arXiv subjects

Nan Chen

Publications and source records attributed to Nan Chen.

At least 37 records · Page 2Linked to original sources

arXiv2026

DomainShuttle: Freeform Open Domain Subject-driven Text-to-video Generation

Open domain subject-driven text-to-video (S2V) generation has drawn significant interest in academia and industry. Open domain S2V mainly involves two scenarios: in-domain, which requires retaining the reference subject features as much as possible, and cross-domain, which preserves the intrinsic features of the subject while allowing subject-irrelevant properties to vary flexibly according to the text prompt. Existing methods primarily focus on maximizing subject fidelity in in-domain scenarios, which limits their editability and adaptability in cross-domain scenarios, such as novel styles, semantic combinations, or domain attributes. In this study, we propose that an ideal S2V method should flexibly shuttle between different domains, achieving strong performance in both in-domain and cross-domain scenarios. To this end, we propose DomainShuttle, which could achieve high fidelity and generative flexibility for open domain video personalization. Specifically, we introduce Domain-MoT, which decouples videos and reference features and introduces the domain-aware AdaLN for domain-specific modeling of reference images. We then introduce the Video-Reference DualRoPE scheme, which places reference image tokens and video tokens in separate RoPE spaces to enable precise subject-level spatial modeling, and Cross-Pair Consistent Loss, which aims to extract intrinsic subject features unaffected by irrelevant features. Extensive experiments demonstrate that DomainShuttle achieves significant performance improvements over existing methods, exhibiting high subject fidelity and generative flexibility across diverse open domain application scenarios.

cs.CV↗

arXiv2026

An Adaptive Online Smoother with Closed-Form Solutions and Information-Theoretic Lag Selection for Conditional Gaussian Nonlinear Systems

Data assimilation (DA) combines partial observations with dynamical models to improve state estimation. Filter-based DA uses only past and present data and is the prerequisite for real-time forecasts. Smoother-based DA exploits both past and future observations. It aims to fill in missing data, provide more accurate estimations, and develop high-quality datasets. However, the standard smoothing procedure requires using all historical state estimations, which is storage-demanding, especially for high-dimensional systems. This paper develops an adaptive-lag online smoother for a large class of complex dynamical systems with strong nonlinear and non-Gaussian features, which has important applications to many real-world problems. The adaptive lag allows the utilization of observations only within a nearby window, thus reducing computational complexity and storage needs. Online lag adjustment is essential for tackling turbulent systems, where temporal autocorrelation varies significantly over time due to intermittency, extreme events, and nonlinearity. Based on the uncertainty reduction in the estimated state, an information criterion is developed to systematically determine the adaptive lag. Notably, the mathematical structure of these systems facilitates the use of closed analytic formulae to calculate the online smoother and adaptive lag, avoiding empirical tunings as in ensemble-based DA methods. The adaptive online smoother is applied to studying three important scientific problems. First, it helps detect online causal relationships between state variables. Second, the advantage of reduced computational storage expenditure is illustrated via Lagrangian DA, a high-dimensional nonlinear problem. Finally, the adaptive smoother advances online parameter estimation with partial observations, emphasizing the role of the observed extreme events in accelerating convergence.

eess.SY↗

arXiv2026

A Martingale-Free Introduction to Conditional Gaussian Nonlinear Systems

The conditional Gaussian nonlinear system (CGNS) is a broad class of nonlinear stochastic dynamical systems. Given the trajectories for a subset of state variables, the remaining follow a Gaussian distribution. Despite the conditionally linear structure, the CGNS exhibits strong nonlinearity, thus capturing many non-Gaussian characteristics observed in nature through its joint and marginal distributions. Desirably, it enjoys closed analytic formulae for the time evolution of its conditional Gaussian statistics, which facilitate the study of data assimilation and other related topics. In this paper, we develop a martingale-free approach to improve the understanding of CGNSs. This methodology provides a tractable approach to proving the time evolution of the conditional statistics by deriving results through time discretization schemes, with the continuous-time regime obtained via a formal limiting process as the discretization time-step vanishes. This discretized approach further allows for developing analytic formulae for optimal posterior sampling of unobserved state variables with correlated noise. These tools are particularly valuable for studying extreme events and intermittency and apply to high-dimensional systems. Moreover, the approach improves the understanding of different sampling methods in characterizing uncertainty. The effectiveness of the framework is demonstrated through a physics-constrained, triad-interaction climate model with cubic nonlinearity and state-dependent cross-interacting noise.

