arXiv subjects

Xuan Yang

Publications and source records attributed to Xuan Yang.

At least 19 recordsLinked to original sources

arXiv2026

NodeSyn: Socially Aligned Synthetic Data for AI Evaluation

Recent advancements in generative AI facilitate large-scale synthetic data generation for model evaluation. However, without targeted approaches, these datasets often lack the sociotechnical nuance required for sensitive domains. We introduce NodeSyn, a methodology that pairs a fine-tuned taxonomy generator (TaG) with credible published sources and extracted sensitive attributes to generate queries reflecting authentic societal contexts and nuanced harms. Evaluated across ten open-weight and proprietary LLMs in Medical Advice and Self-Harm using human-validated autoraters, NodeSyn ranked first on 8 of 10 models in each domain. NodeSyn datasets induced average failure rates 3.1 times higher than generic synthetic baselines, 2.4 times higher than structured synthetic benchmarks, and 1.6 times higher than authentic human-authored queries. Ablation studies confirm that, controlling for structural confounds, our granular Level 3 taxonomic expansion significantly drives these failure rates over Level 2 across 9 of 10 models (+29.2% relative lift in Medical Advice and +24.0% in Self-Harm). Variance attribution analysis reveals that failures are driven not only by what topics are queried (taxonomic nodes and lexical specificity), but also by their sociotechnical grounding (affected demographics, occupations, geographies, and querying personas). We embed this method into a research prototype, and open-source (https://github.com/google-research/nodesynth) the code and datasets to enable scalable, high-stakes model evaluation.

cs.LG↗

arXiv2026

DynaWeave: Learning to Adapt through Dynamic Scenario Synthesis

Large language model (LLM) agents increasingly use tools to complete tasks in stateful environments. However, many training environments remain largely static apart from changes caused by the agent's own actions, whereas external events introduced by other users, services, or sudden contingencies in real environments can invalidate ongoing plans. To address this gap, we introduce DynaWeave, an automated synthesis framework that transforms existing static tasks into verifiable dynamic scenarios. The framework generates exogenous events relevant to each task, covering changes in task requirements, environment state, and external information. It uses directed acyclic graphs (DAGs) to represent dependencies between events and an execution engine to activate them during agent interaction. A dynamic verifier revises only the evaluation checks affected by events that actually occur and retains all other checks. Using DynaWeave, we collect 2,095 verified interaction trajectories across 36 environments and use them for supervised fine-tuning of Qwen3.5 models. Experiments on Gaia2, CostBench, and ToolMaze show that DynaWeave achieves consistent improvements in overall performance across all three evaluated model sizes, increasing the task completion rate on CostBench by 67.34 percentage points and even exceeding the performance of Gemini-3.1-Pro-Preview on ToolMaze. These results support dynamic scenario synthesis as a practical approach to training agents to adapt to changing environments. Our code is available at https://github.com/Miaow-Lab/DynaWeave.

cs.CL↗

arXiv2026

PHRBench: A Behavioral Evaluation of Post-Hallucination Reasoning in LLMs

Hallucinated information can propagate through multi-stage LLM systems and become part of the context for subsequent reasoning. Existing studies of post-hallucination reasoning (PHR) mainly characterize changes in final outcomes and aggregate reasoning dynamics, leaving how models resolve hallucinated premises at the response level insufficiently understood. In this work, we introduce PHRBench, a controlled benchmark for behaviorally structured PHR across four domains and 18 large language models. PHRBench characterizes each reasoning trajectory independently of final-answer correctness through Hallucination Compliance, Hallucination Avoidance, and Heuristic Correction, and defines an insightful trajectory as successful correction that ultimately reaches the correct answer. Across 4820 controlled instances, we find that successful recovery remains relatively rare and is associated with more frequent belief updates along the reasoning trajectory. We further find that properties of the hallucinated prompt contain substantial predictive signal for successful recovery, with a lightweight predictor achieving an AUROC of 0.847. These findings provide a behavioral view of post-hallucination reasoning, characterizing how LLMs resolve erroneous context and when successful recovery is likely to occur.

cs.CL↗

arXiv2026

GraftOOD: Grafting Out-of-Distribution Exemplars into Advantage-Weighted Rectification for Video Reward Optimization

