arXiv subjects

Jian Wang

Publications and source records attributed to Jian Wang.

At least 73 records · Page 4Linked to original sources

arXiv2026

Quantitative Asymptotics for Time-Inhomogeneous Lévy-Driven SDEs with Asymptotically Vanishing Drifts

In this work, we are concerned with a class of multi-dimensional time-inhomogeneous stochastic differential equations (SDEs) on $\R^d$ driven by pure-jump Lévy processes, where the drift coefficient $b(t,x)$ satisfies $\lim_{t\to \infty}b(t,x) =0$ for every $x\in \R^d$. On account of three regimes associated with the index $α$ of large jumps corresponding to the driven Lévy noise, we investigate the quantitative asymptotics of the corresponding rescaled processes. More precisely, for $α\in (0,2)$, we prove that the rescaled processes, governed by time-inhomogeneous SDEs subject to additive processes, converge with respect to suitably chosen Wasserstein distances to time-homogeneous SDEs driven by symmetric $α$-stable processes. Notably, the driven noise in the limiting SDEs depends only on large jumps of the underlying additive processes. In case of $α\ge2$, a phase transition occurs and a diffusive phenomenon arises. In particular, in the setting $α>2$, we establish the ergodicity of the rescaled process by means of asymptotic pseudotrajectories. The resulting time-homogeneous limiting SDEs are driven by Brownian motions, even though the transformed time-inhomogeneous SDEs are driven by (discontinuous) additive processes, and the effective noise intensity is determined by the entire Lévy measure of the original pure-jump process. As far as the critical case $α=2 $ is concerned, we demonstrate that the noise intensity of the limiting SDEs driven by Brownian motions relies merely on the large-jump part of the Lévy measure.

math.PR↗

arXiv2026

Kairos: A Dataset for Fine-Grained Video-Language Modeling over Space, Time, and Dynamics

Many emerging video language modeling tasks require systems to move beyond clip-level abstraction and model visual content as it unfolds over extended time horizons. However, most existing video datasets rely on coarse or sparsely aligned supervision, which compresses temporal variation and limits the ability of models to learn reusable representations of continuous visual dynamics. We introduce Kairos, a video dataset for video-language modeling with time-resolved annotations. Kairos consists of long-duration videos, ranging from ten minutes to half an hour, annotated with fine-grained temporal alignment. The annotations capture ongoing actions, entity appearances and attributes, interactions, and evolving contextual cues along the video timeline. This time-resolved structure supports fine-grained evaluation, long-range modeling and reasoning, instruction data construction, representation learning, and video generation. Kairos provides a general-purpose foundation for modeling visual experiences over time.

cs.CV↗

arXiv2026

Jointly Optimizing Debiased CTR and Uplift for Coupons Marketing: A Unified Causal Framework

In online advertising, marketing interventions such as coupons introduce significant confounding bias into Click-Through Rate (CTR) prediction. Observed clicks reflect a mixture of users' intrinsic preferences and the uplift induced by these interventions. This causes conventional models to miscalibrate base CTRs, which distorts downstream ranking and billing decisions. Furthermore, marketing interventions often operate as multi-valued treatments with varying magnitudes, introducing additional complexity to CTR prediction. To address these issues, we propose the Unified Multi-Valued Treatment Network (UniMVT). Specifically, UniMVT disentangles confounding factors from treatment-sensitive representations, enabling a full-space counterfactual inference module to jointly reconstruct the debiased base CTR and intensity-response curves. To handle the complexity of multi-valued treatments, UniMVT employs an auxiliary intensity estimation task to capture treatment propensities and devise a unit uplift objective that normalizes the intervention effect. This ensures comparable estimation across the continuous coupon-value spectrum. UniMVT simultaneously achieves debiased CTR prediction for accurate system calibration and precise uplift estimation for incentive allocation. Extensive experiments on synthetic and industrial datasets demonstrate UniMVT's superiority in both predictive accuracy and calibration. Furthermore, real-world A/B tests confirm that UniMVT significantly improves business metrics through more effective coupon distribution.