math.DS↗

arXiv2026

$Ω$: Operator-based Mixture Ensemble for Generative Assimilation

Characterizing non-Gaussian posterior distributions in partially observed high-dimensional nonlinear systems remains a fundamental challenge in data assimilation. Ensemble Kalman filters rely on Gaussian approximations that can be inaccurate for strongly non-Gaussian posteriors, whereas particle filters suffer from severe scalability limitations. Recent score-based generative approaches improve posterior characterization but typically require supervised training with ground-truth posterior samples, which are unavailable in most practical applications. We introduce $Ω$ (Operator-based Mixture Ensemble for Generative Assimilation), a scalable framework that integrates conditional Gaussian surrogate modeling, unsupervised score learning, and generative sampling. The conditional Gaussian surrogate provides a nonlinear non-Gaussian baseline approximation while admitting closed-form conditional posterior distributions for the unresolved variables. First, $Ω$ exploits these closed-form conditional distributions to analytically recover the high-dimensional unobserved component, reducing computational cost and mitigating the curse of dimensionality. Second, $Ω$ learns only the residual discrepancy beyond an analytical baseline through denoising score matching using ensemble trajectories alone, eliminating the need for ground-truth posterior samples and substantially reducing the learning burden. Third, $Ω$ reconstructs the full non-Gaussian posterior distribution of both observed and unobserved variables via a Gaussian mixture representation, capturing multimodal, skewed, and heavy-tailed statistics. Finally, $Ω$ employs annealed Langevin sampling to iteratively refine ensemble members from the baseline toward the target posterior. $Ω$ is validated on several turbulent models with intermittency and extreme events, consistently improving posterior accuracy.

cs.LG↗

arXiv2026

TianJi-Environ: An Autonomous AI Scientist for Atmospheric Environmental Research

As atmospheric environmental prediction continues to improve, interpretable validation of pollution mechanisms and feedback processes has become a main challenge in atmospheric chemistry. Yet mechanism validation based on complex numerical models still relies heavily on expert knowledge: mechanistic hypotheses must be operationalized into executable experiments, and model outputs must be organized into traceable evidence. We present TianJi-Environ, an auditable AI Scientist for atmospheric-chemistry mechanism validation. TianJi-Environ establishes the first WRF-Chem-based multi-agent framework that autonomously drives complex atmospheric-chemistry simulations, converting mechanistic hypotheses into executable configurations, testing experiments, and evidence criteria. Using ozone response and particulate-matter feedback as two representative examples, we demonstrate TianJi-Environ's capability for mechanism validation. In a summertime ozone case over the North China Plain, the system detects directionally consistent aerosol-radiation-interaction signals in shortwave radiation and boundary-layer height, but judges the evidence for ozone response to NOx control to be incomplete. In a wintertime PM2.5 case over the Guanzhong Basin, it localizes the unsupported link to insufficient propagation from black-carbon perturbation to particulate response and missing diagnostics of vertical absorptive heating. These results show that TianJi-Environ makes expert-driven mechanism validation explicit, structured, and auditable, offering a reproducible paradigm for multi-agent systems coupled with complex atmospheric-chemistry models.

physics.ao-ph↗

arXiv2026

SPLIT-PINN: Separable Probability Learning Technique via Physics-Informed Neural Networks for High-Dimensional Probabilistic Modeling

