arXiv subjects

Bin Qian

Publications and source records attributed to Bin Qian.

At least 19 recordsLinked to original sources

arXiv2026

VLA-Corrector: Stage-Aware Observable State Understanding for Prompt-Based Closed-Loop Recovery of Vision-Language-Action Policies

Long-horizon robot manipulation with Vision-Language-Action (VLA) policies remains vulnerable to execution-time deviations, as final task success provides little information for diagnosing and correcting failures caused by action noise, object displacement, or goal misalignment. We introduce a stage-aware failure verification and Prompt Recovery framework that enables closed-loop correction of a fixed VLA policy without parameter updates or privileged simulator states. The framework introduces an observable-history-based Learned Verifier that jointly estimates manipulation progress and execution risk by temporally modeling multi-view visual observations, proprioceptive states, and executed actions. To provide interpretable task understanding, we represent manipulation execution through semantic progress stages, including approach, alignment, grasp, transport, and placement, and identify stage-specific failure patterns. Upon detecting abnormal execution, the framework preserves the original instruction and generates a stage-conditioned recovery prompt, allowing the same frozen VLA policy to produce corrective actions. Extensive multi-round evaluations on LIBERO and LIBERO Plus demonstrate that the proposed approach substantially improves closed-loop reliability under diverse perturbations. Without access to privileged object or goal coordinates, the Learned Verifier achieves recovery performance close to that of the privileged rule-based verifier in the evaluated settings. These results show that observable visual-proprioceptive-action history is sufficient to infer latent task states and enable practical failure recovery for existing VLA policies.

cs.RO↗

arXiv2026

EMERGE-Policy: A Robot Mind Emerges Beyond a Single Policy

A robot's effective ``mind'' need not reside in a single policy. It can emerge when specialized components perceive, reason, predict, act, verify, and remember within a shared orchestration process. EMERGE-Policy turns this perspective into a graph-structured agentic framework that coordinates both capability invocation and information exchange. A Main Agent retains task-level state within an active context window, while role-specific Sub Agents process perception, execution monitoring, verification, and memory consolidation in isolated contexts and return structured, task-relevant evidence. Role-specific contexts control information load by exposing only decision-relevant evidence to the Main Agent, while the functional Skill interface composes heterogeneous backends as Operational, Imagination, and Evaluation Skills. Criterion-grounded verification, textual failure diagnosis, and Branch Stack recovery provide localized correction, with token-aware external memory preserving task-relevant state. Together, their closed-loop interaction realizes the system-level policy captured by the name EMERGE-Policy. Without additional fine-tuning, we achieved outstanding performance on several public benchmark that have had a wide-reaching impact, and conducted a series of real robot experiments. These system-level results suggest that through the division of different functional sub-tasks among multiple agents and their concurrent collaboration, as well as the technical paradigm where the model is regarded as a skill and called within the framework, EMERGE-Policy can extend the robust robot policies beyond isolated runs.

cs.RO↗

arXiv2026

ST-WAM: Semantic-Temporal World Action Model for Robust Manipulation under Visual Distribution Shifts

World Action Models (WAMs) have emerged as a promising paradigm by jointly modeling robot actions and future visual dynamics. However, their reliance on pixel-generative future supervision can entangle action-relevant state transitions with task-irrelevant visual content, limiting robustness under visual distribution shifts. We identify Training-Distribution Hallucination, a recurring phenomenon in which futures conditioned on visually shifted observations hallucinate training-domain content rather than remain faithful to the current scene. A controlled frame-triplet diagnosis further shows that DINOv3 features remain more stable across visual shifts while better preserving task-state distinctions than Wan-VAE latents. Rather than correcting the predicted futures, we propose Semantic-Temporal WAM (ST-WAM) to improve action robustness by using DINOv3 as a shared semantic representation for future prediction and history retrieval while retaining fine-grained VAE dynamics. Its Dual-Space Future Experts (DSFE) jointly predict future VAE latents and DINO features, while Current-Anchored Intent Retrieval (CAIR) retrieves task-relevant evidence from recent DINO history under the current visual-language context. ST-WAM is trained end-to-end without additional embodied pretraining or task-specific annotations, and requires no explicit future generation at inference. It achieves 98.7% on LIBERO and 92.8% on RoboTwin 2.0; more importantly, compared with Fast-WAM, it improves zero-shot LIBERO-Plus performance by 21.3 percentage points and more than doubles real-world success under visual shifts from 25.8% to 61.5%. These results demonstrate that semantic-temporal modeling effectively complements pixel-generative dynamics for robust manipulation.

