arXiv Science⌕ Search

arXiv subjects

Zhigang Wang

Publications and source records attributed to Zhigang Wang.

At least 19 recordsLinked to original sources
arXiv2026

OpenFlyV2: An onboard vision-language-action model for aerial perception and navigation

Human cognition flexibly draws on a shared perceptual foundation to understand environments, attend to referenced entities, reason about spatial relationships, and act. Endowing aerial robots with an analogous functional organization remains challenging, as existing systems are often specialized for individual capabilities, dependent on offboard computation, or unable to sustain closed-loop operation on resource-constrained platforms. We present OpenFlyV2, a unified onboard vision-language-action architecture that uses a shared vision-language representation to support semantic understanding, language-grounded perception, spatial reasoning, and continuous aerial control within a single model. A compact action expert converts the shared multimodal context into short-horizon flight waypoints, while motion-aware temporal memory and action priors reduce the computational cost of closed-loop execution. We construct \dataset{}, a simulation-based dataset comprising approximately 70,000 navigation trajectories and 60,000 vision-language instruction--response pairs. OpenFlyV2 generates action chunks, each comprising 12 waypoints, entirely onboard at 10 Hz using an NVIDIA Jetson Orin NX. It transfers zero-shot from simulation to physical quadrotor flight, achieving a 48\% navigation success rate across four real-world environments without real-world aerial training data. Extensive results demonstrate that broad vision-language competence and continuous closed-loop flight can coexist within a single model on a resource-constrained aerial platform. Code will be made publicly available at https://github.com/Ideal-ljl/AeroBrain.

cs.RO↗
arXiv2026

NAKO: A Benchmarking Framework for Evaluating and Diagnosing Neutral-Atom Quantum Compilers

Reconfigurable neutral-atom processors couple gate scheduling, atom placement, and parallel atom transport, so compiler decisions jointly determine physical execution costs. Evaluating these compilers requires attainable resource references and a way to trace excess costs to their causes. We present NAKO, a benchmarking framework that parameterizes circuits by qubit count, logical depth, and requested one- and two-qubit gate densities, and assigns attainable targets for Rydberg stages, routing intervals, and movement layers. Each benchmark is paired with an executable reference schedule; under the stated execution model, construction guarantees minimum Rydberg-stage and routing-interval counts, and minimum movement among schedules with the minimum Rydberg-stage count. Comparing compiler outputs with these references quantifies resource gaps, while reference mappings and intermediate physical states enable controlled interventions that trace excess costs to specific compiler decisions. In a 648-run evaluation of Enola and OLSQ-DPQA, NAKO identifies distinct bottlenecks. For Enola, placement scoring and cross-stage layout management contribute to excess movement; diagnosis-guided interventions reduce mean movement by 78.9% across five matched circuits. For OLSQ-DPQA, premature candidate selection and limited exploration of preparatory states produce avoidable entangling stages and routing intervals; four fixed-checkpoint continuations each eliminate one Rydberg stage and one routing interval without increasing movement. NAKO thus links resource targets backed by constructive optimality guarantees to controlled diagnosis and validation of compiler-specific resource savings.

quant-ph↗
arXiv2026

Physical AI: Concepts, Techniques, and Challenges

In recent years, physical artificial intelligence (physical AI) has experienced rapid development, capturing significant attention in academia and industry. However, the field has yet to reach consensus on its definition, technical landscape, and core challenges. To address this gap, this monograph seeks to define physical AI, present its technical landscape, and examine each key technical subfield in depth. In doing so, it establishes a unified framework for Physical AI: artificial systems that perceive, model, and act upon the physical world through continuous feedback under real-world constraints. Further, this monograph distinguishes physical AI from conventional digital AI, embodied AI, cyber-physical systems, and digital twins. It develops the foundations of physical AI across physics, control, cognition, learning, and reasoning. It reviews key methodologies, including multimodal perception, physical data and simulation, world models, embodied foundation models, motor-control systems, multi-agent coordination, and hardware-software co-design. Furthermore, it examines applications spanning robotics, manufacturing, infrastructure, scientific experimentation, and autonomous services. Particular attention is given to sim-to-real transfer, reliability, real-time execution, evaluation, safety, governance, and long-term deployment. By connecting perception, prediction, decision-making, action, and adaptation in a closed loop, this monograph identifies the technical challenges and research directions required to advance from narrow automation toward more generalizable physical AI and, ultimately, physical artificial general intelligence (Physical AGI).