We present a probabilistic modeling framework for incorporating small-scale spatial heterogeneity into macroscopic descriptions of material behavior for polycrystalline metallic materials. Spatially heterogeneous material state fields are represented using probability density functions (PDFs), providing a principled statistical description of microstructural variability and state evolution across different computational polycrystalline realizations. The framework is built on the inverse identification of a probabilistic transport model, formulated as a Liouville equation with an unknown drift term. To enable accurate, stable, and interpretable inference of this drift field in high-dimensional, transport-dominated settings, we develop a Separable Probability Learning Technique via Physics-Informed Neural Networks (SPLIT-PINN). This method incorporates a marginal-correction drift decomposition, orthogonality constraints, and residual-based adaptive training to enhance well-posedness, numerical stability, and physical consistency without imposing restrictive parametric assumptions. Using SPLIT-PINN, the drift field governing the temporal evolution of joint state PDFs is inferred directly from data. After benchmark validation, the framework is applied to physical computational datasets describing the evolution of polycrystalline microstructural states, including von Mises stress, dislocation density, and equivalent plastic strain rate. The learned Liouville model, trained on a single dataset, is subsequently used in forward predictions of the temporal evolution of joint and marginal PDFs for multiple unseen polycrystal realizations. Quantitative comparisons with reference PDFs demonstrate that the proposed framework yields accurate and robust probabilistic predictions and generalizes effectively across datasets.

cond-mat.mtrl-sci↗

arXiv2026

Mechanisms and Pathways of Extreme Events in Partially-Observed Stochastic Dynamical Systems

Extreme events occur across the natural, engineering, and socioeconomic sciences, where rare but high-impact episodes can lead to disproportionate consequences that pose major challenges for prediction and risk management. Existing studies have mainly focused on the statistics, sampling, forecasting, and attribution of extremes from observable variables. In this paper, we develop a mathematical framework for studying the mechanisms and pathways of extreme events in partially-observed stochastic dynamical systems with hidden variables. By integrating data assimilation with information-theoretic and trajectory-based diagnostics, we infer latent precursor dynamics from observations, quantify their uncertainty, and determine how their influence propagates toward observed extreme events. Conditional Gaussian models provide a tractable analytical setting for deriving closed-form diagnostics, while the framework extends through numerical methods. The analysis proceeds from two complementary perspectives. From a trajectory-wise viewpoint, we compare filtering and smoothing distributions to identify the onset of hidden precursors and quantify temporal influence. From a statistical viewpoint, we construct event-conditioned hidden-state distributions to identify sensitive triggering directions, latent pathways, and multiple classes of extreme-event mechanisms through clustering. Three numerical examples illustrate the methodology. In an intermittent stochastic model, hidden damping dynamics emerge before observed bursts, where discrepancies between the filter and smoother provide an onset diagnostic. In a stochastic model with damping and forcing, separate damping-induced, forcing-driven, and mixed pathways to extremes are identified. In a nonlinear topographic-flow model, distinct mechanisms and pathways for blocking and unblocking patterns associated with observed extreme events are revealed.

math.DS↗

arXiv2026

3D Primitives are a Spatial Language for VLMs

Vision-language models (VLMs) exhibit a striking paradox: they can generate executable code that reconstructs a 3D scene from geometric primitives with correct object counts, classes, and approximate positions, yet the same models fail at simpler spatial questions on the same image. We show that 3D geometric primitives (cubes, spheres, cylinders, expressed in executable code) serve as a powerful intermediate representation for spatial understanding, and exploit this through three contributions. First, we introduce \textbf{\textsc{SpatialBabel}}, a benchmark evaluating fourteen VLMs on primitive-based 3D scene reconstruction across six \emph{scene-code languages} (programming languages and declarative formats for 3D primitive scenes), revealing that a single model's object-detection F1 can vary by up to $5.7\times$ across languages. Second, we propose \textbf{Code-CoT} (Code Chain-of-Thought), a training-free inference strategy that routes spatial reasoning through primitive-based code generation. Code-CoT lifts the SpatialBabel-QA-Score by up to $+6.4$\% on primitive scenes and real-photo CV-Bench-3D accuracy by $+5.0$\% for VLMs with strong coding capabilities. Third, we propose \textbf{S$^{3}$-FT} (Self-Supervised Spatial Fine-Tuning), which self-supervisedly distills primitive spatial knowledge into general visual reasoning by parsing the model's own Three.js primitive-reconstructions into structured annotations and fine-tuning on the result, with \emph{no human labels and no teacher model}. Training on primitive images alone, S$^3$-FT improves Qwen3-VL-8B by $+4.6$ to $+8.6$\% on SpatialBabel-Primitive-QA, $+9.7$\% on CV-Bench-2D, and $+17$\% on HallusionBench; the recipe transfers across model families. These results establish geometric primitives in code as both a diagnostic and a transferable spatial vocabulary for VLMs. We will release all artifacts upon publication.