cs.RO↗

arXiv2026

Temporal properties of the stochastic fractional heat equation with rough dependence in space

This paper investigates the nonlinear stochastic fractional heat equation driven by a Gaussian noise that is white in time and fractional in space with a Hurst parameter $H \in \big(\frac{3-\alpha}{4}, \frac{1}{2}\big)$. Specifically, the driving operator is the fractional Laplacian of order $\alpha/2 \in (1/2, 1)$. We characterize the asymptotic behavior of the temporal increment $u(t+\varepsilon,x)-u(t,x)$ for fixed $t\ge 0$ and $x\in\mathbb{R}$ as $\varepsilon\downarrow 0$. Utilizing these precise asymptotic estimates, we establish Khinchin's and Chung's laws of the iterated logarithm for the temporal process $t \mapsto u(t,x)$.

math.PR↗

arXiv2026

Large deviation principles for SPDEs with locally Lipschitz coefficients

Consider the stochastic partial differential equation, \begin{align*} \partial_t u^{\varepsilon}(t\,,x) = \frac{1}{2} \partial^2_x u^{\varepsilon}(t\,,x) + b(t\,,u^{\varepsilon}(t\,,x)) + \sqrt{\varepsilon}\sigma(t\,,u^{\varepsilon}(t\,,x)) \dot{W}(t\,,x), \end{align*} where $(t\,,x)\in(0\,,\infty)\times\mathbb{R}$, and $\dot{W}$ denotes space-time white noise. Foondun, Khoshnevisan, and Nualart \cite{FKN24} showed that this stochastic partial differential equation is well-posed under the assumptions that the initial condition $u(0)$ is bounded and measurable, while $b$ and $\sigma$ are locally Lipschitz continuous functions with at most linear growth. A Freidlin-Wentzell large deviation principle for the stochastic partial differential equation is established by a weak convergence approach in this paper.

math.PR↗

arXiv2026

Transportation cost inequalities for mean reflection SPDEs with white noise

We establish a quadratic transportation cost inequality under the uniform norm for solutions to mean reflected stochastic partial differential equations, a new type of equation in which the compensating reflection part depends not on the paths but on the law of the solution.

math.PR↗

arXiv2026

CF-VLA: Efficient Coarse-to-Fine Action Generation for Vision-Language-Action Policies

Flow-based vision-language-action (VLA) policies offer strong expressivity for action generation, but suffer from a fundamental inefficiency: multi-step inference is required to recover action structure from uninformative Gaussian noise, leading to a poor efficiency-quality trade-off under real-time constraints. We address this issue by rethinking the role of the starting point in generative action modeling. Instead of shortening the sampling trajectory, we propose CF-VLA, a coarse-to-fine two-stage formulation that restructures action generation into a coarse initialization step that constructs an action-aware starting point, followed by a single-step local refinement that corrects residual errors. Concretely, the coarse stage learns a conditional posterior over endpoint velocity to transform Gaussian noise into a structured initialization, while the fine stage performs a fixed-time refinement from this initialization. To stabilize training, we introduce a stepwise strategy that first learns a controlled coarse predictor and then performs joint optimization. Experiments on CALVIN and LIBERO show that our method establishes a strong efficiency-performance frontier under low-NFE (Number of Function Evaluations) regimes: it consistently outperforms existing NFE=2 methods, matches or surpasses the NFE=10 $\pi_{0.5}$ baseline on several metrics, reduces action sampling latency by 75.4%, and achieves the best average real-robot success rate of 83.0%, outperforming MIP by 19.5 points and $\pi_{0.5}$ by 4.0 points. These results suggest that structured, coarse-to-fine generation enables both strong performance and efficient inference. Our code is available at https://github.com/EmbodiedAI-RoboTron/CF-VLA.

