arXiv subjects

Siyang Wang

Publications and source records attributed to Siyang Wang.

At least 19 recordsLinked to original sources

arXiv2026

From microscopic fluctuations to susceptibility spectra: single-molecule relaxation in glassy media

Single-molecule (SM) rotational dynamics of fluorescent probes in polystyrene near the glass transition temperature ($T_g$) are investigated over long times to reconstruct susceptibility spectra. The loss spectrum, commonly recorded using external field-driven (frequency-domain) spectroscopy, such as dielectric spectroscopy, is reconstructed from purely thermal SM rotational fluctuations. The results reproduce time-temperature superposition typically seen in dielectric spectroscopy for materials near $T_g$ and show that the ensemble spectrum is comprised of individual molecular responses to distinct environments.

cond-mat.soft↗

arXiv2026

Fracture energy of 6H-SiC at the microscale: effects of testing geometry and notch preparation

Micromechanical testing enables small-scale fracture energy measurements, but values depend strongly on geometry and specimen preparation. Here, the fracture energy of the single-crystal 6H-SiC {10-10} plane was measured using microscale double cantilever beam (DCB) and single cantilever beam (SCB) geometries. DCBs showed stable crack growth under displacement control and obtained 7.5 +- 0.3 J/m2. In contrast, SCBs notched by a Ga focused ion beam gave fracture energies over twice this value, indicating Ga implantation and near-notch residual stresses. Increasing the final notching current increased the measured fracture energy further. Although near-cryogenic notching limited ion-beam-induced damage, it did not reconcile SCB-derived values with DCB test results. Vacuum annealing substantially lowered the fracture energy and brought SCB results into close agreement with DCB measurements, whereas annealing in argon was less effective. Our findings highlight the importance of careful sample preparation and testing geometry selection for reliable fracture property measurement in ceramic materials.

cond-mat.mtrl-sci↗

arXiv2026

Enriched higher-order multiscale approaches with applications to wave propagation

We consider the numerical solution of partial differential equations with coefficients that are strongly heterogeneous in space. We provide an overview of higher-order localized orthogonal decomposition (LOD) methods for the elliptic setting, including recent advancements, and then present a generalization of the strategy to linear hyperbolic multiscale problems. We address the limitations of earlier constructions for the wave equation, which only achieve second-order convergence in space, independent of the chosen polynomial degree. Building on the methodology of enriched corrections recently developed for parabolic multiscale problems, we motivate and propose an enriched higher-order LOD method for the wave equation. The enriched corrections exhibit exponential decay and can be computed on patches. Under minimal assumptions on the coefficient and standard well-preparedness conditions on the data, we derive a priori error estimates that achieve optimal high-order convergence rates, thereby overcoming the previously observed saturation of the convergence rate. With the fifth-order Rosenbrock-Wanner (ROW) time integrator, we conduct a series of numerical examples to verify our theoretical results. We provide examples showing the optimal spatial convergence of the method including the localization errors for different polynomial orders. We also present examples showing the optimal convergence rates of the time discretization.

math.NA↗

arXiv2026

Optimal higher-order convergence rates for parabolic multiscale problems

In this paper, we introduce a higher-order multiscale method for time-dependent problems with highly oscillatory coefficients. Building on the localized orthogonal decomposition (LOD) framework, we construct enriched correction operators to enrich the multiscale spaces, ensuring higher-order convergence without requiring assumptions on the coefficient beyond boundedness. This approach addresses the challenge of a reduction of convergence rates when applying higher-order LOD methods to time-dependent problems. Addressing a parabolic equation as a model problem, we prove the exponential decay of these enriched corrections and establish rigorous a priori error estimates. Numerical experiments confirm our theoretical results.

math.NA↗

arXiv2026

Microstructural characteristics, atomic-scale features, and growth mechanisms of deuterides (hydrides) in hafnium

