Artificial Intelligence
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 10 September
We present EXYGEN (EXplore Your Graphs ENgine), a framework for knowledge graph (KG) understanding that enables conversational access to KGs at scale. We address two questions in sequence. First, how effectively can LLMs perform text-to-SPARQL generation given only automatically derived structured metadata and small graph samples, rather than task-specific fine-tuning? We integrate VoID descriptions and ShEx schemas into a retrieval-augmented generation (RAG) pipeline and ablate KG-derived context on the SciQA benchmark. Our best configuration – combining ShEx schemas, retrieved triples, and example question-query pairs – reaches an exact match of 0.419 on execution results without any LLM fine-tuning. We further find that lexical metrics such as F1 poorly predict query correctness, and that larger general-purpose LLMs can outperform smaller code-specialized ones once given sufficient context. Second, we ask how to generate the structured metadata that this method relies on from very large KGs, where KG metadata generation becomes computationally intractable. We introduce a predicate-coverage-aware parallel graph sampling strategy that preserves structural diversity while remaining computationally tractable. On OpenCitations Meta and GESIS, it retains high predicate coverage with minimal triple loss and reduces runtime by over 80x; on ORKG, sampling is not just faster but the only tractable path to obtain complete metadata. Together, these results show that structured schema context and lightweight prompting can substantially reduce reliance on fine-tuning for scalable conversational access to KGs, though closing the remaining gap to fully fine-tuned approaches will likely require reducing dependence on curated question-query exemplars – whether through synthetic generation or an execution-feedback-driven approach – and validating these findings beyond a single benchmark.
Quantum Gases
Finite-time effects in periodically kicked systems
In this work, we study finite-time effects in ultracold atomic systems by considering time-dependent modulations with variable waveforms and durations. These two characteristics can be controlled by adjusting only a single parameter. For arbitrarily short pulses, our model recovers the paradigmatic kicked rotor while maintaining the impulse transmitted per period and unit amplitude constant. Furthermore, we demonstrate that finite-time effects have a profound impact on dynamical localization, a result that cannot be captured by the {\delta}-kicked-rotor model. Through a detailed analysis of the effects of different modulation amplitudes, periods, and waveforms, we identify the conditions for which dynamical localization is significantly enhanced. We show that the strength of dynamical localization increases sharply as the system approaches the {\delta}-kicked-rotor limiting case. Moreover, we establish the existence of an optimal value of the period that maximizes dynamical localization for given values of the amplitude and shape parameter.
Read the full abstract
Optimization and Control
Non-monotone direct-search methods for deterministic and stochastic derivative-free optimization
In derivative-free optimization (DFO), one minimizes functions for which the gradient is unavailable or expensive to compute. In many applications, objective function values and gradients are noisy due to simulations or system randomness. A class of standard direct-search methods for DFO accept a trial point when it decreases the objective function by an amount proportional to the squared stepsize. However, when applied to complex landscapes, such a requirement may trap the algorithm in a neighborhood of sub-optimal solutions. We study a non-monotone direct-search alternative where the trial function value is compared with the largest objective function obtained through the $M$ most recent distinct iterates. This max-$M$ non-monotone condition permits temporary increases in the objective function and can help navigate narrow curved valleys; however, its theoretical analysis is significantly more challenging due to the lack of monotonic decrease. In this paper, we develop a comprehensive complexity theory for the max-$M$ non-monotone direct-search in both deterministic and stochastic DFO problems. For deterministic objectives, we establish a worst-case iteration bound for a complete poll based on a positive spanning set and an expected iteration bound for a probabilistic-descent poll. We then analyze a stochastic variant using independent function estimates and show the expected iteration complexity under tail-bound assumptions of the stochastic errors. All three results have the standard complexity of $\mathcal{O}(\epsilon^{-2})$, which matches the iteration complexity of monotone direct-search methods. Our theory is enabled by a new family of merit functions that correct the stored objective values by ordered multiples of the squared stepsize, together with a renewal-reward stopping-time argument for the probabilistic methods.