cs.AI↗
arXiv2026

WAND: Learning Robust Navigation under Complex Wind Disturbances and Dense Obstacles for Quadrotors

Robust navigation in cluttered environments remains a fundamental challenge for quadrotors, particularly when strong wind disturbances arise, which perturb vehicle dynamics, limit control authority, and substantially increase collision risk. Existing learning-based navigation policies typically rely on obstacle perception and proprioceptive observations, requiring the policy to infer time-varying disturbance effects implicitly and thereby limiting robustness under partial observability. This paper proposes WAND (Wind-Aware Navigation with Disturbance Estimation), a reinforcement learning framework for navigation under time-varying wind disturbances in dense obstacle fields. Specifically, WAND estimates wind-induced disturbance acceleration from historical proprioceptive states using a Temporal Convolutional Network (TCN). This estimation is integrated into the policy via a zero-initialized residual module, \emph{WindAdapter}, while simultaneously providing feedforward compensation for low-level control. The dual use of the estimate couples disturbance-conditioned navigation with feedforward disturbance rejection. Across 12 wind-disturbed simulation settings, WAND improved the observed success rate by 8.3 percentage points on average relative to feedforward compensation alone. Controlled opposite-crosswind experiments further showed wind-direction-dependent trajectory adaptation. In indoor fan-induced flight tests, WAND succeeded in 18 of 20 trials, demonstrating the feasibility of real-time onboard navigation.

cs.RO↗
arXiv2026

MixVLA: Adaptive Mixing of Non-Invariant Information for Generalizable Vision-Language-Action Models

Vision-Language-Action (VLA) models have achieved remarkable advances in robotic manipulation, yet their zero-shot generalization under out-of-distribution (OOD) conditions remains limited. These models often entangle task-relevant invariant structure with environment-specific non-invariant factors, causing policies to rely on spurious appearance cues during action prediction. In this work, we propose \textbf{MixVLA}, a model-agnostic training framework that improves the generalization of VLA models without requiring additional OOD data or architectural modifications. The key component of MixVLA is \textbf{Adaptive Mixing of Non-Invariant Information (AMI)}. AMI stochastically mixes non-invariant representations to regularize distribution-specific variability while preserving complementary predictive cues. The mixed non-invariant features are then fused with invariant representations for final action prediction, resulting in improved robustness without sacrificing policy expressiveness. Extensive experiments across challenging manipulation settings, including LIBERO, LIBERO-Plus, the RoboTwin perturbation suite, and real-world tasks, demonstrate that MixVLA improves overall zero-shot robustness while retaining strong in-domain performance.

cs.RO↗
arXiv2026

OpenFlyScan: A Quality-Guided Aerial Reconstruction System for Consumer Drones

3D Gaussian Splatting (3DGS) provides high-fidelity scenes for large-scale embodied simulation, but constructing large-scale urban assets remains constrained by expensive equipment and delayed quality feedback. Preset surveys can leave complex surfaces insufficiently observed, with defects discovered only after reconstruction, requiring return visits and repeated processing. We present OpenFlyScan, a quality-guided aerial reconstruction system for consumer drones that integrates a GS quality model, a reacquisition planner, and a custom-designed mobile app. The model learns from GS rendering errors to predict regional reconstruction quality. Based on these predictions, the planner then generates complementary reacquisition strips to be executed through the app, which also supports automated oblique surveys and data transfer without additional hardware on board. Across real aerial scenes, the model effectively identifies regions that are likely to be poorly reconstructed. In the Expo West field experiment, targeted reacquisition improves PSNR at additional views by 10.95 dB. With consumer drones, OpenFlyScan integrates capture, targeted reacquisition, and reconstruction to support rapid, low-cost urban asset creation. Code and models will be made publicly available at https://openflyscan.github.io/.

cs.RO↗
arXiv2026

WOLF: World Model Guided LiDAR Exploration with Predictive Frontiers

LiDAR-based unmanned aerial vehicle (UAV) exploration builds maps by continually selecting where to observe next. However, decisions based on the measured map provide limited foresight into spatial continuations behind occlusions, leaving potentially informative directions unrecognized. We present WOLF, a world-model-guided framework that predicts future observations to enhance autonomous exploration. In the training stage, a recurrent world model learns observation dynamics from exploration trajectories, with recurrent memory retaining the spatial context needed to interpret partial observations across successive views. Building on this context, the model combines observation history with candidate motions during exploration to predict local occupancy and visibility. To guide further sensing, a predictive frontier generation mechanism then aligns and fuses these predictions using confidence, branch agreement, and observation quality to identify promising regions. The resulting predictive frontiers join measured ones to guide geometric viewpoint selection and trajectory generation, while new scans update subsequent predictions. In simulations, our method reduces mean terminal time by 10.9% relative to EPIC in Garage at comparable coverage and increases mean coverage from 42.12% to 98.35% in Tunnel. Real-world experiments further demonstrate onboard deployment of the learned model for online inference during physical flight.

cs.RO↗
arXiv2026

What is superatom?