cs.CV↗

arXiv2026

Stochastic fractional heat equation with general rough noise

Consider the following nonlinear one-dimensional stochastic fractional heat equation $$\frac{\partial }{\partial t}u(t, x)= -(-\Delta)^{\alpha/2}u(t, x) +\sigma(t,x,u(t,x)) \dot{W}(t, x), $$ where $-(-\Delta)^{\alpha/2}$ is the fractional Laplacian on $\mathbb R$ for $1<\alpha<2$, and $\dot{W}$ is a Gaussian noise that is white in time and behaves in space as a fractional Brownian motion with Hurst index $H$ satisfying $\frac{3-\alpha}{4}<H<\frac12$. When $\alpha=2$, Hu and Wang ({\it Ann. Inst. Henri Poincar\'e Probab. Stat.} {\bf 58} (2022) 379-423) studied the well-posedness of the solution and its H\"older continuity, removing the technical condition $\sigma(0)=0$ that was previously assumed in Hu et al. ({\it Ann. Probab.} {\bf 45} (2017) 4561-4616). Their approach relied on working in a weighted space with a suitable power decay function. For the case $\alpha\in (1,2)$, inspired by Hu and Wang, we investigate the well-posedness of the stochastic fractional heat equation without imposing the technical condition of $\sigma(0)=0$, which was required in the earlier work of Liu and Mao ({\it Bull. Sci. Math.} {\bf181} (2022) 103207). In our analysis, precise estimates of the heat kernel associated with the fractional Laplacian $-(-\Delta)^{\alpha/2}$ play a crucial role.

math.PR↗

arXiv2026

Enhancement of metallicity by Na doping in La$_3$Ni$_2$O$_{7+{\delta}}$

The observation of high-$T_c$ superconductivity in bilayer nickelate La$_3$Ni$_2$O$_7$ under high pressure provides a new venue for exploring novel unconventional superconductors and elucidating the mechanism of high-$T_c$ superconductivity. Subsequently, numerous chemical substitution studies have been reported, aiming to stabilize superconductivity at ambient pressure, or significantly reduce the pressure threshold required for its occurrence. Here, we report the comprehensive study on sodium (Na) doping in the Ruddlesden-Popper nickelate La$_3$Ni$_2$O$_{7+{\delta}}$, where Na$^+$ substitutes for La$^{3+}$ at the A-site with varying doping concentrations. The structural, thermal, magnetic, and electronic transport properties of as-synthesized polycrystalline samples were systematically investigated. X-ray diffraction (XRD) analysis reveals that Na doping induces a structural transition from the '327' Amam phase to the '4310' Bmab phase when $x\geq0.075$, which is further corroborated by thermogravimetric analysis (TGA) measurements. Substitution of La$^{3+}$ with Na$^+$ gives rise to a gradual expansion of the '327' phase lattice. Meanwhile, resistivity measurements indicate that the density wave (DW) transition is marginally suppressed and metallicity is significantly enhanced. Upon the application of pressure, DW transition can be further suppressed, whereas the low-$T$ insulating behaviors remain insensitive to pressure. These results offer critical insights into the roles of elemental substitution and charge carrier doping in steering the competing electronic phases in layered nickelates.

cond-mat.supr-con↗

arXiv2025

Reverse Thinking Enhances Missing Information Detection in Large Language Models

Large Language Models (LLMs) have demonstrated remarkable capabilities in various reasoning tasks, yet they often struggle with problems involving missing information, exhibiting issues such as incomplete responses, factual errors, and hallucinations. While forward reasoning approaches like Chain-of-Thought (CoT) and Tree-of-Thought (ToT) have shown success in structured problem-solving, they frequently fail to systematically identify and recover omitted information. In this paper, we explore the potential of reverse thinking methodologies to enhance LLMs' performance on missing information detection tasks. Drawing inspiration from recent work on backward reasoning, we propose a novel framework that guides LLMs through reverse thinking to identify necessary conditions and pinpoint missing elements. Our approach transforms the challenging task of missing information identification into a more manageable backward reasoning problem, significantly improving model accuracy. Experimental results demonstrate that our reverse thinking approach achieves substantial performance gains compared to traditional forward reasoning methods, providing a promising direction for enhancing LLMs' logical completeness and reasoning robustness.

