arXiv subjects

Jialin Li

Publications and source records attributed to Jialin Li.

At least 37 records · Page 2Linked to original sources

arXiv2026

Comparing realtime optical gain measurement and methods on MagAO-X

A lingering technical challenge for pyramid wavefront sensors (PyWFS) is their change in response between calibration and correction residuals, a quantity known as optical gain (OG). Given the prevalent use of PyWFSs in current and planned high contrast adaptive optics (AO), understanding and reliably measuring OG for realtime control unlocks advanced correction and post processing techniques. The OG quantity as an unknown inhibits a system's ability to stably correct non common path errors, reconstructing wavefronts, and PSF reconstruction. This work compares kinds of optical gain measurement techniques on MagAO-X, a visible light extreme AO instrument on the 6.5m Magellan Clay telescope. We present a set of on-sky measurements of OG across three techniques: 1) An on-sky calibration that acquires OG per spatial mode, 2) realtime measurements of the instantaneous Strehl Ratio (SR) on the pyramid tip, and 3) realtime measurement of known, high-frequency probe signal on the WFS itself. We compare these on-sky results with performance diagnostics to asses how faithfully OG is returned. We conclude with future steps for active control of OG on MagAO-X.

astro-ph.IM↗

arXiv2026

Sensor fusion on MagAO-X: real time vibration control using accelerometers

Mechanical vibrations are a significant source of residual wavefront error (WFE) in adaptive optics (AO) systems, limiting the performance of high-contrast imaging instruments. We present the design and on-sky deployment of a low-cost, modular accelerometer telemetry system for the MagAO-X extreme AO instrument on the 6.5 m Magellan Clay Telescope, consisting of piezoelectric accelerometers and a Raspberry Pi-based acquisition system that streams synchronized data to the real-time control computer with microsecond-level timing stability. The system is used to identify dominant telescope vibration sources and quantify their coupling to AO telemetry, revealing that several narrow-band modes originate from subsystems including the primary mirror glycol pump, secondary mirror actuation system, and telescope autofocus system. Coherence analysis between the synchronized accelerometer and wavefront sensor telemetry demonstrates that approximately one-third of the residual tip and tilt WFE is correlated with structural vibrations, indicating that accelerometer telemetry provides a promising foundation for future predictive control implementations. These results demonstrate that low-cost accelerometer telemetry provides a practical approach for vibration identification and a foundation for predictive control in current and future AO systems.

astro-ph.IM↗

arXiv2026

windsoCC: reconstructing the wind-driven halo in MagAO-X images using wavefront sensor telemetry

The wind-driven halo (WDH) is a persistent, low spatial frequency noise artifact that arises due to the servo-lag error inherent to all adaptive optics (AO) instruments. Spatial filtering may be employed to overcome this artifact, however, filtering out the WDH while simultaneously preserving signal from an extended astrophysical object of interest is exceptionally challenging. Additionally, since the WDH changes in intensity and position angle through an observation, data-driven algorithms (e.g., KLIP) that are commonly used to subtract the starlight need to be overly-aggressive to remove both the static and dynamic noise components. Since wavefront sensors (WFSs) continuously track the closed-loop residual wavefront error, WFS telemetry presents the ideal resource for combating this type of noise artifact through postprocessing. Using archival WFS telemetry from MagAO-X, which is the ``extreme" AO instrument for the 6.5-meter Magellan-Clay telescope, we demonstrate a novel workflow for WDH reconstruction and removal in individual coronagraphic science images. MagAO-X is equipped with a pyramid WFS capable of recording wavefront telemetry at a high-cadence which is saved during data acquisition. Given this, we detail how our WFS data processing pipeline, windsoCC, cross-correlates the recorded closed-loop wavefront to measure the wind vectors of several turbulent layers of the atmosphere above Las Campanas Observatory. We then make use of the wind parameters learned through windsoCC to reconstruct the WDH footprint by leveraging a parametric model. Notably, we demonstrate a dramatic improvement in object recovery using on-sky MagAO-X images of the disk around HR~4796A at visible wavelengths.

astro-ph.IM↗

arXiv2026

Closed-loop Atmospheric Dispersion Correction for High-Contrast Imaging with MagAO-X