cs.SI↗

arXiv2026

SEGA: Drivable 3D Gaussian Head Avatar from a Single Image

Creating photorealistic 3D head avatars from limited input has become increasingly important for applications in virtual reality, telepresence, and digital entertainment. While recent advances like neural rendering and 3D Gaussian splatting have enabled high-quality digital human avatar creation and animation, most methods rely on multiple images or multi-view inputs, limiting their practicality for real-world use. In this paper, we propose SEGA, a novel approach for Single-imagE-based 3D drivable Gaussian head Avatar creation that combines generalized prior models with a new hierarchical UV-space Gaussian Splatting framework. SEGA seamlessly combines priors derived from large-scale 2D datasets with 3D priors learned from multi-view, multi-expression, and multi-ID data, achieving robust generalization to unseen identities while ensuring 3D consistency across novel viewpoints and expressions. We further present a hierarchical UV-space Gaussian Splatting framework that leverages FLAME-based structural priors and employs a dual-branch architecture to disentangle dynamic and static facial components effectively. The dynamic branch encodes expression-driven fine details, while the static branch focuses on expression-invariant regions, enabling efficient parameter inference and precomputation. This design maximizes the utility of limited 3D data and achieves real-time performance for animation and rendering. Additionally, SEGA performs person-specific fine-tuning to further enhance the fidelity and realism of the generated avatars. Experiments show our method outperforms state-of-the-art approaches in generalization ability, identity preservation, and expression realism, advancing one-shot avatar creation for practical applications.

cs.GR↗

arXiv2026

Unveiling Hidden Threats: Using Fractal Triggers to Boost Stealthiness of Distributed Backdoor Attacks in Federated Learning

Traditional distributed backdoor attacks (DBA) in federated learning improve stealthiness by decomposing global triggers into sub-triggers, which however requires more poisoned data to maintian the attck strength and hence increases the exposure risk. To overcome this defect, This paper proposes a novel method, namely Fractal-Triggerred Distributed Backdoor Attack (FTDBA), which leverages the self-similarity of fractals to enhance the feature strength of sub-triggers and hence significantly reduce the required poisoning volume for the same attack strength. To address the detectability of fractal structures in the frequency and gradient domains, we introduce a dynamic angular perturbation mechanism that adaptively adjusts perturbation intensity across the training phases to balance efficiency and stealthiness. Experiments show that FTDBA achieves a 92.3% attack success rate with only 62.4% of the poisoning volume required by traditional DBA methods, while reducing the detection rate by 22.8% and KL divergence by 41.2%. This study presents a low-exposure, high-efficiency paradigm for federated backdoor attacks and expands the application of fractal features in adversarial sample generation.

cs.CR↗

arXiv2026

Fine-grained Distributed Backdoor Attacks in Federated Learning

Federated learning, as a privacy-preserving distributed machine learning paradigm, faces significant threats from backdoor attacks. Compared to centralized attacks, distributed backdoor attacks are more harmful but require more poisoned samples to compensate for the loss of trigger strength due to decomposition. Fixed trigger patterns are also easily detected by robust aggregation algorithms, increasing the risk of attack exposure. To address these challenges, we propose a fine-grained distributed backdoor attack framework (FDBA). This framework uses dynamic trigger generation and embedding vector optimization to perform attacks with fewer poisoned samples. First, we design a dynamic trigger generation method based on image edge structures using the Canny algorithm to extract edge features, which are then injected with Laplacian noise. RGB channel decomposition is applied for covert adaptation of the distributed trigger, reducing detection chances. Second, we introduce an embedding vector contrastive learning strategy that forces poisoned samples to approach the target class center in the feature space, enhancing attack effectiveness. On CIFAR-10, piecewise-linear estimates for target ASRs between 70\% and 90\% show that FDBA reduces the required poisoning ratio by 37.4\%--48.4\% compared with DBA. In non-independent and identically distributed (Non-IID) scenarios, FDBA retains 84.7\% of its IID attack performance under extreme heterogeneity, whereas DBA drops to 73.5\%, and the framework successfully bypasses mainstream defense mechanisms. This study offers new insights into federated learning security and emphasizes the potential threats and defense challenges posed by fine-grained distributed attacks.

