Search arXiv⌕ Search

arXiv subjects

Hang Zhou

Publications and source records attributed to Hang Zhou.

At least 73 records · Page 4Linked to original sources
arXiv2026

MOMA: Masked Orthogonal Matrix Alignment for Zero-Additional-Parameter Model Merging

Model merging offers a scalable alternative to multi-task learning but often yields suboptimal performance on classification tasks. We attribute this degradation to a geometric misalignment between the merged encoder and static task-specific classifier heads. Existing methods typically rely on auxiliary parameters to enforce strict representation alignment. We challenge this approach by revealing that the misalignment is predominantly an orthogonal transformation, rendering such strict alignment unnecessary. Leveraging this insight, we propose MOMA (Masked Orthogonal Matrix Alignment), which rectifies the misalignment by jointly optimizing a global multi-task vector mask and task-specific orthogonal transformations. Crucially, MOMA absorbs corresponding new parameters directly into the existing model weights, achieving performance comparable to state-of-the-art baselines with zero additional parameters and zero added inference cost.

cs.LG↗
arXiv2026

An Empirical Study of World Model Quantization

World models learn an internal representation of environment dynamics, enabling agents to simulate and reason about future states within a compact latent space for tasks such as planning, prediction, and inference. However, running world models rely on hevay computational cost and memory footprint, making model quantization essential for efficient deployment. To date, the effects of post-training quantization (PTQ) on world models remain largely unexamined. In this work, we present a systematic empirical study of world model quantization using DINO-WM as a representative case, evaluating diverse PTQ methods under both weight-only and joint weight-activation settings. We conduct extensive experiments on different visual planning tasks across a wide range of bit-widths, quantization granularities, and planning horizons up to 50 iterations. Our results show that quantization effects in world models extend beyond standard accuracy and bit-width trade-offs: group-wise weight quantization can stabilize low-bit rollouts, activation quantization granularity yields inconsistent benefits, and quantization sensitivity is highly asymmetric between encoder and predictor modules. Moreover, aggressive low-bit quantization significantly degrades the alignment between the planning objective and task success, leading to failures that cannot be remedied by additional optimization. These findings reveal distinct quantization-induced failure modes in world model-based planning and provide practical guidance for deploying quantized world models under strict computational constraints. The code will be available at https://github.com/huawei-noah/noah-research/tree/master/QuantWM.

cs.LG↗
arXiv2026

VersatileFFN: Achieving Parameter Efficiency in LLMs via Adaptive Wide-and-Deep Reuse

The rapid scaling of Large Language Models (LLMs) has achieved remarkable performance, but it also leads to prohibitive memory costs. Existing parameter-efficient approaches such as pruning and quantization mainly compress pretrained models without enhancing architectural capacity, thereby hitting the representational ceiling of the base model. In this work, we propose VersatileFFN, a novel feed-forward network (FFN) that enables flexible reuse of parameters in both width and depth dimensions within a fixed parameter budget. Inspired by the dual-process theory of cognition, VersatileFFN comprises two adaptive pathways: a width-versatile path that generates a mixture of sub-experts from a single shared FFN, mimicking sparse expert routing without increasing parameters, and a depth-versatile path that recursively applies the same FFN to emulate deeper processing for complex tokens. A difficulty-aware gating dynamically balances the two pathways, steering "easy" tokens through the efficient width-wise route and allocating deeper iterative refinement to "hard" tokens. Crucially, both pathways reuse the same parameters, so all additional capacity comes from computation rather than memory. Experiments across diverse benchmarks and model scales demonstrate the effectiveness of the method. The code is available at https://github.com/huawei-noah/noah-research/tree/master/VersatileFFN.

cs.CL↗
arXiv2026

Physics-informed Diffusion Mamba Transformer for Real-world Driving

Autonomous driving systems demand trajectory planners that not only model the inherent uncertainty of future motions but also respect complex temporal dependencies and underlying physical laws. While diffusion-based generative models excel at capturing multi-modal distributions, they often fail to incorporate long-term sequential contexts and domain-specific physical priors. In this work, we bridge these gaps with two key innovations. First, we introduce a Diffusion Mamba Transformer architecture that embeds mamba and attention into the diffusion process, enabling more effective aggregation of sequential input contexts from sensor streams and past motion histories. Second, we design a Port-Hamiltonian Neural Network module that seamlessly integrates energy-based physical constraints into the diffusion model, thereby enhancing trajectory predictions with both consistency and interpretability. Extensive evaluations on standard autonomous driving benchmarks demonstrate that our unified framework significantly outperforms state-of-the-art baselines in predictive accuracy, physical plausibility, and robustness, thereby advancing safe and reliable motion planning.

