arXiv subjects

Yuhang Li

Publications and source records attributed to Yuhang Li.

At least 37 records · Page 2Linked to original sources

arXiv2026

Mamba Learns in Context: Structure-Aware Domain Generalization for Multi-Task Point Cloud Understanding

While recent Transformer and Mamba architectures have advanced point cloud representation learning, they are typically developed for single-task or single-domain settings. Directly applying them to multi-task domain generalization (DG) leads to degraded performance. Transformers effectively model global dependencies but suffer from quadratic attention cost and lack explicit structural ordering, whereas Mamba offers linear-time recurrence yet often depends on coordinate-driven serialization, which is sensitive to viewpoint changes and missing regions, causing structural drift and unstable sequential modeling. In this paper, we propose Structure-Aware Domain Generalization (SADG), a Mamba-based In-Context Learning framework that preserves structural hierarchy across domains and tasks. We design structure-aware serialization (SAS) that generates transformation-invariant sequences using centroid-based topology and geodesic curvature continuity. We further devise hierarchical domain-aware modeling (HDM) that stabilizes cross-domain reasoning by consolidating intra-domain structure and fusing inter-domain relations. At test time, we introduce a lightweight spectral graph alignment (SGA) that shifts target features toward source prototypes in the spectral domain without updating model parameters, ensuring structure-preserving test-time feature shifting. In addition, we introduce MP3DObject, a real-scan object dataset for multi-task DG evaluation. Comprehensive experiments demonstrate that the proposed approach improves structural fidelity and consistently outperforms state-of-the-art methods across multiple tasks including reconstruction, denoising, and registration.

cs.CV↗

arXiv2026

Interleaved diffractive networks for information transfer through random diffusers

Transferring optical information through random diffusers is a critical yet challenging task. In this work, we introduce a cascaded diffractive optical network for information transfer through random and unknown diffusers, achieved through a series of passive, structured layers physically interleaved within the scattering medium. These interleaved diffractive layers are optimized to mitigate the scattering process without requiring digital computing. The performance of this all-optical system was quantified as a function of several physical parameters, including the diffractive processor's depth, the physical layout of the diffractive layers, and the statistical properties of the scattering medium. To further enhance the performance and robustness of information transfer through a scattering medium, we also developed a hybrid (optical-digital) system that coupled the diffractive processor with a jointly trained digital neural network, which was shown to achieve superior reconstruction fidelity even when the input object information was subjected to unknown random rotations, shifts, and scaling. We also experimentally validated this system using a fabricated multi-layer diffractive system in the visible spectrum, demonstrating reliable information recovery through random diffuser layers. Our numerical and experimental results demonstrate the capabilities of an interleaved diffractive processor architecture to recover optical information through volumetric diffusive media, which can find applications in biomedical imaging, telecommunications, and remote sensing.

physics.optics↗

arXiv2026

TT-SNN: Tensor Train Decomposition for Efficient Spiking Neural Network Training

Spiking Neural Networks (SNNs) have gained significant attention as a potentially energy-efficient alternative for standard neural networks with their sparse binary activation. However, SNNs suffer from memory and computation overhead due to spatio-temporal dynamics and multiple backpropagation computations across timesteps during training. To address this issue, we introduce Tensor Train Decomposition for Spiking Neural Networks (TT-SNN), a method that reduces model size through trainable weight decomposition, resulting in reduced storage, FLOPs, and latency. In addition, we propose a parallel computation pipeline as an alternative to the typical sequential tensor computation, which can be flexibly integrated into various existing SNN architectures. To the best of our knowledge, this is the first of its kind application of tensor decomposition in SNNs. We validate our method using both static and dynamic datasets, CIFAR10/100 and N-Caltech101, respectively. We also propose a TT-SNN-tailored training accelerator to fully harness the parallelism in TT-SNN. Our results demonstrate substantial reductions in parameter size (7.98X), FLOPs (9.25X), training time (17.7%), and training energy (28.3%) during training for the N-Caltech101 dataset, with negligible accuracy degradation.

