arXiv subjects

Yuan Xu

Publications and source records attributed to Yuan Xu.

At least 19 recordsLinked to original sources

arXiv2026

GazeVLA: Learning Human Intention for Robotic Manipulation

Embodied foundation models have significantly propelled the advancement of artificial general intelligence, yet they rely heavily on large-scale data. Recent studies identify human behavior as one of the most scalable sources for fostering robotic physical intelligence. However, the inherent embodiment gap remains a fundamental challenge for effective human-to-robot learning. In this work, we model human intention to enable efficient learning from human demonstrations. Among human behavioral cues, gaze simultaneously encodes high-level task semantics and exhibits a synergistic coupling with low-level action. By modeling intention through gaze, we derive a mid-level representation that facilitates cross-embodiment knowledge transfer. We demonstrate that robots can internalize human intentions to enhance manipulation, even without robotic intention annotations. To capture the synergy between intention and action, we adopt a Chain-of-Thought reasoning paradigm during inference. Our approach enables the execution of long-horizon and fine-grained tasks across diverse embodiments. Extensive experiments in both simulation and real-world demonstrate that our method consistently outperforms baseline approaches and generalizes to unseen settings. Project page: https://gazevla.github.io .

cs.RO↗

arXiv2026

WeSearch: A Unified Model-Centric Retrieval Framework for WeChat Search

We present WeSearch, a unified, model-centric retrieval framework for large-scale, continuously evolving content search. Prevailing methods assign documents fixed hierarchical Semantic IDs, implicitly assuming each document follows a single coherent semantic path-an assumption that often breaks down for multi-topic documents. WeSearch instead retrieves query-relevant clusters and then ranks atomic documents within the resulting query-specific candidate space, decoupling candidate pruning from fine-grained relevance modeling. Combined with relevance-aware training, compact document representations, and incremental index updates, this design enables weekly-or daily when needed-model refreshes. Deployed as the primary semantic retrieval path in WeChat Search alongside a lexical syntax tree, WeSearch shows consistent one-week A/B gains on corpora from hundreds of millions to billions of documents, and a one-month full-traffic A/B test confirms stable improvements of +0.68% CTR and +2.27% per-user article clicks, together with a 41% reduction in retrieval resource cost and a 19 ms reduction in graphic-text retrieval latency. WeSearch is now fully deployed in WeChat Search, serving 100% of production traffic.

cs.IR↗

arXiv2026

P^2-Gaussian: Prior-Driven and Projection-Guided Gaussian Splatting for Sparse-View Cone-Beam CT Reconstruction

Sparse-view cone-beam computed tomography (CBCT) reconstruction reduces imaging dose and scan time, but severe angular undersampling makes the inverse problem highly ill-posed. Existing 3D Gaussian Splatting (3DGS) methods initialize Gaussian primitives from Feldkamp-Davis-Kress (FDK) reconstructions that are contaminated by streak artifacts. The limited angular coverage also leaves the Gaussian parameters underconstrained, producing needle-like artifacts. We propose P2-Gaussian, a prior-driven and projection-guided 3DGS framework. In this framework, a latent rectified flow model, Sparse2Prior, first restores the artifact-corrupted FDK volume into an anatomical prior in approximately 5 s. Gaussian primitives are then initialized from this prior in a tissue-adaptive manner, which yields a compact representation with 62%-65% fewer optimized primitives than random sampling. During optimization, virtual projections synthesized at intermediate angles between the measured projections are activated in rotation, while measured projections remain dominant in the objective function. Experiments on 5, 10 and 20 measured views showed that the proposed P2-Gaussian achieved the highest PSNR and SSIM among all methods. Compared with the optimization-based Gaussian baseline, our method increases PSNR by 5.98, 4.25 and 2.49 dB and accelerates reconstruction by 2.6-3.0 times. These results indicate that coupling a learned anatomical prior with measurement-constrained Gaussian optimization is effective for sparse-view CBCT reconstruction. Code is publicly available at: https://github.com/Ballbo5354/P-2-Gaussian.

physics.med-ph↗

arXiv2026

AWARE: Action-Entropy-Weighted and Attention-Guided Retrieval of Observation Evidence for Vision-Language-Action Models