Hafnium hydride is a promising material for next-generation nuclear reactors, particularly as control rods for fast fission and shielding in fusion systems. The material's intrinsic brittleness encourages its use in the form of hydride-metal composites, where the functional and mechanical performance is strongly influenced by the multiscale structure of hydride-matrix interfaces. In this study, we employ a suite of microscopy techniques, including scanning electron microscopy with electron backscatter diffraction, transmission electron microscopy with electron energy-loss spectroscopy, and atom probe tomography, to investigate the deuteride-matrix interfaces in a deuterium-charged Hf alloy. We characterise their structure and chemistry, extracting key information including the deuteride-matrix crystallographic orientation relationship, microstructural features, misfit-induced dislocation distributions, electron energy-loss characteristics, and the segregation of oxygen during deuteride growth. These findings help clarify the mechanisms of interface evolution and may contribute to improved understanding of hydride-metal systems, with potential relevance for their processing, performance, and design in nuclear applications.

cond-mat.mtrl-sci↗

arXiv2025

From Sequential to Spatial: Reordering Autoregression for Efficient Visual Generation

Inspired by the remarkable success of autoregressive models in language modeling, this paradigm has been widely adopted in visual generation. However, the sequential token-by-token decoding mechanism inherent in traditional autoregressive models leads to low inference efficiency.In this paper, we propose RadAR, an efficient and parallelizable framework designed to accelerate autoregressive visual generation while preserving its representational capacity. Our approach is motivated by the observation that visual tokens exhibit strong local dependencies and spatial correlations with their neighbors--a property not fully exploited in standard raster-scan decoding orders. Specifically, we organize the generation process around a radial topology: an initial token is selected as the starting point, and all other tokens are systematically grouped into multiple concentric rings according to their spatial distances from this center. Generation then proceeds in a ring-wise manner, from inner to outer regions, enabling the parallel prediction of all tokens within the same ring. This design not only preserves the structural locality and spatial coherence of visual scenes but also substantially increases parallelization. Furthermore, to address the risk of inconsistent predictions arising from simultaneous token generation with limited context, we introduce a nested attention mechanism. This mechanism dynamically refines implausible outputs during the forward pass, thereby mitigating error accumulation and preventing model collapse. By integrating radial parallel prediction with dynamic output correction, RadAR significantly improves generation efficiency.

cs.CV↗

arXiv2025

Anomalous Translational Dynamics of Molecular Probes Near the Polymer Glass Transition

The origin of the dramatic slowdown of dynamics near the glass transition temperature (Tg) remains a long-standing fundamental and unresolved issue in soft condensed matter. While single-molecule (SM) experiments using fluorescent probes have provided critical insight for molecular and polymeric glass formers through rotational measurements, translational dynamics remain largely unexplored in such systems at the molecular length scale. Here, we report SM translational dynamics of molecular probes in high molecular weight polystyrene at three temperatures near Tg. The probes exhibit quasi-stationary position fluctuations, non-Gaussian displacement distributions, sub-diffusive transport with anti-correlated displacements, and a characteristic translational relaxation time. The observations are quantitatively described using a microscopic framework based on the generalized Langevin equation and supported by numerical modeling for heterogeneous transport. The translational dynamics of the probes provides direct microscopic evidence of dynamic heterogeneity and suggests a pathway to more fully understand glassy dynamics in glass formers near Tg.

cond-mat.soft↗

arXiv2025

LR-WaveHoltz: A Low-Rank Helmholtz Solver

We propose a low-rank method for solving the Helmholtz equation. Our approach is based on the WaveHoltz method, which computes Helmholtz solutions by applying a time-domain filter to the solution of a related wave equation. The wave equation is discretized by high-order multiblock summation-by-parts finite differences. In two dimensions we seek to compress the solution in matrix form, and in three dimensions using tensor trains. To control rank growth we use step-truncation during time stepping and a low-rank Anderson acceleration for the WaveHoltz fixed point iteration. We have carried out extensive numerical experiments demonstrating the convergence and efficacy of the iterative scheme for free- and half-space problems in two and three dimensions with constant and piecewise constant wave speeds.

math.NA↗

arXiv2025

Towards Context-Aware Human-like Pointing Gestures with RL Motion Imitation

Pointing is a key mode of interaction with robots, yet most prior work has focused on recognition rather than generation. We present a motion capture dataset of human pointing gestures covering diverse styles, handedness, and spatial targets. Using reinforcement learning with motion imitation, we train policies that reproduce human-like pointing while maximizing precision. Results show our approach enables context-aware pointing behaviors in simulation, balancing task performance with natural dynamics.