cs.CV↗

arXiv2026

Vertical Structure of Local Disk Galaxies revealed by DESI Imaging Data

The vertical structure of galactic disks is an important probe of disk assembly history. We investigate a sample of 79 local disk galaxies within 50 Mpc using data from the DESI Legacy Imaging Surveys. Vertical luminosity profiles as a function of radius in the g, r, and z bands are extracted and fitted with a single-component sech^2 model to determine the scale height and its radial variation (flaring). Our results show that local galactic disks are generally thin with negligible flaring. The median scale heights at one effective radius are 0.21, 0.22, and 0.22 kpc in the g, r, and z bands, respectively, while the corresponding median radial gradients are -0.006, 0.003, and 0.001. These values are consistent with those of the geometric thin disk of the Milky Way, represented by metal-rich, low-[alpha/Fe] populations, supporting the conclusion of weak flaring. We also find a clear positive correlation between scale height and stellar mass, extending down to 10^7 solar masses. These results provide a homogeneous benchmark for the vertical structure of nearby disk galaxies and support the cosmological representativeness of the Milky Way thin disk.

astro-ph.GA↗

arXiv2026

Power laws in the sea ice floe size distribution: a stochastic theory

Sea ice is a complex system, and observations have shown that ice segments (i.e., floes) have a wide range of sizes, with a floe size distribution that follows a power law. However, a theory for the power law and its exponent have remained elusive. Here, floe-resolving numerical simulations are investigated with a discrete element model, in order to gain further information by gathering statistics of fracture and welding events. Then, based on the insights from the floe-resolving simulations, a stochastic fragmentation-coagulation theory is proposed. Exact solutions are found with a power law. The power-law exponent can take a variety of values, and it depends on the fracture and welding rates. Such behavior is reminiscent of seasonal changes in the power-law exponent, which have been reported in past analyses of observational data.

physics.ao-ph↗

arXiv2026

TianJi:An autonomous AI meteorologist for discovering physical mechanisms in atmospheric science

Artificial intelligence (AI) has achieved breakthroughs comparable to traditional numerical models in data-driven weather forecasting, yet it remains essentially statistical fitting and struggles to uncover the physical causal mechanisms of the atmosphere. Physics-oriented mechanism research still heavily relies on domain knowledge and cumbersome engineering operations of human scientists, becoming a bottleneck restricting the efficiency of Earth system science exploration. Here, we propose TianJi - the first "AI meteorologist" system capable of autonomously driving complex numerical models to verify physical mechanisms. Powered by a large language model-driven multi-agent architecture, TianJi can autonomously conduct literature research and generate scientific hypotheses. We further decouple scientific research into cognitive planning and engineering execution: the meta-planner interprets hypotheses and devises experimental roadmaps, while a cohort of specialized worker agents collaboratively complete data preparation, model configuration, and multi-dimensional result analysis. In two classic atmospheric dynamic scenarios (squall-line cold pools and typhoon track deflections), TianJi accomplishes expert-level end-to-end experimental operations with zero human intervention, compressing the research cycle to a few hours. It also delivers detailed result analyses and autonomously judges and explains the validity of the hypotheses from outputs. TianJi reveals that the role of AI in Earth system science is transitioning from a "black-box predictor" to an "interpretable scientific collaborator", offering a new paradigm for high-throughput exploration of scientific mechanisms.

cs.AI↗

arXiv2026

ReMe: Scaffolding Personalized Cognitive Training via Controllable LLM-Mediated Conversations