cs.AI↗

arXiv2025

Temporal regularity for the stochastic heat equation with rough dependence in space

Consider the nonlinear stochastic heat equation $$ \frac{\partial u (t,x)}{\partial t}=\frac{\partial^2 u (t,x)}{\partial x^2}+ σ(u (t,x))\dot{W}(t,x),\quad t> 0,\, x\in \mathbb{R}, $$ where $\dot W$ is a Gaussian noise which is white in time and has the covariance of a fractional Brownian motion with Hurst parameter $H\in(\frac 14,\frac 12)$ in the space variable. When $σ(0)=0$, the well-posedness of the solution and its Hölder continuity have been proved by Hu et al. \cite{HHLNT2017}. In this paper, we study the asymptotic properties of the temporal gradient $u(t+\varepsilon, x)-u(t, x)$ at any fixed $t \ge 0$ and $x\in \mathbb R$, as $\varepsilon\downarrow 0$. As applications, we deduce Khintchine's law of iterated logarithm, Chung's law of iterated logarithm, and a result on the $q$-variations of the temporal process $\{u(t, x)\}_{t \ge 0}$, where $x\in \mathbb R$ is fixed.

math.PR↗

arXiv2025

Growth rates for the H\"older coefficients of the linear stochastic fractional heat equation with rough dependence in space

We study the linear stochastic fractional heat equation $$ \frac{\partial}{\partial t}u(t,x)=-(-\Delta)^{\frac{\alpha}2}u (t,x)+\dot{W}(t,x),\ \ t> 0,\ \ x\in\RR, $$ where $-(-\Delta)^{\frac{\alpha}{2}}$ denotes the fractional Laplacian with power $\alpha\in (1, 2)$, and the driving noise $\dot W$ is a centered Gaussian field which is white in time and has the covariance of a fractional Brownian motion with Hurst parameter $H\in\left(\frac {2-\alpha}2,\frac 12\right)$. We establish exact asymptotics for the solution as both time and space variables tend to infinity and derive sharp growth rates for the H\"older coefficients. The proofs are based on Talagrand's majorizing measure theorem and Sudakov's minoration theorem.

math.PR↗

arXiv2025

Q-learning-based Hierarchical Cooperative Local Search for Steelmaking-continuous Casting Scheduling Problem

The steelmaking continuous casting scheduling problem (SCCSP) is a critical and complex challenge in modern steel production, requiring the coordinated assignment and sequencing of steel charges across multiple production stages. Efficient scheduling not only enhances productivity but also significantly reduces energy consumption. However, both traditional heuristics (e.g., two-stage local search) and recent metaheuristics often struggle to adapt to the dynamic characteristics of practical SCCSP instances. To address these limitations, this paper introduces a novel Q learning based hierarchical cooperative local search framework, termed HierC_Q, aimed at minimizing the weighted sum of the maximum completion time and the average waiting time in SCCSP. The core contributions of HierC_Q are twofold. First, considering the intrinsic coupling properties of the SCCSP, a dedicated reward function is proposed based on a novel coupling measure (CM), guiding the exploration process towards promising regions of the solution space. Second, a hierarchical architecture is devised, comprising two distinct tiers: the learn to improve (L2I) tier and the "disturb to renovate" (D2R) tier. The L2I tier performs deep exploitation within promising regions using two independent Q-learning-based local search frameworks (QLSFs) tailored for subproblems, along with a synergy QLSF designed for the main problem. To enhance the effectiveness of local search, a validity evaluation approach and a speed-up evaluation method are also intro-duced, grounded in a detailed study of the problem's structure. Meanwhile, the D2R tier incorporates a perturbation and construction based solution renewal strategy to mitigate the risk of premature convergence. The superiority and effectiveness of HierC_Q are demonstrated through extensive comparisons with eleven local search frameworks and nine state-of-the-art algorithms.