cs.RO↗

arXiv2025

Learning to Generate Pointing Gestures in Situated Embodied Conversational Agents

One of the main goals of robotics and intelligent agent research is to enable natural communication with humans in physically situated settings. While recent work has focused on verbal modes such as language and speech, non-verbal communication is crucial for flexible interaction. We present a framework for generating pointing gestures in embodied agents by combining imitation and reinforcement learning. Using a small motion capture dataset, our method learns a motor control policy that produces physically valid, naturalistic gestures with high referential accuracy. We evaluate the approach against supervised learning and retrieval baselines in both objective metrics and a virtual reality referential game with human users. Results show that our system achieves higher naturalness and accuracy than state-of-the-art supervised models, highlighting the promise of imitation-RL for communicative gesture generation and its potential application to robots.

cs.RO↗

arXiv2025

An energy-based discontinuous Galerkin method for the wave equation with nonsmooth solutions

We develop a stable and high-order accurate discontinuous Galerkin method for the second order wave equation, specifically designed to handle nonsmooth solutions. Our approach integrates the energy-based discontinuous Galerkin method with the oscillation-free technique to effectively suppress spurious oscillations near solution discontinuities. Both stability analysis and apriori error estimates are established for common choices of numerical fluxes. We present a series of numerical experiments to confirm the optimal convergence rates for smooth solutions and its robustness in maintaining oscillation-free behavior for nonsmooth solutions in wave equations without or with nonlinear source terms.

math.NA↗

arXiv2025

On the stability analysis of perfectly matched layer for the elastic wave equation in layered media

In this paper, we present the stability analysis of the perfectly matched layer (PML) in two-space dimensional layered elastic media. Using normal mode analysis we prove that all interface wave modes present at a planar interface of bi-material elastic solids are dissipated by the PML. Our analysis builds upon the ideas presented in [SIAM Journal on Numerical Analysis 52 (2014) 2883-2904] and extends the stability results of boundary waves (such as Rayleigh waves) on a half-plane elastic solid to interface wave modes (such as Stoneley waves) transmitted into the PML at a planar interface separating two half-plane elastic solids. Numerical experiments in two-layer and multi-layer elastic solids corroborate the theoretical analysis, and generalise the results to complex elastic media. Numerical examples using the Marmousi model demonstrates the utility of the PML and our numerical method for seismological applications.

math.NA↗

arXiv2025

Navigating Image Restoration with VAR's Distribution Alignment Prior

Generative models trained on extensive high-quality datasets effectively capture the structural and statistical properties of clean images, rendering them powerful priors for transforming degraded features into clean ones in image restoration. VAR, a novel image generative paradigm, surpasses diffusion models in generation quality by applying a next-scale prediction approach. It progressively captures both global structures and fine-grained details through the autoregressive process, consistent with the multi-scale restoration principle widely acknowledged in the restoration community. Furthermore, we observe that during the image reconstruction process utilizing VAR, scale predictions automatically modulate the input, facilitating the alignment of representations at subsequent scales with the distribution of clean images. To harness VAR's adaptive distribution alignment capability in image restoration tasks, we formulate the multi-scale latent representations within VAR as the restoration prior, thus advancing our delicately designed VarFormer framework. The strategic application of these priors enables our VarFormer to achieve remarkable generalization on unseen tasks while also reducing training computational costs. Extensive experiments underscores that our VarFormer outperforms existing multi-task image restoration methods across various restoration tasks.

cs.CV↗

arXiv2025

A hybrid numerical method for elastic wave propagation in discontinuous media with complex geometry

We develop a high order accurate numerical method for solving the elastic wave equation in second-order form. We hybridize the computationally efficient Cartesian grid formulation of finite differences with geometrically flexible discontinuous Galerkin methods on unstructured grids by a penalty based technique. At the interface between the two methods, we construct projection operators for the pointwise finite difference solutions and discontinuous Galerkin solutions based on piecewise polynomials. In addition, we optimize the projection operators for both accuracy and spectrum. We prove that the overall discretization conserves a discrete energy, and verify optimal convergence in numerical experiments.

