guarantees on the learning rates that hold for any fixed distribution, which can be much faster than the ones uniformly hold over all the distributions. Given that the Empirical Risk Minimization (ERM) principle being fundamental in the PAC theory and ubiquitous in practical machine learning, the recent work of arXiv:2412.02810 studied the universal rates of ERM for binary classification under the realizable setting. However, the assumption of realizability is too restrictive to hold in practice. Indeed, the majority of the literature on universal learning has focused on the realizable case, leaving the non-realizable case barely explored.
In this paper, we consider the problem of universal learning by ERM for binary classification under the agnostic setting, where the ''learning curve" reflects the decay of the excess risk as the sample size increases. We explore the possibilities of agnostic universal rates and reveal a compact trichotomy: there are three possible agnostic universal rates of ERM, being either $e^{-n}$, $o(n^{-1/2})$, or arbitrarily slow. We provide a complete characterization of which concept classes fall into each of these categories. Moreover, we also establish complete characterizations for the target-dependent universal rates as well as the Bayes-dependent universal rates. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:02:11 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hanneke",
"Steve",
""
],
[
"Xu",
"Mingyue",
""
]
] |
2506.14111 | Andrew Hojel | Essential AI: Andrew Hojel, Michael Pust, Tim Romanski, Yash Vanjani, Ritvik Kapila, Mohit Parmar, Adarsh Chaluvaraju, Alok Tripathy, Anil Thomas, Ashish Tanwer, Darsh J Shah, Ishaan Shah, Karl Stratos, Khoi Nguyen, Kurt Smith, Michael Callahan, Peter Rushton, Philip Monk, Platon Mazarakis, Saad Jamal, Saurabh Srivastava, Somanshu Singla, Ashish Vaswani | Essential-Web v1.0: 24T tokens of organized web data | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Data plays the most prominent role in how language models acquire skills and knowledge. The lack of massive, well-organized pre-training datasets results in costly and inaccessible data pipelines. We present Essential-Web v1.0, a 24-trillion-token dataset in which every document is annotated with a twelve-category taxonomy covering topic, format, content complexity, and quality. Taxonomy labels are produced by EAI-Distill-0.5b, a fine-tuned 0.5b-parameter model that achieves an annotator agreement within 3% of Qwen2.5-32B-Instruct. With nothing more than SQL-style filters, we obtain competitive web-curated datasets in math (-8.0% relative to SOTA), web code (+14.3%), STEM (+24.5%) and medical (+8.6%). Essential-Web v1.0 is available on HuggingFace: https://huggingface.co/datasets/EssentialAI/essential-web-v1.0 | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:03:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"AI",
"Essential",
""
],
[
":",
"",
""
],
[
"Hojel",
"Andrew",
""
],
[
"Pust",
"Michael",
""
],
[
"Romanski",
"Tim",
""
],
[
"Vanjani",
"Yash",
""
],
[
"Kapila",
"Ritvik",
""
],
[
"Parmar",
"Mohit",
""
],
[
"Chaluvaraju",
"Adarsh",
""
],
[
"Tripathy",
"Alok",
""
],
[
"Thomas",
"Anil",
""
],
[
"Tanwer",
"Ashish",
""
],
[
"Shah",
"Darsh J",
""
],
[
"Shah",
"Ishaan",
""
],
[
"Stratos",
"Karl",
""
],
[
"Nguyen",
"Khoi",
""
],
[
"Smith",
"Kurt",
""
],
[
"Callahan",
"Michael",
""
],
[
"Rushton",
"Peter",
""
],
[
"Monk",
"Philip",
""
],
[
"Mazarakis",
"Platon",
""
],
[
"Jamal",
"Saad",
""
],
[
"Srivastava",
"Saurabh",
""
],
[
"Singla",
"Somanshu",
""
],
[
"Vaswani",
"Ashish",
""
]
] |
2506.14112 | Si Li | Hengyu Liu, Yanhong Luo, Congcong Wu, Yin Guan, Ahmed Lotfy Elrefai, Andreas Elombo, Si Li, Sahban Wael Saeed Alnaser, Mingyu Yan | Considering the multi-time scale rolling optimization scheduling method of micro-energy network connected to electric vehicles | 7 pages,9 figures,1 table,conference | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The large-scale access of electric vehicles to the power grid not only provides flexible adjustment resources for the power system, but the temporal uncertainty and distribution complexity of their energy interaction pose significant challenges to the economy and robustness of the micro-energy network. In this paper, we propose a multi-time scale rolling optimization scheduling method for micro-energy networks considering the access of electric vehicles. In order to solve the problem of evaluating the dispatchable potential of electric vehicle clusters, a charging station aggregation model was constructed based on Minkowski summation theory, and the scattered electric vehicle resources were aggregated into virtual energy storage units to participate in system scheduling. Integrate price-based and incentive-based demand response mechanisms to synergistically tap the potential of source-load two-side regulation; On this basis, a two-stage optimal scheduling model of day-ahead and intra-day is constructed. The simulation results show that the proposed method reduces the scale of "preventive curtailment" due to more accurate scheduling, avoids the threat of power shortage to the safety of the power grid, and has more advantages in the efficiency of new energy consumption. At the same time, intra-day scheduling significantly reduces economic penalties and operating costs by avoiding output shortages, and improves the economy of the system in an uncertain forecasting environment. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:05:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Liu",
"Hengyu",
""
],
[
"Luo",
"Yanhong",
""
],
[
"Wu",
"Congcong",
""
],
[
"Guan",
"Yin",
""
],
[
"Elrefai",
"Ahmed Lotfy",
""
],
[
"Elombo",
"Andreas",
""
],
[
"Li",
"Si",
""
],
[
"Alnaser",
"Sahban Wael Saeed",
""
],
[
"Yan",
"Mingyu",
""
]
] |
2506.14113 | Yitian Zhang | Yitian Zhang, Liheng Ma, Antonios Valkanas, Boris N. Oreshkin, Mark Coates | SKOLR: Structured Koopman Operator Linear RNN for Time-Series Forecasting | null | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Koopman operator theory provides a framework for nonlinear dynamical system analysis and time-series forecasting by mapping dynamics to a space of real-valued measurement functions, enabling a linear operator representation. Despite the advantage of linearity, the operator is generally infinite-dimensional. Therefore, the objective is to learn measurement functions that yield a tractable finite-dimensional Koopman operator approximation. In this work, we establish a connection between Koopman operator approximation and linear Recurrent Neural Networks (RNNs), which have recently demonstrated remarkable success in sequence modeling. We show that by considering an extended state consisting of lagged observations, we can establish an equivalence between a structured Koopman operator and linear RNN updates. Building on this connection, we present SKOLR, which integrates a learnable spectral decomposition of the input signal with a multilayer perceptron (MLP) as the measurement functions and implements a structured Koopman operator via a highly parallel linear RNN stack. Numerical experiments on various forecasting benchmarks and dynamical systems show that this streamlined, Koopman-theory-based design delivers exceptional performance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:11:06 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Yitian",