cs.RO↗
arXiv2026

Discovery of a gas-enshrouded broad-line AGN at z $\sim$ 7

The Lyman-alpha (Ly$α$) absorption profile in star-forming galaxies serves as a powerful tracer of the extended, dense neutral hydrogen in their surroundings during the Epoch of Reionization (EoR). We report a unique galaxy, A2744-z7DLA, at $z\approx 6.87$ gravitationally lensed by the foreground galaxy cluster Abell 2744, which exhibits both moderate Ly$α$ emission and damped Ly$α$ absorption, suggesting the presence of a dense neutral hydrogen environment. Our analysis suggests that the UV continuum turnover near Ly$α$ is more likely shaped by a damped Ly$α$ system rather than nebular continuum from two photon process. We analyze the physical properties of A2744-z7DLA with spectroscopic and photometric data from the JWST and the HST. The galaxy shows a compact morphology ($r_e \sim 0.3\ {\rm kpc}$) and a broadened H$α$ emission line, suggesting possible AGN activity. The broad component of H$α$ has a FWHM of $2721 \pm 200\ {\rm km\ s^{-1}}$, corresponding to a black hole mass of $M_{\rm BH}=2.90^{+2.35}_{-1.28}\times 10^7 M_\odot$ and a black hole-to-stellar mass ratio of $\log (M_{\rm BH}/M_{\rm stellar}) = -1.58^{+0.45}_{-0.34}$. The Balmer decrement ($\rm Hα/Hβ$) yields a dust attenuation of $\rm A_V \approx 1.15 \pm 0.23$, indicating that this system is less dust-rich than some "little red dots". Furthermore, we perform SED fitting using both stellar and AGN models. The results show that the UV and optical wavelengths are dominated by star-forming regions, while the AGN component contributes primarily at longer wavelengths. This work provides new insights into the interplay between star formation, neutral gas, and potential AGN activity in galaxies during the EoR.

astro-ph.GA↗
arXiv2026

Competent Discrete Time Modeling For analogue controlled PWM Converter Considering State-Feedback

Ever since R.D.Middlebrook proposed the state space averaging notion. The small signal model has been widely used as a design tool to tune control parameters. As Moore's law is continuing and the AI chip's high demand for power consumption and dynamic response, the control bandwidth needs to be boosted. However, the average model has two basic assumptions: the low-frequency assumption, the small ripple assumption. In high-bandwidth design, these two assumptions are violated. In order to solve this, various methods have been proposed. This paper gives a comprehensive overview of the existing small signal model for PWM converters from the following perspectives: 1. model fidelity, 2. analytical tractability. 3. complexity of the derivation process and result 4.generality.

eess.SY↗
arXiv2026

GA-S$^3$: Comprehensive Social Network Simulation with Group Agents

Social network simulation is developed to provide a comprehensive understanding of social networks in the real world, which can be leveraged for a wide range of applications such as group behavior emergence, policy optimization, and business strategy development. However, billions of individuals and their evolving interactions involved in social networks pose challenges in accurately reflecting real-world complexities. In this study, we propose a comprehensive Social Network Simulation System (GA-S3) that leverages newly designed Group Agents to make intelligent decisions regarding various online events. Unlike other intelligent agents that represent an individual entity, our group agents model a collection of individuals exhibiting similar behaviors, facilitating the simulation of large-scale network phenomena with complex interactions at a manageable computational cost. Additionally, we have constructed a social network benchmark from 2024 popular online events that contains fine-grained information on Internet traffic variations. The experiment demonstrates that our approach is capable of achieving accurate and highly realistic prediction results. Code is open at https://github.com/AI4SS/GAS-3.

cs.SI↗
arXiv2025

GestureHYDRA: Semantic Co-speech Gesture Synthesis via Hybrid Modality Diffusion Transformer and Cascaded-Synchronized Retrieval-Augmented Generation

