arXiv subjects

Xudong Sun

Publications and source records attributed to Xudong Sun.

At least 19 recordsLinked to original sources

arXiv2026

AstronOS: A Unified Execution Model and Runtime for Long-Horizon Agentic Systems

Agentic systems often organize execution and state around a single conversation, model invocation, or agent instance, even when real work spans many calls and stages. We introduce a unified execution model that maintains a work item's persistent identity and versioned authoritative state across calls. Each step receives input scoped to a specific state version and new material; a result advances state only after validation and recording. We implement selected paths of this model in AstronOS using Cases, Tasks, and Scenario Packs across central and local execution. We compare five complete strategies for carrying an established software-version update plan into a fresh model session: rereading original materials, replaying full history, deterministic text summary, deterministic JSON, and the AstronOS runtime-mediated handoff. Ten controlled tasks are run under all five strategies with three repetitions, yielding 150 included executions. On the single-stage reference family, strategies perform similarly. In the primary three-stage A-C batch, AstronOS passes the frozen scorer in 14 of 15 executions, compared with 0 of 15 for rereading and 2 of 15 for full-history replay; later non-interleaved summary and JSON batches each pass 0 of 15. AstronOS has lower attempt-accounted model-token cost per passing execution, while requiring more execution-window time per attempt. These results associate the complete AstronOS condition with higher end-to-end pass rates across fresh sessions in this benchmark, at a measurable time cost.

cs.AI↗

arXiv2026

Massless scalar scattering by Kerr-Bertotti-Robinson black holes:transparent-end channels and superradiance

We formulate and numerically solve the real-frequency scattering problem for a neutral, minimally coupled, massless scalar on the rotating Kerr--Bertotti--Robinson (Kerr--BR) black-hole geometry, using a specified transparent boundary condition at the coordinate end. Since the Maxwell stress tensor of the background is traceless, the minimally coupled wave equation reduces to the four-dimensional conformal wave equation, enabling exact Carter-like separation after scaling the scalar field by the conformal factor. Unlike the asymptotically flat case, the coordinate end $r\to\infty$ lies at a finite tortoise distance. We show that the resulting reflection data are conditional on this boundary prescription rather than defining a unique, observer-independent cross section. Within the transparent-end model,open-channel superradiance is governed by a double-gate mechanism requiring both the local horizon condition and the outer propagation condition $q_\infty^2>0$. At the benchmark spin $a/M=0.9$, the co-rotating dipole amplification decreases as the external magnetic field increases. Crucially, the field narrows and closes the open superradiant window at $BM\simeq0.243$. Near the propagation threshold, the amplification coefficient vanishes linearly with the outer wave number.

gr-qc↗

arXiv2026

Convergence of Consensus-Based Particle Methods for Nonconvex Bi-Level Optimization

In this paper, we study a consensus-based optimization method for nonconvex bi-level optimization, where the objective is to minimize an upper-level function over the set of global minimizers of a lower-level problem. The proposed approach is derivative-free, and constructs its consensus point via smooth quantile selection combined with a Gibbs-type Laplace approximation. We establish convergence guarantees for both the associated \textit{mean-field} dynamics and its \textit{finite-particle} approximation. In particular, under suitable assumptions on smooth quantile localization, error bounds, and stability, we show that the mean-field law reaches any arbitrary prescribed Wasserstein neighborhood of the target bi-level solution with an explicit exponential rate up to the hitting time. Numerical experiments on a two-dimensional constrained problem and neural network training further support the theoretical results.

math.OC↗

arXiv2026

Scattering and absorption of a charged massive scalar field by a Reissner-Nordstr\"om black hole surrounded by perfect fluid dark matter

We study the scattering of charged massive particles impinging on a Reissner-Nordstr\"{o}m (RN) black hole immersed in perfect fluid dark matter (PFDM). We obtain an approximation of absorption cross section in the low-frequency regime via the matching method. In the high-frequency regime, we derive the weak-field deflection angle up to second order. The numerical results are in excellent agreement with classical approximation and glory scattering. The effects of dark matter, particle charge, and mass upon scattering and absorption are examined in detail. The results show that as the dark matter parameter $\lambda$ increases, the absorption cross section of the black hole is strongly suppressed, and its high-frequency limit depends only on the black hole charge $Q$ and $\lambda$. The scattering cross section also decreases overall. In the superradiant regime, the amplification factor of the PFDM black hole is much larger than that of the RN black hole. Finally, we discuss the behavior of the absorption cross section as $\omega/m\rightarrow1$, as well as the scattering cross section at small scattering angles.