""
],
[
"Ma",
"Liheng",
""
],
[
"Valkanas",
"Antonios",
""
],
[
"Oreshkin",
"Boris N.",
""
],
[
"Coates",
"Mark",
""
]
] |
2506.14114 | Khushnood Abbas | Khushnood Abbas, Ruizhe Hou, Zhou Wengang, Dong Shi, Niu Ling, Satyaki Nan, Alireza Abbasi | Evaluating Loss Functions for Graph Neural Networks: Towards Pretraining and Generalization | ACM single column 633 pages | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph Neural Networks (GNNs) became useful for learning on non-Euclidean data. However, their best performance depends on choosing the right model architecture and the training objective, also called the loss function. Researchers have studied these parts separately, but a large-scale evaluation has not looked at how GNN models and many loss functions work together across different tasks. To fix this, we ran a thorough study - it included seven well-known GNN architectures. We also used a large group of 30 single plus mixed loss functions. The study looked at both inductive and transductive settings. Our evaluation spanned three distinct real-world datasets, assessing performance in both inductive and transductive settings using 21 comprehensive evaluation metrics. From these extensive results (detailed in supplementary information 1 \& 2), we meticulously analyzed the top ten model-loss combinations for each metric based on their average rank. Our findings reveal that, especially for the inductive case: 1) Hybrid loss functions generally yield superior and more robust performance compared to single loss functions, indicating the benefit of multi-objective optimization. 2) The GIN architecture always showed the highest-level average performance, especially with Cross-Entropy loss. 3) Although some combinations had overall lower average ranks, models such as GAT, particularly with certain hybrid losses, demonstrated incredible specialized strengths, maximizing the most top-1 results among the individual metrics, emphasizing subtle strengths for particular task demands. 4) On the other hand, the MPNN architecture typically lagged behind the scenarios it was tested against. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:12:19 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Abbas",
"Khushnood",
""
],
[
"Hou",
"Ruizhe",
""
],
[
"Wengang",
"Zhou",
""
],
[
"Shi",
"Dong",
""
],
[
"Ling",
"Niu",
""
],
[
"Nan",
"Satyaki",
""
],
[
"Abbasi",
"Alireza",
""
]
] |
2506.14115 | Shu-Min Wu | Shu-Min Wu, Yu-Xuan Wang, Wentao Liu | Entangled Unruh-DeWitt detectors amplify quantum coherence | 22 pages, 4 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We explore the quantum coherence between a pair of entangled Unruh-DeWitt detectors, interacting with a quantum field, using a nonperturbative approach in a (3+1)-dimensional Minkowski spacetime with instantaneous switching ($\delta$-switching). It is intriguing to observe that for a maximally entangled state, increasing the coupling strength enhances the detectors' initial quantum coherence while simultaneously causing a monotonic decrease in their initial entanglement. This reveals a remarkable phenomenon: through nonperturbative interactions, entangled Unruh-DeWitt detectors can exhibit a dual effect-amplifying quantum coherence while degrading quantum entanglement. This finding stands in stark contrast to previous studies based on perturbative methods or Gaussian switching functions, which generally concluded that interactions between detectors and the field lead to a simultaneous degradation of quantum coherence and entanglement due to environmental decoherence. Notably, while initially separable detectors successfully harvest quantum coherence from the vacuum, entanglement extraction remains fundamentally prohibited. These contrasting behaviors underscore the fundamental distinction between coherence and entanglement as quantum resources, and highlight their complementary roles in field-detector interactions. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:16:36 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wu",
"Shu-Min",
""
],
[
"Wang",
"Yu-Xuan",
""
],
[
"Liu",
"Wentao",
""
]
] |
2506.14116 | Rongyu Yu | Rongyu Yu, Kan Chen, Zeyu Deng, Chen Wang, Burak Kizilkaya, Liying Emma Li | Haptic-Based User Authentication for Tele-robotic System | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Tele-operated robots rely on real-time user behavior mapping for remote tasks, but ensuring secure authentication remains a challenge. Traditional methods, such as passwords and static biometrics, are vulnerable to spoofing and replay attacks, particularly in high-stakes, continuous interactions. This paper presents a novel anti-spoofing and anti-replay authentication approach that leverages distinctive user behavioral features extracted from haptic feedback during human-robot interactions. To evaluate our authentication approach, we collected a time-series force feedback dataset from 15 participants performing seven distinct tasks. We then developed a transformer-based deep learning model to extract temporal features from the haptic signals. By analyzing user-specific force dynamics, our method achieves over 90 percent accuracy in both user identification and task classification, demonstrating its potential for enhancing access control and identity assurance in tele-robotic systems. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:16:39 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Yu",
"Rongyu",
""
],
[
"Chen",
"Kan",
""
],
[
"Deng",
"Zeyu",
""
],
[
"Wang",
"Chen",
""
],
[
"Kizilkaya",
"Burak",
""
],
[
"Li",
"Liying Emma",
""
]
] |