The term "superatom" was introduced over three decades ago to describe clusters that emulate elemental atoms. The field has long been guided by the spherical jellium model, where magic numbers arise from shell closure of delocalized electrons. This Perspective argues that delocalization, not near-sphericity, is what makes a system atom-like. It shows that superatomic shell structure persists under arbitrary point-group symmetry, that superatomicity survives as a tunable quantum state across pressurized, ionized, and chemically precompressed systems, and that the symmetry rules governing superatoms are conditional, deeper than the jellium picture admits. The future of this field lies not in finding more magic numbers, but in exploiting superatomic states as artificial quantum systems at the atomic level.

physics.atm-clus↗
arXiv2026

Auto-RecSys: Harnessing Autonomous Research Agents for Industry-Scale Recommender System

Auto-research agents have shown the potential to automate hypothesis generation, experiment execution, and iterative refinement. However, scaling this paradigm to industry-scale recommendation models introduces two challenges: (1) long feedback loops, where model training can take days, making serial iteration prohibitively slow and requiring parallel exploration across multiple research directions; and (2) system complexity, where large configurations, fragile infrastructure dependencies, and multi-day GPU jobs require robust and recoverable execution. We present Auto-RecSys, an autonomous research system for long-horizon experimentation on industry-scale recommendation models. Auto-RecSys addresses these challenges through three harness designs: (1) distributed asynchronous execution for running multiple experiments in parallel across servers, (2) centralized cross-server memory for persistent and recoverable execution across sessions and failures, and (3) cognitive-procedural separation, where natural-language skill files guide LLM reasoning while deterministic scripts enforce operational correctness. Auto-RecSys further employs a dual-loop self-evolving architecture: an Execution Evolution Loop in which model-specific playbooks accumulate operational knowledge by recording failed attempts and crystallizing successful pipelines, and an Idea Evolution Loop in which experimental outcomes inform subsequent ideation. Evaluated on recommendation models, Auto-RecSys significantly reduces the human time required per experiment cycle and improves execution reliability as its playbooks mature.

cs.CL↗
arXiv2026

S2Dialog: Multimodal Dialogue Retrieval with Semantic and Acoustic-Style Modeling

Multimodal dialogue retrieval aims to retrieve dialogues from multimodal dialogue banks that are similar to a target dialogue in terms of both textual semantics and acoustic conversational styles. Such dialogue-level retrieval is crucial for many dialogue-related tasks, including Emotion Recognition in Conversation, Spoken Dialogue Systems, and Conversational Speech Synthesis, where external dialogue examples can provide valuable semantic and stylistic references. However, existing retrieval methods are still largely limited to utterance-level or unimodal matching, and often fail to capture the global semantic coherence and stylistic consistency of an entire dialogue. To address this gap, we propose S2Dialog, a unified framework for dialogue-level semantic-style retrieval from multimodal dialogue banks. Specifically, S2Dialog consists of a Dialogue-level Textual Retriever and a Dialogue-level Acoustic Retriever, which encode the textual and acoustic modalities of a dialogue into dialogue-level representations, respectively. To further enhance multimodal retrieval, we introduce Dialogue-level Textual-Acoustic Contrastive Learning, which aligns semantically and stylistically similar dialogues while distinguishing unrelated ones. Extensive experiments on the multimodal dialogue dataset DailyTalk demonstrate that S2Dialog achieves outstanding retrieval performance.

cs.CL↗
arXiv2026

Physical prior guided cooperative learning framework for joint turbulence degradation estimation and infrared video restoration

