arXiv subjects

Wei Duan

Publications and source records attributed to Wei Duan.

At least 19 recordsLinked to original sources

arXiv2026

FreKoo++: Learning Continuous Spectral Dynamics for Temporal Domain Generalization

Temporal Domain Generalization (TDG) aims to learn from historical domains and generalize to unseen future distributions under concept drift. Nevertheless, prevailing TDG methods struggle with complex real-world streaming scenarios involving both multi-scale drift patterns (e.g., long-term periodicity intertwined with short-term incremental changes) and local uncertainties, especially in continuous settings where observations arrive irregularly. To address this limitation, we propose FreKoo++, a novel continuous spectral-dynamical framework that pioneers the unification of continuous Koopman modal dynamics with adaptive spectral disentanglement. Specifically, FreKoo++ maps source-domain parameters into a compact latent space, modeling their evolution as a superposition of learnable continuous modes where complex eigenvalues jointly encode oscillatory frequency and temporal growth or decay. This formulation naturally accommodates irregular timestamps and supports arbitrary horizon extrapolation without rigid discrete stepping. Furthermore, we propose a new adaptive soft spectral weighting mechanism backed by stability and spectral regularization, which automatically isolates persistent dominant dynamics from transient noise without relying on manual frequency thresholds. We derive modal approximation and generalization bounds that characterize how amplitude and eigenvalue estimation errors propagate with the prediction horizon. Extensive experiments on both discrete and continuous TDG benchmarks demonstrate that FreKoo++ achieves state-of-the-art performance under complex multi-scale drifts and irregular sampling.

cs.LG↗

arXiv2026

Design and Implementation of a Microservice-Architecture Master Control System for AIMS

The mid-infrared solar magnetic field telescope AIMS (An Infrared System for the Accurate Measurement of Solar Magnetic Field) is the first ground-based telescope designed to directly measure solar magnetic fields via Zeeman splitting in the 8-14 um band, overcoming the century-long bottleneck of model-dependent indirect measurements. Its remote high-altitude site, heterogeneous multi-institute components, and complex observation modes comprising Fourier Transform Infrared (FTIR) spectropolarimetry and broadband imaging demand a highly autonomous Master Control System (MCS). We present the design and implementation of the AIMS MCS, featuring three key contributions: (1) an L0-L5 telescope automation classification inspired by the SAE J3016 autonomous driving standard, providing well-defined boundaries and a progressive evolution roadmap; (2) a three-layer system framework device control, autonomy support, and central decision-implemented with a microservice software architecture that achieves loose coupling, high cohesion, and continuous integration of heterogeneous components; and (3) a suite of key enabling tech-nologies including automatic pointing/tracking, autofocus via lucky-frame selection combined with power spectral ratio analysis, and environment-adaptive observation integrating auto-exposure, cloud detection, and power/thermal monitoring. The MCS has been validated across three telescopes at progressive automation levels: AIMS itself, the WenQuan Solar Magnetic Field Telescope, and the Solar Full-disk Multi-layer Magnetograph (SFMM). Collectively, these deployments demonstrate the feasibility and stability of the proposed architecture for progressive telescope automation.

astro-ph.IM↗

arXiv2026

Heterogeneous Information-Bottleneck Coordination Graphs for Multi-Agent Reinforcement Learning

Coordination graphs are a central abstraction in cooperative multi-agent reinforcement learning (MARL), yet existing sparse-graph learners lack a theoretically grounded mechanism to decide which edges should exist and how much information each edge should carry. Current methods rely on heuristic criteria that offer no formal guarantee on the learned topology, and no principled way to allocate different communication capacities to structurally different agent relationships. To address this, we propose Heterogeneous Information-Bottleneck Coordination Graphs (HIBCG), which learns a group-aware sparse graph in which both edge existence and message capacity are theoretically justified. With the graph information bottleneck (GIB) serving as the underlying tool, HIBCG first constructs a group-aligned block-diagonal prior that provides a closed-form criterion for edge retention -- determining which edges should exist and at what density per group block -- and then controls per-agent feature bandwidth on the resulting topology, compressing messages to retain only task-relevant content. We prove that the group-aligned prior strictly tightens the variational bound on topology learning, that the objective decomposes per group block, enabling differential edge control, and that capacity allocation follows a water-filling principle.