While increasing attention has been paid to co-speech gesture synthesis, most previous works neglect to investigate hand gestures with explicit and essential semantics. In this paper, we study co-speech gesture generation with an emphasis on specific hand gesture activation, which can deliver more instructional information than common body movements. To achieve this, we first build a high-quality dataset of 3D human body movements including a set of semantically explicit hand gestures that are commonly used by live streamers. Then we present a hybrid-modality gesture generation system GestureHYDRA built upon a hybrid-modality diffusion transformer architecture with novelly designed motion-style injective transformer layers, which enables advanced gesture modeling ability and versatile gesture operations. To guarantee these specific hand gestures can be activated, we introduce a cascaded retrieval-augmented generation strategy built upon a semantic gesture repository annotated for each subject and an adaptive audio-gesture synchronization mechanism, which substantially improves semantic gesture activation and production efficiency. Quantitative and qualitative experiments demonstrate that our proposed approach achieves superior performance over all the counterparts. The project page can be found at https://mumuwei.github.io/GestureHYDRA/.

cs.MM↗
arXiv2025

State-Space Averaging Revisited via Reconstruction Operators

This paper presents an operator-theoretic reconstruction of an equivalent continuous-time LTI model from an exact sampled-data (Poincaré-map) baseline of a piecewise-linear switching system. The rebuilding is explicitly expressed via matrix logarithms. By expanding the logarithm of a product of matrix exponentials using the Baker--Campbell--Hausdorff (BCH) formula, we show that the classical state-space averaging (SSA) model can be interpreted as the leading-order truncation of this exact reconstruction when the switching period is small and the ripple is small. The same view explains why SSA critically relies on low-frequency and small-ripple assumptions, and why the method becomes fragile for converters with more than two subintervals per cycle. Finally, we provide a complexity-reduced, SSA-flavoured implementation strategy for obtaining the required spectral quantities and a real-valued logarithm without explicitly calling eigen-decomposition or complex matrix logarithms, by exploiting $2\times 2$ invariants and a minimal real-lift construction.

eess.SY↗
arXiv2025

Microwave Signature of the Emerging Abrikosov Lattice Above $H_{c2}$

The emergence of the Abrikosov lattice in the normal phase of type-II superconducting films when the magnetic field approaches the critical field $H_{c2}$ from above was predicted in Ref.~\cite{GVV2011}. In the quantum fluctuation regime \cite{GL2001} it is characterized by the formation of relatively large (with sizes of order $ξ_{\mathrm{QF}} \sim ξ_{\mathrm{BCS}}\sqrt{H_{c2}/(H-H_{c2})}$) ``long lived'' (lifetime of order $τ_{\mathrm{QF}} \sim \hbar Δ^{-1} H_{c2}/(H-H_{c2})$) clusters of rotating fluctuation Cooper pairs - signatures of developing Abrikosov vortices. We demonstrate that these fluctuation-induced vortex clusters, previously considered unobservable due to their ultrafast dynamics and weak (only logarithmically singular) contribution to the dc-conductivity, can in fact be detected through their distinct electromagnetic signature. By analyzing the high-frequency electromagnetic response of these rotating fluctuation Cooper pairs above the second critical field in superconducting film, we predict a pronounced and measurable enhancement in the imaginary part of the ac-conductivity arising directly from quantum fluctuations. This enhancement is expected to occur at characteristic frequencies $ω_{QF} \sim \hbar^{-1}Δ(H-H_{c2})/H_{c2}$, which are well below the superconducting threshold at $2\hbar^{-1}Δ$, where a similar increase in imaginary conductivity occurs in the superconducting phase. For niobium, a prototypical type II superconductor, $ω_{QF}$ lies in the experimentally accessible microwave range, making the effect directly testable with modern microwave spectroscopy.

