Peer Revenue Equity in science

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Friday 28 August

Computation and Language

BanglaMed-QA: A Question Answering System for Healthcare Support in Bangla

Medical question answering (QA) systems have become crucial tools for providing reliable health information. But they remain very unexplored for low-resource languages like Bangla due to limited datasets and systems tailored to these languages. To address this, we introduce BanglaMed-QA, a robust QA system specifically designed for the Bangla medical domain. The process begins with building a structured medical knowledge base that includes 4,493 QA pairs in 9 categories under 506 diseases. To improve semantic comprehension, domain-specific root word dictionaries and synonym sets are proposed, in addition to part-of-speech tagging for anaphora resolution. We adopt supervised machine learning models in which SVM is found to be the best model to categorize questions. Multiple similarity metrics, including cosine, Jaccard, BM25, and Levenshtein, are applied with soft and hard voting methods for query matching. The performance of the QA system has been evaluated in two aspects, with a 95% F1 score in an automated evaluation and an average human satisfaction rating of 0.9 out of 1.0. This validates the real-world application of BanglaMed-QA in closing the healthcare information gap for Bangla speakers.
Read the full abstract

All authors Rowzatul Zannat, Abdullah Al Shafi, K. M. Azharul Hasan, Atia Shahnaz Ipa

Probability

Critical Gaussian Multiplicative Chaos on the Circle Is Rajchman

We prove that the Fourier coefficients of the canonical critical Gaussian multiplicative chaos on the circle vanish almost surely at infinity. More precisely, let $M_\phi^{\mathrm{crit}}$ be the canonical critical chaos associated with the centered circle field $\phi$ of covariance $\mathbb{E}[\phi(\theta)\phi(\theta')] = \log\frac{1}{\lvert e^{i\theta}-e^{i\theta'}\rvert}$. Then, almost surely, $\widehat{M_\phi^{\mathrm{crit}}}(n)\longrightarrow0$ as $\lvert n\rvert\to\infty$. This resolves the almost-sure critical Rajchman problem for the canonical circle field. Since critical chaos has Fourier dimension zero almost surely, no positive polynomial Fourier-decay rate can hold; the theorem therefore exhibits qualitative Fourier cancellation beyond the regime of positive Fourier dimension. The proof addresses two coupled difficulties: the heavy, nonuniform cell masses of critical chaos and the need to control exponentially many frequencies in each dyadic annulus. For an auxiliary periodized compact-range star-scale field, a derivative-rooted Bessel regression yields weighted small-cell summability and moving-tail control of exceptional large cells. After conditioning at a coarse scale below the Fourier scale, finite-range independence and conditional Bernstein concentration reduce uniform control of the terminal Fourier coefficients over each dyadic annulus to a spatial-variation estimate for a coarse predictable measure. A smooth positive-definite covariance correction and critical-chaos uniqueness then transfer the Rajchman property to the canonical critical chaos of the exact circle field.
Read the full abstract

All authors Yin Cai, Bonan Chen, Xiang Fang, Feng Guo

Cryptography and Security

How Diverse Are Layered Safeguards? Estimands and Measurement for Ensembles of LLM Jailbreak Defenses

Practitioners protect large language models (LLMs) by stacking safeguards: a likelihood filter, a guard classifier, hidden-state probes, a decoding-time intervention. Such a stack is an ensemble of learned detectors, and ensemble theory states when combination pays: when members fail on different inputs. Stacking is recommended throughout the LLM security literature, but this condition has not been measured for any pair of safeguards. The condition is three quantities, not one, and conflating them is why it has resisted measurement. Separately optimized attacks license only a claim about which behaviors are jointly vulnerable; the identity that governs a deployed stack is written over shared inputs; and neither predicts what an adversary attacking the assembled pipeline achieves. We define the three estimands, give an estimator for each, and supply two controls: the correlation between two independent runs of the same attack, which sets the level a pair must clear, and a multilevel model separating detector diversity from behavior difficulty. On a seven-layer stack and two open-weight targets, no pair of safeguards is distinguishable from the correlation that two independent runs of the attack produce with no safeguard present, and behavior difficulty explains much of what remains. A second attack family, an attacker model rewriting its own prompts, reaches that verdict independently on a thirtieth of the budget and overturns the practical one: the only member that blocked the first adversary admits 0.19 under the second, against 0.03. Diversity between neural safeguards is measurable, but it and any member's strength are properties of a (stack, adversary, target) triple.
Read the full abstract

2608.28327 · 28 Aug · 40 pages, 4 figures. Submitted to Machine Learning. v2: major revision with a new title. The survey in v1 is removed and the experiment is rebuilt; v1 results are superseded. Authors changed: v2 is by A. Alotaibi and M. Ahmed. M. S. Jabbar and S. Al-Azani, authors of v1, took part in writing parts of the survey, which is not in v2. Code: https://github.com/AbrarAlotaibi/defense-correlation