cs.NE↗

arXiv2026

RAIN: Secure and Robust Aggregation under Shuffle Model of Differential Privacy

Secure aggregation is a foundational building block of privacy-preserving learning, yet achieving robustness under adversarial behavior remains challenging. Modern systems increasingly adopt the shuffle model of differential privacy (Shuffle-DP) to locally perturb client updates and globally anonymize them via shuffling for enhanced privacy protection. However, these perturbations and anonymization distort gradient geometry and remove identity linkage, leaving systems vulnerable to adversarial poisoning attacks. Moreover, the shuffler, typically a third party, can be compromised, undermining security against malicious adversaries. To address these challenges, we present Robust Aggregation in Noise (RAIN), a unified framework that reconciles privacy, robustness, and verifiability under Shuffle-DP. At its core, RAIN adopts sign-space aggregation to robustly measure update consistency and limit malicious influence under noise and anonymization. Specifically, we design two novel secret-shared protocols for shuffling and aggregation that operate directly on additive shares and preserve Shuffle-DP's tight privacy guarantee. In each round, the aggregated result is verified to ensure correct aggregation and detect any selective dropping, achieving malicious security with minimal overhead. Extensive experiments across comprehensive benchmarks show that RAIN maintains strong privacy guarantees under Shuffle-DP and remains robust to poisoning attacks with negligible degradation in accuracy and convergence. It further provides real-time integrity verification with complete tampering detection, while achieving up to 90x lower communication cost and 10x faster aggregation compared with prior work.

cs.CR↗

arXiv2026

Memba: Membrane-driven Parameter-Efficient Fine-Tuning for Mamba

State Space Models (SSMs) have emerged as powerful alternatives to attention-based Transformers, with Mamba demonstrating impressive efficiency and scalability. As these models grow increasingly larger, the need for Parameter-Efficient Fine-Tuning (PEFT) methods becomes critical to adapt pre-trained Mamba to downstream tasks without prohibitive computational costs. However, previous approaches simply apply traditional Transformer-tailored PEFT methods without addressing the unique temporal processing dynamics of SSMs. To address this limitation, we propose Memba, a membrane-driven PEFT approach specifically designed for Mamba. Memba introduces Leaky Integrate Membrane (LIM) neurons as bio-inspired gating mechanisms that naturally accumulate membrane potentials over time, enhancing selective information retention. By strategically combining LIM neurons with Low-Rank Adaptations (LoRA) and cross-layer membrane transfer, our approach significantly improves Mamba's temporal modeling capabilities. Extensive experiments across language and vision tasks demonstrate that Memba achieves substantial improvements over existing PEFT methods. The code is available at https://github.com/Intelligent-Computing-Lab-Yale/Memba.

cs.LG↗

arXiv2026

FuxiShuffle: An Adaptive and Resilient Shuffle Service for Distributed Data Processing on Alibaba Cloud

Shuffle exchanges intermediate results between upstream and downstream operators in distributed data processing and is usually the bottleneck due to factors such as small random I/Os and network contention. Several systems have been designed to improve shuffle efficiency, but from our experiences of running ultra-large clusters at Alibaba Cloud MaxCompute platform, we observe that they can not adapt to highly dynamic job characteristics and cluster resource conditions, and their fault tolerance mechanisms are passive and inefficient when failures are inevitable. To tackle their limitations, we design and implement FuxiShuffle as a general data shuffle service for the ultra-large production environment of MaxCompute, featuring good adaptability and efficient failure resilience. Specifically, to achieve good adaptability, FuxiShuffle dynamically selects the shuffle mode based on runtime information, conducts progress-aware scheduling for the downstream workers, and automatically determines the most suitable backup strategy for each shuffle data chunk. To make failure resilience efficient, FuxiShuffle actively ensures data availability with multi-replica failover, prevents memory overflow with careful memory management, and employs an incremental recovery mechanism that does not lose computation progress. Our experiments show that, compared to baseline systems, FuxiShuffle significantly reduces not only end-to-end job completion time but also aggregate resource consumption. Micro experiments suggest that our designs are effective in improving adaptability and failure resilience.