cs.AI↗

arXiv2026

MD-Face: MoE-Enhanced Label-Free Disentangled Representation for Interactive Facial Attribute Editing

GAN-based facial attribute editing is widely used in virtual avatars and social media but often suffers from attribute entanglement, where modifying one face attribute unintentionally alters others. While supervised disentangled representation learning can address this, it relies heavily on labeled data, incurring high annotation costs. To address these challenges, we propose MD-Face, a label-free disentangled representation learning framework based on Mixture of Experts (MoE). MD-Face utilizes a MoE backbone with a gating mechanism that dynamically allocates experts, enabling the model to learn semantic vectors with greater independence. To further enhance attribute entanglement, we introduce a geometry-aware loss, which aligns each semantic vector with its corresponding Semantic Boundary Vector (SBV) through a Jacobian-based pushforward method. Experiments with ProGAN and StyleGAN show that MD-Face outperforms unsupervised baselines and competes with supervised ones. Compared to diffusion-based methods, it offers better image quality and lower inference latency, making it ideal for interactive editing.

cs.CV↗

arXiv2026

Towards Robust Endogenous Reasoning: Unifying Drift Adaptation in Non-Stationary Tuning

Reinforcement Fine-Tuning (RFT) has established itself as a critical paradigm for the alignment of Multi-modal Large Language Models (MLLMs) with complex human values and domain-specific requirements. Nevertheless, current research primarily focuses on mitigating exogenous distribution shifts arising from data-centric factors, the non-stationarity inherent in the endogenous reasoning remains largely unexplored. In this work, a critical vulnerability is revealed within MLLMs: they are highly susceptible to endogenous reasoning drift, across both thinking and perception perspectives. It manifests as unpredictable distribution changes that emerge spontaneously during the autoregressive generation process, independent of external environmental perturbations. To adapt it, we first theoretically define endogenous reasoning drift within the RFT of MLLMs as the multi-modal concept drift. In this context, this paper proposes Counterfactual Preference Optimization ++ (CPO++), a comprehensive and autonomous framework adapted to the multi-modal concept drift. It integrates counterfactual reasoning with domain knowledge to execute controlled perturbations across thinking and perception, employing preference optimization to disentangle spurious correlations. Extensive empirical evaluations across two highly dynamic and safety-critical domains: medical diagnosis and autonomous driving. They demonstrate that the proposed framework achieves superior performance in reasoning coherence, decision-making precision, and inherent robustness against extreme interference. The methodology also exhibits exceptional zero-shot cross-domain generalization, providing a principled foundation for reliable multi-modal reasoning in safety-critical applications.

cs.LG↗

arXiv2026

IGU-LoRA: Adaptive Rank Allocation via Integrated Gradients and Uncertainty-Aware Scoring

As large language models (LLMs) scale to billions of parameters, full-parameter fine-tuning becomes compute- and memory-prohibitive. Parameter-efficient fine-tuning (PEFT) mitigates this issue by updating only a small set of task-specific parameters while keeping the base model frozen. Among PEFT approaches, low-rank adaptation (LoRA) is widely adopted; however, it enforces a uniform rank across layers despite substantial variation in layer importance, motivating {layerwise} rank allocation. Recent adaptive-rank variants (e.g., AdaLoRA) allocate ranks based on importance scores, yet typically rely on instantaneous gradients that capture only local sensitivity, overlooking non-local, pathwise effects within the same layer, which yields unstable and biased scores. To address this limitation, we introduce IGU-LoRA, an adaptive-rank LoRA that (i) computes within-layer Integrated Gradients (IG) sensitivities and aggregates them into a layer-level score for rank allocation, and (ii) applies an uncertainty-aware scheme using exponential moving averages with deviation tracking to suppress noisy updates and calibrate rank selection. Theoretically, we prove an upper bound on the composite trapezoidal rule approximation error for parameter-space IG under a pathwise Hessian-Lipschitz condition, which informs the quadrature budget. Across diverse tasks and architectures, IGU-LoRA consistently outperforms strong PEFT baselines at matched parameter budgets, improving downstream accuracy and robustness. Ablations confirm the contributions of pathwise within-layer sensitivity estimates and uncertainty-aware selection to effective rank allocation. Our code is publicly available at https://github.com/withyou12/igulora.git