Video reward optimization can increase proxy rewards while reducing video dynamics, a mismatch we identify as \emph{motion collapse}, including under a video-level motion-quality reward. We propose \textbf{GraftOOD}, which replaces part of the online rollouts in selected prompt groups with out-of-distribution video exemplars. A shared velocity-regression objective combines group-relative reward rectification with sequence-level exemplar supervision, requiring neither exemplar generation trajectories nor an additional supervised-loss coefficient. We characterize its pointwise optimal field as a source-posterior-weighted combination of the two source-specific target fields. We construct \textbf{GraftSet-Video}, containing 9,588 format-matched exemplars for 2,397 prompts, to supply external supervision. On Wan2.1-1.3B, GraftOOD raises Dynamic Degree from 4.86 to 44.41 relative to reward-only FlowAWR, while HPSv3 decreases from 8.10 to 7.32 and VisionReward remains close. At the main setting, GraftOOD leads nine of ten metrics among three exemplar-augmented methods. Ablations of exemplar frequency and within-group count reveal dynamics--reward trade-offs rather than uniform gains from increasing exemplar usage. These results support external exemplar supervision as a complement to scalar reward feedback, without equating higher measured dynamics with coherent, visually plausible motion in the generated videos. Code is available at \href{https://github.com/Fuzm7/FlowAWR}{GraftOOD}.

cs.CV↗

arXiv2026

Anatomy-Aligned Surface Field Learning for Myocardial Reconstruction from Sparse Short-Axis Cine MRI

Patient-specific 4D myocardial reconstruction from cine MRI supports quantitative functional assessment, regional motion analysis, and simulation-based modeling. However, routinely acquired short-axis (SAX) cine MRI is sparsely sampled along the through-plane direction, making dense and anatomically consistent surface reconstruction challenging. In this study, we propose an anatomy-aligned surface learning framework that parameterizes the epicardial and endocardial surfaces on a shared circumferential-longitudinal UV domain. This formulation converts irregular 3D reconstruction into structured coordinate-field completion with explicit correspondence across subjects and cardiac phases. Sparse SAX contours are encoded as UV observation fields, coverage-aware sampling improves robustness to incomplete slice coverage, and topology- and distortion-aware learning preserves circumferential continuity and local surface quality. Experiments on three public cine MRI datasets showed that the proposed method consistently outperformed representative mesh-based and implicit reconstruction approaches, achieving overall Chamfer distances of $2.887$~mm on ACDC, $2.641$~mm on M\&Ms, and $2.810$~mm on M\&Ms-2. The reconstructed sequences also preserved ventricular function, with end-diastolic volume and ejection fraction errors of $3.3$~mL and $1.1 \%$, respectively. These results demonstrate that anatomy-aligned UV learning provides an accurate, efficient, and correspondence-aware representation for sparse cine MRI reconstruction and myocardial modeling. The source code will be available at https://github.com/yuan-xiaohan/SAX2MyoSurf.

cs.CV↗

arXiv2026

Detecting Extragalactic Axion-like Dark Matter with Polarization Measurements of Fast Radio Bursts

Axions or axion-like particles (ALPs) are one of the promising dark matter (DM) candidates. A prevalent method to detect axion-like DM is to seek periodic oscillation in the polarization angles (PAs) of linearly polarized light emitted from astrophysical sources. In this work, we use the time-resolved polarization measurements of the hyperactive repeating fast radio burst, FRB 20220912A, detected by the Five-hundred-meter Aperture Spherical radio Telescope (FAST) to search for extragalactic axion-like DM. Given a DM density profile of FRB 20220912A's host, we obtain upper limits on the ALP-photon coupling constant of $g_{a γ}<(3.4 \times 10^{-11}-1.9\times 10^{-9})\,\mathrm{GeV}^{-1}$ for the ALP masses $m_a \sim (1.4\times10^{-21}-5.2\times10^{-20})\,\mathrm{eV}$. Persistent polarimetric observations with FAST would extend the constraints to lower masses. Although the $g_{a γ}$ constraints derived from FRBs are less competitive than those from other methods, FRBs offer an alternative way to detect axion-like DM on extragalactic distance scales, complementary to galactic DM probes.