Incoming starlight is refracted as it enters Earth's atmosphere from the vacuum of space. The wavelength-dependence of atmospheric refraction causes elongation of the broadband PSF of ground-based telescopes, especially in the visible spectrum. The result is degraded image quality alongside reduced coronagraph light-blocking efficiency, both of which limit high-contrast observations. An atmospheric dispersion corrector (ADC) is a dispersive optic used to compensate for this effect. Current methods for dispersion compensation use analytical models to anticipate dispersion strength based on parameters such as site altitude and telescope zenith angle; however, dispersion strength is also dictated by a number of factors that cannot be measured, including instantaneous humidity, temperature, and pressure along the line of sight to the star. This leads to constant over- or under-correction of the true atmospheric dispersion by the ADC. In this work, we use the Magellan extreme adaptive optics system MagAO-X at Las Campanas Observatory to measure and correct residual atmospheric dispersion in real-time. The amount of residual dispersion is encoded in the orientation of satellite spots generated by using MagAO-X's deformable mirror as a diffraction grating. We have used these real-time measurements as feedback for closed-loop control of the ADCs on-sky at visible and NIR wavelengths, reducing residual atmospheric dispersion down to sub-mas/$μ$m levels. Active atmospheric dispersion correction on MagAO-X is a precursor to high-contrast imaging with Extreme AO for the upcoming Extremely Large Telescopes, where high-precision dispersion compensation will be required to image exoplanets in reflected light.

astro-ph.IM↗

arXiv2026

Visible-Light High-Contrast Polarimetry with MagAO-X: Characterization and Initial Results

MagAO-X is a visible-light extreme adaptive optics instrument on the 6.5 meter Magellan Clay Telescope, recently upgraded to enable high-contrast polarimetric differential imaging (PDI) in r', i', and z' filters. Polarimetry is a powerful technique for suppressing unpolarized starlight and isolating the faint, polarized signal scattered by circumstellar dust, but it demands precise calibration of instrumental polarization effects introduced by the telescope and instrument optics. We present an overview of the MagAO-X polarimeter and characterize its polarimetric response using a purpose-built polarization generator that injects light of a known polarization state. From these measurements, we fit a Mueller-matrix model of the instrument and quantify its polarimetric efficiency and instrumental polarization as a function of the k-mirror image rotator angle and observing filter. The initial characterization revealed significant, dynamic inefficiencies driven by the image rotator, motivating the deployment of a dual rotating quarter-wave plate (DQWP) compensator that dynamically reorients the input polarization to the instrument's eigenpolarization. Following installation of the DQWP, we measured an average increase in polarimetric efficiency of +17.5% (to 87.4%) and a reduction in instrumental polarization of -5.4% (to 8.3%) across all filters. Finally, we demonstrate the on-sky performance of the polarimeter with i' imaging of the debris disk around HR 4796, producing one of the closest inner-working-angle views of the bright, forward-scattering side of the disk. These results help pave the way for polarimeters on future extremely large telescopes such as GMT and ELT.

astro-ph.IM↗

arXiv2026

AuroraRL: Fast, Fault-Tolerant, and Cost-Efficient Reinforcement Learning over Decentralized Network

LLM reinforcement learning (RL) requires frequent synchronization of large model parameters between the trainer and distributed rollout actors. High-throughput RL post-training therefore relies on dedicated RDMA HPC/cloud clusters, an infrastructure cost most organizations cannot absorb. A natural alternative is to aggregate loosely-coupled GPUs over standard Ethernet and WAN links, but this commodity connectivity cannot sustain full-weight broadcasts: synchronizing an 8B model can take over 100~seconds on bandwidth-limited links, while rollout generation typically takes tens of seconds. Toward making RL practical in this regime, we observe that RL fine-tuning yields highly sparse per-step updates, with only around 1\% of parameter elements changing. On top of this insight, we present AuroraRL, a novel high-performance RL training system that preserves bit-exact updates without dropping or quantizing information, designed for commodity-networked, loosely-coupled GPU resources. AuroraRL represents each step as a sparse delta checkpoint, pipelines delta extraction with multi-stream transmission, overlaps transfer with rollout generation, and coordinates heterogeneous workers with throughput- and bandwidth-aware scheduling plus lease-based fault tolerance. Across Qwen3 4B--14B models deployed in up to four geographic regions, AuroraRL shrinks per-step weight transfer by 79$\times$ on Qwen3-8B, delivers 1.3--9.5$\times$ higher throughput than dense-broadcast baselines (PrimeRL-Full, async-tolerant, multi-stream variants), and brings end-to-end training within 8.91\% of an ideal RDMA single-datacenter baseline, while transparently tolerating common failures and preserving training accuracy. By leveraging on-demand, cross-cloud GPUs over commodity links, AuroraRL delivers 1.21--1.59$\times$ higher tokens per dollar than reserved RDMA clusters at comparable throughput.