Infrared imaging and turbulence strength measurements are in widespread demand in many fields. This paper introduces a Physical Prior Guided Cooperative Learning (P2GCL) framework to jointly enhance atmospheric turbulence strength estimation and infrared image restoration. P2GCL involves a cyclic collaboration between two models, i.e., a TMNet measures turbulence strength and outputs the refractive index structure constant (Cn2) as a physical prior, a TRNet conducts infrared image sequence restoration based on Cn2 and feeds the restored images back to the TMNet to boost the measurement accuracy. A novel Cn2-guided frequency loss function and a physical constraint loss are introduced to align the training process with physical theories. Experiments demonstrate P2GCL achieves the best performance for both turbulence strength estimation (improving Cn2 MAE by 0.0156, enhancing R2 by 0.1065) and image restoration (enhancing PSNR by 0.2775 dB), validating the significant impact of physical prior guided cooperative learning.

eess.IV↗
arXiv2026

ARGON: A GNN-Empowered Compilation Framework for Scalable Neutral Atom Computing

Neutral atom quantum systems offer a promising pathway to large-scale quantum computing due to high qubit uniformity and flexible connectivity. To exploit this architecture, compilers must coordinate dynamic atom transport alongside highly parallel entangling gates. As circuits scale, the interplay between these operations becomes a system bottleneck, introducing denser logical interactions and longer temporal dependencies. Compilers must simultaneously satisfy rigid spatial constraints and complex movement schedules. Existing joint spatiotemporal compilation methods face an exponentially expanding search space, incurring substantial overheads or compromising fidelity as circuit size grows. In this work, we propose ARGON, a scalable compilation framework that introduces a spatiotemporal decoupling paradigm for neutral atom processors. Our key novelty is offloading static geometric conflict resolution to an offline phase, precomputing a library of hardware-certified, high-parallelism spatial layouts. To guide temporal routing, we deploy a Graph Neural Network (GNN) predictor to evaluate candidate layouts against deep temporal horizons, proactively evading downstream kinematic bottlenecks. Finally, a heuristic router translates the selected sequence into collision-free physical transport. Evaluations show ARGON completes compilation in under 10 seconds, delivering up to a >10^4x and 600x average speedup over state-of-the-art baselines. ARGON also minimizes routing decoherence and reduces Rydberg stages, improving execution fidelity by up to 10^2x on dense circuits.

cs.ET↗
arXiv2026

Resonant-impurity scanning tunneling spectroscopy in altermagnets: dual Fano resonance and Landau-quantization-induced nodal spin contrast

Using a Green's-function formalism, we study the spin-resolved local spectral function of a resonant impurity coupled to a two-dimensional $d$% -wave altermagnetic substrate. It is found that the interplay between direct tunneling from the impurity to the scanning tunneling microscopy (STM) tip and altermagnet-mediated tunneling gives rise to a dual Fano resonance in the absence of an external magnetic field. Moreover, the anisotropic spin-dependent oscillations of the local density of states and the corresponding Fano factors provide information on the altermagnetic splitting strength from complementary local and global perspectives. In addition, spin-selective tunneling can be achieved by tuning the Fermi energy and the tip position. In the presence of a strong magnetic field with Landau-level quantization, the dominant scanning tunneling spectroscopy (STS) signature appears as a spin-dependent nodal structure in real space: the nodal mismatch between opposite spin channels produces a large local spin contrast. These results establish resonant-impurity STM/STS as a phase-sensitive local probe of altermagnetic band anisotropy.

cond-mat.mes-hall↗
arXiv2026

Q-GeoMem: Question-Guided Geometric Memory for Video Spatial Reasoning

Video spatial reasoning requires accumulating viewpoint-dependent evidence over time while retaining information useful to the question being asked. Existing spatial video-language models improve geometric perception and long-range context modeling, but often treat memory as a generic temporal cache, which can introduce redundant or irrelevant evidence and weaken long-horizon reasoning. We propose Q-GeoMem, a question-guided geometric memory framework for video spatial reasoning. Q-GeoMem injects camera-conditioned geometry into visual tokens and maintains two complementary memories: a Fine-Grained Context Bank for recent dense features and camera states, and a Semantic-Geometric Evidence Bank for compact long-range evidence. For each candidate frame, a calibrated Q-Former estimates question relevance, while novelty and evidence utility are recomputed with respect to the active evidence bank. The resulting relevance-novelty utility controls capacity-based replacement and serves as an attention bias during memory reading. During reasoning, both memories are read before update and adaptively fused with the current frame representation. Extensive experiments across two in-domain and five out-of-distribution benchmarks, and controlled memory analyses show that Q-GeoMem achieves state-of-the-art performance in the evaluated settings and validate the effectiveness of question-guided geometric evidence selection.

cs.CV↗
arXiv2026

Hydrogen s-electrons as the origin of crystal magnetism beyond spin-orbit coupling