astro-ph.HE↗

arXiv2026

Hardware-aware quantum attention for fast radio burst identification

Fast radio bursts (FRBs) are millisecond-duration extragalactic pulses whose discovery requires searches over large signal-parameter spaces and the rejection of candidate sets dominated by radio-frequency interference and noise. Integrating quantum processors into FRB searches requires a workflow connecting telescop data to hardware-compatible models. Here we develop a pipeline that converts raw search-mode PSRFITS data into dynamic-spectrum segments and identifies FRB-like signals using a hybrid Quantum Vision Transformer (QViT). On labelled FAST data, the selected QViT achieves a mean accuracy of 94.00% and recall of 98.30%, with similar performance to a compact classical Vision Transformer. Recall reaches 97.11% on an FRB source excluded from training and model selection. Real-device execution on 82 segments yields a mean simulator-hardware Hellinger fidelity of 0.9161 +/- 0.0074. The workflow connects raw telescope data to quantum-assisted identification and provides an experimental basis for assessing the measurement requirements that must be addressed before survey-scale deployment.

astro-ph.HE↗

arXiv2026

A fully integrated dispersion-managed femtosecond mode-locked laser

Femtosecond lasers underpin applications ranging from material processing to corneal surgery, while their regular pulse trains form optical frequency combs that have revolutionized timekeeping, spectroscopy, and metrology. On-chip optical frequency combs, such as Kerr microcombs, have enabled high-repetition-rate applications in optical communications and microwave photonics. However, integrated chip-scale sources operating at low repetition rates (100 MHz to 1 GHz), crucial for high peak intensities, remain elusive, as existing devices typically operate well beyond 10 GHz. Here, we demonstrate a self-starting, photonic integrated mode-locked laser based on a dispersion-managed architecture that accesses this regime. The laser combines erbium-implanted silicon nitride gain waveguides, integrated chirped Bragg gratings, and a semiconductor saturable absorber mirror to generate optical pulses with repetition rates from 0.5 to 1.2 GHz, pulse durations as short as 300 fs, and mode-locking thresholds down to 27.3 mW. The output forms a passively stable optical frequency comb with a comb-line drift below 1% of the repetition rate, surpassing the stability of commercial fiber lasers by two orders of magnitude. Leveraging this ultra-low threshold, we achieve complete hybrid integration by co-packaging the laser with a telecom-grade 980-nm III-V pump diode chip inside a compact photonic module. The resulting electrical-in/optical-out module delivers turnkey, stable mode-locked pulses, providing a compact, low-power, and vibration-insensitive foundry-compatible platform for field-deployable optical metrology and precision sensing.

physics.optics↗

arXiv2026

Learning to Beat: Phenotype-Guided Latent Flow with Regional Motion Priors for Biventricular Motion Synthesis

Full-cycle biventricular geometry is essential for characterizing cardiac function. However, dense and temporally consistent 3D+t biventricular meshes are not routinely available, whereas end-diastolic (ED) anatomy can often be obtained reliably. We therefore investigate full-cycle biventricular motion synthesis from a single ED mesh. This task is challenging because cardiac deformation is spatially heterogeneous and phenotype dependent, while conventional global generative models often obscure localized motion patterns. In this study, we propose a region-specific and phenotype-adaptive framework that integrates motion-informed functional parcellation with conditional latent flow. A functional partition learned from reconstructed motion organizes the ventricular surface into regions with coherent dynamics and enables topology-aware regional feature exchange. A phenotype-conditioned rectified-flow model subsequently maps the ED anatomy to full-cycle motion latents through fine-grained conditioning and prototype-routed motion adapters. An optional control branch further incorporates available motion descriptors for controllable synthesis. Experiments on ACDC, M\&Ms, and M\&Ms-2 demonstrate consistent improvements in geometric accuracy and functional fidelity. Under ED-only synthesis, our method achieves biventricular ASSD, HD95, and vRMSE of $1.49\pm0.34$~mm, $3.77\pm1.06$~mm, and $3.31\pm1.03$~mm, respectively, outperforming all competing methods. Complementary functional and robustness evaluations further demonstrate that the synthesized sequences preserve physiologically plausible ventricular dynamics and generalize across cohorts and disease phenotypes. The code will be released publicly upon acceptance of the manuscript for publication.