Global aging calls for scalable and engaging cognitive interventions. Computerized cognitive training (CCT) is a promising non-pharmacological approach, yet many unsupervised programs rely on rigid, hand-authored puzzles that are difficult to personalize and can hinder adherence. Large language models (LLMs) offer more natural interaction, but their open-ended generation complicates the controlled task structure required for cognitive training. We present ReMe, a web-based framework that scaffolds cognitive training through controllable LLM-mediated conversations, addressing both rigidity in conventional CCT content and the need for conversational controllability. ReMe features a modular Puzzle Engine that represents training activities as reusable puzzle groups specified by structured templates and constraint rules, enabling rapid development of dialogue-based word games and personalized tasks grounded in user context. By integrating personal life logs, ReMe supports Life Recall activities for episodic-memory practice through guided retrieval and progressive cues. A community pilot with 32 adults aged 50+ provides initial feasibility signals.

cs.AI↗

arXiv2026

A Lagrangian Conditional Gaussian Koopman Network for Data Assimilation and Prediction

Lagrangian data assimilation aims to recover hidden Eulerian flow fields from sparse, indirect observations of moving tracers. This problem is challenging because tracer trajectories are nonlinearly coupled with the underlying flow, making posterior inference computationally intractable in realistic, high-dimensional systems. In this work, we develop a Lagrangian conditional Gaussian Koopman network (LaCGKN), a structure-preserving, data-driven framework for joint data assimilation and prediction from Lagrangian observations. LaCGKN embeds Eulerian flow dynamics into a low-dimensional latent space governed by a nonlinear stochastic system with conditional Gaussian structure, enabling analytic posterior updates without ensemble forecasting. Unlike existing conditional Gaussian Koopman formulations that assume direct Eulerian observations, the Lagrangian setting imposes additional demands on the latent representation, which must simultaneously encode the flow dynamics and mediate nonlinear tracer-flow interactions. To address these challenges, the LaCGKN incorporates three key components: (i) tracer homogenization to enforce permutation equivariance and generalize across varying numbers of tracers; (ii) Fourier positional encoding to capture spatial dependence and reconstruct local flow features at moving tracer locations; and (iii) an SVD-inspired low-rank parameterization of the latent transition operator, which reduces model complexity while retaining expressiveness. An application to a two-layer quasi-geostrophic flow with surface tracer observations shows that LaCGKN achieves accurate and efficient Lagrangian data assimilation and prediction, without reliance on ensemble methods or the governing physical model. These results establish the LaCGKN as a unified and computationally tractable alternative to both traditional model-based approaches and purely black-box data-driven methods.

cs.CE↗

arXiv2026

Assimilative Causal Inference

Causal inference is fundamental across scientific disciplines, yet existing methods struggle to capture instantaneous, time-evolving causal relationships in complex, high-dimensional systems. In this paper, assimilative causal inference (ACI) is developed, which is a methodological framework that leverages Bayesian data assimilation to trace causes backward from observed effects. ACI solves the inverse problem rather than quantifying forward influence. It uniquely identifies dynamic causal interactions without requiring observations of candidate causes, accommodates short datasets, and, in principle, can be implemented in high-dimensional settings by employing efficient data assimilation algorithms. Crucially, it provides online tracking of causal roles that may reverse intermittently and facilitates a mathematically rigorous criterion for the causal influence range, revealing how far effects propagate. The effectiveness of ACI is demonstrated by complex dynamical systems showcasing intermittency and extreme events. ACI opens valuable pathways for studying complex systems, where transient causal structures are critical.