Nuclear Theory

Spin-State Teleportation and Tests of EPR Correlations Using 151 MeV Entangled Protons

We discuss the feasibility of quantum spin-state teleportation in a three-proton system at an energy of 151 MeV, where, similarly to the low-energy case, a single Bell-state term dominates the proton-proton scattering matrix. We find that, in contrast to the low-energy regime, where unpolarized proton-proton scattering produces strongly entangled outgoing proton pairs, at higher energies a pair of protons, each with an energy of $151$~MeV, must be produced in an unpolarized, exclusive proton-deuteron breakup reaction under complete final-state-interaction kinematics. The subsequent interaction of one of the entangled protons with a polarized hydrogen target triggers, as in the low-energy case, the teleportation process, whereby the polarization of the target proton is transferred to the second member of the entangled pair within the very narrow angular region of strong entanglement centered around a laboratory scattering angle of $45^\circ$. We also find that scattering one member of a strongly correlated proton pair forming a Bell state from an unpolarized hydrogen target leads to Einstein-Podolsky-Rosen-like correlations between the polarization of the scattered proton and that of the unscattered second entangled proton. The polarization of the scattered proton varies with its scattering angle. An identical polarization, following the angular dependence of thepolarization of the first proton, is induced in the second member of the pair, whose initial polarization was zero and whose momentum remains unchanged. The sign of the polarization of the second proton is determined by the sign of the spin correlation in the Bell state.
Read the full abstract

Statistical Methodology

Response Propensity Estimation and Cross-Fitting

This paper investigates whether five fold cross fitting improves nonresponse adjustment in survey estimation when flexible machine learning methods are used to estimate response propensities. We conduct a finite population Monte Carlo simulation with 90 experimental configurations and 2,000 replications per configuration, varying sample size, response rate, and the structure of the response mechanism. Logistic regression is used as a conventional parametric benchmark, while Random Forest and Gradient Boosting Machine are considered as flexible nonparametric models.
Read the full abstract

Astrophysics of Galaxies

Deep VISTA observations of the Magellanic Clouds (VMCDeep)

The central regions of galaxies challenge observers with a high density of stars, gas and dust. The LMC hosts a dense stellar bar whereas the SMC hosts a dense core, which have eluded even the most sensitive large-scale wide-field surveys of individual stars from the ground and from space. This work describes deep imaging observations obtained with VISTA of the inner regions of the Clouds. They were obtained to resolve stars fainter than the red clump, that due to crowding were missed in previous observations, and to quantify features associated with the star formation, kinematics and structure of the galaxies. These studies will allow us to establish how the dynamical interactions between the LMC and the SMC, and the interaction between the Clouds and the Milky Way have influenced the most gravitationally bound regions. Multiple near-infrared images were obtained in the J and Ks filters during excellent weather conditions. The data were reduced using the VDFS which provides single-filter and band-merged calibrated images and catalogues, with associated confidence maps, for individual observations and combined observations for deep tiles. Approximately one million sources, with only minor quality issues and about 40 epochs per filter, are detected in the central 1.77 sq. deg. They provide a more contiguous spatial area coverage than sources in the VMC survey, which covers the same regions by overlapping two VISTA tiles per galaxy. The catalogues, which are made available with this work, contain attributes to select unique good quality detections for which we provide examples of ongoing scientific applications. These data represent valuable complementary information to deep imaging observations at other wavelengths and provide additional epochs for proper motion and photometric variability studies extending the time baseline to 11 yr, as well as targets for spectroscopic observations.
Read the full abstract

All authors M.-R. L. Cioni, A. O. Omkumar, N. J. G. Cross, F. Cusano, F. Dresbach, F. Ficara, F. Niederhofer, J. M. Oliveira, V. Ripepi, S. Vijayasree, G. Clementini, R. de Grijs, L. Girardi, M. A. T. Groenewegen, V. D. Ivanov, N. Kacharov, M. Marconi, A. Mazzi, G. Pastorelli, C. M. Pennock, M. Petr-Gotzens, J. Storm, S. Subramanian, J. Th. van Loon, R. P. Blake, M. J. Holliman, J. Irwin, M. J. Irwin, R. G. Mann, M. Read, E. Sutorius

Instrumentation and Methods for Astrophysics

JWST telemetry combined with active coronagraphy: raw contrast predictions for exoplanet imaging with the Habitable Worlds Observatory