Read the full abstract
Analysis of PDEs
Diffeomorphism Invariance of the Favre-Filtered Compressible Navier–Stokes System under Spacetime Dilation
We prove that the system of Favre-filtered compressible Navier–Stokes equations with subgrid-scale closure terms, written in the coordinates $(Lt,Lx_1,Lx_2,Lx_3)$ for an arbitrary parameter $L>0$, is obtained from the system written in standard coordinates $(t,x_1,x_2,x_3)$ by the action of a smooth diffeomorphism of the underlying spacetime manifold. The proof begins with a self-contained treatment of infinite-dimensional manifolds and the Fréchet manifold of smooth field configurations. We then define the one-parameter dilation group $\{\phi_L\}_{L>0}$, verify that each $\phi_L$ is a diffeomorphism, and compute its tangent and cotangent actions. The induced pullback action on sections of the configuration bundle is shown to commute with the total derivative operator, which implies that the first-jet prolongation of $\phi_L$ maps the equation submanifold of the system onto itself. Equation-by-equation verification—continuity, momentum, and energy—confirms that every residual is multiplied by the common nonzero factor $L^{-1}$ (or, equivalently, is identically zero in the transformed coordinates), so that the solution set is preserved. We conclude that the family of systems parameterised by $L>0$ satisfies the diffeomorphism transformation expression and that solution spaces are canonically isomorphic for all $L$.
Read the full abstract
Instrumentation and Detectors
High-Speed Semi-FE Readout Module for ATLAS MDT at HL-LHC: Design and Production-Level Characterization
The High-Luminosity upgrade of the Large Hadron Collider (HL-LHC) introduces increased demands on the ATLAS Muon Spectrometer, particularly in terms of data throughput, timing distribution and system reliability. The Phase-II Chamber Service Module (CSM) is a key component of the upgraded Monitored Drift Tube (MDT) trigger and readout system, providing a high-speed interface between the front-end electronics and the backend systems. This paper describes the design and implementation of the Phase-II CSM, together with its validation. The results show that the CSM supports two independent optical uplinks, each operating at a line rate of 10.24 Gbps, together with clock distribution and slow control in the expected operating environment. Integration with small-diameter MDT (sMDT) chambers and tests with the prototype L0MDT trigger system are also presented. The CSM boards are now in production and will be used for installation and integration during the upcoming LHC Long Shutdown.
Read the full abstract
Systems and Control
Critic-Free Policy Iteration for Continuous-Time Zero-Sum Games: A Policy-Space Riccati Approach
This paper develops a critic-free policy iteration (PI) method for continuous-time linear zero-sum games. The central idea is to characterize the saddle-point policies directly in the joint policy space, rather than treating the quadratic value matrix as an iterative variable. A policy game Riccati equation (PGRE) is introduced whose unknowns are policy gains only. Its solutions are shown to be in one-to-one correspondence with the symmetric solutions of the game algebraic Riccati equation. Based on a stabilizing anchor, PI is performed directly in the actor space. The actor-space Jacobian is nonsingular at every stabilizing policy, and the resulting policy sequence coincides with that of simultaneous PI. For unknown dynamics, a data-driven algorithm uses a single batch of data and nullspace projection of endpoint increments to eliminate the value matrix, yielding an actor-only regression. A necessary and sufficient rank condition for unique policy recovery is established and shown to be iteration-invariant, demonstrating that critic identifiability is unnecessary. A power systems frequency-regulation example verifies convergence and policy recovery, while scalability tests demonstrate substantial reductions in computational and memory requirements.
Read the full abstract
Plasma Physics
kobra: a new Vlasov code intended for plasma-wall modeling
In a fusion device plasma-wall interactions \edit{on the sheath scale} can be modeled as a collisionless problem. When modeling these regions particle-in-cell codes suffer from statistical error originating from undersampling the velocity space. On the other hand, Vlasov codes do not have this issue as they evolve the full distribution function. Here, we present a new finite-volume Vlasov code, kobra, equipped with adaptive-mesh refinement to reduce computational effort. Currently, the code solves the Vlasov-Poisson equations. We validate our code in 1d1v and 1d2v using established benchmarks, i.e. the two-stream instability, Landau damping, the Dory-Guest-Harris instability, and also a classical electrostatic plasma sheath. We find that the code reproduces the theoretical properties of these problems well. More importantly, the adaptive grid provides a computational gain that is likely to scale to higher dimensional, plasma-wall simulations.