cs.LG↗

arXiv2026

U-Face: An Efficient and Generalizable Framework for Unsupervised Facial Attribute Editing via Subspace Learning

Latent space-based facial attribute editing methods have gained popularity in applications such as digital entertainment, virtual avatar creation, and human-computer interaction systems due to their potential for efficient and flexible attribute manipulation, particularly for continuous edits. Among these, unsupervised latent space-based methods, which discover effective semantic vectors without relying on labeled data, have attracted considerable attention in the research community. However, existing methods still encounter difficulties in disentanglement, as manipulating a specific facial attribute may unintentionally affect other attributes, complicating fine-grained controllability. To address these challenges, we propose a novel framework designed to offer an effective and adaptable solution for unsupervised facial attribute editing, called Unsupervised Facial Attribute Controllable Editing (U-Face). The proposed method frames semantic vector learning as a subspace learning problem, where latent vectors are approximated within a lower-dimensional semantic subspace spanned by a semantic vector matrix. This formulation can also be equivalently interpreted from a projection-reconstruction perspective and further generalized into an autoencoder framework, providing a foundation that can support disentangled representation learning in a flexible manner. To improve disentanglement and controllability, we impose orthogonal non-negative constraints on the semantic vectors and incorporate attribute boundary vectors to reduce entanglement in the learned directions. Although these constraints make the optimization problem challenging, we design an alternating iterative algorithm, called Alternating Iterative Disentanglement and Controllability (AIDC), with closed-form updates and provable convergence under specific conditions.

cs.CV↗

arXiv2025

Bandwidth-constrained Variational Message Encoding for Cooperative Multi-agent Reinforcement Learning

Graph-based multi-agent reinforcement learning (MARL) enables coordinated behavior under partial observability by modeling agents as nodes and communication links as edges. While recent methods excel at learning sparse coordination graphs-determining who communicates with whom-they do not address what information should be transmitted under hard bandwidth constraints. We study this bandwidth-limited regime and show that naive dimensionality reduction consistently degrades coordination performance. Hard bandwidth constraints force selective encoding, but deterministic projections lack mechanisms to control how compression occurs. We introduce Bandwidth-constrained Variational Message Encoding (BVME), a lightweight module that treats messages as samples from learned Gaussian posteriors regularized via KL divergence to an uninformative prior. BVME's variational framework provides principled, tunable control over compression strength through interpretable hyperparameters, directly constraining the representations used for decision-making. Across SMACv1, SMACv2, and MPE benchmarks, BVME achieves comparable or superior performance while using 67--83% fewer message dimensions, with gains most pronounced on sparse graphs where message quality critically impacts coordination. Ablations reveal U-shaped sensitivity to bandwidth, with BVME excelling at extreme ratios while adding minimal overhead.

cs.LG↗

arXiv2025

Turning Drift into Constraint: Robust Reasoning Alignment in Non-Stationary Multi-Stream Environments

This paper identifies a critical yet underexplored challenge in reasoning alignment from multiple multi-modal large language models (MLLMs): In non-stationary environments, the diverse reasoning distributions of source models often evolve unpredictably, transmitting systematic biases and drift to the target model. To address this, we formulate multi-source reasoning alignment as a constraint satisfaction problem under concept drift theory. We propose Autonomous Preference Optimization (APO), a novel framework that treats inter-model divergences not as noise, but as dynamic negative constraints. APO operates via a two-stage protocol: first, supervised bootstrapping projects the target model into the capability union of source models; second, constraint-aware optimization synthesizes a consistent consensus manifold by explicitly suppressing drifting trajectories via a multi-negative Plackett-Luce objective. Extensive experiments on chest X-ray interpretation demonstrate that our 7B model achieves superior robustness, outperforming even proprietary source models in average accuracy. Furthermore, we release CXR-MAX, a large-scale benchmark comprising 170,982 reasoning trajectories from seven large-scale MLLMs to facilitate research on reasoning alignment under drift. Code and data are available at: https://github.com/XiaoyuYoung/APO.