cs.DC↗

arXiv2026

QuRL: Efficient Reinforcement Learning with Quantized Rollout

Reinforcement learning with verifiable rewards (RLVR) has become a trending paradigm for training reasoning large language models (LLMs). However, due to the autoregressive decoding nature of LLMs, the rollout process becomes the efficiency bottleneck of RL training, consisting of up to 70\% of the total training time. In this work, we propose Quantized Reinforcement Learning (QuRL) that uses a quantized actor for accelerating the rollout. We address two challenges in QuRL. First, we propose Adaptive Clipping Range (ACR) that dynamically adjusts the clipping ratio based on the policy ratio between the full-precision actor and the quantized actor, which is essential for mitigating long-term training collapse. Second, we identify the weight update problem, where weight changes between RL steps are extremely small, making it difficult for the quantization operation to capture them effectively. We mitigate this problem through the invariant scaling technique that reduces quantization noise and increases weight update. We evaluate our method with INT8 and FP8 quantization experiments on DeepScaleR and DAPO, and achieve 20% to 80% faster rollout during training.

cs.LG↗

arXiv2026

OpenWorldSAM: Extending SAM2 for Universal Image Segmentation with Language Prompts

The ability to segment objects based on open-ended language prompts remains a critical challenge, requiring models to ground textual semantics into precise spatial masks while handling diverse and unseen categories. We present OpenWorldSAM, a framework that extends the prompt-driven Segment Anything Model v2 (SAM2) to open-vocabulary scenarios by integrating multi-modal embeddings extracted from a lightweight vision-language model (VLM). Our approach is guided by four key principles: i) Unified prompting: OpenWorldSAM supports a diverse range of prompts, including category-level and sentence-level language descriptions, providing a flexible interface for various segmentation tasks. ii) Efficiency: By freezing the pre-trained components of SAM2 and the VLM, we train only 4.5 million parameters on the COCO-stuff dataset, achieving remarkable resource efficiency. iii) Instance Awareness: We enhance the model's spatial understanding through novel positional tie-breaker embeddings and cross-attention layers, enabling effective segmentation of multiple instances. iv) Generalization: OpenWorldSAM exhibits strong zero-shot capabilities, generalizing well on unseen categories and an open vocabulary of concepts without additional training. Extensive experiments demonstrate that OpenWorldSAM achieves state-of-the-art performance in open-vocabulary semantic, instance, and panoptic segmentation across multiple benchmarks. Code is available at https://github.com/GinnyXiao/OpenWorldSAM.

cs.CV↗

arXiv2026

SVIP: Towards Verifiable Inference of Open-source Large Language Models

The ever-increasing size of open-source Large Language Models (LLMs) renders local deployment impractical for individual users. Decentralized computing has emerged as a cost-effective solution, allowing individuals and small companies to perform LLM inference for users using surplus computational power. However, a computing provider may stealthily substitute the requested LLM with a smaller, less capable model without consent from users, thereby benefiting from cost savings. We introduce SVIP, a secret-based verifiable LLM inference protocol. Unlike existing solutions based on cryptographic or game-theoretic techniques, our method is computationally effective and does not rest on strong assumptions. Our protocol requires the computing provider to return both the generated text and processed hidden representations from LLMs. We then train a proxy task on these representations, effectively transforming them into a unique model identifier. With our protocol, users can reliably verify whether the computing provider is acting honestly. A carefully integrated secret mechanism further strengthens its security. We thoroughly analyze our protocol under multiple strong and adaptive adversarial scenarios. Our extensive experiments demonstrate that SVIP is accurate, generalizable, computationally efficient, and resistant to various attacks. Notably, SVIP achieves false negative rates below 5% and false positive rates below 3%, while requiring less than 0.01 seconds per prompt query for verification.

cs.LG↗

arXiv2026

Electric field switching of altermagnetic spin-splitting in multiferroic skyrmions

Magnetic skyrmions are localized magnetic structures that retain their shape and stability over time, thanks to their topological nature. Recent theoretical and experimental progress has laid the groundwork for understanding magnetic skyrmions characterized by negligible net magnetization and ultrafast dynamics. Notably, skyrmions emerging in materials with altermagnetism, a novel magnetic phase featuring lifted Kramers degeneracy-have remained unreported until now. In this study, we demonstrate that BiFeO3, a multiferroic renowned for its strong coupling between ferroelectricity and magnetism, can transit from a spin cycloid to a Neel-type skyrmion under antidamping spin-orbit torque at room temperature. Strikingly, the altermagnetic spin splitting within BiFeO3 skyrmion can be reversed through the application of an electric field, revealed via the Circular photogalvanic effect. This quasiparticle, which possesses a neutral topological charge, holds substantial promise for diverse applications-most notably, enabling the development of unconventional computing systems with low power consumption and magnetoelectric controllability.