cs.CV↗

arXiv2026

Local Shapley: Model-Induced Locality and Optimal Reuse in Data Valuation

The Shapley value provides a principled foundation for data valuation, but exact computation is #P-hard due to the exponential coalition space. Existing accelerations remain global and ignore a structural property of modern predictors: for a given test instance, only a small subset of training points influences the prediction. We formalize this model-induced locality through support sets defined by the model's computational pathway (e.g., neighbors in KNN, leaves in trees, receptive fields in GNNs), showing that Shapley computation can be projected onto these supports without loss when locality is exact. This reframes Shapley evaluation as a structured data processing problem over overlapping support-induced subset families rather than exhaustive coalition enumeration. We prove that the intrinsic complexity of Local Shapley is governed by the number of distinct influential subsets, establishing an information-theoretic lower bound on retraining operations. Guided by this result, we propose LSMR (Local Shapley via Model Reuse), an optimal subset-centric algorithm that trains each influential subset exactly once via support mapping and pivot scheduling. For larger supports, we develop LSMR-A, a reuse-aware Monte Carlo estimator that remains unbiased with exponential concentration, with runtime determined by the number of distinct sampled subsets rather than total draws. Experiments across multiple model families demonstrate substantial retraining reductions and speedups while preserving high valuation fidelity.

cs.LG↗

arXiv2026

Searching for Gamma Ray Bursts associated with CHIME Fast Radio bursts

Fast radio bursts (FRBs) and gamma-ray bursts (GRBs) are both linked to compact-object activity, yet their possible connection remains unclear. Here we perform a systematic search for spatial and temporal associations between FRBs in the second CHIME/FRB catalog and Swift GRBs. Instead of using the positional ellipses reported in the catalog, the full CHIME localization probability maps are adopted for spatial cross-matching. This yields 130 candidate pairs and increases the number of spatially consistent matches by a factor of several. Applying a distance-consistency criterion based on DM-inferred FRB redshifts and GRB distances inferred via the Amati relation reduces the sample to 37 pairs, including 26 GRB-preceding-FRB candidates (24 LGRB--FRB and 2 SGRB--FRB). Monte Carlo simulations show that the overall excess of associations is not statistically significant, and the distribution of matches across localization confidence levels is consistent with random expectations. These pairs are therefore not claimed as secure associations, but are used to constrain a possible subdominant FRB--GRB connection. These results place constraints on any FRB--GRB connection and highlight the need for improved localization and larger samples.

astro-ph.HE↗

arXiv2026

Astra: a generalizable report generation foundation model for 3D computed tomography

Interpreting computed tomography (CT) requires review of hundreds of volumetric slices and remains time-intensive and expertise-dependent. Automated CT report generation offers a promising route to improving clinical efficiency, yet the field still lacks a generalizable CT report generation foundation model that supports multi-region reporting and remains robust across external real-world cohorts. Intrinsic inconsistencies in reporting style and diagnostic terminology across cohorts make naive joint training difficult. Here we present Astra, a generalizable CT report generation foundation model developed on 90,678 thoracoabdominal CT-report pairs collected from five sites worldwide (CTRgDB), comprising 353,671 abnormalities spanning eight organ systems. By harmonizing report style and further refining diagnostic consistency via reinforcement learning, Astra achieves style-consistent and diagnostically accurate report generation across diverse anatomical regions and institutions. Evaluated on CTRgDB and six external cohorts, Astra achieves state-of-the-art performance with a 38.4% average improvement in fine-grained diagnostic metrics (P<0.001). Deployed at external clinical sites without any site-specific fine-tuning, Astra accelerated chest report drafting by 29.6% and improved abdominal report completeness by 11.3% among junior and mid-level radiologists (P<0.001). Furthermore, Astra demonstrates broad utility as a foundation for CT AI development, improving downstream diagnostic performance and scaling vision-language pretrain through high-quality report synthesis. Overall, Astra serves as a broadly accessible clinical assistant and a pivotal infrastructure for the next generation of AI-powered healthcare. The code for Astra is publicly available at https://github.com/zh-Wang-Med/Astra.