cond-mat.supr-con↗
arXiv2025

ROOT: Robust Orthogonalized Optimizer for Neural Network Training

The optimization of large language models (LLMs) remains a critical challenge, particularly as model scaling exacerbates sensitivity to algorithmic imprecision and training instability. Recent advances in optimizers have improved convergence efficiency through momentum orthogonalization, but suffer from two key robustness limitations: dimensional fragility in orthogonalization precision and vulnerability to outlier-induced noise. To address these robustness challenges, we introduce ROOT, a Robust Orthogonalized Optimizer that enhances training stability through dual robustness mechanisms. First, we develop a dimension-robust orthogonalization scheme using adaptive Newton iterations with fine-grained coefficients tailored to specific matrix sizes, ensuring consistent precision across diverse architectural configurations. Second, we introduce an optimization-robust framework via proximal optimization that suppresses outlier noise while preserving meaningful gradient directions. Extensive experiments demonstrate that ROOT achieves significantly improved robustness, with faster convergence and superior final performance compared to both Muon and Adam-based optimizers, particularly in noisy and non-convex scenarios. Our work establishes a new paradigm for developing robust and precise optimizers capable of handling the complexities of modern large-scale model training. The code will be available at https://github.com/huawei-noah/noah-research/tree/master/ROOT.

cs.LG↗
arXiv2025

Real-Time Lane-Level Crash Detection on Freeways Using Sparse Telematics Data

Real-time traffic crash detection is critical in intelligent transportation systems because traditional crash notifications often suffer delays and lack specific, lane-level location information, which can lead to safety risks and economic losses. This paper proposes a real-time, lane-level crash detection approach for freeways that only leverages sparse telematics trajectory data. In the offline stage, the historical trajectories are discretized into spatial cells using vector cross-product techniques, and then used to estimate a vehicle intention distribution and select an alert threshold by maximizing the F1-score based on official crash reports. In the online stage, incoming telematics records are mapped to these cells and scored for three modules: transition anomalies, speed deviations, and lateral maneuver risks, with scores accumulated into a cell-specific risk map. When any cell's risk exceeds the alert threshold, the system issues a prompt warning. Relying solely on telematics data, this real-time and low-cost solution is evaluated on a Wisconsin dataset and validated against official crash reports, achieving a 75% crash identification rate with accurate lane-level localization, an overall accuracy of 96%, an F1-score of 0.84, and a non-crash-to-crash misclassification rate of only 0.6%, while also detecting 13% of crashes more than 3 minutes before the recorded crash time.

eess.SY↗
arXiv2025

SRA-CP: Spontaneous Risk-Aware Selective Cooperative Perception

Cooperative perception (CP) offers significant potential to overcome the limitations of single-vehicle sensing by enabling information sharing among connected vehicles (CVs). However, existing generic CP approaches need to transmit large volumes of perception data that are irrelevant to the driving safety, exceeding available communication bandwidth. Moreover, most CP frameworks rely on pre-defined communication partners, making them unsuitable for dynamic traffic environments. This paper proposes a Spontaneous Risk-Aware Selective Cooperative Perception (SRA-CP) framework to address these challenges. SRA-CP introduces a decentralized protocol where connected agents continuously broadcast lightweight perception coverage summaries and initiate targeted cooperation only when risk-relevant blind zones are detected. A perceptual risk identification module enables each CV to locally assess the impact of occlusions on its driving task and determine whether cooperation is necessary. When CP is triggered, the ego vehicle selects appropriate peers based on shared perception coverage and engages in selective information exchange through a fusion module that prioritizes safety-critical content and adapts to bandwidth constraints. We evaluate SRA-CP on a public dataset against several representative baselines. Results show that SRA-CP achieves less than 1% average precision (AP) loss for safety-critical objects compared to generic CP, while using only 20% of the communication bandwidth. Moreover, it improves the perception performance by 15% over existing selective CP methods that do not incorporate risk awareness.

cs.AI↗
arXiv2025

GRAM-R$^2$: Self-Training Generative Foundation Reward Models for Reward Reasoning

Significant progress in reward modeling over recent years has been driven by a paradigm shift from task-specific designs towards generalist reward models. Despite this trend, developing effective reward models remains a fundamental challenge: the heavy reliance on large-scale labeled preference data. Pre-training on abundant unlabeled data offers a promising direction, but existing approaches fall short of instilling explicit reasoning into reward models. To bridge this gap, we propose a self-training approach that leverages unlabeled data to elicit reward reasoning in reward models. Based on this approach, we develop GRAM-R$^2$, a generative reward model trained to produce not only preference labels but also accompanying reward rationales. GRAM-R$^2$ can serve as a foundation model for reward reasoning and can be applied to a wide range of tasks with minimal or no additional fine-tuning. It can support downstream applications such as response ranking and task-specific reward tuning. Experiments on response ranking, task adaptation, and reinforcement learning from human feedback demonstrate that GRAM-R$^2$ consistently delivers strong performance, outperforming several strong discriminative and generative baselines.