cond-mat.mtrl-sci↗

arXiv2025

Snapshot 3D image projection using a diffractive decoder

3D image display is essential for next-generation volumetric imaging; however, dense depth multiplexing for 3D image projection remains challenging because diffraction-induced cross-talk rapidly increases as the axial image planes get closer. Here, we introduce a 3D display system comprising a digital encoder and a diffractive optical decoder, which simultaneously projects different images onto multiple target axial planes with high axial resolution. By leveraging multi-layer diffractive wavefront decoding and deep learning-based end-to-end optimization, the system achieves high-fidelity depth-resolved 3D image projection in a snapshot, enabling axial plane separations on the order of a wavelength. The digital encoder leverages a Fourier encoder network to capture multi-scale spatial and frequency-domain features from input images, integrates axial position encoding, and generates a unified phase representation that simultaneously encodes all images to be axially projected in a single snapshot through a jointly-optimized diffractive decoder. We characterized the impact of diffractive decoder depth, output diffraction efficiency, spatial light modulator resolution, and axial encoding density, revealing trade-offs that govern axial separation and 3D image projection quality. We further demonstrated the capability to display volumetric images containing 28 axial slices, as well as the ability to dynamically reconfigure the axial locations of the image planes, performed on demand. Finally, we experimentally validated the presented approach, demonstrating close agreement between the measured results and the target images. These results establish the diffractive 3D display system as a compact and scalable framework for depth-resolved snapshot 3D image projection, with potential applications in holographic displays, AR/VR interfaces, and volumetric optical computing.

physics.optics↗

arXiv2025

From Code Foundation Models to Agents and Applications: A Comprehensive Survey and Practical Guide to Code Intelligence

Large language models (LLMs) have fundamentally transformed automated software development by enabling direct translation of natural language descriptions into functional code, driving commercial adoption through tools like Github Copilot (Microsoft), Cursor (Anysphere), Trae (ByteDance), and Claude Code (Anthropic). While the field has evolved dramatically from rule-based systems to Transformer-based architectures, achieving performance improvements from single-digit to over 95\% success rates on benchmarks like HumanEval. In this work, we provide a comprehensive synthesis and practical guide (a series of analytic and probing experiments) about code LLMs, systematically examining the complete model life cycle from data curation to post-training through advanced prompting paradigms, code pre-training, supervised fine-tuning, reinforcement learning, and autonomous coding agents. We analyze the code capability of the general LLMs (GPT-4, Claude, LLaMA) and code-specialized LLMs (StarCoder, Code LLaMA, DeepSeek-Coder, and QwenCoder), critically examining the techniques, design decisions, and trade-offs. Further, we articulate the research-practice gap between academic research (e.g., benchmarks and tasks) and real-world deployment (e.g., software-related code tasks), including code correctness, security, contextual awareness of large codebases, and integration with development workflows, and map promising research directions to practical needs. Last, we conduct a series of experiments to provide a comprehensive analysis of code pre-training, supervised fine-tuning, and reinforcement learning, covering scaling law, framework selection, hyperparameter sensitivity, model architectures, and dataset comparisons.

cs.SE↗

arXiv2025

Structured Light at the Extreme: Harnessing Spatiotemporal Control for High-Field Laser-Matter Interactions

This review charts the emerging paradigm of intelligent structured light for high-field laser-matter interactions, where the precise spatiotemporal and vectorial control of light is a critical degree of freedom. We outline a transformative framework built upon three synergistic pillars. First, we survey the advanced electromagnetic toolkit, moving beyond conventional spatial light modulators to include robust static optics and the promising frontier of plasma light modulators. Second, we detail the optimization engine for this high-dimensional design space, focusing on physics-informed digital twins and AI-driven inverse design to automate the discovery of optimal light structures. Finally, we explore the groundbreaking applications enabled by this integrated approach, including programmable electron beams, orbital-angular-momentum-carrying γ-rays, compact THz accelerators, and robust communications. The path forward necessitates overcoming grand challenges in material science, real-time adaptive control at MHz rates, and the extension of these principles to the quantum realm. This review serves as a call to action for a coordinated, interdisciplinary effort to command, rather than merely observe, light-matter interactions at the extreme.