gr-qc↗

arXiv2026

Energy Evolution from the Chromosphere to the Heliosphere in the 2021 October 28 Solar Eruption

We perform a detailed study of the energetics for a well-observed solar eruption and flare that occurred on 28 October 2021. This event included a GOES class X1.0 flare, a global EUV wave, and a coronal mass ejection that reached speeds of >2000 km/s. The event was observed from a variety of spacecraft in NASA's Heliophysics System Observatory, including multiple missions near Earth, STEREO-A off the Sun-Earth line, and Solar Orbiter, near the Sun-Earth line at about 0.8 au. Using remote sensing, in situ observations, and in some cases scaling laws based on previous observations, we characterize the following quantities: free magnetic energy, energy in non-thermal electrons, energy in non-thermal ions, bolometric energy, energy deposited in the chromosphere, thermal energy radiated in the flare loops, energy dissipated by the EUV wave, CME kinetic and gravitational potential energy, CME energy flux in the heliosphere, and the energy partition in the CME shock. We find that the total energy released during the event is consistent with estimates of the pre-event stored magnetic energy, and the CME kinetic + potential energy dominates the energy partition.

astro-ph.SR↗

arXiv2026

Towards a Robust Estimate of the Solar Photospheric Poynting Flux and Helicity Flux

The observed solar photospheric magnetic fields and Doppler velocities are frequently used to quantify the Poynting flux and helicity flux. Multiple methods have been developed for this purpose, but their estimates of the Poynting flux and helicity flux often differ from one another. Here we study the performance of three widely used methods on NOAA active region 12673: "PTD-Doppler-FLCT Ideal" (PDFI), "Differential Affine Velocity Estimator for Vector Magnetograms" (DAVE4VM), and an extension of the latter with Doppler velocity constraint (DAVE4VMwDV). We find that the values of the accumulated energy and helicity differ significantly between the three methods, even in signs. Using the Helmholtz-Hodge decomposition, we show that Doppler velocity can contribute significantly to the Poynting flux and helicity flux through the non-inductive (curl-free) electric field. The different, ad hoc treatments of the Doppler and transverse velocities in three methods are directly responsible for the discrepancies. We discuss the desired future observations that can better constrain these methods.

astro-ph.SR↗

arXiv2026

Quasi-Periodic Fast-Mode Wave Trains Associated with the 2015-Jun-22 M6.5 Flare in AR~12371: Observations and 3D MHD Modeling

Quasi-periodic fast-propagating (QFP) wave trains are a distinctive form of magnetohydrodynamic disturbance frequently observed in the solar corona. Yet their excitation mechanism and propagation characteristics are not well understood. In this study, we investigate a well-observed QFP wave event associated with an M6.5-class flare and coronal mass ejection that occurred in active region (AR) 12371 on 2015 June 22 by combining multi-wavelength observations from SDO/AIA and HMI with data-inspired 3D MHD simulations. The QFP wave trains propagating at high speeds of 1140$-$1760 km~s$^{-1}$ are detected in the AIA 171 \AA\ channel, following global EUV wave fronts visible at 171 and 193~\AA\ traveling at considerably lower speeds of 300$-$510 km~s$^{-1}$. Wavelet analysis reveals consistent 2--4 minutes periodicities in both the QFPs and flare quasi-periodic pulsations (QPPs) observed in UV/EUV and hard X-ray emissions, suggesting a common origin likely linked to intermittent magnetic reconnection. Guided by these observations, we construct realistic 3D MHD models incorporating dense fan-loop structures and periodic drivers applied at different locations. The simulations reproduce the key characteristics of the observed wave trains. Comparison between cases with and without a coronal background (non-loop plasma emission) indicates that coronal density structuring significantly modifies the detected wave amplitudes and propagation patterns. Our results highlight the importance of realistic coronal magnetic configurations in modeling QFP dynamics and suggest that their observed association with fan loops in AIA 171 \AA\ may represent a temperature-dependent visibility effect rather than a genuine confinement of the waves.

astro-ph.SR↗

arXiv2026

Refinements to the Solar Polar Magnetic Flux: Implications from Inversion Methodologies