2506.14117 | Ian Roberts | Ian D. Roberts, Michael L. Balogh, Visal Sok, Adam Muzzin, Michael J. Hudson, Pascale Jablonka | JWST Reveals a Likely Jellyfish Galaxy at z=1.156 | 12 pages, 8 figures, 1 appendix, submitted to AAS Journals | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | We report the discovery of COSMOS2020-635829 as a likely jellyfish galaxy undergoing to ram pressure stripping in a (proto)cluster at $z > 1$. High-resolution imaging from the James Webb Space Telescope reveals a symmetric stellar disk coupled to a unilateral tail of star-forming knots to the south. We show that these extra-planar continuum sources are embedded within an ionized gas tail that is kinematically connected to the disk of COSMOS2020-635829. Likely representing the highest-redshift discovery of a ram pressure stripped ionized gas tail. The tail sources are characterized by extremely young stellar populations ($\lesssim 100\,\mathrm{Myr}$), have stellar masses of ${\sim}10^8\,\mathrm{M_\odot}$, and star formation rates of $0.1\text{--}1\,\mathrm{M_\odot\,yr^{-1}}$. This work reinforces the notion that ram pressure stripping can perturb group and cluster galaxies at $z > 1$ and likely contributes to environmental quenching even near Cosmic Noon. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:18:17 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Roberts",
"Ian D.",
""
],
[
"Balogh",
"Michael L.",
""
],
[
"Sok",
"Visal",
""
],
[
"Muzzin",
"Adam",
""
],
[
"Hudson",
"Michael J.",
""
],
[
"Jablonka",
"Pascale",
""
]
] |
2506.14118 | Tran-Trung Nghiem | Ronan J. Conlon, Tran-Trung Nghiem | Examples of complete Calabi--Yau metrics on affine smoothings of irregular toric Calabi--Yau cones | 45 pages, 9 figures | null | null | null | math.DG math.AG math.CO | http://creativecommons.org/licenses/by/4.0/ | We present new examples of affine Calabi--Yau manifolds of Euclidean volume growth and quadratic curvature decay, whose tangent cones at infinity are irregular and have smooth links. In the process, we demonstrate (and provide the relevant computer code) how to explicitly compute the Reeb field and all Minkowski decompositions of a given toric Calabi--Yau cone with smooth link from the data of its toric polytope. Minkowski decompositions of this polytope into lattice segments and/or triangles give rise to smoothings of the given cone. Furthermore, we propose an effective strategy to generate smoothable Calabi--Yau cones from a given non-smoothable one by taking Minkowski sums of certain toric diagrams, and provide an example to illustrate the method. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:22:05 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Conlon",
"Ronan J.",
""
],
[
"Nghiem",
"Tran-Trung",
""
]
] |
2506.14119 | Meng Zhao | Meng Zhao | Level-3 large deviations for the white-forced 2D Navier-Stokes system in a bounded domain | null | null | null | null | math.AP math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the large deviations principle (LDP) of Donsker-Varadhan type for the white-forced Navier-Stokes system in a bounded domain. Under the assumption that the noise is non-degenerate, we establish level-2 and level-3 LDPs with rate functions given by the Donsker-Varadhan formulas. The proof relies on an improved version of Kifer's criterion, a lift argument inspired from [DV83], an improved abstract result on the large-time asymptotics of generalized Markov semigroups, and a delicate approximation scheme utilizing the resolvent operators of the Markov semigroup. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:26:50 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhao",
"Meng",
""
]
] |
2506.14120 | An Do | Ziheng Chen, Po T. Wang, Mina Ibrahim, Shivali Baveja, Rong Mu, An H. Do, and Zoran Nenadic | Leveraging Transfer Learning and User-Specific Updates for Rapid Training of BCI Decoders | 6 page conference proceeding preprint | null | null | null | q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Lengthy subject- or session-specific data acquisition and calibration remain a key barrier to deploying electroencephalography (EEG)-based brain-computer interfaces (BCIs) outside the laboratory. Previous work has shown that cross subject, cross-session invariant features exist in EEG. We propose a transfer learning pipeline based on a two-layer convolutional neural network (CNN) that leverages these invariants to reduce the burden of data acquisition and calibration. A baseline model is trained on EEG data from five able-bodied individuals and then rapidly updated with a small amount of data from a sixth, holdout subject. The remaining holdout data were used to test the performance of both the baseline and updated models. We repeated this procedure via a leave-one-subject out (LOSO) validation framework. Averaged over six LOSO folds, the updated model improved classification accuracy upon the baseline by 10.0, 18.8, and 22.1 percentage points on two binary and one ternary classification tasks, respectively. These results demonstrate that decoding accuracy can be substantially improved with minimal subject-specific data. They also indicate that a CNN-based decoder can be personalized rapidly, enabling near plug-and-play BCI functionality for neurorehabilitation and other time-critical EEG applications. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:28:31 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chen",
"Ziheng",
""
],
[
"Wang",
"Po T.",
""
],
[
"Ibrahim",
"Mina",
""
],
[
"Baveja",
"Shivali",
""
],
[
"Mu",
"Rong",
""
],
[
"Do",
"An H.",
""
],
[
"Nenadic",
"Zoran",
""
]
] |
2506.14121 | Guangwei Gao | Siyu Xu, Wenjie Li, Guangwei Gao, Jian Yang, Guo-Jun Qi, Chia-Wen Lin | FADPNet: Frequency-Aware Dual-Path Network for Face Super-Resolution | 12 pages, 11 figures, 6 tales | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Face super-resolution (FSR) under limited computational costs remains an open problem. Existing approaches typically treat all facial pixels equally, resulting in suboptimal allocation of computational resources and degraded FSR performance. CNN is relatively sensitive to high-frequency facial features, such as component contours and facial outlines. Meanwhile, Mamba excels at capturing low-frequency features like facial color and fine-grained texture, and does so with lower complexity than Transformers. Motivated by these observations, we propose FADPNet, a Frequency-Aware Dual-Path Network that decomposes facial features into low- and high-frequency components and processes them via dedicated branches. For low-frequency regions, we introduce a Mamba-based Low-Frequency Enhancement Block (LFEB), which combines state-space attention with squeeze-and-excitation operations to extract low-frequency global interactions and emphasize informative channels. For high-frequency regions, we design a CNN-based Deep Position-Aware Attention (DPA) module to enhance spatially-dependent structural details, complemented by a lightweight High-Frequency Refinement (HFR) module that further refines frequency-specific representations. Through the above designs, our method achieves an excellent balance between FSR quality and model efficiency, outperforming existing approaches. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:33:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Siyu",