cs.CV↗

arXiv2026

Cross-validation of six dispersion measure estimation methods for FRB 20240114A

Fast Radio Bursts (FRBs) are important cosmological probes, but their applications depend critically on accurate dispersion measure (DM) determinations. We present a systematic comparison of six DM estimation methods using 2,874 bursts from FRB20240114A, the most active repeating FRB currently known, observed by FAST during a single 4.4-hr session on 2024 March 12. This large, homogeneous sample over a short timescale, during which the propagation environment is expected to be nearly static, provides an ideal benchmark for isolating algorithmic effects on DM determination. We investigate the dependence of inter-method consistency on signal-to-noise ratio (S/N), burst morphology, and radio frequency interference (RFI). Low-S/N bursts exhibit significantly larger inter-method deviations, while single-component bursts produce highly consistent DM values across methods. In contrast, complex double- and multiple-component bursts with drifting substructures lead to substantial inter-method scattering, indicating that DM discrepancies are primarily driven by algorithmic responses to burst morphology. RFI does not significantly alter the global statistical behavior of DM deviations, but it affects density-filtering methods through morphology distortion caused by frequency-channel masking. Even after imposing strict inter-method consistency constraints, FRB20240114A still exhibits notable apparent DM fluctuations spanning $\sim$528-534~pc~cm$^{-3}$ over 15,780s. For morphologically simple bursts these variations far exceed the measurement uncertainty and, on second-to-minute timescales, cannot arise from any plausible change in the line-of-sight electron column, pointing instead to a frequency-dependent emission-time structure intrinsic to the bursts that mimics dispersion.

astro-ph.HE↗

arXiv2026

Step-Level Sparse Autoencoder for Reasoning Process Interpretation

Large Language Models (LLMs) have achieved strong complex reasoning capabilities through Chain-of-Thought (CoT) reasoning. However, their reasoning patterns remain too complicated to analyze. While Sparse Autoencoders (SAEs) have emerged as a powerful tool for interpretability, existing approaches predominantly operate at the token level, creating a granularity mismatch when capturing more critical step-level information, such as reasoning direction and semantic transitions. In this work, we propose step-level sparse autoencoder (SSAE), which serves as an analytical tool to disentangle different aspects of LLMs' reasoning steps into sparse features. Specifically, by precisely controlling the sparsity of a step feature conditioned on its context, we form an information bottleneck in step reconstruction, which splits incremental information from background information and disentangles it into several sparsely activated dimensions. Experiments on multiple base models and reasoning tasks show the effectiveness of the extracted features. By linear probing, we can easily predict surface-level information, such as generation length and first token distribution, as well as more complicated properties, such as the correctness and logicality of the step. These observations indicate that LLMs should already at least partly know about these properties during generation, which provides the foundation for the self-verification ability of LLMs. Our code is available at https://github.com/Miaow-Lab/SSAE.

cs.LG↗

arXiv2026

TOA_SP: A Multi-Strategy Framework for Single-Pulse Timing

Precision pulsar timing typically relies on the stability of average pulse profiles, enabling time-of-arrival (TOA) estimation through template cross-correlation. This assumption breaks down for highly variable radio sources such as Rotating Radio Transients (RRATs) and fast radio bursts (FRBs), where individual pulses could exhibit strong variability in morphology and amplitude, and no single averaged profile may represent the underlying emission process. We present toa_sp, an open-source Python package for extracting TOAs directly from PSRFITS search-mode data without requiring profile folding into a stable template. The framework implements a suite of complementary single-pulse timing strategies, including parametric profile fitting, non-parametric estimators, and adaptive sub-band and time-resolution optimisation, together with empirical diagnostics for assessing model consistency. We apply toa_sp to 688 single pulses from a 3-hour FAST observation of RRAT~J1913+1330. The resulting TOAs residual achieve a weighted RMS residual of 1.33\,ms, a 24\% improvement over a standard template-based PSRCHIVE pipeline, while retaining all pulses without statistical outlier rejection. A set of bright FRB 20220529 bursts provides a controlled test of the framework across regimes of increasing pulse complexity, revealing frequency-dependent substructure not captured by band-integrated profiles. We introduce an empirical convergence diagnostic that identifies well-constrained pulses and guides the transition between parametric and non-parametric regimes. Full multi-strategy processing of 688 pulses requires approximately 7.6\,s per pulse on a 10-thread CPU. The package is publicly available via pip install toa_sp.