The magnetic fields in the solar polar region are important to our understanding of the internal dynamo process, the global coronal structure, and the origin of the solar wind. The inference of polar fields based on spectropolarimetric observation is highly model-dependent and can suffer from various systematic effects. Here we analyze a raster map of the southern polar region taken by the Hinode Spectro-Polarimeter, utilizing the Stokes Inversion based on Response functions code. The inversions provide height-dependent vector magnetic field maps between optical depths $\log_{10}\tau = -2$ and $0$. We examine the impact on the total magnetic flux estimate from adopting (1) 1- vs 2-component atmospheric models via a "filling factor" parameter and (2) different analysis schemes. At $\log_{10}\tau = -1.5$, the polar magnetic flux is estimated to be $(1.84 \pm 0.03) \times 10^{21}$ Mx and $(1.38 \pm 0.02) \times 10^{21}$ Mx under the 1- and 2-component atmosphere assumption, respectively. The magnetic flux is approximately constant or increases slightly with height, respectively. We find that the 2-component (1-component) configuration is preferred for 58.3% (32.3%) of the pixels. Different initial guesses, including the input atmosphere model and the filling factor, as well as different inversion settings, can significantly affect the results, especially for locations with weaker polarization signals. Our work highlights the importance of including unresolved magnetic structures or stray light into consideration. Model degeneracy and the convergence to local minima limit the precision of the polar magnetic flux inference (no better than several tens of percent in this case). Higher-resolution observations and advanced inversion and disambiguation algorithms may alleviate these limitations.

astro-ph.SR↗

arXiv2026

SURE: Safe Uncertainty-Aware Robot-Environment Interaction using Trajectory Optimization

Robotic tasks involving contact interactions pose significant challenges for trajectory optimization due to discontinuous dynamics. Conventional formulations typically assume deterministic contact events, which limit robustness and adaptability in real-world settings. In this work, we propose SURE, a robust trajectory optimization framework that explicitly accounts for contact timing uncertainty. By allowing multiple trajectories to branch from possible pre-impact states and later rejoin a shared trajectory, SURE achieves both robustness and computational efficiency within a unified optimization framework. We evaluate SURE on two representative tasks with unknown impact times. In a cart-pole balancing task involving uncertain wall location, SURE achieves an average improvement of 21.6% in success rate when branch switching is enabled during control. In an egg-catching experiment using a robotic manipulator, SURE improves the success rate by 40%. These results demonstrate that SURE substantially enhances robustness compared to conventional nominal formulations.

cs.RO↗

arXiv2026

Consensus-based optimization (CBO): Towards Global Optimality in Robotics

Zero-order optimization has recently received significant attention for designing optimal trajectories and policies for robotic systems. However, most existing methods (e.g., MPPI, CEM, and CMA-ES) are local in nature, as they rely on gradient estimation. In this paper, we introduce consensus-based optimization (CBO) to robotics, which is guaranteed to converge to a global optimum under mild assumptions. We provide theoretical analysis and illustrative examples that give intuition into the fundamental differences between CBO and existing methods. To demonstrate the scalability of CBO for robotics problems, we consider three challenging trajectory optimization scenarios: (1) a long-horizon problem for a simple system, (2) a dynamic balance problem for a highly underactuated system, and (3) a high-dimensional problem with only a terminal cost. Our results show that CBO is able to achieve lower costs with respect to existing methods on all three challenging settings. This opens a new framework to study global trajectory optimization in robotics.

cs.RO↗

arXiv2025

Spectropolarimetric Inversion in Four Dimensions with Deep Learning (SPIn4D): II. A Physics-Informed Machine Learning Method for 3D Solar Photosphere Reconstruction

Inferring the three-dimensional (3D) solar atmospheric structures from observations is a critical task for advancing our understanding of the magnetic fields and electric currents that drive solar activity. In this work, we introduce a novel, Physics-Informed Machine Learning method to reconstruct the 3D structure of the lower solar atmosphere based on the output of optical depth sampled spectropolarimetric inversions, wherein both the fully disambiguated vector magnetic fields and the geometric height associated with each optical depth are returned simultaneously. Traditional techniques typically resolve the 180-degree azimuthal ambiguity assuming a single layer, either ignoring the intrinsic non-planar physical geometry of constant optical-depth surfaces (e.g., the Wilson depression in sunspots), or correcting the effect as a post-processing step. In contrast, our approach simultaneously maps the optical depths to physical heights, and enforces the divergence-free condition for magnetic fields fully in 3D. Tests on magnetohydrodynamic simulations of quiet Sun, plage, and a sunspot demonstrate that our method reliably recovers the horizontal magnetic field orientation in locations with appreciable magnetic field strength. By coupling the resolutions of the azimuthal ambiguity and the geometric heights problems, we achieve a self-consistent reconstruction of the 3D vector magnetic fields and, by extension, the electric current density and Lorentz force. This physics-constrained, label-free training paradigm is a generalizable, physics-anchored framework that extends across solar magnetic environments while improving the understanding of various solar puzzles.