""
],
[
"Li",
"Wenjie",
""
],
[
"Gao",
"Guangwei",
""
],
[
"Yang",
"Jian",
""
],
[
"Qi",
"Guo-Jun",
""
],
[
"Lin",
"Chia-Wen",
""
]
] |
2506.14122 | Renbo Zhang | Tianming Zhang, Renbo Zhang, Zhengyi Yang, Yunjun Gao, Bin Cao, Jing Fan | CLGNN: A Contrastive Learning-based GNN Model for Betweenness Centrality Prediction on Temporal Graphs | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Temporal Betweenness Centrality (TBC) measures how often a node appears on optimal temporal paths, reflecting its importance in temporal networks. However, exact computation is highly expensive, and real-world TBC distributions are extremely imbalanced. The severe imbalance leads learning-based models to overfit to zero-centrality nodes, resulting in inaccurate TBC predictions and failure to identify truly central nodes. Existing graph neural network (GNN) methods either fail to handle such imbalance or ignore temporal dependencies altogether. To address these issues, we propose a scalable and inductive contrastive learning-based GNN (CLGNN) for accurate and efficient TBC prediction. CLGNN builds an instance graph to preserve path validity and temporal order, then encodes structural and temporal features using dual aggregation, i.e., mean and edge-to-node multi-head attention mechanisms, enhanced by temporal path count and time encodings. A stability-based clustering-guided contrastive module (KContrastNet) is introduced to separate high-, median-, and low-centrality nodes in representation space, mitigating class imbalance, while a regression module (ValueNet) estimates TBC values. CLGNN also supports multiple optimal path definitions to accommodate diverse temporal semantics. Extensive experiments demonstrate the effectiveness and efficiency of CLGNN across diverse benchmarks. CLGNN achieves up to a 663.7~$\times$ speedup compared to state-of-the-art exact TBC computation methods. It outperforms leading static GNN baselines with up to 31.4~$\times$ lower MAE and 16.7~$\times$ higher Spearman correlation, and surpasses state-of-the-art temporal GNNs with up to 5.7~$\times$ lower MAE and 3.9~$\times$ higher Spearman correlation. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:34:09 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Tianming",
""
],
[
"Zhang",
"Renbo",
""
],
[
"Yang",
"Zhengyi",
""
],
[
"Gao",
"Yunjun",
""
],
[
"Cao",
"Bin",
""
],
[
"Fan",
"Jing",
""
]
] |
2506.14123 | Jonathan Hayase | Jonathan Hayase, Alisa Liu, Noah A. Smith, Sewoong Oh | Sampling from Your Language Model One Byte at a Time | 23 pages, 8 figures | null | null | null | cs.CL cs.FL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tokenization is used almost universally by modern language models, enabling efficient text representation using multi-byte or multi-character tokens. However, prior work has shown that tokenization can introduce distortion into the model's generations. For example, users are often advised not to end their prompts with a space because it prevents the model from including the space as part of the next token. This Prompt Boundary Problem (PBP) also arises in languages such as Chinese and in code generation, where tokens often do not line up with syntactic boundaries. Additionally mismatching tokenizers often hinder model composition and interoperability. For example, it is not possible to directly ensemble models with different tokenizers due to their mismatching vocabularies. To address these issues, we present an inference-time method to convert any autoregressive LM with a BPE tokenizer into a character-level or byte-level LM, without changing its generative distribution at the text level. Our method efficient solves the PBP and is also able to unify the vocabularies of language models with different tokenizers, allowing one to ensemble LMs with different tokenizers at inference time as well as transfer the post-training from one model to another using proxy-tuning. We demonstrate in experiments that the ensemble and proxy-tuned models outperform their constituents on downstream evals. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:37:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Hayase",
"Jonathan",
""
],
[
"Liu",
"Alisa",
""
],
[
"Smith",
"Noah A.",
""
],
[
"Oh",
"Sewoong",
""
]
] |
2506.14124 | Joachim Neu | Jovan Komatovic, Andrew Lewis-Pye, Joachim Neu, Tim Roughgarden, Ertem Nusret Tas | From Permissioned to Proof-of-Stake Consensus | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents the first generic compiler that transforms any permissioned consensus protocol into a proof-of-stake permissionless consensus protocol. For each of the following properties, if the initial permissioned protocol satisfies that property in the partially synchronous setting, the consequent proof-of-stake protocol also satisfies that property in the partially synchronous and quasi-permissionless setting (with the same fault-tolerance): consistency; liveness; optimistic responsiveness; every composable log-specific property; and message complexity of a given order. Moreover, our transformation ensures that the output protocol satisfies accountability (identifying culprits in the event of a consistency violation), whether or not the original permissioned protocol satisfied it. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:37:40 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Komatovic",
"Jovan",
""
],
[
"Lewis-Pye",
"Andrew",
""
],
[
"Neu",
"Joachim",
""
],
[
"Roughgarden",
"Tim",
""
],
[
"Tas",
"Ertem Nusret",
""
]
] |
2506.14125 | Libo Zhang | Libo Zhang, Yang Chen, Toru Takisaka, Kaiqi Zhao, Weidong Li, Jiamou Liu | Situational-Constrained Sequential Resources Allocation via Reinforcement Learning | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential Resource Allocation with situational constraints presents a significant challenge in real-world applications, where resource demands and priorities are context-dependent. This paper introduces a novel framework, SCRL, to address this problem. We formalize situational constraints as logic implications and develop a new algorithm that dynamically penalizes constraint violations. To handle situational constraints effectively, we propose a probabilistic selection mechanism to overcome limitations of traditional constraint reinforcement learning (CRL) approaches. We evaluate SCRL across two scenarios: medical resource allocation during a pandemic and pesticide distribution in agriculture. Experiments demonstrate that SCRL outperforms existing baselines in satisfying constraints while maintaining high resource efficiency, showcasing its potential for real-world, context-sensitive decision-making tasks. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:40:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Libo",