cs.CV↗

arXiv2025

Bayesian Ego-graph Inference for Networked Multi-Agent Reinforcement Learning

In networked multi-agent reinforcement learning (Networked-MARL), decentralized agents must act under local observability and constrained communication over fixed physical graphs. Existing methods often assume static neighborhoods, limiting adaptability to dynamic or heterogeneous environments. While centralized frameworks can learn dynamic graphs, their reliance on global state access and centralized infrastructure is impractical in real-world decentralized systems. We propose a stochastic graph-based policy for Networked-MARL, where each agent conditions its decision on a sampled subgraph over its local physical neighborhood. Building on this formulation, we introduce BayesG, a decentralized actor-framework that learns sparse, context-aware interaction structures via Bayesian variational inference. Each agent operates over an ego-graph and samples a latent communication mask to guide message passing and policy computation. The variational distribution is trained end-to-end alongside the policy using an evidence lower bound (ELBO) objective, enabling agents to jointly learn both interaction topology and decision-making strategies. BayesG outperforms strong MARL baselines on large-scale traffic control tasks with up to 167 agents, demonstrating superior scalability, efficiency, and performance.

cs.MA↗

arXiv2025

Trojan Horse Prompting: Jailbreaking Conversational Multimodal Models by Forging Assistant Message

The rise of conversational interfaces has greatly enhanced LLM usability by leveraging dialogue history for sophisticated reasoning. However, this reliance introduces an unexplored attack surface. This paper introduces Trojan Horse Prompting, a novel jailbreak technique. Adversaries bypass safety mechanisms by forging the model's own past utterances within the conversational history provided to its API. A malicious payload is injected into a model-attributed message, followed by a benign user prompt to trigger harmful content generation. This vulnerability stems from Asymmetric Safety Alignment: models are extensively trained to refuse harmful user requests but lack comparable skepticism towards their own purported conversational history. This implicit trust in its "past" creates a high-impact vulnerability. Experimental validation on Google's Gemini-2.0-flash-preview-image-generation shows Trojan Horse Prompting achieves a significantly higher Attack Success Rate (ASR) than established user-turn jailbreaking methods. These findings reveal a fundamental flaw in modern conversational AI security, necessitating a paradigm shift from input-level filtering to robust, protocol-level validation of conversational context integrity.

cs.AI↗

arXiv2025

Resilient Contrastive Pre-training under Non-Stationary Drift

The remarkable success of large-scale contrastive pre-training has been largely driven by by vast yet static datasets. However, as the scaling paradigm evolves, this paradigm encounters a fundamental challenge when applied to dynamic data streams characterized by concept drift - unpredictable changes in the underlying data distribution. This paper aims to advance robust pre-training under such non-stationary environments. We begin by revealing that conventional contrastive pre-training methods are highly susceptible to concept drift, resulting in significant substantial bias and instability within the learned feature representations. To systematically analyze these effects, we develop a structural causal model that elucidates how drift acts as a confounder, distorting the learned representations. Based on these causal insights, we propose Resilient Contrastive Pre-training (RCP), a novel method that incorporates causal intervention. RCP formulates a causally-informed objective to mitigate drift-induced biases through targeted interventions. The method is designed for simple and scalable implementation and exhibits notable adaptability, promoting robust and autonomous pre-training on non-stationary data. Comprehensive experiments across various downstream tasks consistently demonstrate that RCP effectively alleviates the detrimental impact of concept drift, yielding more resilient and generalizable representations.