Read the full abstract
Distributed Computing
Entwine: Coordinating Tiled Computation and Fine-Grained Communication across GPUs
Modern high-performance GPU computations partition tensors into tiles to exploit data reuse and parallelism. Individual tile computations complete earlier than the full tensor computation, creating opportunities to overlap computation and communication. However, a mismatch between computation and communication progress can limit these opportunities. Communication stalls when no data is ready, and may lag when data arrives in bursts. Communication can also slow computation by consuming shared resources, offsetting the benefits of overlap. We present Entwine, which coordinates tile computation order, fine-grained communication, and SM resource allocation to minimize overall completion time. Entwine reorders tile computation to produce data for communication at a more regular pace. Entwine couples this schedule with fine-grained SM-based communication to process tile results with low latency and low overhead. Since the communication kernel also consumes SM resources, Entwine coordinates their allocation to balance communication progress against computation slowdown. Across representative tensor-parallel LLM workloads, Entwine achieves a geomean speedup of 1.232x (up to 1.433x) over cuBLAS+NCCL, and outperforms state-of-the-art overlap baselines by 3.1-9.8% in geomean. We will open-source our implementation upon publication.
Read the full abstract
Robotics
Memory as Plans: World-Action Modeling with Memory-Grounded Planning
Mainstream robotic policies often adopt a Markovian formulation, but many complex real-world manipulation tasks are inherently non-Markovian, requiring long-horizon memory beyond the current observation. Existing memory mechanisms often rely on language summaries, growing visual windows, or their combinations, and may therefore lose fine-grained visual evidence or face a trade-off between history coverage and execution efficiency. We introduce MaP-WAM, a Memory-as-Plans framework that decomposes memory-dependent world-action modeling into memory-grounded planning and plan-conditioned execution, and uses long-term multimodal episodic context as planning-time evidence rather than repeatedly conditioning the executor on the full history. MaP-WAM represents memory as completed segment records containing language instructions and sparse visual context, and converts this episodic memory into compact plans comprising the next segment-level language plan and corresponding visual guidance. A World-Action-Progress (WAP) model executes each plan over an unknown duration by jointly predicting action chunks and corresponding execution progress at inference time, calibrating predicted progress through plan-observation alignment for adaptive segment transitions and closed-loop context updates. MaP-WAM keeps the executor context length fixed, while structured attention further enables key-value caching in both planning and execution. MaP-WAM achieves state-of-the-art performance on RMBench with an 83.3% success rate and attains 78.0% success on real-robot tasks, while maintaining approximately constant executor inference latency as task history grows.
Read the full abstract
Functional Analysis
The Local Embedding Problem for Hardy Spaces of Dirichlet Series
We solve the local embedding problem for Hardy spaces of Dirichlet series, which is a dimension-free trace problem asking whether the global $\mathscr{H}^p$-norm controls local $L^p$-mass on the critical line $\operatorname{Re}s=1/2$. More precisely, for every $2<p<\infty$, there exists a constant $C_p<\infty$ such that every Dirichlet polynomial $P$ satisfies $$ \sup_{\theta\in\mathbb{R}}\int_{\theta}^{\theta+1}\left|P\left(\frac12+it\right)\right|^p\,\mathrm{d}t\le C_p\left\lVert P\right\rVert_{\mathscr{H}^p}^{p}, $$ with $C_p$ independent of the number and choice of prime variables on which $P$ depends. Before the present work, the embedding was known at $p=2$ and, by taking integer powers, at the even exponents $p=2k$; it had been conjectured that these exhaust the finite positive cases above $2$. Together with the known failure for $0<p<2$, our theorem gives the sharp finite-exponent classification: the local embedding property holds exactly for $p\ge2$. Thus, the true threshold is $p=2$, rather than even integrality. The proof passes to the dual exponent $q=p/(p-1)\in(1,2)$, where an exact frequency decomposition isolates a single resonant Euler-product term. A covariance-preserving replacement of the shared prime factors reduces the resulting fractional-moment estimate to a log-correlated Gaussian field, and a critical branching-random-walk bound supplies the required multiscale decay. A finite-cyclic square-function estimate assembles the resonant scales, and Hardy-quotient duality converts the resulting vector-valued bound into the critical-line trace. For $1\le p<\infty$, known equivalences give the same sharp threshold in several classical problems, including the conformally invariant half-plane embedding, the reverse local Carleson-measure transfer, and boundedness of all characteristic-zero Gordon–Hedenmalm composition operators.
Read the full abstract
Software Engineering
PRISMA-LLM: An Empirical Reporting Framework for AI-Assisted Systematic Reviews
Large language models (LLMs) and AI-enabled software increasingly participate in systematic-review decisions, yet the information needed to audit these workflows is reported inconsistently. We analyze SciLitBench, a corpus of 888 review-automation papers with 14,726 annotations, to characterize changes in methods, review-stage use, evaluation and reported limitations. Automation has shifted toward LLM- and software-facing workflows, including stages that can alter the evidence base. Since 2023, 38.0% of software/product papers reported no evaluation, compared with 9.3% of LLM papers. Reporting coverage increased with LLM workflow complexity, yet 52% of positive-only LLM evaluations still reported an unmet reliability or performance requirement. From these patterns, we introduce PRISMA-LLM, an empirically grounded framework separating implementation disclosure from consequence-sensitive evaluation and limitation reporting.