astro-ph.SR↗

arXiv2025

Path representations in multiparameter persistent homology

Multiparameter persistence module can capture more topological differences across data instances compared to using a single parameter, where the well-studied matching distance investigates the distance along a straight line in the multiparameter space that gives the biggest difference. We propose to generalize the straight line to a monotone path filtration and offer software implementations.

math.AT↗

arXiv2025

Analyzing the Morphology of Late-phase Stellar Flares From G-, K-, and M-type Stars

Stellar flares occasionally present a $\textit{peak-bump}$ light curve morphology, consisting of an initial impulsive phase followed by a gradual late phase. Analyzing this specific morphology can uncover the underlying physics of stellar flare dynamics, particularly the plasma heating-evaporation-condensation process. While previous studies have mainly examined peak-bump occurrences on M-dwarfs, this report extends the investigation to G-, K-, and M-type stars. We utilize the flare catalog published by arXiv:2212.00993, encompassing 12,597 flares, detected by using $\textit{Transiting Exoplanet Survey Satellite}$ (TESS) observations. Our analysis identifies 10,142 flares with discernible classical and complex morphology, of which 197 ($\sim1.9\%$) exhibit the peak-bump feature. We delve into the statistical properties of these TESS late-phase flares, noting that both the amplitude and full-width-half-maximum (FWHM) duration of both the peaks and bumps show positive correlations across all source-star spectral types, following a power law with indices 0.69 $\pm$ 0.09 and 1.0 $\pm$ 0.15, respectively. Additionally, a negative correlation between flare amplitude and the effective temperature of their host stars is observed. Compared to the other flares in our sample, peak-bump flares tend to have larger and longer initial peak amplitudes and FWHM durations, and possess energies ranging from $10^{31}$ to $10^{36}$ erg.

astro-ph.SR↗

arXiv2025

Addressing pitfalls in implicit unobserved confounding synthesis using explicit block hierarchical ancestral sampling

Unbiased data synthesis is crucial for evaluating causal discovery algorithms in the presence of unobserved confounding, given the scarcity of real-world datasets. A common approach, implicit parameterization, encodes unobserved confounding by modifying the off-diagonal entries of the idiosyncratic covariance matrix while preserving positive definiteness. Within this approach, we identify that state-of-the-art protocols have two distinct issues that hinder unbiased sampling from the complete space of causal models: first, we give a detailed analysis of use of diagonally dominant constructions restricts the spectrum of partial correlation matrices; and second, the restriction of possible graphical structures when sampling bidirected edges, unnecessarily ruling out valid causal models. To address these limitations, we propose an improved explicit modeling approach for unobserved confounding, leveraging block-hierarchical ancestral generation of ground truth causal graphs. Algorithms for converting the ground truth DAG into ancestral graph is provided so that the output of causal discovery algorithms could be compared with. We draw connections between implicit and explicit parameterization, prove that our approach fully covers the space of causal models, including those generated by the implicit parameterization, thus enabling more robust evaluation of methods for causal discovery and inference.

stat.ML↗

arXiv2024

What Can DKIST/DL-NIRSP Tell Us About Quiet-Sun Magnetism?