cs.LG↗

arXiv2024

All-electric mimicking synaptic plasticity based on the noncollinear antiferromagnetic device

Neuromorphic computing, which seeks to replicate the brain's ability to process information, has garnered significant attention due to its potential to achieve brain-like computing efficiency and human cognitive intelligence. Spin-orbit torque (SOT) devices can be used to simulate artificial synapses with non-volatile, high-speed processing and endurance characteristics. Nevertheless, achieving energy-efficient all-electric synaptic plasticity emulation using SOT devices remains a challenge. We chose the noncollinear antiferromagnetic Mn3Pt as spin source to fabricate the Mn3Pt-based SOT device, leveraging its unconventional spin current resulting from magnetic space breaking. By adjusting the amplitude, duration, and number of pulsed currents, the Mn3Pt-based SOT device achieves nonvolatile multi-state modulated by all-electric SOT switching, enabling emulate synaptic behaviors like excitatory postsynaptic potential (EPSP), inhibitory postsynaptic potential (IPSP), long-term depression (LTD) and the long-term potentiation (LTP) process. In addition, we show the successful training of an artificial neural network based on such SOT device in recognizing handwritten digits with a high recognition accuracy of 94.95 %, which is only slightly lower than that from simulations (98.04 %). These findings suggest that the Mn3Pt-based SOT device is a promising candidate for the implementation of memristor-based brain-inspired computing systems.

physics.app-ph↗

arXiv2024

Optical optimization of a multi-slit extreme ultraviolet spectrograph for global solar corona diagnostics

The spatial-temporal evolution of coronal plasma parameters of the solar outer atmosphere at global scales, derived from solar full-disk imaging spectroscopic observation in the extreme-ultraviolet band, is critical for understanding and forecasting solar eruptions. We propose a multi-slits extreme ultraviolet imaging spectrograph for global coronal diagnostics with high cadence and present the preliminary instrument designs for the wavelength range from 18.3 to 19.8 nm. The instrument takes a comprehensive approach to obtain global coronal spatial and spectral information, improve the detected cadence and avoid overlapping. We first describe the relationship between optical properties and structural parameters, especially the relationship between the overlapping and the number of slits, and give a general multi-slits extreme-ultraviolet imaging spectrograph design process. Themultilayer structure is optimized to enhance the effective areas in the observation band. Five distantly-separated slits are set to divide the entire solar field of view, which increase the cadence for raster scanning the solar disk by 5 times relative to a single slit. The spectral resolving power of the optical system with an aperture diameter of 150 mm are optimized to be greater than 1461. The spatial resolution along the slits direction and the scanning direction are about 4.4''and 6.86'', respectively. The Al/Mo/B4C multilayer structure is optimized and the peak effective area is about 1.60 cm2 at 19.3 nm with a full width at half maximum of about 1.3 nm. The cadence to finish full-disk raster scan is about 5 minutes. Finally, the instrument performance is evaluated by an end-to-end calculation of the system photon budget and a simulation of the observational image and spectra. Our investigation shows that this approach is promising for global coronal plasma diagnostics.

astro-ph.SR↗

arXiv2024

Opacity Enforcement by Edit Functions Under Incomparable Observations

As an information-flow privacy property, opacity characterizes whether a malicious external observer (referred to as an intruder) is able to infer the secret behavior of a system. This paper addresses the problem of opacity enforcement using edit functions in discrete event systems modeled by partially observed deterministic finite automata. A defender uses the edit function as an interface at the output of a system to manipulate actual observations through insertion, substitution, and deletion operations so that the intruder will be prevented from inferring the secret behavior of the system. Unlike existing work which usually assumes that the observation capabilities of the intruder and the defender are identical, we consider a more general setting where they may observe incomparable subsets of events generated by the system.To characterize whether the defender has the ability to enforce opacity of the system under this setting, the notion of \emph{$ic$-enforceability} is introduced. Then, the opacity enforcement problem is transformed to a two-player game, with imperfect information between the system and the defender, which can be used to determine a feasible decision-making strategy for the defender. Within the game scheme, an edit mechanism is constructed to enumerate all feasible edit actions following system behavior. We further show that an $ic$-enforcing edit function (if one exists) can be synthesized from the edit mechanism to enforce opacity.