cs.DC↗

arXiv2026

Computational Study of Water Adsorption and Dissociative Mechanisms Impacting g-C3N4's Optical and Electronic Properties

In the quest for sustainable energy solutions, water splitting emerges as a crucial process for generating clean hydrogen a versatile and renewable fuel essential for energy storage, emissions reduction, and achieving sustainability goals. This study employs a comprehensive computational approach, utilizing atomistic simulations to systematically investigate the effects of water absorption on the electronic and optical properties of g-C3N4 nanosheets. Our methodology integrates ab initio computations grounded in density functional theory (DFT), which allows for a detailed characterization of the nanosheet and serves as a benchmark for self-consistent charge density functional tight binding (SCC-DFTB) simulations. This approach provides valuable insights into the behavior of the nanosheet under the influence of absorbed OH and H2O molecules by considering calculated parameters for photocatalytic efficiency. Additionally, we extend our investigation to classical molecular dynamics simulations within the ReaxFF framework, modeling the emission of multiple H2O molecules and assessing the subsequent rate of H2 evolution. A key finding of our study reveals that the dissociation of H2O into HO and O molecules significantly enhances both the optical absorbance and conductance of the nanosheet compared to its pristine state. These results underscore the potential of g-C3N4 nanosheets as effective materials for water splitting applications.

cond-mat.mtrl-sci↗

arXiv2026

Dual-Mode SERS and Colorimetric Sensor for Lung Cancer VOC-Biomarker Detection Using Hydrogel Patches

Hexanal, a volatile organic compound (VOC), is a potential biomarker for the early detection of lung cancer. In this study, we developed a dual-mode flexible biosensor that integrates surface-enhanced Raman scattering (SERS) and colorimetric detection for the quantitative analysis of hexanal in human exhaled breath. The biosensor employs Ag nanocubes wrapped with Co-Ni layered double hydroxide (AgNCs@Co-Ni LDH) as a functional matrix, offering both superior SERS enhancement and efficient VOC adsorption properties. To achieve selective detection, AgNCs@Co-Ni LDH were incorporated into agarose hydrogels along with 3-methyl-2-benzothiazolinone hydrazone (MBTH). The resulting MBTH-AgNCs@Co-Ni LDH/hydrogel-patch facilitate the oxidation of hexanal, enabling simultaneous colorimetric and SERS signal generation, while producing acrizine, a blue-colored reaction product. This hydrogel-based dual-mode sensing platform exhibits high selectivity, excellent stability, and precision in SERS-based hexanal detection. The detection limit for the SERS method was determined to be as low as 3.34x10-13 M. Furthermore, developed and optimized compact CNN-based multi-terminal intelligent recognition system for enhanced hydrogel-patch detection through AI-driven, portable, and real-time colorimetric analysis. Therefore, this work not only enables the effective detection of hexanal in the exhaled breath of suspected lung cancer patients, underscoring its potential for early lung cancer screening, but also establishes a foundation for the development of multimodal hydrogel biosensors for broader applications in disease diagnosis.