""
],
[
"Chen",
"Yang",
""
],
[
"Takisaka",
"Toru",
""
],
[
"Zhao",
"Kaiqi",
""
],
[
"Li",
"Weidong",
""
],
[
"Liu",
"Jiamou",
""
]
] |
2506.14126 | Stefan Horoi | Stefan Horoi, Guy Wolf, Eugene Belilovsky and Gintare Karolina Dziugaite | Less is More: Undertraining Experts Improves Model Upcycling | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Modern deep learning is increasingly characterized by the use of open-weight foundation models that can be fine-tuned on specialized datasets. This has led to a proliferation of expert models and adapters, often shared via platforms like HuggingFace and AdapterHub. To leverage these resources, numerous model upcycling methods have emerged, enabling the reuse of fine-tuned models in multi-task systems. A natural pipeline has thus formed to harness the benefits of transfer learning and amortize sunk training costs: models are pre-trained on general data, fine-tuned on specific tasks, and then upcycled into more general-purpose systems. A prevailing assumption is that improvements at one stage of this pipeline propagate downstream, leading to gains at subsequent steps. In this work, we challenge that assumption by examining how expert fine-tuning affects model upcycling. We show that long fine-tuning of experts that optimizes for their individual performance leads to degraded merging performance, both for fully fine-tuned and LoRA-adapted models, and to worse downstream results when LoRA adapters are upcycled into MoE layers. We trace this degradation to the memorization of a small set of difficult examples that dominate late fine-tuning steps and are subsequently forgotten during merging. Finally, we demonstrate that a task-dependent aggressive early stopping strategy can significantly improve upcycling performance. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:42:10 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Horoi",
"Stefan",
""
],
[
"Wolf",
"Guy",
""
],
[
"Belilovsky",
"Eugene",
""
],
[
"Dziugaite",
"Gintare Karolina",
""
]
] |
2506.14127 | Peng Gao | Xiaowen Zhang, Jiade Li, Xiaoyue Gao, Ruochen Shi, Bo Han, Xiaomei Li, Jinlong Du, Jinsheng Wen, Genda Gu, Shichong Wang, Wentao Zhang, Peng Gao | Resolving Phonons in Superconductor Bi2Sr2CaCu2O8+{\delta} at Sub-Unit-Cell Resolution | null | null | null | null | cond-mat.supr-con cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | The role of phonons in cuprates remains controversial, with their complex lattice structure complicating the investigation. Here, we identify phonon modes originating from charge reservoir and superconducting layers of Bi2Sr2CaCu2O8+{\delta} using sub-unit-cell resolved electron energy loss spectroscopy in a scanning transmission electron microscope. We reveal several phonon modes exhibiting layer-specific localization: ~78 meV in-plane modes localized in the CuO2 planes, ~42 meV hybrid (in-plane and out-of-plane) modes in the CuO2 planes, and ~38 meV hybrid modes in the BiO layers. We also observe a periodic modulation of phonon frequencies induced by the superstructure, spatially correlating with the superconducting gap variation. Our findings offer new insights into the electron-phonon coupling in cuprates, fostering deeper exploration of the microscopic linkage between lattice dynamics and superconductivity. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:43:45 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Xiaowen",
""
],
[
"Li",
"Jiade",
""
],
[
"Gao",
"Xiaoyue",
""
],
[
"Shi",
"Ruochen",
""
],
[
"Han",
"Bo",
""
],
[
"Li",
"Xiaomei",
""
],
[
"Du",
"Jinlong",
""
],
[
"Wen",
"Jinsheng",
""
],
[
"Gu",
"Genda",
""
],
[
"Wang",
"Shichong",
""
],
[
"Zhang",
"Wentao",
""
],
[
"Gao",
"Peng",
""
]
] |
2506.14128 | Wen Zheng | Jianwen Xu, Xiang Deng, Wen Zheng, Wenchang Yan, Tao Zhang, Zhenchuan Zhang, Wanli Huang, Xiaoyu Xia, Xudong Liao, Yu Zhang, Jie Zhao, Shaoxiong Li, Xinsheng Tan, Dong Lan, and Yang Yu | Tunable Hybrid-Mode Coupler Enabling Strong Interactions between Transmons at Centimeter-Scale Distance | 12 pages, 7 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The transmon, a fabrication-friendly superconducting qubit, remains a leading candidate for scalable quantum computing. Recent advances in tunable couplers have accelerated progress toward high-performance quantum processors. However, extending coherent interactions beyond millimeter scales to enhance quantum connectivity presents a critical challenge. Here, we introduce a hybrid-mode coupler exploiting resonator-transmon hybridization to simultaneously engineer the two lowest-frequency mode, enabling high-contrast coupling between centimeter-scale transmons. For a 1-cm coupler, our framework predicts flux-tunable $XX$ and $ZZ$ coupling strengths reaching 23 MHz and 100 MHz, with modulation contrasts exceeding $10^2$ and $10^4$, respectively, demonstrating quantitative agreement with an effective two-channel model. This work provides an efficient pathway to mitigate the inherent connectivity constraints imposed by short-range interactions, enabling transmon-based architectures compatible with hardware-efficient quantum tasks. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:45:16 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Xu",
"Jianwen",
""
],
[
"Deng",
"Xiang",
""
],
[
"Zheng",
"Wen",
""
],
[
"Yan",
"Wenchang",
""
],
[
"Zhang",
"Tao",
""
],
[
"Zhang",
"Zhenchuan",
""
],
[
"Huang",
"Wanli",
""
],
[
"Xia",
"Xiaoyu",
""
],
[
"Liao",
"Xudong",
""
],
[
"Zhang",
"Yu",
""
],
[
"Zhao",
"Jie",
""
],
[
"Li",
"Shaoxiong",
""
],
[
"Tan",
"Xinsheng",
""
],
[
"Lan",
"Dong",
""
],
[
"Yu",
"Yang",
""
]
] |