cs.FL↗

arXiv2024

Group-Aware Coordination Graph for Multi-Agent Reinforcement Learning

Cooperative Multi-Agent Reinforcement Learning (MARL) necessitates seamless collaboration among agents, often represented by an underlying relation graph. Existing methods for learning this graph primarily focus on agent-pair relations, neglecting higher-order relationships. While several approaches attempt to extend cooperation modelling to encompass behaviour similarities within groups, they commonly fall short in concurrently learning the latent graph, thereby constraining the information exchange among partially observed agents. To overcome these limitations, we present a novel approach to infer the Group-Aware Coordination Graph (GACG), which is designed to capture both the cooperation between agent pairs based on current observations and group-level dependencies from behaviour patterns observed across trajectories. This graph is further used in graph convolution for information exchange between agents during decision-making. To further ensure behavioural consistency among agents within the same group, we introduce a group distance loss, which promotes group cohesion and encourages specialization between groups. Our evaluations, conducted on StarCraft II micromanagement tasks, demonstrate GACG's superior performance. An ablation study further provides experimental evidence of the effectiveness of each component of our method.

cs.LG↗

arXiv2024

Event Concealment and Concealability Enforcement in Discrete Event Systems Under Partial Observation

Inspired by privacy problems where the behavior of a system should not be revealed to an external curious observer, we investigate event concealment and concealability enforcement in discrete event systems modeled as non-deterministic finite automata under partial observation. Given a subset of secret events in a given system, concealability holds if the occurrence of all secret events remains hidden to a curious observer (an eavesdropper). A secret event is said to be (at least under some executions) unconcealable (inferable) if its occurrence can be indirectly determined with certainty after a finite number of observations. When concealability of a system does not hold (i.e., one or more secret events are unconcealable), we analyze how a defender, placed at the interface of the system with the eavesdropper, can be used to enforce concealability. The defender takes as input each observed event of the system and outputs a carefully modified event sequence (seen by the eavesdropper) using event deletion, insertion, or replacement. The defender is said to be C-enforceable if, following the occurrence of the secret events and regardless of subsequent activity generated by the system, it can always deploy a strategy to manipulate observations and conceal the events perpetually. We discuss systematic procedures to detect the presence of unconcealable secret events and verify C-Enforceability using techniques from state estimation and event diagnosis. We also propose a polynomial complexity construction for obtaining one necessary and one sufficient condition for C-Enforceability.

cs.CR↗

arXiv2024

Inferring Latent Temporal Sparse Coordination Graph for Multi-Agent Reinforcement Learning

Effective agent coordination is crucial in cooperative Multi-Agent Reinforcement Learning (MARL). While agent cooperation can be represented by graph structures, prevailing graph learning methods in MARL are limited. They rely solely on one-step observations, neglecting crucial historical experiences, leading to deficient graphs that foster redundant or detrimental information exchanges. Additionally, high computational demands for action-pair calculations in dense graphs impede scalability. To address these challenges, we propose inferring a Latent Temporal Sparse Coordination Graph (LTS-CG) for MARL. The LTS-CG leverages agents' historical observations to calculate an agent-pair probability matrix, where a sparse graph is sampled from and used for knowledge exchange between agents, thereby simultaneously capturing agent dependencies and relation uncertainty. The computational complexity of this procedure is only related to the number of agents. This graph learning process is further augmented by two innovative characteristics: Predict-Future, which enables agents to foresee upcoming observations, and Infer-Present, ensuring a thorough grasp of the environmental context from limited data. These features allow LTS-CG to construct temporal graphs from historical and real-time information, promoting knowledge exchange during policy learning and effective collaboration. Graph learning and agent training occur simultaneously in an end-to-end manner. Our demonstrated results on the StarCraft II benchmark underscore LTS-CG's superior performance.

cs.LG↗