Vision-Language-Action (VLA) models combine multimodal perception with robotic control, demonstrating remarkable potential for generalizable manipulation. Yet, enhancing their execution typically relies on auxiliary sensory inputs (e.g., depth, 3D point clouds) or costly fine-tuning. Inspired by the insight that Feed-Forward Networks (FFNs) act as key-value memory, we propose Action-Entropy-Weighted and Attention-Guided Retrieval of Observation Evidence (AWARE), an effective, training-free, and plug-and-play module for VLAs. Specifically, AWARE tracks layer-wise uncertainty via an element-wise Action Entropy metric, and retrieves observation evidence into the element's FFN through associative memory lookup. Crucially, the retrieved features are weighted by instruction-guided cross-modal attention to isolate task-relevant observation cues, while injection strength is adaptively scaled by the element's entropy and relative network depth. This selectively channels observation evidence to high-entropy action dimensions and deep layers where grounding decays, while leaving converged representations intact. Across LIBERO, SIMPLER, CALVIN, and real-world robot manipulation, AWARE consistently improves diverse VLA architectures with minimal computational overhead, achieving the largest gains on long-horizon tasks without architectural changes or additional training. Project page: https://jiabingyang01.github.io/AWARE/.

cs.CV↗

arXiv2026

Understanding Human-Robot Collaboration through Joint Action

Human-robot collaboration requires partners to coordinate interdependent contributions toward a shared goal. Joint Action theory provides concepts for understanding this coordination, but how these concepts are instantiated and combined across empirical research remains fragmented. We conducted a systematic review following PRISMA 2020, identifying 83 empirical papers published between 2021 and 2026. Through theory-driven framework analysis, we organized the literature into nine constructs across three capacities: Representation, Prediction, and Integration. We analyze how these constructs are distributed and co-occur across studies, revealing where empirical evidence concentrates and where it is absent. By mapping empirical HRC research onto a unified Joint Action framework, this review offers a coherent overview of HRC studies, identifies gaps in current evidence, and provides guidelines for future research.

cs.HC↗

arXiv2026

Out of the box age estimation through facial imagery: A Comprehensive Benchmark of Vision-Language Models vs. out-of-the-box Traditional Architectures

Facial age estimation plays a critical role in content moderation, age verification, and deepfake detection. However, no prior benchmark has systematically compared modern vision-language models (VLMs) with specialized age estimation architectures. We present the first large-scale cross-paradigm benchmark, evaluating 34 models - 22 specialized architectures with publicly available pretrained weights and 12 general-purpose VLMs - across eight standard datasets (UTKFace, IMDB-WIKI, MORPH, AFAD, CACD, FG-NET, APPA-REAL, and AgeDB), totaling 1,100 test images per model. Our key finding is striking: zero-shot VLMs significantly outperform most specialized models, achieving an average mean absolute error (MAE) of 5.65 years compared to 9.88 years for non-LLM models. The best-performing VLM (Gemini 3 Flash Preview, MAE 4.32) surpasses the strongest non-LLM model (MiVOLO, MAE 5.10) by 15%. MiVOLO - unique in combining face and body features using Vision Transformers - is the only specialized model that remains competitive with VLMs. We further analyze age verification at the 18-year threshold and find that most non-LLM models exhibit false adult rates between 39% and 100% for minors, whereas VLMs reduce this to 16%-29%. Additionally, coarse age binning (8-9 classes) consistently increases MAE beyond 13 years. Stratified analysis across 14 age groups reveals that all models struggle most at extreme ages (under 5 and over 65). Overall, these findings challenge the assumption that task-specific architectures are necessary for high-performance age estimation and suggest that future work should focus on distilling VLM capabilities into efficient specialized models.

cs.CV↗

arXiv2026

Semantic Anchoring for Robotic Action Representations

Vision-Language-Action (VLA) models inherit rich semantic representations from pretrained Vision-Language Models, yet fine-tuning on limited robot demonstrations degrades this structure and undermines generalization. A fundamental question therefore arises: what constitutes a good action representation? Inspired by the mirror neuron theory's insight that observation and execution share an intention-level encoding, we examine whether a robot's action representations preserve the semantic structure captured by pretrained encoders. Systematic probing confirms that this structure erodes during finetuning, and that its quality synchronizes with both task success and out-of-distribution generalization. We further introduce a plug-and-play method that anchors action representations to a semantic manifold while decomposing representations into a shared semantic channel and a private channel, all discarded at inference, leaving the deployed model unchanged. Validated on different VLA backbones across simulation and real-world benchmarks, our method yields up to +18.7% on real-world in-distribution tasks and +21.5% on out-of-distribution generalization.