We provide a quantification of the technological gap between the James Webb Space Telescope (JWST) and the Habitable Worlds Observatory (HWO) for the goal of exo-Earth imaging around Sun-like stars at the $10^{-10}$ raw contrast level. We use JWST's in-flight telemetry of the primary segmented mirror to simulate a JWST-like telescope equipped with a modern coronagraph instrument, inspired by the Roman Space Telescope (RST) Coronagraphic Instrument (CGI), featuring an Apodized Pupil Lyot Coronagraph and active deformable mirror wavefront control on a segmented, unobstructed, off-axis telescope. We show that it can achieve around $10^{-10}$ raw contrast for very bright stars (brighter than magnitude 4) for fast control frequencies of 100 Hz, but that this level of control still lacks sufficient signal to correct JWST-amplitude errors for fainter stars. We show that an improvement of a factor of ten in wavefront stability is sufficient to extend this capability to a $10^{-10}$ raw contrast across all considered control frequencies (1 Hz to 100 Hz), assuming no reaction wheel vibrations, for stars up to a magnitude of 11. These results establish a new quantitative benchmark linking JWST's demonstrated thermo-mechanical stability to HWO's requirements, showing that active wavefront control relaxes the structural stability demands on the observatory, and identifying wavefront stability as the critical technological gap that must be closed for HWO to achieve its exo-Earth imaging goals.
Read the full abstract

All authors Raphaël Pourcelot, Laurent A. Pueyo, Emiel H. Por, Marshall D. Perrin, Iva Laginja, Bryony Nickson, Sarah Steiger, Rémi Soummer, Rendal Telfer

Optics

Structural perturbation theory for bound states in the continuum via bifurcation of zeros and extrema of the dispersion relation

In a lossless periodic structure, a bound state in the continuum (BIC) corresponds to a real zero and a local maximum of the imaginary part of a complex dispersion relation $k = k(\beta)$, where $\beta$ is the Bloch wave number. A perturbation of the structure deforms the dispersion curve and may destroy, move or split the BIC, as demonstrated in existing studies involving lossless symmetry-preserving perturbations, symmetry-breaking perturbations and dissipative perturbations. We present a comprehensive perturbation theory, emphasizing the evolution of the real zeros and extreme points of $\mathrm{Im}[k(\beta)]$ under various structural perturbations. In particular, our theory reveals the existence of lasing threshold modes (LTMs), which are also zeros of $\mathrm{Im}[k(\beta)]$ when the perturbation involves gain with or without balanced loss. Using local Taylor expansions and Puiseux series, we determine the number, locations, and leading-order scaling of real zeros and extreme points for various types of BICs under different types of structural perturbations. The theory recovers known results and predicts new behavior for super-BICs under $\mathcal{PT}$-symmetric perturbations. Specifically, a propagating super-BIC corresponding to a fourth-order zero of $\mathrm{Im}[k(\beta)]$ splits into two real zeros representing either two BICs or two LTMs, and a symmetric standing wave splits into two BIC-LTM pairs. Our theory provides a general framework for studying BICs and nearby resonant modes in both lossless and non-Hermitian periodic structures.
Read the full abstract

Physics and Society

Early warning signals for synchronization transitions from partial observations

Anticipating the onset of collective synchronization is important in many networked systems, yet observing every oscillator is often impractical. We investigate whether synchronization transitions can be detected from a small set of monitored, or sentinel, nodes. Using a stochastic Kuramoto model on networks, we numerically compare three early warning signals: the local order parameter, its temporal variance, and the variance of individual oscillator phases after removing their mean rotational trends. We also compare sentinel-selection strategies based on node dynamics, degree, and random sampling. We show that, under partial observation, the local order parameter and the variance of detrended phases provide substantially stronger warning signals than the variance of the local order parameter. Selecting nodes according to their dynamical behavior near the transition consistently improves performance over other sentinel-selection methods. With only $\lfloor\ln N\rfloor$ dynamically selected sentinels, two success indicators approach the performance obtained by observing all $N$ nodes. These results demonstrate that synchronization transitions can be anticipated from sparse observations when the warning signal and monitored nodes are chosen appropriately.
Read the full abstract

Optimization and Control

First-order approximations of multifunctions defined implicitly by "split" feasibility problems with applications to optimization

In the present paper, a study is made of first-order differential properties of certain multifunctions, which are defined as a solution map associated with a family of parameterized “split" feasibility problems. The latter are a class of convex feasibility problems, whose specific structure makes them suitable for well-established applications in several areas of engineering and systems biology. As a result, inner and outer approximations of the graphical (contingent) derivative of the solution map are provided, which are expressed in terms of derivatives of the problem data. As an application, a perspective is also discussed for employing some of these achievements in the formulation of optimality conditions for mathematical programs with “split" feasibility constraints.
Read the full abstract