cs.CL↗
arXiv2025

CATS-V2V: A Real-World Vehicle-to-Vehicle Cooperative Perception Dataset with Complex Adverse Traffic Scenarios

Vehicle-to-Vehicle (V2V) cooperative perception has great potential to enhance autonomous driving performance by overcoming perception limitations in complex adverse traffic scenarios (CATS). Meanwhile, data serves as the fundamental infrastructure for modern autonomous driving AI. However, due to stringent data collection requirements, existing datasets focus primarily on ordinary traffic scenarios, constraining the benefits of cooperative perception. To address this challenge, we introduce CATS-V2V, the first-of-its-kind real-world dataset for V2V cooperative perception under complex adverse traffic scenarios. The dataset was collected by two hardware time-synchronized vehicles, covering 10 weather and lighting conditions across 10 diverse locations. The 100-clip dataset includes 60K frames of 10 Hz LiDAR point clouds and 1.26M multi-view 30 Hz camera images, along with 750K anonymized yet high-precision RTK-fixed GNSS and IMU records. Correspondingly, we provide time-consistent 3D bounding box annotations for objects, as well as static scenes to construct a 4D BEV representation. On this basis, we propose a target-based temporal alignment method, ensuring that all objects are precisely aligned across all sensor modalities. We hope that CATS-V2V, the largest-scale, most supportive, and highest-quality dataset of its kind to date, will benefit the autonomous driving community in related tasks.

cs.CV↗
arXiv2025

Unveiling Uniform Shifted Power Law in Stochastic Human and Autonomous Driving Behavior

Accurately simulating rare but safety-critical driving behaviors is essential for the evaluation and certification of autonomous vehicles (AVs). However, current models often fail to reproduce realistic collision rates when calibrated on real-world data, largely due to inadequate representation of long-tailed behavioral distributions. Here, we uncover a simple yet unifying shifted power law that robustly characterizes the stochasticity of both human-driven vehicle (HV) and AV behaviors, especially in the long-tail regime. The model adopts a parsimonious analytical form with only one or two parameters, enabling efficient calibration even under data sparsity. Analyzing large-scale, micro-level trajectory data from global HV and AV datasets, the shifted power law achieves an average R2 of 0.97 and a nearly identical tail distribution, uniformly fits both frequent behaviors and rare safety-critical deviations, significantly outperforming existing Gaussian-based baselines. When integrated into an agent-based traffic simulator, it enables forward-rolling simulations that reproduce realistic crash patterns for both HVs and AVs, achieving rates consistent with real-world statistics and improving the fidelity of safety assessment without post hoc correction. This discovery offers a unified and data-efficient foundation for modeling high-risk behavior and improves the fidelity of simulation-based safety assessments for mixed AV/HV traffic. The shifted power law provides a promising path toward simulation-driven validation and global certification of AV technologies.

eess.SY↗
arXiv2025

Benchmarking Tesla's Traffic Light and Stop Sign Control: Field Dataset and Behavior Insights

Understanding how Advanced Driver-Assistance Systems (ADAS) interact with Traffic Control Devices (TCDs) is critical for assessing their influence on traffic operations, yet this interaction has received little focused empirical study. This paper presents a field dataset and behavioral analysis of Tesla's Traffic Light and Stop Sign Control (TLSSC), a mature ADAS that perceives traffic lights and stop signs. We design and execute experiments across varied speed limits and TCD types, collecting synchronized high-resolution vehicle trajectory data and driver-perspective video. From these data, we develop a taxonomy of TLSSC-TCD interaction behaviors (i.e., stopping, accelerating, and car following) and calibrate the Full Velocity Difference Model (FVDM) to quantitatively characterize each behavior mode. A novel empirical insight is the identification of a car-following threshold (~90 m). Calibration results reveal that stopping behavior is driven by strong responsiveness to both desired speed deviation and relative speed, whereas accelerating behavior is more conservative. Intersection car-following behavior exhibits smoother dynamics and tighter headways compared to standard car-following behaviors. The established dataset, behavior definitions, and model characterizations together provide a foundation for future simulation, safety evaluation, and design of ADAS-TCD interaction logic. Our dataset is available at GitHub.

cs.RO↗