astro-ph.IM↗

arXiv2026

TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents

As large language models and harness frameworks continue to advance, agents operating in terminals are increasingly capable of performing a broader range of general computer-use tasks beyond coding. However, existing benchmarks do not adequately evaluate general-purpose terminal computer-use agents (TUAs): general computer-use benchmarks primarily target graphical user interfaces (GUIs), whereas terminal-based benchmarks largely emphasize technical and programming-centric workflows historically native to the shell. We introduce TUA-Bench, a general-purpose benchmark for terminal-use agents. TUA-Bench includes 120 real-world tasks across five task families, covering routine digital activities-including document editing, email management, and live-web information seeking-as well as scientific and engineering workflows co-designed with PhD-level domain experts that require specialized software. This breadth distinguishes TUA-Bench from prior shell-focused or domain-specific benchmarks. Each task is manually designed, runs in a real terminal with a deterministic setup script, and is evaluated by an execution-based scoring protocol. We find that the strongest frontier agent, Claude Code with Claude Opus 4.8 max reasoning effort, achieves 65.8% overall performance, with substantial gaps across both tracks. By providing a broad and realistic evaluation of terminal-use capabilities, TUA-Bench aims to accelerate the transition from narrow, task-specific assistants to general-purpose agents capable of operating reliably across diverse digital environments.

cs.SE↗

arXiv2026

Beyond Ideal Instruction: A Comprehensive Framework for Evaluating LLMs in Realistic Interactions

Despite great advances in tool-use capabilities of large language models (LLMs), existing evaluation benchmarks struggle to fully align with real-world scenarios. Such benchmarks mostly rely on simulated idealized user assumptions and lacks experience-oriented evaluation. These limitations fail to account for the ambiguity, uncooperative behaviors, and shifting intentions characteristic of real-world users. To fill this gap, we propose RUT-Bench, a dedicated benchmark designed to assess LLMs under diverse Real-world User Tool calling scenarios. RUT-Bench supports high-fidelity simulations covering both ideal rational patterns and heterogeneous non-ideal behaviors across single-turn and multi-turn dialogues. We conduct comprehensive evaluations on 19 widely adopted open-source and proprietary LLMs using our benchmark. Experimental results reveal that no tested LLMs achieve an overall success rate above 40%, and nearly all of them experience noticeable performance drops when facing more complicated non-ideal user inputs. Our code and data is available at https://github.com/Miaow-Lab/RUT-Bench.

cs.CL↗

arXiv2026

Spin Dynamics from Niu-Kleinman Adiabatic Approach and Slave Boson Mean Field Theory

Spin-wave excitations provide a central probe of magnetic order and electronic correlations in strongly correlated materials. In this work, we develop an adiabatic theory of spin dynamics by combining the Niu-Kleinman formalism with Kotliar-Ruckenstein slave-boson theory (NK+KRSB). For each frozen spin configuration, the constrained slave-boson saddle point is solved self-consistently, allowing the Berry-curvature matrix and energy Hessian entering the linearized adiabatic equations of motion to be extracted directly. Applied to the half-filled single-orbital Hubbard model, the resulting spin-wave dispersion shows substantially improved agreement with determinant quantum Monte Carlo benchmarks compared with the random phase approximation and closely approaches results from the time-dependent Gutzwiller approximation. We further extend the method to a two-orbital model of $\mathrm{La}_2\mathrm{NiO}_4$, demonstrating its applicability to realistic multi-orbital correlated systems. Because the approach only requires saddle-point solutions near the magnetic ground state, it remains computationally efficient while incorporating strong-correlation effects beyond conventional weak-coupling descriptions, providing a practical framework for studying low-energy spin excitations in correlated quantum materials.

cond-mat.str-el↗