physics.med-ph↗

arXiv2026

Concurrency Testing in the Linux Kernel via eBPF

Concurrency is indispensable for modern software systems to meet performance and scalability demands, yet concurrency bugs remain notoriously difficult to detect and reproduce. Controlled Concurrency Testing (CCT) mitigates this challenge by systematically exploring thread interleavings through scheduling control. However, existing CCT approaches for OS kernels largely rely on external enforcement mechanisms, such as custom hypervisors or invasive kernel patches, resulting in substantial overhead and limited maintainability and extensibility. In this work, we present SECT, the first kernel-native concurrency fuzzing framework that rethinks scheduling as a first-class exploration mechanism. SECT introduces a novel CCT scheduler with temporal isolation scheduling and embeds programmable scheduling policies directly into the kernel dispatch path via eBPF, enabling fine-grained control over thread interleavings without customized hypervisors or extensive kernel core modification. In addition, SECT provides a preemption-safe instrumentation mechanism for injecting scheduling points at critical kernel events and incorporates a two-phase fuzzing workflow to jointly explore both sequential and concurrent behaviors. Our evaluation demonstrates that SECT achieves 38% more branches, 57% overhead reduction and 11.4$\times$ speed-up in bug exposure compared to a leading state-of-the-art kernel concurrency fuzzer. Moreover, SECT discovers eight previously unknown concurrency-related bugs in the Linux kernel, six of which have already been confirmed and fixed by developers.

cs.OS↗

arXiv2026

CausalGame: Benchmarking Causal Thinking of LLM Agents in Games

Building AI Scientist agents with Large Language Models (LLMs) has recently attracted growing attention. Since scientific discovery fundamentally relies on uncovering causal relationships from observations, the capability of causal thinking, i.e., distinguishing causation from correlation and recognizing hidden biases, is essential to LLM agents. Although a number of benchmarks exist for AI Scientists, none explicitly incorporate challenges from selection bias, measurement error, and hidden confounders that widely exist in real-world scientific discovery. To this end, we present CausalGame, a benchmark that evaluates the causal thinking capabilities of LLM agents through interactive games. CausalGame asks LLM agents to actively design experimental protocols, collect observation data, and derive a final solution with an explanation report. To emulate realistic scientific discovery challenges, we design 14 scenarios that incorporate selection bias, measurement error, and hidden confounders. Across 30 LLM agents, none demonstrates reliable causal thinking: the best model reaches only 68.0% survival against analytical optima of 78-85%, and merely 5-7% of sessions receive credits on the causal-reasoning rubrics. CausalGame provides a scalable and controlled testbed for evaluating the causal thinking of AI Scientist agents.

cs.CL↗

arXiv2026

ffortissimo: A Freeform Forward-Modeling Pipeline for High-Contrast Images of Circumstellar Disks Based on Automatic Differentiation

Modeling circumstellar disks in the traditional sense carries the assumption that the dust density distribution can be accurately described with a fixed parametric form. Furthermore, commonly-used algorithms for subtracting the stellar point-spread function (PSF) distort the true morphology of the faint underlying disk structure, especially dusty features that are located at small angular separations. These phenomena often lead to significant residuals with parametric disk models and make it difficult to measure the full realizable range of the scattering function of the dust. We address these challenges with ffortissimo, a novel, pixel-based freeform forward modeling pipeline designed to characterize extended objects in KLIP-reduced images. We built this pipeline within the framework of JAX, which is a machine learning library in Python that enables efficient optimization through automatic differentiation ("autodiff") and GPU-accelerated array computations. Using visible light images of the disk around HR 4796A taken by the "extreme" Magellan Adaptive Optics instrument (MagAO-X), we show that our data-driven freeform models excel at fitting a complex dust distribution and can infer the dust scattering properties even through PSF subtraction artifacts. Additionally, we demonstrate the potential for retrieving spatial dust features beyond the diffraction limit of the telescope. We note that there are remaining challenges to address before precision photometry using these freeform models is advised. These include better background, wind-driven halo, and speckle characterization as preventing the freeform models from learning these noise artifacts is currently difficult.