eess.SY↗

arXiv2025

EMO: Edge Model Overlays to Scale Model Size in Federated Learning

Federated Learning (FL) trains machine learning models on edge devices with distributed data. However, the computational and memory limitations of these devices restrict the training of large models using FL. Split Federated Learning (SFL) addresses this challenge by distributing the model across the device and server, but it introduces a tightly coupled data flow, leading to computational bottlenecks and high communication costs. We propose EMO as a solution to enable the training of large models in FL while mitigating the challenges of SFL. EMO introduces Edge Model Overlay(s) between the device and server, enabling the creation of a larger ensemble model without modifying the FL workflow. The key innovation in EMO is Augmented Federated Learning (AFL), which builds an ensemble model by connecting the original (smaller) FL model with model(s) trained in the overlay(s) to facilitate horizontal or vertical scaling. This is accomplished through three key modules: a hierarchical activation replay cache to decouple AFL from FL, a convergence-aware communication controller to optimize communication overhead, and an ensemble inference module. Evaluations on a real-world prototype show that EMO improves accuracy by up to 17.77% compared to FL, and reduces communication costs by up to 7.17x and decreases training time by up to 6.9x compared to SFL.

cs.LG↗

arXiv2025

A Review on Edge Large Language Models: Design, Execution, and Applications

Large language models (LLMs) have revolutionized natural language processing with their exceptional understanding, synthesizing, and reasoning capabilities. However, deploying LLMs on resource-constrained edge devices presents significant challenges due to computational limitations, memory constraints, and edge hardware heterogeneity. This survey provides a comprehensive overview of recent advancements in edge LLMs, covering the entire lifecycle: from resource-efficient model design and pre-deployment strategies to runtime inference optimizations. It also explores on-device applications across various domains. By synthesizing state-of-the-art techniques and identifying future research directions, this survey bridges the gap between the immense potential of LLMs and the constraints of edge computing.

cs.DC↗

arXiv2024

Gradient bounds and Liouville property for a class of hypoelliptic diffusion via coupling

In this paper, we obtain the reverse Bakry-Émery type estimates for a class of hypoelliptic diffusion operator by coupling method. The (right and reverse) Poincaré inequalities and the (right and reverse) logarithmic Sobolev inequalities are presented as consequences of such estimates. Wang-Harnack inequality, Hamilton's gradient estimate and Liouville property are also presented by reverse logarithmic Sobolev inequality.

math.PR↗

arXiv2024

FLMarket: Enabling Privacy-preserved Pre-training Data Pricing for Federated Learning

Federated Learning (FL), as a mainstream privacy-preserving machine learning paradigm, offers promising solutions for privacy-critical domains such as healthcare and finance. Although extensive efforts have been dedicated from both academia and industry to improve the vanilla FL, little work focuses on the data pricing mechanism. In contrast to the straightforward in/post-training pricing techniques, we study a more difficult problem of pre-training pricing without direct information from the learning process. We propose FLMarket that integrates a two-stage, auction-based pricing mechanism with a security protocol to address the utility-privacy conflict. Through comprehensive experiments, we show that the client selection according to FLMarket can achieve more than 10% higher accuracy in subsequent FL training compared to state-of-the-art methods. In addition, it outperforms the in-training baseline with more than 2% accuracy increase and 3x run-time speedup.

cs.LG↗

arXiv2024

Generalized and high-efficiency arbitrary-positioned buffer for smoothed particle hydrodynamics

This paper develops an arbitrary-positioned buffer for the smoothed particle hydrodynamics (SPH) method, whose generality and high efficiency are achieved through two techniques. First, with the local coordinate system established at each arbitrary-positioned in-/outlet, particle positions in the global coordinate system are transformed into those in it via coordinate transformation. Since one local axis is located perpendicular to the in-/outlet boundary, the position comparison between particles and the threshold line or surface can be simplified to just this coordinate dimension. Second, particle candidates subjected to position comparison at one specific in-/outlet are restricted to those within the local cell-linked lists nearby the defined buffer zone, which significantly enhances computational efficiency due to a small portion of particles being checked. With this developed buffer, particle generation and deletion at arbitrary-positioned in- and outlets of complex flows can be handled efficiently and straightforwardly. We validate the effectiveness and versatility of the developed buffer through 2-D and 3-D non-/orthogonal uni-/bidirectional flows with arbitrary-positioned in- and outlets, driven by either pressure or velocity boundary conditions.

cs.CE↗