Read the full abstract
Probability
The radial part of $p$-Brownian motion
We initiate a geometric theory of $p$-Brownian motion, the nonlinear Markov process associated with the $p$-Laplacian introduced by Barbu-Rehmeier-Röckner. More precisely, we analyze its radial processes thoroughly, relative to an arbitrary center and in every dimension. On the one hand, we show explicit Tanaka–Meyer semimartingale formulas; our consequential characterization of nontriviality of the associated local times reveals notable differences to classical Brownian motion. In parallel, we establish sharp exit time estimates as well as scaling estimates for the self-similarly rescaled radial process. We also prove isometry of the corresponding marginal laws in each Wasserstein distance. Our contributions equally cover the Leibenson process - the nonlinear Markov process associated with the "porous medium equation" with $p$-Laplacian - recently introduced by Barbu-Grube-Rehmeier-Röckner.
Read the full abstract
Probability
Martingale central limit theorems in $p$-Wasserstein distance
We obtain multivariate martingale central limit theorems in $p$-Wasserstein distance with respect to the $\ell_r$ norm in $\mathbb{R}^d$ for $p\geq 1$ and $r\in [1,\infty]$, which generalize the results for $p=1$ and $r=2$ in the literature. As corollaries, we obtain the Yurinskii coupling and Cramér-type moderate deviation results. We also provide an illustrative application to the stochastic gradient descent algorithm. To prove our main results, we combine Lindeberg's swapping argument with a new Gaussian convolution inequality controlling the $p$-Wasserstein distance between a Gaussian convolved with a perturbation and the Gaussian with matching mean and covariance matrix. The latter is obtained by developing the recent line of research on $p$-Wasserstein bounds.
Read the full abstract
Complex Variables
A counterexample to an open problem of Dorff
The classical Pólya-Schoenberg conjecture, proved by Ruscheweyh-Sheil-Small, asserts that the convolution of two normalized convex univalent functions is again convex. This property fails to carry over to planar harmonic mappings. In 2001, Dorff posed the open problem whether the self-convolution of a normalized convex harmonic mapping with bounded image must remain in the same class. We construct a normalized sense-preserving harmonic diffeomorphism that maps the unit disk onto an ellipse; its self-convolution has vanishing Jacobian at some interior point of the unit disk, which provides a negative answer to Dorff's open problem.
Read the full abstract
Astrophysics of Galaxies
Testing Chemical Tagging with LAMOST: Intrinsic Abundance Dispersion of Subgiant Stars in the Galactic Disk
The scatter in elemental abundances among stars of similar age and metallicity reflects chemical inhomogeneity in their birth environments, making abundance scatter a key observable for chemical-tagging studies of Galactic formation and evolution. Using a large sample of subgiant stars with precise ages and elemental abundances derived from LAMOST low-resolution spectra, we investigate the intrinsic chemical abundance scatter of the low-$\alpha$ thin disk near the solar neighborhood ($7 < R < 10$ kpc). We model the abundance ratio [X/Fe], for each of the 19 elements of concern, as a function of age, [Fe/H], and [Mg/Fe], and deduce the intrinsic dispersions with a forward modelling technique. Our results confirm previous findings that the intrinsic scatters are small, typically $\lesssim0.05$~dex, for light elements (C, Al), $\alpha$-elements (O, Mg, Si, Ca, Ti), and iron-peak elements (Mn, Ni). A dedicated analysis of M67 yields similarly small scatter values for these elements, implying limited discriminatory power from light-element abundances alone. In contrast, neutron-capture elements exhibit substantially larger scatters, typically $\gtrsim$0.1 dex, which are significantly larger than those of M67 member stars ($\sim$0.07~dex). In particular, our analysis suggests that the abundance variations of individual neutron-capture elements cannot be explained by a single tracer such as [Ba/Fe]. These findings clarify the utility of neutron-capture elements for chemical tagging and highlight the potential of low-resolution spectroscopy in such studies.