Databases

VeriTS: Verifiable Model-Enhanced Time-Series Queries on Blockchain Systems

Every blockchain transaction carries a timestamp, and the chain imposes a total order. On-chain data therefore forms per-source time-series streams. However, existing systems support only basic lookups on blocks and transactions, and cannot answer time-series queries such as time-range retrieval and windowed aggregation. Offloading queries off-chain restores expressiveness, but the off-chain query layer is untrusted, so results must be verifiable. To this end, we propose VeriTS, the first verifiable time-series query framework for blockchain systems. It supports efficient range and aggregation queries. VeriTS maintains an off-chain query layer. In this layer, each stream is kept under one tree whose nodes carry authenticated aggregates, so the query index is itself the authenticated data structure. A light client thus verifies a windowed aggregate from a logarithmic number of authenticated nodes rather than from every record. VeriTS further answers error-tolerant queries from compact model representations of a stream, and extends the completeness and soundness guarantees to such approximate answers. As VeriTS never trusts the model behind a representation, a faulty or adversarial model can only widen an answer's certified interval, never falsify it. Experiments offer evidence that on windowed aggregation, VeriTS improves verification efficiency by more than two orders of magnitude over per-record proofs. On range retrieval, proofs shrink by up to 14.5x.
Read the full abstract

All authors Zhongming Yao, Jun Pang, Chenxu Wang, Qian Ma, Peiyuan Guan, Shiliang Zhang

Probability

Subcritical non-linear heat equations via spectral gap

We establish local well-posedness for a class of non-linear heat equations on the torus $\mathbb{T}^n$ whose highest order non-linearities have scaling-critical dimension $-1/k, \, k\in \mathbb{N}$, when the initial condition is a random field satisfying a spectral gap inequality involving a suitable Lebesgue norm. The corresponding subcriticality condition matches that of the Gaussian free field in dimension $d<2+2/k$. This extends previous subcritical well-posedness results beyond the Gaussian setting. Moreover, our proof is simpler and avoids diagrammatic arguments.
Read the full abstract

Computer Vision

Conditional Visual Evidence Utility: State-Dependent Rank Reversals in Frozen Vision-Language Encoders

Static importance scores compress visual evidence into a single ranking, but the value of remaining evidence can change after one cue has been observed. We study this possibility in controlled compositional visual search, where color, shape, and texture evidence can be independently exposed and their conditional marginal utility measured across acquisition states. In a held-out confirmation on 800 scenes, frozen OpenCLIP and SigLIP exhibit robust state-dependent rank reversals that concentrate in candidate-overlap regimes designed to induce ordering changes, persist across two evidence-accumulation constructions and ten equivalent query wordings, and collapse to near-chance-scale behavior under query-scene derangement. A subsequent role-balanced follow-up on 1,200 scenes rotates the abstract roles of initially strong, redundancy-inducing, and comparator attributes; the positive-minus-negative reversal contrast remains positive across all 24 role-permutation, backbone, and evidence-mode cells, although residual attribute-identity effects remain. We further distinguish measured replanning opportunity from prospective predictability. Matched-first-action utility analyses show substantial opportunity to rerank remaining evidence, but lightweight predictors using posterior-based or acquired-embedding state representations do not establish a robust incremental advantage of acquired-state information over legal static controls on the role-balanced benchmark. Together, these results show that conditional visual evidence utility is reliably state dependent in this controlled setting, while separating the existence of changing utility from the stronger claim that those changes are prospectively predictable by a learned selector.
Read the full abstract

Artificial Intelligence

MAIL: Memory-driven, Adaptive, Incremental, and Literature-grounded Framework for Hypothesis Generation in Chemistry

The ever-expanding volume of the chemical literature offers unprecedented opportunities to generate novel and impactful hypotheses. However, the bottleneck lies in efficiently navigating this vast knowledge base to formulate high-quality, experimentally meaningful insights. While Large Language Models (LLMs) show promise for this task, existing methods often rely on static inspiration corpora, predefined heuristics, or laborious human-in-the-loop pipelines and decision-support frameworks that limit scalability and novelty. In this work, we propose an automated approach, a Memory-augmented, Adaptive, Incremental, and Literature-grounded (MAIL) framework for hypothesis generation in chemistry. Our MAIL method formulates hypothesis generation as a temporally grounded, memory-driven reasoning process, where hypotheses emerge from an evolving conceptual path that continuously accumulates and reinterprets prior knowledge. We evaluated the MAIL framework on a public TOMATO-Chem dataset and a newly curated and disseminated high-novelty nature/science challenge (HN-NS) dataset. Across both datasets, MAIL generates structurally coherent and mechanistically plausible hypotheses, achieves the highest MIOS and MPOS by more effectively recovering the central ideas and methodological elements of the historical target hypotheses, and obtains the highest overall expert-evaluation scores for scientific quality. These results demonstrate the potential of LLMs to autonomously explore chemical domains and generate hypotheses that are both innovative and chemically plausible.
Read the full abstract