cs.RO↗

arXiv2026

Coupling Execution Horizon and Sampling Depth in Flow-Matching VLA Policies

Flow-matching Vision-Language-Action (VLA) models generate expressive continuous trajectories for robotic manipulation, yet balancing real-time responsiveness with execution stability remains a persistent challenge. Standard deployment workflows typically treat numerical solver depth and execution horizon as decoupled choices. The horizon amortizes inference latency, while solver depth is kept high to preserve integration fidelity. In this paper, we reveal a fundamental compute-execution coupling in flow-matching VLAs governed by a systematic sign reversal. While coarse numerical integration degrades precision under rapid replanning, it markedly improves task performance under extended open-loop commitment. Geometrically, numerical truncation contracts lateral action dispersion around an invariant trajectory centroid, serving as an implicit regularizer that suppresses compounding open-loop drift. Across LIBERO, RoboTwin 2.0, and real-world Franka manipulation on representative VLA backbones including $π_0$, $π_{0.5}$, and GR00T, exploiting this joint coupling consistently achieves superior task success while drastically lowering inference latency, entirely training-free and without architectural modifications. These findings establish numerical sampling depth as an operational variance regulator, shifting its role from a static precision budget to an active deployment knob in robot learning. Project page: https://jiabingyang01.github.io/Horizon-Depth-VLA/

cs.RO↗

arXiv2026

Relational Geometry Forcing for World Action Models

World Action Models (WAMs) have shown strong potential for robot manipulation by jointly learning visual dynamics and executable actions. However, their video and action objectives do not explicitly constrain the 3D geometry that is crucial for precise manipulation. Recently, geometry-aware WAMs introduce depth or 4D prediction branches, whereas representation alignment methods supervise intermediate WAM features with pretrained 3D foundation models. Yet we find that such pointwise alignment benefits the video features but degrades action learning, since dense geometric feature targets overly restrict compact action features. To address this limitation, we propose Relational Geometry Forcing (RGF), which transfers geometric priors from a frozen 3D foundation model by aligning pairwise feature relations instead of pointwise features. Specifically, RGF applies three complementary relational objectives, where intra-frame and cross-view relations transfer scene geometry and multi-view correspondence to the video features, while action-temporal relations couple action dynamics to geometric scene evolution. Applied only during training, RGF requires no explicit geometry branch or extra inference module. Extensive experiments in simulation and the real world demonstrate the effectiveness of Relational Geometry Forcing, showing consistent gains across diverse manipulation scenarios and WAM paradigms.

cs.RO↗

arXiv2026

Negative Sobolev regularity and optimal decay for the fractional Navier--Stokes--Poisson system

We establish global well-posedness and sharp higher-order decay estimates for the fractional compressible Navier--Stokes--Poisson system in $\R^{3}$ with fractional Lamé--Navier dissipation $\Lame^αu$ with $α\in[1/2,1]$, and Riesz coupling $Λ^{2β}ϕ=ρ-\barρ$ with $β\in[0,1]$, where $Λ:=(-Δ)^{1/2}$ and $\barρ>0$ denotes the constant background density. For small perturbations of a constant equilibrium in $H^{Nα}$ with $N\geq3$ and $Nα>5/2$, we prove the global existence and uniqueness of classical solutions. Furthermore, we propagate the negative Sobolev norms with uniform-in-time bounds and derive the higher-order decay estimates, provided that the initial perturbation additionally belongs to $\dot{H}^{-\vartheta}$ for some $\vartheta\in[0,3/2)$. Moreover, we establish the optimality of these decay rates within the prescribed initial data class for the velocity and the Riesz potential in their natural energy norms. The density decay exponents are also optimal whenever the available regularity permits the full Riesz-induced gain. The analysis relies on a nonlinear energy structure adapted to both fractional dissipation and Riesz interaction, together with Sobolev interpolation. The optimality of the decay is proved by constructing smooth initial data localized at low frequencies and proving matching upper and lower decay bounds for the resulting solutions. These solutions show that the low-frequency linear dynamics including the Riesz-induced density gain, persists under nonlinear evolution.

math.AP↗

arXiv2026

Quantum-enhanced Search for Galactic Dark Photons Using Cat-State Cavities

We present a quantum-enhanced search for the Galactic dark-photon background (GDPB), a relativistic and anisotropic vector field that can be produced by dark-matter decay or annihilation in the Milky Way halo. We reinterpret an existing quantum-enhanced dark-photon dataset obtained with a superconducting circuit QED platform originally developed for wave-like dark-photon dark-matter searches, consisting of a high-$Q$ niobium cavity in the $\mathrm{TM}_{010}$ mode dispersively coupled to a transmon qubit. By preparing four-component cat states and performing repeated quantum non-demolition parity measurements, analyzed with a hidden Markov model, we extract photon-transition probabilities for multiple cat amplitudes and integration times. The coupling to the GDPB is governed by a polarization- and angle-dependent cavity overlap function, which predicts a distinctive diurnal modulation of the excitation rate for both longitudinal and transverse polarizations. From the absence of such modulation, we set 90\% confidence-level bounds on the kinetic-mixing parameter $ε$ for two benchmark production mechanisms (parametric and cascade decay) over the relativistic mass range $m_X \ll ω_c$.

hep-ph↗

arXiv2026

What Matters for World Modeling in Vision-Language-Action Models?