Read the full abstract
Solar and Stellar Astrophysics
Evidence for CME–CME Interaction in a Magnetic-Cloud-Like Ejecta: Insights from Multipoint Observations and Polytropic Analysis
Using in-situ observations from Solar Orbiter, STEREO-A, and Wind, we investigate the heliospheric evolution of an interplanetary coronal mass ejection (ICME). The magnetic ejecta (ME) shows a magnetic-cloud-like (MCL) configuration: a front region in which the magnetic field rotates and its magnitude declines, followed by a weakly rotating "back region" of nearly constant field magnitude. Near-Sun EUV and white-light observations reveal two fast CMEs launched in rapid succession and interacting at low coronal heights, providing direct evidence that an MCL ejecta can arise from interaction between two closely spaced CMEs sampled near their apex. Using near-radially aligned Solar Orbiter and STEREO-A observations, we find that the sheath expands more rapidly than the ME, consistent with the "snow-plow" effect, while the ME properties broadly follow trends reported in previous ICME studies. Comparison of STEREO-A and Wind observations, separated by only 9.2 degrees in longitude, reveals measurable mesoscale variability within both the sheath and the ME. We identify dual polytropic behavior within both substructures, consistent with the merging of two interacting CMEs to form the observed MCL ejecta. This dual behavior in the ME is most pronounced in the inner heliosphere and weakens with heliocentric distance, indicating thermodynamic homogenization during radial expansion, whereas the sheath maintains thermodynamic contrasts from the inner heliosphere to 1 au. Our findings suggest that polytropic diagnostics can help reveal thermodynamically distinct plasma populations associated with interacting CMEs, providing additional evidence for CME-CME interaction in MCL ejecta where conventional in-situ signatures show no clear evidence of such interactions.
Read the full abstract
Robotics
CAP: Continuously Adaptive Perception-Blind Humanoid Locomotion via Learned Denoising
Humanoid locomotion across complex terrain demands forward-looking exteroception to anticipate obstacles, yet this signal is unreliable in real-world deployment, failing partially and intermittently. Existing perceptive policies often assume that depth observations remain clean and in-distribution, while recent attempts to unify perceptive and blind control typically route or switch between separate sub-policies, leaving recoverable information in partially corrupted depth unexploited. We instead propose CAP, a single-stage humanoid locomotion policy that recovers this signal with a perceptive world-model encoder trained as a learned denoiser to reconstruct clean depth from a corrupted input, together with a co-active proprioceptive variational encoder that supplies depth-free body-state information. A coupled training recipe pairs a depth-noise curriculum on the world-model input with world-model feature dropout on the policy-facing latent, exposing the policy to failures across the entire perception-quality spectrum. In simulation, CAP matches or improves upon perceptive baselines when depth remains informative, and degrades more smoothly than a binary-switching baseline as perception worsens. On the Unitree G1, controlled trials and indoor-outdoor deployments demonstrate perception-robust locomotion under intermittent occlusion, real-sensor corruption, and outdoor depth artifacts.
Read the full abstract
Cryptography and Security
Accountability in Certificate Transparency and Variants
Certificate Transparency (CT) aims to reduce the trust required in Certificate Authorities (CAs) within the TLS certificate ecosystem. It is supported by all major browsers. The protocol obliges all CAs to record the certificates they issue in a public log, which itself is monitored for compliance and consistency by third parties. Given this complex set of checks between the four roles-CA, loggers, monitor but also the end user's client-it is very hard to provide a precise account of how CT eliminates trust assumptions in exchange for complex infrastructure. Analyses both in the Dolev-Yao paradigm and the computational paradigm only regard a very simplified model and feature definitions adapted specifically to CAs, essentially capturing design features rather than the target property. The present paper posits accountability as the main goal of CT and presents a thorough analysis in the Dolev-Yao model. We start with the vanilla PKI and, step by step, move to CT, finally analyzing proposed extensions for SCT Auditing and Gossiping. We show that plain CT relies on an honest log, but provides accountability under this assumption. Furthermore, we show that the SCT Auditing extension can eliminate this assumption, while the Gossiping extension cannot.