All authors Mahdi Babaei, Xueshen Li, Yutao Kuang, Jolene P. Reid, Yu Gan

High Energy Physics (Experiment)

Vertex reconstruction for a search for neutron-antineutron conversions with HIBEAM

The HIBEAM/NNBAR programme (incorporated into the FINESSE/NNBAR programme) at the European Spallation Source is proposed to search for neutrons converting to antineutrons. An important observable is the reconstructed vertex arising from charged particles produced by an antineutron annihilating on a thin target foil and which pass through a time projection chamber. This paper studies track clustering and foil-plane vertex reconstruction for this topology. Both non-machine-learning methods and graph-neural-network methods are tested and compared with each other, including deterministic clustering, trackless projection and a hybrid clustering/graph-neural-network chain with track classification and vertex refinement. We conclude that, for the geometrically simple events in the HIBEAM TPC, classical reconstruction methods perform on-par with machine-learning based methods in terms of vertex coordinate reconstruction. Custom machine-learning based methods can, however, deliver event-shape information that may be important in downstream analyses.
Read the full abstract

All authors Alexander Burgman, Sze Chun Yiu, Yamna Shaikh, David Milstead, Eily Merjy, Lucas Åstrand, Kenneth Österberg, Fredrik Oljemark, Bernhard Meirose, Matthias Holl, Valentina Santoro, Jorge Amaral, Ernesto Kemp, André Nepomuceno, Tiago Quirino, Joshua L. Barrow

Data Structures and Algorithms

Multi-tier Flexible Graph Connectivity

Motivated by non-uniform edge failures in network design, we introduce a multi-tier model of flexible graph connectivity. In k-tier Flexible Graph Connectivity (k-tier FGC), the input is an undirected graph G=(V, E) with non-negative edge costs, along with a classification of the edges into nested tiers T_1 subseteq T_2 subseteq ... subseteq T_k = E and non-negative integral tier requirements q_1 <= q_2 <= ... <= q_k. A non-empty proper subset R of vertices is safe if it is safe along one of the tiers, i.e., there exists i in [k] such that |delta(R) cap T_i| >= q_i. The goal is to find a minimum cost subset F subseteq E of edges such that the subgraph (V, F) has no unsafe cuts. The case of k=1 corresponds to the min-cost p-edge-connected spanning subgraph problem which is APX-hard. We design approximation algorithms for every fixed constant k for three variants of k-tier FGC: (i) for k-tier FGC, we design an LP-based logarithmic approximation, (ii) for min-cardinality k-tier FGC, we design a combinatorial approximation whose factor depends only on the tier requirements q_1 and q_k, and (iii) for k-tier Flexible Multi-Graph Connectivity, where we are allowed to use multiple copies of each edge while paying the cost of the edge for each chosen copy of the edge, we design an LP-based 2-approximation.
Read the full abstract

Computer Vision

AIM: Anchor Identity Features, Then Match for Multimodal Large Language Model Unlearning

Multimodal large language models (MLLMs) can memorize identity-specific facts about people in their fine-tuning data, creating privacy risks when a person requests deletion. Existing MLLM unlearning methods often assume access to retain images or ground-truth answers during deletion, which is unrealistic in many practical scenarios. We study identity unlearning when retain images are unavailable at deletion time. Our analysis shows that identity and visual-perception questions occupy distinct regions in fine-tuned hidden states and are organized differently: identity questions cluster by person, whereas perception questions cluster by question type. This suggests that identity knowledge can be suppressed without erasing general visual perception. Building on this observation, we propose AIM, a two-stage method that anchors an identity-forgetting target with a universal visual prompt and then matches the vision encoder to that target under a Fisher-based constraint. Extensive experiments show that AIM achieves competitive identity forgetting while preserving non-deleted identities, prior knowledge, and visual perception on the same images.
Read the full abstract

All authors Wonjun Lee, Jaehyuk Jang, Kangwook Ko, Hee-Seon Kim, Changick Kim

High Energy Physics (Theory)

Scattering Equations as the lowest order K-identities in the calculation of Stringy Scaling of Hard String Scattering Amplitudes