astro-ph.IM↗

arXiv2026

A Multiband Study of the HR 4796A Disk in the Optical Using MagAO-X

We present total intensity images of the debris disk around HR 4796A from observations spanning 2023 to 2025 with the Magellan extreme adaptive optics instrument (MagAO-X). We detected the disk at high signal-to-noise ratios at $g' (527$ nm), $r' (615$ nm), $i' (762$ nm), and $z' (909 $ nm). Additionally, we present images collected using the "star-hopping" technique that show the entirety of the disk, including the dramatic forward-scattering at the minor axis. We subjected our images to a battery of modeling techniques to constrain the geometry and photometry of the disk. Leveraging our clear detections of the disk's minor axis, we modeled the scattering phase function (SPF) using a basis of the Legendre polynomials. To mitigate self-subtraction artifacts in our angular differential imaging, we implemented a forward-modeling pipeline that generates a pixel-based freeform disk forward model leading to a deconvolved image of the disk. Our best-fit disk models reveal: (1) highly forward-scattering SPFs with a minimum at the $\sim65^{\circ}$ scattering angle, (2) a faint halo of dust just exterior to the spine of the disk that is not well-described by a broken power law density profile, (3) a red spectral slope for the dust, and finally (4) a compact, clump-like feature in the freeform disk models. Our empirically-measured SPFs suggest that the scattering is dominated by large, highly-absorptive grains. However, we emphasize the need for testing advanced irregular grain models using our SPFs to learn more about the physical and chemical properties of this complex system.

astro-ph.EP↗

arXiv2026

Cordon: Semantic Transactions for Tool-Using LLM Agents

Tool-using LLM agents are shifting the unit of computation from explicit human-issued commands to model-driven tasks with stateful consequences. Yet today's agent runtimes still expose tools as isolated RPCs. This interface gives runtimes a convenient integration point, but it lacks a task-scoped execution boundary for commit, rollback, recovery, and audit across multi-step agent workflows. We argue that this mismatch calls for a runtime containment boundary rather than another per-call guardrail. This paper introduces Cordon, a transactional runtime system for staging and validating irreversible agent effects before commit. A semantic transaction is a task-level execution boundary that binds tool intents and runtime-tracked result lineage to reversible local state, staged external effects, delegated authority, and audit metadata. Cordon implements this abstraction with a transaction manager that tracks derived result objects, executes reversible mutations in shadow state, stages outward-facing actions in an effect outbox, and records recovery metadata. The runtime then validates the composed execution flow before it commits state or releases external effects. Our evaluation across adversarial and benign workflows shows that Cordon exposes cross-step violations missed by existing defenses. It also reduces irreversible-effect failures while preserving benign task completion with modest approval and latency overhead.

cs.OS↗

arXiv2026

Amortized Asynchronous Byzantine Reliable Broadcast with Optimal Resilience

Byzantine Reliable Broadcast (BRB) is a fundamental primitive in distributed computing and cryptographic systems; reducing the communication cost of BRB thus remains an important research direction. However, most existing works either focus strictly on the synchronous network model or utilize computationally impractical erasure codes. Therefore, to achieve a practical yet network-robust algorithm, one must turn toward committee sampling techniques. However, Committee sampling techniques often forgo optimal resilience ($f < \lfloor\frac{n}{3} \rfloor$) in the face of asynchrony. This work produces two interesting results: Firstly, we propose a \textit{randomly asynchronous} BRB protocol that can achieve both optimal resilience and asymptotically optimal communication complexity ($O(n|m|)$) through an underutilized technique: \textit{amortization}; and does not utilize computationally expensive \textit{erasure codes}. Next, we show that an optimally resilient BRB protocol utilizing sampled committees cannot exist in a \textit{fully asynchronous} network.

cs.DC↗

arXiv2026

VDSB-GWSyn: Diffusion Schrödinger Bridge for Controllable and Anatomically Feasible Guidewire Synthesis in Coronary Angiography

Coronary guidewire endpoint localization is a fundamental capability for computer-assisted PCI, and its importance increases as robot-assisted PCI is progressively adopted to reduce operator radiation exposure. However, the scarcity of annotated CAG images with guidewires and the limited adaptability of existing guidewire synthesis models remain key bottlenecks for guidewire endpoint localization. To address this issue, we propose VDSB-GWSyn, a Diffusion Schrödinger Bridge (DSB) model-based framework, enabling synthesis of controllable, high-fidelity guidewire samples under complex anatomical backgrounds. VDSB-GWSyn first uses our shape prior algorithm to learn the basic guidewire geometry. It then generates guidewire masks under constraints imposed by the vessel segmentation masks and outputs the corresponding endpoint coordinates. Finally, it synthesizes realistic guidewire samples on real CAG images using DSB conditioned with SPADE. Experimental results show that the guidewire samples synthesized by VDSB-GWSyn achieve favorable ROI-FID and ROI-KID, as well as high IPR scores. In addition, incorporating our synthesized data for synthetic pre-training followed by real fine-tuning substantially improves downstream guidewire endpoint localization, reducing MPE from 16.01~px to 7.71~px and increasing PCK at 3~px from 52.63\% to 86.27\%, leading to more clinically reliable deployment of robot-assisted guidewire delivery systems. Moreover, the core design philosophy of controllable device synthesis with strict background preservation and anatomical feasibility constraints has the potential to transfer to other interventional device perception tasks where annotated data are scarce.