Read the full abstract
Materials Science
Skyrmion nucleus resolves the Landauer paradox
Ferroelectric switching is among the most fundamental and widely studied symmetry-breaking processes in physics. However, the Landauer paradox asserts that an ideal single-domain ferroelectric should be kinetically unswitchable, because a three-dimensional reversed nucleus would carry a prohibitive depolarization penalty, leading to unphysically large coercive field (the electric field required to switch the polarization). Here we show that this long-standing paradox is resolved by topology. In defect-free PbTiO3, large-scale molecular dynamics simulations reveal that the intrinsic critical nucleus is a three-dimensional polar skyrmion. Its continuous polarization rotation and Néel-type boundary walls self-compensate bound charge, suppressing the depolarization energy by orders of magnitude and bringing the predicted coercive field into agreement with experiment. This topological nucleus further overturns the 60-year-old Janovec-Kay-Dunn law. We derive an analytical thickness-dependent coercive-field law governed by field-induced softening of the parent ferroelectric state and controlled by a single material descriptor. This framework captures diverse experimental trends across ferroelectric families and recasts the Kay-Dunn exponent of -2/3 as an effective finite-window behavior. These results establish polar topology as an organizing principle for intrinsic switching and suggest that hidden topological transition states may broadly shape nonequilibrium phase transformations traditionally understood as Landau-type symmetry breaking.
Read the full abstract
Computer Vision
A Comparative Evaluation of Pre-trained Convolutional Neural Networks for Melanoma Detection
Early diagnosis of melanoma is critical for improving patient survival rates. However, accurately distinguishing melanoma from other skin lesions remains a significant clinical challenge due to the high visual similarity among lesion types and variability in image acquisition conditions. Artificial intelligence, particularly machine learning, has emerged as a promising tool to support dermatological diagnosis by automating feature extraction from medical images. Among the available approaches, convolutional neural networks (CNNs) have demonstrated strong performance in image classification tasks, making them well-suited for analyzing both dermatoscopic and histopathological images, given their ability to capture hierarchical visual patterns relevant to lesion characterization. Nevertheless, despite numerous pre-trained CNN architectures having been proposed, selecting the most appropriate one for a given imaging modality remains an open challenge. In this study, we evaluate pre-trained convolutional neural networks (CNNs) for skin lesion classification using dermatoscopic and histopathological image datasets. Experiments were conducted on the HAM10000, ISIC 2018, and CR-AI4SkIN datasets, evaluating the ResNet50, VGG16, VGG19, MobileNet, and InceptionV3 architectures under the same training protocol. The experimental evaluation showed that the models achieved accuracies ranging from 71% (InceptionV3 on ISIC 2018) to 84% (ResNet50 on HAM10000) on dermatoscopic images. For histopathological images, accuracies ranged from 72% (VGG19) to 83% (ResNet50) on the CR-AI4SkIN dataset. The results demonstrate that model performance differs between dermatoscopic and histopathological image modalities, showing that architectures exhibiting similar performance on dermatoscopic images exhibit different performance on histopathological data.
Read the full abstract
Robotics
Using Automated Vehicles Operational Data to Confirm Safety and Anticipate Threats
European Union (EU) policymakers adopted revolutionary data collection provisions for Automated Driving Systems (ADS) in the recently approved regulation that allows driverless vehicles to be operated on public roads. The framework is inspired by best practices developed at the United Nations Economic Commission for Europe(UNECE) level: the In-Service Monitoring and Reporting (ISMR); and by similar operational data collection regulatory approaches in nuclear energy production and transportation fields. The collection of real-world data will enable the competent safety authorities to gather the information needed to confirm the homologation safety target. Safety-relevant driving scenarios discovered during the real-world operation of a given ADS can also be stored in a scenario catalogue to investigate how other ADS types might have addressed such a traffic conflict. Moreover, lessons learnt deriving from the data collected can be shared among original equipment manufacturers (OEMs) and safety authorities. Ultimately, the ISMR is recognised as a necessary tool to properly tackle the challenges associated with ADS safety assessment given the number of unknowns that might remain undisclosed by leveraging the traditional homologation validation scheme only.
Read the full abstract
Computer Vision
World in World: Explore the World with World Models
Autoregressive video world models enable interactive, long-horizon exploration, but flexible control remains challenging. Exploring a source video from new viewpoints requires the generated rollout to remain synchronised with the recorded event, place observed content in the requested view, plausibly complete newly exposed regions, and recover previously generated appearance on revisits. Existing methods typically address these requirements through task-specific modules or additional training. We present World in World, a training-free inference-time interface that converts heterogeneous control evidence into camera- and time-labelled clean visual states, which are read through the native self attention of a frozen causal video model. The evidence comprises source-video observations, target-view scene projections, geometry renderings that guide completion of newly exposed subject regions, and retrieved generated states beyond the rolling cache. Each evidence source carries token-level support and its own availability schedule. A correspondence router combines persistent point identities with geometry to establish token correspondences, guiding supported queries towards matching source-video tokens. Evidence-wise attention CFG (EWA) then independently regulates each auxiliary channel's additional contribution using attention responses from the same denoising forward pass. The shared interface supports camera-controlled rerendering, long-horizon revisiting, and human-motion transfer with the same frozen backbone. We evaluate World in World on camera-controlled video rerendering under diverse viewpoint changes, assessing perceptual quality, temporal consistency, and camera-following accuracy.