2506.14129 | Shuchang Wang | Shuchang Wang, Xiaopeng Qiu, Yingxing Xue, Yanfu Li and Wei Yang | A Quantum Annealing Approach for Solving Optimal Feature Selection and Next Release Problems | null | null | null | null | cs.SE quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Search-based software engineering (SBSE) addresses critical optimization challenges in software engineering, including the next release problem (NRP) and feature selection problem (FSP). While traditional heuristic approaches and integer linear programming (ILP) methods have demonstrated efficacy for small to medium-scale problems, their scalability to large-scale instances remains unknown. Here, we introduce quantum annealing (QA) as a subroutine to tackling multi-objective SBSE problems, leveraging the computational potential of quantum systems. We propose two QA-based algorithms tailored to different problem scales. For small-scale problems, we reformulate multi-objective optimization (MOO) as single-objective optimization (SOO) using penalty-based mappings for quantum processing. For large-scale problems, we employ a decomposition strategy guided by maximum energy impact (MEI), integrating QA with a steepest descent method to enhance local search efficiency. Applied to NRP and FSP, our approaches are benchmarked against the heuristic NSGA-II and the ILP-based $\epsilon$-constraint method. Experimental results reveal that while our methods produce fewer non-dominated solutions than $\epsilon$-constraint, they achieve significant reductions in execution time. Moreover, compared to NSGA-II, our methods deliver more non-dominated solutions with superior computational efficiency. These findings underscore the potential of QA in advancing scalable and efficient solutions for SBSE challenges. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:46:53 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Wang",
"Shuchang",
""
],
[
"Qiu",
"Xiaopeng",
""
],
[
"Xue",
"Yingxing",
""
],
[
"Li",
"Yanfu",
""
],
[
"Yang",
"Wei",
""
]
] |
2506.14130 | Chunyu Cao | Chunyu Cao, Jintao Cheng, Zeyu Chen, Linfan Zhan, Rui Fan, Zhijian He, Xiaoyu Tang | KDMOS:Knowledge Distillation for Motion Segmentation | null | null | null | null | cs.CV cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motion Object Segmentation (MOS) is crucial for autonomous driving, as it enhances localization, path planning, map construction, scene flow estimation, and future state prediction. While existing methods achieve strong performance, balancing accuracy and real-time inference remains a challenge. To address this, we propose a logits-based knowledge distillation framework for MOS, aiming to improve accuracy while maintaining real-time efficiency. Specifically, we adopt a Bird's Eye View (BEV) projection-based model as the student and a non-projection model as the teacher. To handle the severe imbalance between moving and non-moving classes, we decouple them and apply tailored distillation strategies, allowing the teacher model to better learn key motion-related features. This approach significantly reduces false positives and false negatives. Additionally, we introduce dynamic upsampling, optimize the network architecture, and achieve a 7.69% reduction in parameter count, mitigating overfitting. Our method achieves a notable IoU of 78.8% on the hidden test set of the SemanticKITTI-MOS dataset and delivers competitive results on the Apollo dataset. The KDMOS implementation is available at https://github.com/SCNU-RISLAB/KDMOS. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:47:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Cao",
"Chunyu",
""
],
[
"Cheng",
"Jintao",
""
],
[
"Chen",
"Zeyu",
""
],
[
"Zhan",
"Linfan",
""
],
[
"Fan",
"Rui",
""
],
[
"He",
"Zhijian",
""
],
[
"Tang",
"Xiaoyu",
""
]
] |
2506.14131 | Beniamino Accattoli | Beniamino Accattoli, Claudio Sacerdoti Coen, Jui-Hsuan Wu | Positive Sharing and Abstract Machines | null | null | null | null | cs.LO cs.PL | http://creativecommons.org/licenses/by-sa/4.0/ | Wu's positive $\lambda$-calculus is a recent call-by-value $\lambda$-calculus with sharing coming from Miller and Wu's study of the proof-theoretical concept of focalization. Accattoli and Wu showed that it simplifies a technical aspect of the study of sharing; namely it rules out the recurrent issue of renaming chains, that often causes a quadratic time slowdown.
In this paper, we define the natural abstract machine for the positive $\lambda$-calculus and show that it suffers from an inefficiency: the quadratic slowdown somehow reappears when analyzing the cost of the machine. We then design an optimized machine for the positive $\lambda$-calculus, which we prove efficient. The optimization is based on a new slicing technique which is dual to the standard structure of machine environments. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:47:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Accattoli",
"Beniamino",
""
],
[
"Coen",
"Claudio Sacerdoti",
""
],
[
"Wu",
"Jui-Hsuan",
""
]
] |
2506.14132 | Ruihu Zhu | Jiangdong Li, Jiancheng Yang, Guoxing Xia, Jie Liu, Wenlong Zhan, Ruihu Zhu | Numerical investigations of heavy ion driven plasma wakefield acceleration | null | null | null | null | physics.acc-ph | http://creativecommons.org/licenses/by/4.0/ | Plasma-Based Acceleration (PBA) has emerged as a promising approach to achieve ultra-high gradient particle acceleration. While extensive PBA studies have been conducted using laser, electron, and proton drivers, significant challenges remain in achieving high efficiency, stable acceleration, and scalable energy gain. Meanwhile, due to their higher beam charge density, heavier particle mass and higher kinetic energy, heavy-ion beam drivers represent an interesting direction in PBA research. In this paper, the plasma wakefield acceleration driven by heavy ion beam is studied for the first time, aiming to find the best mechanism for generating high-amplitude wakefields. Using the high intensity, high energy heavy ion beams provided by the High Intensity heavy-ion Accelerator Facility (HIAF), our simulations show that heavy ions can excite stable, high-amplitude plasma wakefields up to 6 GV/m, suitable for electron acceleration. These results show good performance of heavy ion beam drivers and their potential as a viable and promising approach in the field of PBA. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:48:01 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Li",
"Jiangdong",
""
],
[
"Yang",
"Jiancheng",
""
],
[
"Xia",
"Guoxing",
""
],
[
"Liu",
"Jie",
""
],
[
"Zhan",
"Wenlong",
""
],
[
"Zhu",
"Ruihu",
""
]
] |