cs.LG↗

arXiv2026

Lindbladian Phase Geometry and Hall Transport in Open Bloch Systems

Quantum geometry underlies a wide range of transport phenomena in Bloch systems. How quantum-geometric transport is modified when Bloch electrons are coupled to an environment, however, remains largely unexplored, even though the environment can alter both the electronic state and the physical current operator. Here we formulate dc linear response within a trace-preserving Lindblad kinetic theory by defining the physical velocity as the Liouvillian time derivative of the position operator. This construction reveals an environment-induced contribution to the current vertex whose momentum-space curl generates new Hall responses governed by the gauge-invariant phase geometry encoded by Lindblad jump amplitudes, which characterize electron-environment coupling. To leading order in the dissipative coupling, this geometry gives rise to interband shift-vector and diagonal-vorticity Hall responses arising from off-diagonal and diagonal jump amplitudes, respectively. Remarkably, both mechanisms can produce a finite Hall conductivity even when the conventional Berry-curvature anomalous Hall effect vanishes identically. Our work establishes Lindbladian phase geometry as an independent geometric origin of transverse transport in open quantum matter.

cond-mat.mes-hall↗

arXiv2026

ProtoRAG: Prototype-Based Retrieval Augmentation for Few-Shot Fine-Grained Remote Sensing Object Detection

Few-shot fine-grained object detection (FGOD) in remote sensing imagery is challenging because limited annotations must support both object localization and discrimination among visually similar subcategories. Although multimodal large language models (MLLMs) provide strong coarse object localization, they lack explicit visual evidence for reliable fine-grained recognition. To address this limitation, we propose ProtoRAG, a prototype-based retrieval-augmented framework that decouples coarse localization from fine-grained recognition by equipping MLLMs with an external object-level visual memory. To construct a reliable visual memory from limited support samples, we introduce Discriminative Prototype Space Learning (DPSL), which encourages discriminative and prototype-stable representations through supervised contrastive learning and prototype-consistency regularization. We further develop an uncertainty-guided candidate-constrained reasoning strategy that augments MLLMs with retrieved candidate-specific visual references and invokes multimodal reasoning only for ambiguous instances. Extensive experiments show that ProtoRAG consistently surpasses representative baselines in nine few-shot settings, outperforming the strongest baselines by 14.80, 2.27, and 4.04 mAP$_{50}$ on MAR20, HRSC2016, and FAIR1M-2.0, respectively.

cs.CV↗

arXiv2026

Higgs Boson Pair Production via Gluon Fusion: Higher-Order Corrections and Theoretical Uncertainties

In this contribution, the higher-order QCD and electroweak corrections to Standard Model Higgs boson pair production via the gluon-fusion mechanism, $gg\to hh$, are summarized and the different sources of theoretical uncertainty are assessed. The discussion includes finite top quark mass effects, matching to parton showers, approximate NNLO and N$^3$LO QCD corrections, NLO electroweak effects, and uncertainties associated with the top quark mass scheme and perturbative scale choices. In addition, we provide an updated state-of-the-art recommendation for the inclusive gluon-fusion Higgs boson pair production cross section and the corresponding Higgs boson pair invariant-mass distribution.

hep-ph↗

arXiv2026

Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models

Recent advances in world models have shown promise for modeling future dynamics of environmental states, enabling agents to reason and act without accessing real environments. Current methods mainly perform single-step or fixed-horizon rollouts, leaving their potential for complex task planning under-exploited. We propose Imagine-then-Plan (\texttt{ITP}), a unified framework for agent learning via lookahead imagination, where an agent's policy model interacts with the learned world model, yielding multi-step ``imagined'' trajectories. Since the imagination horizon may vary by tasks and stages, we introduce a novel adaptive lookahead mechanism by trading off the ultimate goal and task progress. The resulting imagined trajectories provide rich signals about future consequences, such as achieved progress and potential conflicts, which are fused with current observations, formulating a partially \textit{observable} and \textit{imaginable} Markov decision process to guide policy learning. We instantiate \texttt{ITP} with both training-free and reinforcement-trained variants. Extensive experiments across representative agent benchmarks demonstrate that \texttt{ITP} significantly outperforms competitive baselines. Further analyses validate that our adaptive lookahead largely enhances agents' reasoning capability, providing valuable insights into addressing broader, complex tasks. Our code and data will be publicly available at https://github.com/loyiv/ITP.