Read the full abstract
Mesoscale and Nanoscale Physics
One-Dimensional Type-IV Collinear Magnetism
Two-dimensional type-IV collinear magnets combine nonrelativistic spin degeneracy with spin-orbit-coupling-enabled magnetic responses. Here we establish a one-dimensional counterpart through altermagnetic confinement. A Lieb-based construction connects three-dimensional altermagnets to two-dimensional type-IV magnets and extends the same principle to one-dimensional ribbons. In a symmetric $[110]$ ribbon, a sublattice-exchange operation preserves the longitudinal momentum and enforces nonrelativistic spin degeneracy. {Spin-orbit coupling (SOC) ties spin to the lattice, so the Néel orientation determines whether the remaining symmetries protect the doublets or allow them to split.} First-principles-derived Wannier calculations for Ta$_2$TeSeO ribbons reveal spin-polarized propagating channels with an ideal right-moving-mode polarization of approximately $35\%$ at finite conductance for the parent easy-axis orientation $\mathbf N\parallel[100]$. Direct DFT calculations of Ta$_2$Te$_2$O ribbons resolve the contrast between spin-split $[100]$ and spin-degenerate $[001]$ SOC spectra. Assembling the same isolated Ta$_2$Te$_2$O ribbons into a transversely periodic array restores spin splitting without SOC. This conversion parallels the recovery of three-dimensional altermagnetism through suitable periodic stacking of two-dimensional type-IV magnets. One-dimensional type-IV magnetism thus provides a route to spin splitting through the interplay of antiferromagnetic order and SOC, distinct from conventional Rashba splitting driven by structural inversion asymmetry.
Read the full abstract
Analysis of PDEs
The Singular Set of Minima of Multiple Integrals
The Hausdorff dimension of the singular set of local minimizers of quasiconvex integrals is strictly less than the ambient dimension.
Read the full abstract
Computation and Language
Complex-Text Robustness Evaluation and Failure Diagnosis for Low-Resource Multilingual Text-to-Speech
Low-resource multilingual text-to-speech (TTS) systems have expanded language coverage, but their robustness under complex text inputs remains insufficiently diagnosed. Existing evaluations mainly focus on naturalness, speaker similarity, and content consistency using regular test sentences, while providing limited insight into how multilingual TTS systems fail when handling challenging inputs such as numbers, dates, named entities, long sentences, code-switched expressions, and punctuation-related structures. This paper proposes a complex-text robustness diagnosis framework for low-resource multilingual TTS. We evaluate robustness from three dimensions: content consistency, language consistency, and generation stability. A multilingual robustness testing scheme is designed for Thai, Vietnamese, Swahili, and Indonesian, covering ordinary sentences and multiple types of complex text inputs. We further introduce automatic diagnostic metrics, including character error rate, language identification accuracy, and duration abnormal rate. To support input-level risk analysis before speech generation, we propose a lightweight Text Risk Score (TRS), which estimates synthesis risk from interpretable text features without manual annotation or model training. Experiments on three representative multilingual TTS systems, including OmniVoice, VoxCPM2, and MMS-TTS, show that complex text inputs expose systematic failure patterns that are not fully reflected by ordinary short-sentence evaluation. Different systems exhibit distinct vulnerabilities in number normalization, named entity handling, long-text generation, and code-switched input processing. Furthermore, TRS shows a positive correlation with content errors and duration abnormalities, demonstrating its usefulness as a low-cost pre-synthesis indicator for complex-text risk diagnosis in low-resource multilingual TTS.