Magnetism has long been attributed to localized d, f, and even p electrons with strong correlations, whereas s electrons exemplified by hydrogen are reactive and tend to have their spins quenched, making s-electron-derived magnetism and long-range ordered magnetic crystals seem unattainable. Here we report a low-Z ferromagnetic crystal H13@(BN)12 using first-principles calculations, where thirteen hydrogen atoms are encapsulated within a (BN)12 cage and magnetism originates from the 1s electron of the central hydrogen atom. The crystal remains stable under ambient pressure owing to chemical precompression. Notably, the central hydrogen atom retains a magnetic moment of 1 μB, with long-range magnetic order established through multicenter bonding within the H13 aggregate and the intercell B-B network, while the zero orbital angular momentum of s electrons renders spin-orbit coupling (SOC) negligible as expected. Electronic structure analyses reveal that the large cavity and central negative electrostatic potential of the (BN)12 cage localize the hydrogen 1s electron, preventing spin quenching. Interestingly, under 16 GPa compression, the system transforms into a nonmagnetic metallic state driven by delocalized electrons of the central hydrogen atom. This study opens a pathway for constructing s-electron-driven magnetic materials and lays the foundation for developing low-energy consumption magnetic devices without SOC.

cond-mat.mtrl-sci↗
arXiv2026

Tunneling phase diagram: A machine-learning framework for multidimensional kinetic isotope effects

The kinetic isotope effect (KIE) is the conventional probe for quantum tunneling, yet its composite nature conflates tunneling with zero-point energy and classical kinetics. Here, we introduce the tunneling phase diagram, a machine-learning framework that decouples true tunneling strength by decoding the nonlinear relationship between KIE and the tunneling factor (\k{appa}). With exceptional fidelity (R^2 > 0.98, RMSE = 0.21), this framework reveals an anomalous high KIE-low \k{appa} spanning 300-600 K, thereby defining a paradigm for the quantitative assessment of quantum tunneling.

quant-ph↗
arXiv2026

Convex Dataset Valuation for Post-Training

Improving LLM performance on downstream tasks sometimes requires leveraging auxiliary datasets during post-training. In practice, however, developers face constraints on compute, labeling, and licensing costs that preclude using all available data, necessitating principled dataset-level selection. These constraints are increasingly shaped by dataset marketplaces, where data acquisition is governed by budgets and negotiation. We study dataset valuation as a subset selection problem during LLM post-training. Our goal is to identify and weight auxiliary datasets so as to maximize target task performance given constrained budgets. We first show that commonly used gradient alignment scores provide a reasonable yet incomplete valuation signal, as they ignore redundancy among datasets. To address this, we propose a scalable convex dataset-level valuation method based on kernel mean matching (KMM) in gradient space, which jointly accounts for alignment with the target task and redundancy across auxiliary datasets. Through extensive experiments across diverse post-training settings and tasks, we show that our approach consistently outperforms existing valuation baselines, achieving stronger performance with low computational overhead. Our results position dataset valuation as a practical decision tool for post-training data selection in market-constrained large language model settings. The code is available at https://github.com/uiuctml/convex_data_valuation.

cs.LG↗
arXiv2026

Efficient Diffusion as Low Light Enhancer

The computational burden of the iterative sampling process remains a major challenge in diffusion-based Low-Light Image Enhancement (LLIE). Current acceleration methods, whether training-based or training-free, often lead to significant performance degradation, highlighting the trade-off between performance and efficiency. In this paper, we identify two primary factors contributing to performance degradation: fitting errors and the inference gap. Our key insight is that fitting errors can be mitigated by linearly extrapolating the incorrect score functions, while the inference gap can be reduced by shifting the Gaussian flow to a reflectance-aware residual space. Based on the above insights, we design Reflectance-Aware Trajectory Refinement (RATR) module, a simple yet effective module to refine the teacher trajectory using the reflectance component of images. Following this, we introduce \textbf{Re}flectance-aware \textbf{D}iffusion with \textbf{Di}stilled \textbf{T}rajectory (\textbf{ReDDiT}), an efficient and flexible distillation framework tailored for LLIE. Our framework achieves comparable performance to previous diffusion-based methods with redundant steps in just 2 steps while establishing new state-of-the-art (SOTA) results with 8 or 4 steps. Comprehensive experimental evaluations on 10 benchmark datasets validate the effectiveness of our method, consistently outperforming existing SOTA methods.

cs.CV↗