cs.CL↗

arXiv2026

Phase-Singularity Control of Topological Acoustic Spin Textures

Acoustic spin textures provide a degree of freedom for programmable topological field states, yet a unified framework for constructing and classifying higher-order textures in open-air acoustics remains lacking. Here we introduce polygonal standing-wave interference as a framework connecting discrete symmetry, angular phase encoding, phase singularities, chiral time-averaged energy flux, and acoustic spin. By controlling the number and relative phases of the standing-wave channels, we experimentally realize D4 meron/anti-meron lattice, D6 double-skyrmion superlattice, and D8 acoustic-spin quasicrystal composed of four meron/anti-meron sublattices. Finite polygonal fields are normalized angular discretizations of continuous standing-wave spectra. Under the m=2 phase encoding, the D6 and D8 fields reproduce the local q=+2 core of a second-order Bessel vortex while retaining distinct long-range order through their discrete interference channels. At fixed m increasing N under normalized angular sampling yields the continuous Bessel limit. These results establish phase encoding, pairwise interference, reciprocal-space composition, and angular discretization as a unified design principle for programmable topological acoustic spin textures.

cond-mat.mtrl-sci↗

arXiv2026

CoMAP: Co-Evolving World Models and Agent Policies for LLM Agents

Equipping language agents with world models enables them to anticipate environment dynamics and evaluate candidate actions before execution. However, existing textual world models are typically fixed after training, preventing them from adapting to the on-policy state-action distributions induced by an evolving agent. Meanwhile, agent-improvement methods often rely on external rewards or verifiers, limiting their applicability in realistic interactive environments. In this paper, we propose COMAP, a novel framework that co-evolves textual world models and agent policies through closed-loop interaction. At each decision step, the world model predicts future state feedback for candidate actions, and the agent performs future-aware reflection by estimating the reliability of this feedback and refining its action accordingly. The resulting on-policy trajectories are then used to update the world model via self-distillation, allowing it to better match the agent's evolving interaction distribution. Across embodied task planning, Web navigation, and tool-use benchmarks, COMAP consistently outperforms competitive baselines, e.g., +16.75% relative improvement with Qwen3-4B. Further analyses show that the co-evolutionary loop improves the world model's prediction accuracy over time and leads to more effective long-horizon decision-making. Our code is available at: https://github.com/loyiv/CoMAP.

cs.AI↗

arXiv2026

$L^\infty$-metrics on tori and Schoen's conjecture

We prove Schoen's conjecture on $L^\infty$-metrics for tori. More precisely, we show that any $L^\infty$-metric on a torus that is smooth and has non-negative scalar curvature away from an embedded submanifold with codimension at least three extends to a smooth flat metric. We also prove a LLarull-type theorem for $L^\infty$-metric. Our proof uses weighted scalar curvature and the relative index theorem.

math.DG↗

arXiv2026

An extensible software platform for automated operational and characterization testing of scientific cameras

Multi-detector programs such as Earth 2.0 (ET) require detector characterization across different technologies, operating conditions, and production batches, together with consistent procedures for operation, characterization, and device selection. Traditional laboratory workflows often distribute camera control, illumination, image analysis, and report generation across detector-specific programs and manual steps, slowing iterative development and hindering standardized characterization. We address this gap with PixelXVision, a layered software platform that combines a common camera SDK, preset-based configuration, and a runtime plugin system. Detector-specific behavior is configured through declarative presets, allowing the same host interface to operate CCD, CMOS, and infrared detectors without model-specific core changes. Characterization procedures are implemented as plugins that use a shared object manager to access camera, illumination, storage, and database services. Each run links raw frames, measured operating conditions, configuration, analysis products, and reports into a traceable archive with a queryable index. The platform supports lightweight deployment for iterative single-detector development and structured workflows for multi-detector campaigns. The platform has been applied in characterization testing across CMOS, CCD, and HgCdTe cameras, and its configurable architecture can accommodate additional camera models for subsequent CCD testing for WFST without model-specific changes to the core host software. For operational troubleshooting, the platform integrates ChatPixel, a knowledge-base assistant that provides answers with cited sources when it has relevant information.

physics.ins-det↗

arXiv2026

Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning

Large Language Models (LLMs) have shown impressive performance on complex tasks through Chain-of-Thought (CoT) reasoning. However, conventional CoT relies on explicitly verbalized intermediate steps, which constrains its broader applicability, particularly in abstract reasoning tasks beyond language. To address this, there has been growing research interest in \textit{latent CoT reasoning}, where the reasoning process is embedded within latent spaces. By decoupling reasoning from explicit language generation, latent CoT offers the promise of richer cognitive representations and facilitates more flexible, faster inference. This paper aims to present a comprehensive overview of this emerging paradigm and establish a systematic taxonomy. We analyze recent advances in methods, categorizing them from token-wise horizontal approaches to layer-wise vertical strategies. We then provide in-depth discussions of these methods, highlighting their design principles, applications, and remaining challenges. We hope that our survey provides a structured foundation for advancing this promising direction in LLM reasoning. The relevant papers will be regularly updated at https://github.com/EIT-NLP/Awesome-Latent-CoT.