We prove explicitly the n-point K-identities we proposed previously in the calculation of one tensor hard string scattering amplitudes (HSSA). These K-identities were the key to obtain the stringy scaling behavior in the saddle point calculation of the n-point HSSA and, on the other hand, to consistently match with the calculation of decoupling of zero norm states. Moreover, we introduce a G function to generate an infinite set of generalized K-identities (GKI). The lowest order set of these GKI is the scattering equations (SE) used in the calculation of field theory amplitudes in the CHY formalism. The next to leading order set of these GKI is the K-identities used previously in the calculation of one tensor HSSA. We expect that the higher order sets of these GKI can find applications in calculating other HSSA.
Read the full abstract

Graphics

Transparency Rendering in Computer-Aided Design: Methodologies, Trade-offs, and Challenges

This paper surveys the state of transparency rendering in Computer-Aided Design (CAD), with a focus on both practical deployment in industrial systems and the underlying algorithms. We first review current approaches to transparency rendering in CAD environments and outline application scenarios in which accurate and performant transparency is critical for design inspection, communication, and decision-making. We then analyze the trade-offs between approximate and exact transparency techniques, comparing their performance-quality balance on desktop and mobile platforms and discussing criteria for selecting appropriate methods. The survey further identifies the need for robust benchmarks, quality metrics, and evaluation methodologies tailored to CAD-specific visualization tasks. We examine techniques for emphasizing important interior components such as importance-driven transparency, silhouette-based methods, and related approaches to support effective spatial understanding in complex assemblies. Finally, we discuss the unique challenges of rendering transparent constructive solid geometry (CSG) objects, including robustness, correctness, and integration with modern rendering pipelines. Collectively, these contributions characterize current capabilities, systematize open problems, and outline future research directions for transparency rendering in CAD visualization.
Read the full abstract

Space Physics

Solar Magnetic Configuration Control over Radiation Belt Electrons

Earth is surrounded by two highly dynamic concentric belts of particle radiation. The outer radiation belt exhibits coherent variability at solar-cycle, seasonal, and 27-day (Carrington) timescales. While the solar-cycle and Carrington variations have been attributed to the recurrence of coronal hole solar wind, the seasonal variation has long been explained by local geometric effects that modulate the solar wind-magnetosphere coupling. Here, we challenge this paradigm by showing that the periodic recurrence of coronal hole solar wind likewise drives the seasonal variation. We further demonstrate that the Alfvénic nature of this solar wind is a key driver of the observed electron flux enhancement in the outer radiation belt. These findings provide a unifying framework linking solar magnetic topology, solar wind properties, and magnetospheric dynamics across multiple timescales at Earth and beyond.
Read the full abstract

All authors Ahmad Lalti, Jonathan Rae, Clare Watt, Stephanie Yardley, Savvas Raptis

Machine Learning

Deriving Scaling Laws for OpenEuroLLM Models: Learning Rate, Batch Size and Loss

We study the scaling behavior of learning rate and batch size in pretraining dense large language models on English-prevalent corpora. Beyond scaling jointly optimal learning rates and batch sizes, we investigate their marginal evolution with model capacity and data scale and develop a model that captures these relationships. As we employ a Warmup-Stable-Decay learning rate schedule, we further investigate the gains from learning rate annealing over a broad range of hyperparameters settings, models and data budgets, and whether the optimal learning rate and batch size transfer between the stable and decay phases. Finally, we characterize the dependence of loss on model capacity and dataset size, evaluating recently proposed scaling forms that explicitly model their interaction. We find these approaches particularly effective at capturing both undertraining and overtraining regimes across our experiments. This study establishes a first baseline and scaling procedure for the development of future OpenEuroLLM models. We open-source the complete collection of pretraining runs used in this study.
Read the full abstract

All authors Niccolò Ajroldi, Diana Alexandra Onutu, Haider Al-Tahan, Jörg Franke, Sampo Pyysalo, Jenia Jitsev, Aaron Klein

Differential Geometry

Rigidity of closed $\lambda$-self-expanders

A $\lambda$-self-expander $x: M^n\to \mathbb{R}^{n+1}$ is the solution of the isoperimetric problem of weight $e^{\frac{|x|^2}{4}}$. In this paper, we prove that the closed $\lambda$-self-expander is a round sphere centered at the origin under some curvature conditions. These curvature conditions mainly include the scalar curvature, the mean curvature, and the squared norm of the second fundamental form.
Read the full abstract

Machine Learning

VISTA: Verifier-Informed Student-to-Teacher Adaptation for On-Policy Self-Distillation