cs.CV↗

arXiv2026

The spectral energy distribution of YSES 1 b and its circumplanetary disc

Context. Direct imaging enables the characterisation of substellar companions on wide orbits. These objects provide a testbed for our formation theories; therefore, it is important to obtain accurate physical parameters for them. One of these objects is YSES 1 b. Aims. Our objective is to improve the spectral energy distribution (SED) modelling of YSES 1 b and determine the bulk and atmospheric parameters. Methods. We obtained observations in the r', i', and z' bands using MagAO-X on the 6.5 metre Magellan Clay telescope at Las Campanas Observatory. We combined this data with archival VLT/SPHERE and VLT/NACO data and used a forward modelling approach to estimate the physical parameters. We tested models both without and with a circumplanetary disc (CPD) model. We represented the CPD by including a dust extinction model and a blackbody radiation component. Using the derived bolometric luminosity, we estimated the mass of YSES 1 b by fitting evolutionary models. Results. Including the CPD model provides a significantly better fit to the photometric data, yielding an object that is considerably warmer (2854+110-94 K vs 1727+172-127 K) and smaller (1.58+0.06-0.07 RJ vs 3.0+0.2-0.7 RJ) than previous estimates. The newly determined radius suggests that the addition of dust extinction could resolve the large radius anomaly identified previously. Depending on the age of the system, the estimated mass increases from 14+-3 MJ (17 Myr) to either 25.7+4.1-3.6 (17 Myr) or 41.6+3.6-3.4 MJ (27 Myr). Conclusions. Dust extinction and blackbody radiation from a CPD can substantially change the estimated physical parameters of an object. For YSES 1 b, this moves it into the brown dwarf regime.

astro-ph.EP↗

arXiv2026

AtelierEval: Agentic Evaluation of Humans & LLMs as Text-to-Image Prompters

Text-to-image (T2I) systems increasingly rely on upstream prompters, either humans or multimodal large language models (MLLMs), to translate user intent into detailed prompts. Yet current benchmarks fix the prompt and only evaluate T2I models, leaving the prompting proficiency of this upstream component entirely unmeasured. We introduce AtelierEval, the first unified benchmark that quantifies prompting proficiency across 360 expert-crafted tasks. Grounded in a cognitive view, it spans three task categories and instantiates tasks using a taxonomy of real-world challenges, with a dual interface for both humans and MLLMs. To enable scalable and reliable evaluation, we propose AtelierJudge, a skill-based, memory-augmented agentic evaluator. It produces subjective and objective scores for prompt-image pairs, achieving a Spearman correlation of 0.79 with human experts, approaching human performance. Extensive experiments benchmark 8 MLLMs against 48 human users across 4 T2I backends, validate AtelierEval as a robust diagnostic tool, and reveal the superiority of mimicry over planning, advocating for an image-augmented direction for future prompters. Our work is released to support future research.

cs.AI↗

arXiv2026

Geometrically Constrained Stenosis Editing in Coronary Angiography via Entropic Optimal Transport

The scarcity of high-quality imaging data for coronary angiography (CAG) stenosis limits the clinical translation of automated stenosis detection. Synthetic stenosis data provides a practical avenue to augment training sets, improving data quality, diversity, and distributional coverage, and enhancing detection precision and generalization. However, diffusion-based editing commonly relies on soft guidance in a noise-initialized reverse process, offering limited pixel-level precision and structure preservation. We propose the OT-Bridge Editor, which reframes localized editing as a constrained entropic optimal transport (OT) problem and leverages geometric information to steer the generation path, enabling stronger geometric control. Extensive experiments show that our synthesized angiograms consistently improve downstream stenosis detection, yielding substantial relative gains of 27.8% on the public ARCADE benchmark and 23.0% on our multi-center dataset, supported by consistent qualitative results.

cs.CV↗