cs.CL↗

arXiv2026

To Retrieve or To Think? Cross-Boundary Context Evolution for Multi-hop Complex Reasoning

Current context augmentation methods, such as retrieval-augmented generation, play a crucial role in bridging a model's internal knowledge boundary and external evidence for multi-hop reasoning. However, they often follow a rigid policy and treat external retrieval as the default action at each step. Such brute-force context expansion incurs unnecessary computational cost and may degrade reasoning performance by saturating the context with redundant or weakly relevant evidence. In this paper, we propose cross-boundary Context Evolution (EvoCtx), a framework that models complex reasoning as an adaptive process of boundary-aware context evolution. EvoCtx dynamically decides whether the next reasoning transition should cross the current evidence boundary through retrieval or refine the reasoning state within the existing context. It estimates the semantic gap between the reasoning state and the accumulated evidence, and strategically alternates between boundary expansion and intra-boundary trajectory refinement. This eliminates redundant retrieval steps and preserves a compact, evidence-supported reasoning trajectory. Extensive experiments on challenging open-domain and multi-hop QA benchmarks demonstrate that EvoCtx significantly outperforms previous methods, offering an effective approach to complex reasoning tasks. The source code can be accessed at https://github.com/Anya-RB-Chen/EvoCtx.

cs.CL↗

arXiv2026

FRAMEWORKERS: A Dynamic Multi-Agent Framework for AI-Generated Video Production

Modern video generators excel at synthesizing individual clips, but complete video production requires coordinating a long sequence of interdependent creative steps, including scripting, storyboarding, generation, and editing. It further demands persistent asset management and dynamic task orchestration as intermediate outputs, dependencies, and execution states evolve over time. Existing automated systems typically rely on rigid pipelines that are difficult to adapt to diverse inputs and changing workflows, while general-purpose large language models (LLMs) remain unreliable for long-horizon orchestration and multimodal asset routing. We introduce FRAMEWORKERS, a task-centric and workspace-grounded multi-agent framework for open-ended video production. A central Director formulates video creation as dynamic task management, continuously editing a Task Stack to determine which subtask to execute next and which sub-agent to invoke. An Assistant serves as the execution layer, grounding each selected task in a shared Workspace, retrieving the required assets and context, invoking the assigned sub-agent, and persisting the resulting artifacts. Execution capabilities are exposed through modular sub-agents with registered descriptors, allowing new sub-agents to be integrated without redesigning the orchestration workflow. To improve orchestration reliability, we fine-tune the Director via supervised fine-tuning (SFT) followed by Group Relative Policy Optimization (GRPO) for descriptor-conditioned task routing. Experiments show that FRAMEWORKERS outperforms strong LLM planners in routing accuracy, recovers reliably from runtime failures, generalizes to unseen sub-agents without retraining, and achieves higher end-to-end video quality and broader task coverage than fixed pipelines, single-agent systems, and prior multi-agent approaches.

cs.AI↗

arXiv2026

Regularity, quantitative deviation, and non-rigidity of a lacunary skew product

Let $α$ be irrational and let $q_j$ be the denominators of its continued-fraction convergents. We study the function \[ h(x)=\sum_{j\geq1}\frac{\cos(2πq_jx)}{q_j} \] and the skew product \[f(x,y)=(x+α,y+h(x))\quad\mathrm{mod}\quad\mathbb{Z}^2.\] The function $h$ is Hölder continuous of every exponent below one. A Fourier argument shows that $h$ is not Lipschitz. The map $f$ is a toral pseudo-rotation with rotation vector $(α,0)$, but it has neither bounded mean motion nor $C^0$-rigidity. Suppose $α$ satisfies the Diophantine condition $\mathrm{DC}(τ)$. Then, $f$ has $(C,1-1/τ)$-deviation when $τ>1$; and it has $(C_δ,δ)$-deviation for every $0<δ<1$, but not for $δ=0$ when $τ=1$.

math.DS↗