On-policy self-distillation (OPSD) improves reasoning by training a problem-only student on its own rollouts using dense token-level supervision from a privileged teacher that also sees a reference solution. However, standard OPSD treats the teacher distribution as a fixed target along the student's rollout and updates only the student %, although – even though privileged conditioning does not guarantee that the teacher always provides the most appropriate target for problem-only reasoning. This one-way supervision can therefore misdirect the student when the teacher distribution is misaligned with valid student reasoning. We therefore introduce Verifier-Informed Student-to-Teacher Adaptation (VISTA), which preserves the standard OPSD student update while using outcome-verified rollouts to adapt the teacher toward the student distribution. Within each verified rollout, VISTA further restricts this adaptation to the top-$k$ positions with the largest teacher–student KL divergence. Notably, VISTA reuses the rollout and loss function from standard OPSD, introducing no additional sampling or separate reward objective. Across AIME24, AIME25, and HMMT25 with Qwen3 models at 1.7B, 4B, and 8B, VISTA achieves the highest Avg@12 at every scale, improving over OPSD by $0.6$, $0.7$, and $2.1$ points, respectively. These results demonstrate the value of student supervision from outcome-verified rollouts and highlight student-to-teacher adaptation as a promising direction for OPSD.
Read the full abstract

All authors Zewen Ding, Zezhong Wu, Zhou Tao, Shida Wang, Shizhuo Hou, YongXiang Hua, Haoyu Cao, Linli Xu

Robotics

PanelShield: Verifiable Closed-Loop Safe Planning for Robotic Industrial Panel Operation

Industrial panel operation is knowledge-intensive and safety-critical. Beyond control recognition and action generation, execution must satisfy constraints in operation manuals and safety regulations. While foundation-model-based planners show strong semantic capability, they typically lack computable, localizable, and reproducible mechanisms for violation detection and repair. To address this, we propose PanelShield, a verifiable closed-loop safety planning framework for manual-guided industrial panel operation. The framework generates parameterized action primitive sequences from task-relevant manual evidence and applies dual formal verification with LTL and a Safety FSM to enforce cross-step temporal correctness and local transition legality. When violations occur, it outputs a structured counterexample with the earliest violating step and cause, enabling targeted repair and re-verification. We build a multi-level long-horizon planning benchmark covering three representative industrial device panels, and evaluate the framework in simulation and real-world robotic experiments. Results show that PanelShield improves complex safety-constrained task performance over foundation-model-only planning baselines while reducing the violation rate to 2.7%, with 4.1 s total latency. Real-world experiments demonstrate end-toend feasibility. Overall, PanelShield offers a verifiable approach to robotic panel operation that balances flexibility, safety, and auditability.
Read the full abstract

All authors Guipeng Xin, Jiahe Xu, Chenhui Wan, Jie Liu, Youmin Hu, Zhongxu Hu

Instrumentation and Methods for Astrophysics

Correction of the influence of rolling shutter detectors on determining the velocity of meteors from video recordings

Modern video cameras use either global or rolling shutter to create individual video frames. In case of global shutter, the whole frames are exposed and read out at the same time. In case of rolling shutter, the exposure and readout proceed sequentially line by line. Such cameras are now commonly used for meteor observations. It is then necessary to perform a correction for this type of image reading in order to correctly and accurately determine the velocity of recorded meteors. We provide here the correct procedure. The correction must include the value of the rolling period of the sensor, which is generally not provided in camera specifications. We discuss four methods of determining the rolling period experimentally. The most accurate method uses the New EXposure Timing Analyser (NEXTA) of [1]. We list rolling periods measured for several cameras.
Read the full abstract

All authors Lukáš Shrbený, Jiří Borovička, Pavel Spurný, Jan Mánek, Jan Fuchs

2608.28304 · 28 Aug · This preprint has not undergone peer review (when applicable) or any post-submission improvements or corrections. The Version of Record of this article is published in Experimental Astronomy, and is available online at https://doi.org/10.1007/s10686-026-10075-7

Operator Algebras

Compact Quantum Metric Spaces from Weakly Geodesic Length Functions on Hyperbolic Groups

We show that length functions on hyperbolic groups whose induced metrics are hyperbolic and weakly geodesic naturally give rise to compact quantum metric spaces, thus generalizing the result of Ozawa and Rieffel. As a result we obtain several new examples of compact quantum metric spaces, in particular arising from Green metrics associated with random walks on hyperbolic groups, and from proper cocompact isometric actions on proper geodesic hyperbolic spaces admitting a point with trivial stabilizer.
Read the full abstract

Computer Vision

FUSED: Forensic-Semantic Mixture-of-Experts for AI Inpainting Detection and Localization