2506.14133 | Van-Hai Bui | Nikhil Pawar, Guilherme Vieira Hollweg, Akhtar Hussain, Wencong Su, Van-Hai Bui | Enhancing Forecasting Accuracy in Dynamic Environments via PELT-Driven Drift Detection and Model Adaptation | 18 pages | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Accurate time series forecasting models are often compromised by data drift, where underlying data distributions change over time, leading to significant declines in prediction performance. To address this challenge, this study proposes an adaptive forecasting framework that integrates drift detection with targeted model retraining to compensate for drift effects. The framework utilizes the Pruned Exact Linear Time (PELT) algorithm to identify drift points within the feature space of time series data. Once drift intervals are detected, selective retraining is applied to prediction models using Multilayer Perceptron (MLP) and Lasso Regressor architectures, allowing the models to adjust to changing data patterns. The effectiveness of the proposed approach is demonstrated on two datasets: a real-world dataset containing electricity consumption and HVAC system data, and a synthetic financial dataset designed to test cross-domain applicability. Initial baseline models were developed without drift detection using extensive feature engineering. After integrating drift-aware retraining, the MLP model achieved a 44% reduction in mean absolute error (MAE) and a 39% increase in R^2 on the real-world dataset, while even greater improvements were observed on the synthetic financial dataset. Similar enhancements were achieved with the Lasso Regressor. These results highlight the robustness and generalizability of incorporating drift detection and adaptive retraining to sustain forecasting accuracy across diverse domains. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:52:07 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Pawar",
"Nikhil",
""
],
[
"Hollweg",
"Guilherme Vieira",
""
],
[
"Hussain",
"Akhtar",
""
],
[
"Su",
"Wencong",
""
],
[
"Bui",
"Van-Hai",
""
]
] |
2506.14134 | Ryoma Sin'ya | Ryoma Sin'ya and Takao Yuyama | Measure-Theoretic Aspects of Star-Free and Group Languages | null | null | null | null | cs.FL | http://creativecommons.org/licenses/by/4.0/ | A language $L$ is said to be ${\cal C}$-measurable, where ${\cal C}$ is a class of languages, if there is an infinite sequence of languages in ${\cal C}$ that ``converges'' to $L$. We investigate the properties of ${\cal C}$-measurability in the cases where ${\cal C}$ is SF, the class of all star-free languages, and G, the class of all group languages. It is shown that a language $L$ is SF-measurable if and only if $L$ is GD-measurable, where GD is the class of all generalised definite languages (a more restricted subclass of star-free languages). This means that GD and SF have the same ``measuring power'', whereas GD is a very restricted proper subclass of SF. Moreover, we give a purely algebraic characterisation of SF-measurable regular languages, which is a natural extension of Schutzenberger's theorem stating the correspondence between star-free languages and aperiodic monoids. We also show the probabilistic independence of star-free and group languages, which is an important application of the former result. Finally, while the measuring power of star-free and generalised definite languages are equal, we show that the situation is rather opposite for subclasses of group languages as follows. For any two local subvarieties ${\cal C} \subsetneq {\cal D}$ of group languages, we have $\{L \mid L \text{ is } {\cal C}\text{-measurable}\} \subsetneq \{ L \mid L \text{ is } {\cal D}\text{-measurable}\}$. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:52:49 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sin'ya",
"Ryoma",
""
],
[
"Yuyama",
"Takao",
""
]
] |
2506.14135 | Litao Deng | Ying Chai, Litao Deng, Ruizhi Shao, Jiajun Zhang, Liangjun Xing, Hongwen Zhang, Yebin Liu | GAF: Gaussian Action Field as a Dvnamic World Model for Robotic Mlanipulation | http://chaiying1.github.io/GAF.github.io/project_page/ | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate action inference is critical for vision-based robotic manipulation. Existing approaches typically follow either a Vision-to-Action (V-A) paradigm, predicting actions directly from visual inputs, or a Vision-to-3D-to-Action (V-3D-A) paradigm, leveraging intermediate 3D representations. However, these methods often struggle with action inaccuracies due to the complexity and dynamic nature of manipulation scenes. In this paper, we propose a V-4D-A framework that enables direct action reasoning from motion-aware 4D representations via a Gaussian Action Field (GAF). GAF extends 3D Gaussian Splatting (3DGS) by incorporating learnable motion attributes, allowing simultaneous modeling of dynamic scenes and manipulation actions. To learn time-varying scene geometry and action-aware robot motion, GAF supports three key query types: reconstruction of the current scene, prediction of future frames, and estimation of initial action via robot motion. Furthermore, the high-quality current and future frames generated by GAF facilitate manipulation action refinement through a GAF-guided diffusion model. Extensive experiments demonstrate significant improvements, with GAF achieving +11.5385 dB PSNR and -0.5574 LPIPS improvements in reconstruction quality, while boosting the average success rate in robotic manipulation tasks by 10.33% over state-of-the-art methods. Project page: http://chaiying1.github.io/GAF.github.io/project_page/ | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:55:20 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Chai",
"Ying",
""
],
[
"Deng",
"Litao",
""
],
[
"Shao",
"Ruizhi",
""
],
[
"Zhang",
"Jiajun",
""
],
[
"Xing",
"Liangjun",
""
],
[
"Zhang",
"Hongwen",
""
],
[
"Liu",
"Yebin",
""
]
] |
2506.14136 | Nafiz Sadman | Nafiz Sadman, Farhana Zulkernine, Benjamin Kwan | Interpreting Biomedical VLMs on High-Imbalance Out-of-Distributions: An Insight into BiomedCLIP on Radiology | GitHub: https://github.com/Nafiz95/BioVLM_Eval_CXR | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we construct two research objectives: i) explore the learned embedding space of BiomedCLIP, an open-source large vision language model, to analyse meaningful class separations, and ii) quantify the limitations of BiomedCLIP when applied to a highly imbalanced, out-of-distribution multi-label medical dataset. We experiment on IU-xray dataset, which exhibits the aforementioned criteria, and evaluate BiomedCLIP in classifying images (radiographs) in three contexts: zero-shot inference, full finetuning, and linear probing. The results show that the model under zero-shot settings over-predicts all labels, leading to poor precision and inter-class separability. Full fine-tuning improves classification of distinct diseases, while linear probing detects overlapping features. We demonstrate visual understanding of the model using Grad-CAM heatmaps and compare with 15 annotations by a radiologist. We highlight the need for careful adaptations of the models to foster reliability and applicability in a real-world setting. The code for the experiments in this work is available and maintained on GitHub. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 02:59:42 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sadman",
"Nafiz",
""
],
[
"Zulkernine",
"Farhana",
""
],
[
"Kwan",
"Benjamin",
""
]
] |