cs.LG↗

arXiv2026

Learning Rate Scaling across LoRA Ranks and Transfer to Full Finetuning

Low-Rank Adaptation (LoRA) is a standard tool for parameter-efficient finetuning of large models. While it induces a small memory footprint, its training dynamics can be surprisingly complex as they depend on several hyperparameters such as initialization, adapter rank, and learning rate. In particular, it is unclear how the optimal learning rate scales with adapter rank, which forces practitioners to re-tune the learning rate whenever the rank is changed. In this paper, we introduce Maximal-Update Adaptation ($μ$A), a theoretical framework that characterizes how the "optimal" learning rate should scale with model width and adapter rank to produce stable, non-vanishing feature updates under standard configurations. $μ$A is inspired from the Maximal-Update Parametrization ($μ$P) in pretraining. Our analysis leverages techniques from hyperparameter transfer and reveals that the optimal learning rate exhibits different scaling patterns depending on initialization and LoRA scaling factor. Specifically, we identify two regimes: one where the optimal learning rate remains roughly invariant across ranks, and another where it scales inversely with rank. We further identify a configuration that allows learning rate transfer from LoRA to full finetuning, drastically reducing the cost of learning rate tuning for full finetuning. Experiments across language, vision, vision--language, image generation, and reinforcement learning tasks validate our scaling rules and show that learning rates tuned on LoRA transfer reliably to full finetuning.

cs.LG↗

arXiv2026

From Struggle to Success: Context-Aware Guidance for Screen Reader Users in Computer Use

Equal access to digital technologies is critical for education, employment, and social participation. However, mainstream interfaces are visually oriented, creating steep learning curves and frequent obstacles for screen reader users, and limiting their independence and opportunities. Existing support is inadequate -- tutorials mainly target sighted users, while human assistance lacks real-time availability. We introduce AskEase, an on-demand AI assistant that provides step-by-step, screen reader user-friendly guidance for computer use. AskEase manages multiple sources of context to infer user intent and deliver precise, situation-specific guidance. Its seamless interaction design minimizes disruption and reduces the effort of seeking help. We demonstrated its effectiveness through representative usage scenarios and robustness tests. In a within-subjects study with 12 screen reader users, AskEase significantly improved task success while reducing perceived workload, including physical demand, effort, and frustration. These results demonstrate the potential of LLM-powered assistants to promote accessible computing and expand opportunities for users with visual impairments.

cs.HC↗

arXiv2025

LEFT-RS: A Lock-Free Fault-Tolerant Resource Sharing Protocol for Multicore Real-Time Systems

Emerging real-time applications have driven the transition to multicore embedded systems, where tasks must share resources due to functional demands and limited availability. These resources, whether local or global, are protected within critical sections to prevent race conditions, with locking protocols ensuring both exclusive access and timing requirements. However, transient faults occurring within critical sections can disrupt execution and propagate errors across multiple tasks. Conventional locking protocols fail to address such faults, and integrating traditional fault tolerance techniques often increases blocking. Recent approaches improve fault recovery through parallel replica execution; however, challenges remain due to sequential accessing, coordination overhead, and susceptibility to common-mode faults. In this paper, we propose a Lock-frEe Fault-Tolerant Resource Sharing (LEFT-RS) protocol for multicore real-time systems. LEFT-RS allows tasks to concurrently access and read global resources while entering their critical sections in parallel. Each task can complete its access earlier upon successful execution if other tasks experience faults, thereby improving the efficiency of resource usage. Our design also limits the overhead and enhances fault resilience. We present a comprehensive worst-case response time analysis to ensure timing guarantees. Extensive evaluation results demonstrate that our method significantly outperforms existing approaches, achieving up to an 84.5% improvement in schedulability on average.

cs.OS↗

arXiv2025

Screen Reader Programmers in the Vibe Coding Era: Adaptation, Empowerment, and New Accessibility Landscape

Generative AI agents are reshaping human-computer interaction, shifting users from direct task execution to supervising machine-driven actions, especially the rise of "vibe coding" in programming. Yet little is known about how screen reader programmers interact with AI code assistants in practice. We conducted a longitudinal study with 16 blind and low-vision programmers. Participants completed a GitHub Copilot tutorial, engaged with a programming task, and provided initial feedback. After two weeks of AI-assisted programming, follow-ups examined how their practices and perceptions evolved. Our findings show that code assistants enhanced programming efficiency and bridged accessibility gaps. However, participants struggled to convey intent, interpret AI outputs, and manage multiple views while maintaining situational awareness. They showed diverse preferences for accessibility features, expressed a need to balance automation with control, and encountered barriers when learning to use these tools. Furthermore, we propose design principles and recommendations for more accessible and inclusive human-AI collaborations.

cs.HC↗