Diffusion-based inpainting models modify only a localized part of an image, while many AI-image detectors rely on global artifacts and do not localize. These artifacts vary across generators, limiting detector transfer under distribution shifts. Recent work shows that restoring the authentic pixels outside the inpainted region removes these cues and can degrade pretrained detectors. To address this, we present FUSED, a unified framework for the joint detection and localization of AI-generated inpainting. FUSED combines low-level forensic cues with high-level semantic features using a sparsely-gated Mixture-of-Experts architecture, enabling the model to adaptively prioritize the most relevant signal for each token. For each input, FUSED predicts both an image-level manipulation score and a pixel-level mask of the inpainted area. On the OpenSDID cross-generator benchmark, FUSED achieves the best average detection and localization, with the largest gains on unseen generators. The same model transfers directly to the held-out AutoSplice and CocoGlide benchmarks, more than doubling localization performance. Evaluating each held-out benchmark with and without the global generator artifact further shows that all evaluated methods, ours included, partly read the artifact as evidence of manipulation, and FUSED remains the strongest under both conditions. Code and pretrained models are available at https://github.com/AntonNuzhdin/FUSED.
Read the full abstract

Accelerator Physics

Wakefield-Dressed Relativistic Vortex Electrons in Plasma Accelerators

Plasma wakefield acceleration is usually regarded as a classical mechanism for producing high- energy charged-particle beams. Here we show that an axisymmetric plasma wakefield can also act as a moving quantum structure that supports relativistic vortex electron states. Starting from the Dirac equation, we derive the electron spinor eigenstates with definite total angular momentum in an ideal bubble-regime wakefield. The transverse focusing field confines and quantizes the electron transverse motion into Laguerre-Gaussian vortex modes, while the longitudinal electric field acceler- ates the electron without destroying the symmetry-protected angular momenta. We further analyze the localized and off-axis vortex electron wave-packets, and non-ideal wakefield perturbations, and identify the conditions for preserving electron-vortex-state purity. These results suggest plasma wakefield as a route toward high-energy vortex electrons and extend plasma-based acceleration from classical beam dynamics to quantum-state control of relativistic particles.
Read the full abstract

Robotics

MaCoPlanner: LLM-Assisted Manual-Compiled Task Planning with Proactive Safety Verification for Robotic Industrial Panel Operation

Robotic industrial panel operation requires not only accurate control localization but also compliance with operating procedures, safety rules, and device-state constraints distributed across heterogeneous manuals. This study presents MaCoPlanner, a task-planning framework built on knowledge compiled from equipment manuals that converts equipment manuals into a typed intermediate representation, retrieves task- and state-relevant evidence, and uses it to support plan generation. Before actuation, candidate plans are symbolically rolled out and checked against procedural and state-transition constraints; detected violations are localized and returned for targeted repair, while unresolved plans are rejected. A separate execution interface grounds verified symbolic actions to physical controls and updates the device state. Under an independent evaluation oracle, MaCoPlanner achieves a final violation rate of 2.7%, and 26.3% of the runs in the repair analysis are rejected after exhausting the refinement budget. Compared with Raw-Manual, task success increases from 62.8% to 84.4% on Level-2 tasks and from 25.9% to 43.2% on Level-3 tasks. Experiments on a controller-panel simulator without an attached industrial load further demonstrate integrated execution feasibility under representative interaction conditions, without claiming industrial deployment readiness.
Read the full abstract

All authors Guipeng Xin, Jiahe Xua, Mohammad Deghat, Chenhui Wan, Jie Liu, Youmin Hu, Zhongxu Hu

Nuclear Theory

Hadronic rescattering effects on net-proton cumulants from functional renormalization group calculations

Net-proton cumulants in the Beam Energy Scan region of heavy-ion collisions are widely used to probe critical fluctuations associated with the conjectured critical endpoint of Quantum Chromodynamics (QCD). Most existing studies, however, concentrate on the initial-state or phase-transition contributions, while the impact of hadronic rescattering on these observables has not been fully quantified. To address this gap, we construct event-by-event proton and antiproton distributions from functional renormalization group (fRG) cumulants using the maximum entropy principle, and propagate the resulting particles through the hadronic transport model SMASH in a simplified spherical evolution setup. We systematically investigate how the hadronic cascade modifies net-proton cumulants at collision energies $\sqrt{s_{NN}}=3.0$, 3.9, 4.9, 7.2, and 7.7~GeV. In the canonical-ensemble framework, which enforces exact net-baryon number conservation, the higher-order cumulant signal—in particular the ratio $C_4/C_2$ at $\sqrt{s_{NN}}=4.9$~GeV—is strongly reduced during the early stage of the cascade; the suppression of $C_4/C_2$ reaches approximately $20\%$. The non-monotonic energy dependence inherited from the fRG input survives the hadronic evolution, but its magnitude is substantially modified. These results demonstrate that hadronic rescattering provides a non-negligible background effect that must be accounted for when extracting QCD critical-point signals from experimental data.
Read the full abstract

All authors Qianru Lin, Shi Yin, Jianing Li, Hannah Elfner, Fabian Rennecke, Long-Gang Pang, Jan M. Pawlowski