2506.14137 | Koji Sugitani | Koji Sugitani, Jan G. A. Wouterloot, Harriet Parsons, Sarah Graves, Takayoshi Kusune and Archana Soam | Structure and Magnetic Field of Bright-rimmed Cloud IC 1396E/SFO 38 | 14 pages, 10 figures, to be submitted to PASJ | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We carried out polarimetric observations of the bright-rimmed cloud IC 1396E/SFO 38 with SCUBA-2/POL-2 to study the effect of ultraviolet (UV) light on its structure and magnetic field. This bright-rimmed cloud appears optically to be one single cloud illuminated by the UV light from the excited star of IC 1396, however our Stokes I image and 13CO(J=3-2) archival data suggest that this cloud is not a simple, single structure, but appears to be composed of two parts on first glance; a head part with wings and a tail, and a north-west extension part.Since molecular clouds are generally filamentary and it seems likely that the initial structures of bright-rimmed clouds are expected to be also generally elongated, we examined the possibility that the structure was created from a single elongated cloud by the UV impact. We compared the cloud structure with a simulation study that investigated the evolution of prolate clouds exposed to the UV radiation from various directions and found that this apparent two-part structure could be reproduced in a situation where a single filamentary cloud is obliquely illuminated by UV light. The magnetic field directions of the cloud are different from the ambient field direction, demonstrating the field reconfiguration due to the UV impact. The distortion or pinch of the magnetic field is seen toward the cloud head, where an intermediate-mass star cluster is located, suggesting gravitational contraction. We roughly estimated the magnetic strength and stability in three parts of the cloud and found that the cloud head is most likely to be supercritical. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 03:00:14 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Sugitani",
"Koji",
""
],
[
"Wouterloot",
"Jan G. A.",
""
],
[
"Parsons",
"Harriet",
""
],
[
"Graves",
"Sarah",
""
],
[
"Kusune",
"Takayoshi",
""
],
[
"Soam",
"Archana",
""
]
] |
2506.14138 | Ashish Gautam | Ashish Gautam, Prasanna Date, Shruti Kulkarni, Robert Patton, Thomas Potok | NeuroCoreX: An Open-Source FPGA-Based Spiking Neural Network Emulator with On-Chip Learning | Neuromorphic computing, FPGA, STDP, Spiking Graph Neural Networks, Spiking Neural Networks, VHDL | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by/4.0/ | Spiking Neural Networks (SNNs) are computational models inspired by the structure and dynamics of biological neuronal networks. Their event-driven nature enables them to achieve high energy efficiency, particularly when deployed on neuromorphic hardware platforms. Unlike conventional Artificial Neural Networks (ANNs), which primarily rely on layered architectures, SNNs naturally support a wide range of connectivity patterns, from traditional layered structures to small-world graphs characterized by locally dense and globally sparse connections. In this work, we introduce NeuroCoreX, an FPGA-based emulator designed for the flexible co-design and testing of SNNs. NeuroCoreX supports all-to-all connectivity, providing the capability to implement diverse network topologies without architectural restrictions. It features a biologically motivated local learning mechanism based on Spike-Timing-Dependent Plasticity (STDP). The neuron model implemented within NeuroCoreX is the Leaky Integrate-and-Fire (LIF) model, with current-based synapses facilitating spike integration and transmission . A Universal Asynchronous Receiver-Transmitter (UART) interface is provided for programming and configuring the network parameters, including neuron, synapse, and learning rule settings. Users interact with the emulator through a simple Python-based interface, streamlining SNN deployment from model design to hardware execution. NeuroCoreX is released as an open-source framework, aiming to accelerate research and development in energy-efficient, biologically inspired computing. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 03:02:04 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Gautam",
"Ashish",
""
],
[
"Date",
"Prasanna",
""
],
[
"Kulkarni",
"Shruti",
""
],
[
"Patton",
"Robert",
""
],
[
"Potok",
"Thomas",
""
]
] |
2506.14139 | Dong Sung Yoon | Ja Kyung Koo, Dong Hwa Shin and Dong Sung Yoon | A simplified algorithmic realization of Galois actions on special values of modular functions | 14 pages | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an explicit and practical algorithm for computing Galois conjugates and irreducible polynomials for special values of modular functions evaluated at CM points associated with imaginary quadratic orders. Our approach builds upon the theory of extended form class groups developed by Jung et al., offering a refinement of earlier methods by Stevenhagen and Cho, respectively. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 03:05:32 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Koo",
"Ja Kyung",
""
],
[
"Shin",
"Dong Hwa",
""
],
[
"Yoon",
"Dong Sung",
""
]
] |
2506.14140 | Guo-Li Wang | Shi-Hang Zhang, Wen-Yuan Ke, Su-Yan Pei, Wei Li, Xiao-Ze Tan, Lili Zhu, Guo-Li Wang | Identification of $D^*_2(3000)$ as the $D_2^*(2^3P_2)$ and exploring potential of undiscovered $2^+$ mesons via $B$ decays | 24 pages, 7 figures | null | null | null | hep-ph hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Following the discovery of the $D^*_2(3000)$, its mass and full width have been extensively studied. Yet its nature remains undetermined to date. Since it was discovered through nonleptonic decay of $B$ meson and the corresponding cascade process, we therefore in this paper investigate the nonleptonic and semileptonic decays of $B$ meson to $J^P = 2^+$ charmed mesons using the Bethe-Salpeter equation approach. Our calculations on nonleptonic $B$ decays reveal that the unconfirmed resonance $D^*_2(3000)$ aligns well with $D^*_2(2^3P_2)$ predictions. Other candidates, including $D^*_2(1^3F_2)$, $D^*_2(3^3P_2)$, and $D^*_2(2^3F_2)$, are excluded due to their very small branching ratios in $B$ decays. Considering that the $D^*_2(1F)$, $D^*_2(3P)$, and $D^*_2(2F)$ have not yet been experimentally observed, we investigate the feasibility of their detection in $B$-meson decays. | [
{
"version": "v1",
"created": "Tue, 17 Jun 2025 03:05:59 GMT"
}
] | 2025-06-18T00:00:00 | [
[
"Zhang",
"Shi-Hang",
""
],
[
"Ke",
"Wen-Yuan",
""
],
[
"Pei",
"Su-Yan",
""
],
[
"Li",
"Wei",
""
],
[
"Tan",
"Xiao-Ze",
""
],
[
"Zhu",
"Lili",
""
],
[
"Wang",
"Guo-Li",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.