Quiet-Sun regions cover most of the Sun's surface; its magnetic fields contribute significantly to the solar chromospheric and coronal heating. However, characterizing the magnetic fields of the quiet Sun is challenging due to their weak polarization signal. The 4-m \textit{Daniel K. Inouye Solar Telescope} (\textit{DKIST}) is expected to improve our understanding of the quiet-Sun magnetism. In this paper, we assess the diagnostic capability of the Diffraction-Limited Near Infrared Spectropolarimeter (DL-NIRSP) instrument on \textit{DKIST} on the energy transport processes in the quiet-Sun photosphere. To this end, we synthesize high-resolution, high-cadence Stokes profiles of the \ion{Fe}{1} 630~nm lines using a realistic magnetohydrodynamic simulation, degrade them to emulate the \textit{DKIST}/DL-NIRSP observations, and subsequently infer the vector magnetic and velocity fields. For the assessment, we first verify that a widely used flow-tracking algorithm, Differential Affine Velocity Estimator for Vector Magnetograms, works well for estimating the large-scale ($> 200$ km) photospheric velocity fields with these high-resolution data. We then examine how the accuracy of inferred velocity depends on the temporal resolution. Finally, we investigate the reliability of the Poynting flux estimate and its dependence on the model assumptions. The results suggest that the unsigned Poynting flux, estimated with existing schemes, can account for about $71.4\%$ and $52.6\%$ of the reference ground truth at $\log \tau =0.0$ and $\log \tau = -1$. However, the net Poynting flux tends to be significantly underestimated. The error mainly arises from the underestimated contribution of the horizontal motion. We discuss the implications on \textit{DKIST} observations.

astro-ph.SR↗

arXiv2024

HardRace: A Dynamic Data Race Monitor for Production Use

Data races are critical issues in multithreaded program, leading to unpredictable, catastrophic and difficult-to-diagnose problems. Despite the extensive in-house testing, data races often escape to deployed software and manifest in production runs. Existing approaches suffer from either prohibitively high runtime overhead or incomplete detection capability. In this paper, we introduce HardRace, a data race monitor to detect races on-the-fly while with sufficiently low runtime overhead and high detection capability. HardRace firstly employs sound static analysis to determine a minimal set of essential memory accesses relevant to data races. It then leverages hardware trace instruction, i.e., Intel PTWRITE, to selectively record only these memory accesses and thread synchronization events during execution with negligible runtime overhead. Given the tracing data, HardRace performs standard data race detection algorithms to timely report potential races occurred in production runs. The experimental evaluations show that HardRace outperforms state-of-the-art tools like ProRace and Kard in terms of both runtime overhead and detection capability -- HardRace can detect all kinds of data races in read-world applications while maintaining a negligible overhead, less than 2% on average.

cs.SE↗

arXiv2024

Ensemble Kalman Filter Data Assimilation Into Surface Flux Transport Model To Infer Surface Flows: An Observing System Simulation Experiment

Knowledge of the global magnetic field distribution and its evolution on the Sun's surface is crucial for modeling the coronal magnetic field, understanding solar wind dynamics, computing the heliospheric open flux distribution and predicting solar cycle strength. As the far side of the Sun cannot be observed directly and high-latitude observations always suffer from projection effects, we often rely on surface flux transport simulations (SFT) to model long-term global magnetic field distribution. Meridional circulation, the large-scale north-south component of the surface flow profile, is one of the key components of the SFT simulation that requires further constraints near high latitudes. Prediction of the photospheric magnetic field distribution requires knowledge of the flow profile in the future, which demands reconstruction of that same flow at the current time so that it can be estimated at a later time. By performing Observing System Simulation Experiments, we demonstrate how the Ensemble Kalman Filter technique, when used with a SFT model, can be utilized to make ``posterior'' estimates of flow profiles into the future that can be used to drive the model forward to forecast photospheric magnetic field distribution.

astro-ph.SR↗

arXiv2024

Multi-Grained Specifications for Distributed System Model Checking and Verification

This paper presents our experience specifying and verifying the correctness of ZooKeeper, a complex and evolving distributed coordination system. We use TLA+ to model fine-grained behaviors of ZooKeeper and use the TLC model checker to verify its correctness properties; we also check conformance between the model and code. The fundamental challenge is to balance the granularity of specifications and the scalability of model checking -- fine-grained specifications lead to state-space explosion, while coarse-grained specifications introduce model-code gaps. To address this challenge, we write specifications with different granularities for composable modules, and compose them into mixed-grained specifications based on specific scenarios. For example, to verify code changes, we compose fine-grained specifications of changed modules and coarse-grained specifications that abstract away details of unchanged code with preserved interactions. We show that writing multi-grained specifications is a viable practice and can cope with model-code gaps without untenable state space, especially for evolving software where changes are typically local and incremental. We detected six severe bugs that violate five types of invariants and verified their code fixes; the fixes have been merged to ZooKeeper. We also improve the protocol design to make it easy to implement correctly.

cs.DC↗