math.NA↗

arXiv2025

Stable and high-order accurate finite difference methods for the diffusive viscous wave equation

The diffusive viscous wave equation describes wave propagation in diffusive and viscous media. Examples include seismic waves traveling through the Earth's crust, taking into account of both the elastic properties of rocks and the dissipative effects due to internal friction and viscosity; acoustic waves propagating through biological tissues, where both elastic and viscous effects play a significant role. We propose a stable and high-order finite difference method for solving the governing equations. By designing the spatial discretization with the summation-by-parts property, we prove stability by deriving a discrete energy estimate. In addition, we derive error estimates for problems with constant coefficients using the normal mode analysis and for problems with variable coefficients using the energy method. Numerical examples are presented to demonstrate the stability and accuracy properties of the developed method.

math.NA↗

arXiv2024

Enhancing Layout Hotspot Detection Efficiency with YOLOv8 and PCA-Guided Augmentation

In this paper, we present a YOLO-based framework for layout hotspot detection, aiming to enhance the efficiency and performance of the design rule checking (DRC) process. Our approach leverages the YOLOv8 vision model to detect multiple hotspots within each layout image, even when dealing with large layout image sizes. Additionally, to enhance pattern-matching effectiveness, we introduce a novel approach to augment the layout image using information extracted through Principal Component Analysis (PCA). The core of our proposed method is an algorithm that utilizes PCA to extract valuable auxiliary information from the layout image. This extracted information is then incorporated into the layout image as an additional color channel. This augmentation significantly improves the accuracy of multi-hotspot detection while reducing the false alarm rate of the object detection algorithm. We evaluate the effectiveness of our framework using four datasets generated from layouts found in the ICCAD-2019 benchmark dataset. The results demonstrate that our framework achieves a precision (recall) of approximately 83% (86%) while maintaining a false alarm rate of less than 7.4\%. Also, the studies show that the proposed augmentation approach could improve the detection ability of never-seen-before (NSB) hotspots by about 10%.

cs.CV↗

arXiv2024

Upwind summation-by-parts finite differences: error estimates and WENO methodology

High order upwind summation-by-parts finite difference operators have recently been developed. When combined with the simultaneous-approximation-term method to impose boundary conditions, the method converges faster than using traditional summation-by-parts operators. We prove the convergence rate by the normal mode analysis for such methods for a class of hyperbolic partial differential equations. Our analysis shows that the penalty parameter for imposing boundary conditions affects the convergence rate for stable methods. In addition, to solve problems with discontinuous data, we extend the method to also have the weighted essentially nonoscillatory property. The overall method is stable, achieves high order accuracy for smooth problems, and is capable of solving problems with discontinuities.

math.NA↗

arXiv2024

Evaluating Text-to-Speech Synthesis from a Large Discrete Token-based Speech Language Model

Recent advances in generative language modeling applied to discrete speech tokens presented a new avenue for text-to-speech (TTS) synthesis. These speech language models (SLMs), similarly to their textual counterparts, are scalable, probabilistic, and context-aware. While they can produce diverse and natural outputs, they sometimes face issues such as unintelligibility and the inclusion of non-speech noises or hallucination. As the adoption of this innovative paradigm in speech synthesis increases, there is a clear need for an in-depth evaluation of its capabilities and limitations. In this paper, we evaluate TTS from a discrete token-based SLM, through both automatic metrics and listening tests. We examine five key dimensions: speaking style, intelligibility, speaker consistency, prosodic variation, spontaneous behaviour. Our results highlight the model's strength in generating varied prosody and spontaneous outputs. It is also rated higher in naturalness and context appropriateness in listening tests compared to a conventional TTS. However, the model's performance in intelligibility and speaker consistency lags behind traditional TTS. Additionally, we show that increasing the scale of SLMs offers a modest boost in robustness. Our findings aim to serve as a benchmark for future advancements in generative SLMs for speech synthesis.

eess.AS↗