Integrating world models into Vision-Language-Action models (VLAs) has emerged as a promising direction, enabling policies to anticipate future dynamics beyond reactive observation-to-action mapping. However, existing approaches vary simultaneously in their world representations, integration paradigms, and policy backbones, making it difficult to determine what truly matters for effective world model integration. To disentangle these factors, we categorize existing efforts into four representative paradigms, namely Implicit Future Alignment, Explicit Future Query, MoT Future Expert, and External World Predictor. We instantiate each within a unified framework and systematically evaluate these paradigms with diverse world representations and policy backbones across in-domain, out-of-domain, long-horizon, dynamic, few-shot, and real-world settings. Our study yields four main findings: 1) world representation effectiveness is paradigm-dependent, whereas 2D semantic representations offer robust, cross-paradigm gains; 2) no single paradigm is universally optimal, yet the decoupled MoT Future Expert offers competitive overall performance, particularly in OOD robustness; 3) world modeling consistently benefits policies built on standard VLMs, while decoupled designs exhibit greater compatibility with pretrained VLAs; and 4) world modeling improves performance over action-only baselines across multiple settings, with notable gains on long-horizon and dynamic tasks. Project page and video demonstrations are available at https://dylanmqs.github.io/WM-in-VLAs/.

cs.RO↗

arXiv2026

On-chip nanoplasma for adaptive electromagnetic protection

Over the past decade, semiconductor diodes have served as the primary switching elements in adaptive electromagnetic (EM) protection, yet their performance has been compromised by parasitic effects and thermal accumulation, rendering them inadequate against the rapidly evolving landscape of high-power microwave (HPM) threats. Here we show that on-chip nanoplasma switches (NPMS), composed of gallium nitride electrodes on silicon carbide substrates, exhibit superior radio frequency (RF) and thermal characteristics, positioning them as ideal field-driven switches in RF front-end protectors. By integrating NPMS into metasurfaces, antennas and circuit limiters, we achieve an adaptive response that ensures low-loss transmission for normal signals and high shielding against HPMs, while offering extended operating bandwidth and substantially higher tolerance than conventional solid-state devices. This robust, nanoscale structure has significant potential for protecting unmanned aerial vehicles, radars, satellites and other highly integrated platforms requiring strength and stability in EM environments. The findings of this study open up new routes to support EM safety of high-precision detection and imaging for next-generation RF front ends, with straightforward scalability to millimetre-wave and terahertz frequencies.

physics.app-ph↗

arXiv2026

$L^{2}-L^{\infty}$ decay estimates and inviscid limits for Global smooth solutions to the compressible Navier-Stokes-Riesz system

We study the Cauchy problem in $\mathbb{R}^{3}$ for the repulsive compressible Navier-Stokes-Riesz system with Riesz exponent $0<s<1$ and viscosity $0<\varepsilon\leq1$, where the Riesz interaction $\nabla(-Δ)^{-s}(ρ-\barρ)$ is a generalization of the Coulomb interaction for electrons. For small perturbations of a constant equilibrium, with the solenoidal component of the initial velocity of order $\mathcal{O}(\varepsilon)$, we prove the global existence and uniqueness of smooth solutions. We derive time-decay estimates in $L^{2}$ norms and $L^{\infty}$ norms that capture both uniform-in-$\varepsilon$ dispersive behavior and viscosity-dependent dissipation. We further establish a global-in-time inviscid limit to the irrotational global solution of the compressible Euler-Riesz system whose initial data consist of the same density and the curl-free component of the velocity, with an explicit convergence rate in $W^{k,p}$ norms. The proof combines viscosity-adapted dispersive estimates, normal-form analysis and nonlinear energy estimates with control of both negative and positive Sobolev norms.

math.AP↗

arXiv2026

LM-X: Explainable Vision--Language--Action Modeling via Progress, Event, and Uncertainty Prediction