Read the full abstract
Atomic Physics
Kinetic modeling of molecular beam formation in a cryogenic buffer-gas cell
Cryogenic buffer-gas cells are widely used to produce cold molecular beams, but the microscopic dynamics governing beam formation remain challenging to model. Here we present fully kinetic simulations of a cryogenic buffer-gas cell using the Direct Simulation Monte Carlo method implemented in the PICLas framework, treating the buffer gas and ablated molecules within a single unified model. We capture characteristic features of cryogenic buffer-gas sources, including plume cooling, directed transport toward the aperture, and the formation of a slow molecular beam, while also resolving energy transfer from the hot ablation plume to the helium buffer gas that is inaccessible to existing approaches relying on the background-gas approximation. Our results demonstrate that fully kinetic simulations can provide detailed insights into buffer-gas cell dynamics and open a route toward a systematic optimization of such sources.
Read the full abstract
Probability
Fourth-Order Fusion Asymptotics for $\mathrm{Sine}_\beta$ Correlation Functions
We compute the fourth-order correction to the full-collision asymptotics of the correlation functions of the $\mathrm{Sine}_\beta$ process. For $m \ge 2$ and $m\beta > 3$, the normalized correlation has an expansion through order $\epsilon^4$, with an explicit rational coefficient depending on the centered profile only through its fourth power sum and the square of its second power sum. The remainder is $o(\epsilon^4)$, locally uniformly in the collision profile. We evaluate the required fourth inverse moments of the Hua-Pickrell environment by finite-dimensional Ward identities and prove their convergence using characteristic-polynomial derivative bounds. A fourth-order expectation-Taylor lemma handles the full range $m\beta > 3$ without requiring fourth moments of every analytic derivative. For general unitary ensembles with a $C^4$ confining potential and a regular bulk point, we prove convergence of the finite-particle fusion coefficients through fourth order and a joint second-order limit, using complex kernel universality and divided differences. This unitary result imposes no fused-environment hypotheses. For arbitrary $\beta$, we retain a conditional quadratic transfer criterion.
Read the full abstract
Artificial Intelligence
Characterizing Job Power Elasticity for Power-Flexible AI Training
Large language model (LLM) training is among the fastest-growing sources of electricity demand in modern data centers, and power availability is a primary bottleneck to continued AI infrastructure growth. Making the power consumption of these workloads flexible could unlock additional power for AI growth, limit increases in electricity prices, and improve the utilization of existing grid infrastructure. However, to realize this flexibility, we must first understand how the performance of training workloads changes when GPU power is reduced. This paper presents the first systematic characterization of job power elasticity (the sensitivity of throughput to power reductions) in LLM training. To quantify elasticity, we introduce the Power Flexibility Index (PFI), a normalized metric that quantifies the performance cost of power reductions and provides a control primitive for SLA-aware power flexibility. We collect data from 131 LLM training runs on H200 (plus 24 H200 validation runs and 34 matched H100 runs), including both dense and mixture-of-experts models, pretraining and fine-tuning tasks, and up to 32 GPUs. We find that LLM training jobs exhibit substantial but variable power elasticity, and we identify telemetry signals that predict PFI at runtime. Finally, we demonstrate that PFI-aware power allocation maximizes total tokens/second throughput under power constraints. Under a 30% power reduction, PFI-aware power allocation recovers ~1.5k tokens/s per job, 63% of the performance gap between an equal-weight allocation and an oracle with perfect information. Our results establish power elasticity as a measurable property of training jobs and provide a foundation for power-aware, grid-responsive AI infrastructure.
Read the full abstract
Differential Geometry
A Complex Structure on A Rational Homology Six-Sphere with Two-Torsion
We construct a compact, smooth, simply connected complex threefold. Its second and third integral homology groups are cyclic of order two, and its remaining homology groups agree with those of the six-sphere.
Dynamical Systems
Basins of attraction and escape in the Lozi map
For the Lozi map $L_{a,b}$, we consider parameter pairs for which the fixed point $X$ in the first quadrant has no homoclinic points and the period-two orbit $\{P,P'\}$ is attracting. For such parameters, let $\ell$ denote the set of accumulation points of the unstable manifold $W_X^u$ that do not belong to $W_X^u$. We completely classify the forward asymptotic behavior of points in the phase space. The forward orbit of every point in the plane either converges to $X$, to the other fixed point $Y$ in the third quadrant, or to $\ell$, or it escapes to infinity. The global phase space is organized by the stable manifolds of the fixed points: $W_Y^s$ separates the basin of $\ell$ from the region of escaping orbits, while $W_X^s$ is the exceptional set of points whose orbits converge to $X$. In particular, if $\mathcal{A}_1$ denotes the component of $\mathbb{R}^2 \setminus W_Y^s$ containing $X$, then $\mathcal{A}_1 \setminus W_X^s$ is precisely the basin of attraction of $\ell$.