Large-scale vision--language--action (VLA) policies have advanced generalist robot control, yet most remain stimulus-to-action black boxes: actions are exposed, but their explanatory state is not. They provide no native account of three explanatory signals: task progress, the next semantic transition, or local command reliability. Prior work shows that progress and event structure aid long-horizon control and that uncertainty supports monitoring; however, such capabilities are typically added or extracted only after action pretraining. The field therefore lacks a VLA foundation model whose explanatory state is jointly pretrained with control. Drawing on biological sensorimotor organization, in which outcome-sensitive, event-segmented, and probabilistic predictions structure behavior, we introduce LM-X. LM-X learns three directly supervised online signals: return-to-go (RTG) estimates visible progress and state quality; event-to-go (ETG) predicts the action sequence to the next semantic event; and heteroscedastic action-flow variance reports local command reliability. RTG conditions ETG and both condition action generation; uncertainty is estimated inside the action expert, making explanation part of control rather than a post-hoc description. We pretrain LM-X on more than 20,000 hours of heterogeneous real-robot trajectories, including over 1,000 hours of failed rollouts. LM-X achieves 74.1\% success on 50 randomized-hard RoboTwin2.0 tasks and 73.5\% on seven real-robot tasks, compared with 55.4\% and 50.7\% for GR00T N1.7. Its signals track progress and regression, anticipate event-scale motion, detect high-error actions, and provide advance failure warning. These results establish LM-X as an explainable VLA foundation model that couples transparent predictive state with stronger generalist control. Github: https://github.com/loongOpen/LoongWu-LM-X-VLA

cs.RO↗

arXiv2026

ViDAL: A Visual Dynamics-Grounded Action Latent Space for Vision-Language-Action Models

Vision-Language-Action (VLA) models have become a central paradigm for robot policy learning, which predict actions in three forms: raw action chunks, discrete action tokens, or continuous action latents. However, existing action representations primarily model action trajectories, with limited consideration of the visual dynamics induced by these actions. We introduce ViDAL, a Visual Dynamics-grounded Action Latent Space that anchors continuous action latents in the future visual dynamics of the scene. Specifically, ViDAL learns action latent space by training an Action Variational Autoencoder (Action VAE) to reconstruct action chunks while aligning its latent with future scene dynamics. When integrated into downstream robot policies, the proposed Action VAE serves as a plug-in action interface compatible with multiple VLA architectures and enables optional future-video prediction as an additional capability. Empirically, ViDAL outperforms competitive baselines on LIBERO with 98.1% average success, improves a multi-task $π_{0.5}$ policy on RoboTwin 2.0 from 54.3% to 65.5% (Clean) and from 33.2% to 43.1% (Random) success rates over 50 dual-arm tasks, and yields 20.0% and 23.4% absolute success-rate gains on real-world single-arm Franka and dual-arm ARX robot platforms.

cs.RO↗

arXiv2026

Addition formula for two families of orthogonal polynomials in several variables

An addition formula is a closed-form formula for reproducing kernels of orthogonal polynomials of several variables; the latter are kernels of orthogonal projection operators. We derive addition formulas for two families of orthogonal polynomials, which consist of wrapped products of classical Gegenbauer polynomials on the unit ball with either classical Jacobi polynomials on the simplex or product Laguerre polynomials, each of which is given explicitly as a multilayer of integrals of a classical orthogonal polynomial of one variable. These formulas provide a powerful tool for analysis and allow us, in particular, to reduce a substantial portion of analysis on the Fourier orthogonal expansions on the domains to that of one variable.

math.CA↗

arXiv2026

Physically Grounded Monocular Depth via Nanophotonic Wavefront Encoding

Depth foundation models (DFMs) offer strong learned priors for 3D perception from single RGB images but lack physical depth cues, leading to ambiguities in metric scale. We introduce metalenses, an emerging class of ultrathin planar optical elements, as a solution to physically encode missing metric depth cues via nanophotonics. In this paper, we bridge the gap between metalens and DFMs to achieve accurate metric monocular depth sensing. In a single monocular shot, our metalens embeds depth-dependent positional shifts into two polarized optical wavefronts. With an input adaptation strategty, we enable direct fine-tuning that aligns a pretrained DFM with the optical signals. To scale the training data, we further develop a comprehensive simulation pipeline that synthesizes metalens responses from RGB-D datasets, incorporating physical factors to minimize the sim-to-real gap. Experiments demonstrate that this approach outperforms both monocular metric depth estimation and depth-from-defocus baselines, showing an effective pathway for accurate monocular metric depth sensing.

physics.optics↗