physics.optics↗

arXiv2025

Detection and imaging of chemicals and hidden explosives using terahertz time-domain spectroscopy and deep learning

Detecting concealed chemicals and explosives remains a critical challenge in global security. Terahertz time-domain spectroscopy (THz-TDS) offers a promising non-invasive and stand-off detection technique owing to its ability to penetrate optically opaque materials without causing ionization damage. While many chemicals exhibit distinct spectral features in the terahertz range, conventional terahertz-based detection methods often struggle in real-world environments, where variations in sample geometry, thickness, and packaging can lead to inconsistent spectral responses. In this study, we present a chemical imaging system that integrates THz-TDS with deep learning to enable accurate pixel-level identification and classification of different explosives. Operating in reflection mode and enhanced with plasmonic nanoantenna arrays, our THz-TDS system achieves a peak dynamic range of 96 dB and a detection bandwidth of 4.5 THz, supporting practical, stand-off operation. By analyzing individual time-domain pulses with deep neural networks, the system exhibits strong resilience to environmental variations and sample inconsistencies. Blind testing across eight chemicals, including pharmaceutical excipients and explosive compounds, resulted in an average classification accuracy of 99.42% at the pixel level. Notably, the system maintained an average accuracy of 88.83% when detecting explosives concealed under opaque paper coverings, demonstrating its robust generalization capability. These results highlight the potential of combining advanced terahertz spectroscopy with neural networks for highly sensitive and specific chemical and explosive detection in diverse and operationally relevant scenarios.

physics.optics↗

arXiv2025

DuoGPT: Training-free Dual Sparsity through Activation-aware Pruning in LLMs

Large language models (LLMs) deliver strong performance but are difficult to deploy due to high memory and compute costs. While pruning reduces these demands, most methods ignore activation sparsity observed at runtime. We reinterpret activation sparsity as dynamic structured weight sparsity and propose DuoGPT, a unified framework that constructs dual-sparse (spMspV) workloads by combining unstructured weight pruning with activation sparsity. To preserve accuracy, we extend the Optimal Brain Compression (OBC) framework with activation-aware calibration and introduce output residuals from the dense model as correction terms. We further optimize the solution for efficient GPU execution, enabling scalability to billion-parameter LLMs. Evaluations on LLaMA-2 and LLaMA-3 show that DuoGPT outperforms state-of-the-art structured pruning methods by up to 9.17% accuracy at an iso-speedup of 1.39$\times$ compared to the baseline dense model. Code is available at Github.

cs.LG↗

arXiv2025

On a nonlinear nonlocal reaction-diffusion system applied to image restoration

This paper deals with a novel nonlinear coupled nonlocal reaction-diffusion system proposed for image restoration, characterized by the advantages of preserving low gray level features and textures.The gray level indicator in the proposed model is regularized using a new method based on porous media type equations, which is suitable for recovering noisy blurred images. The well-posedness, regularity, and other properties of the model are investigated, addressing the lack of theoretical analysis in those existing similar types of models. Numerical experiments conducted on texture and satellite images demonstrate the effectiveness of the proposed model in denoising and deblurring tasks.

math.AP↗

arXiv2025

Maximum principle for optimal control of infinite horizon stochastic difference equations driven by fractional noises

In this paper, infinite horizon stochastic difference equations and backward stochastic difference equations with fractional noises are studied. The main difficulty comes from fractional noises on infinite horizon. Motivated by discrete-time optimal control problem driven by fractional noises and on infinite horizon, the stochastic maximum principle for discrete-time control problem driven by fractional noises in infinite horizon is proved. As an application, an optimal investment problem is solved.

math.OC↗

arXiv2025

SpikePool: Event-driven Spiking Transformer with Pooling Attention

Building on the success of transformers, Spiking Neural Networks (SNNs) have increasingly been integrated with transformer architectures, leading to spiking transformers that demonstrate promising performance on event-based vision tasks. However, despite these empirical successes, there remains limited understanding of how spiking transformers fundamentally process event-based data. Current approaches primarily focus on architectural modifications without analyzing the underlying signal processing characteristics. In this work, we analyze spiking transformers through the frequency spectrum domain and discover that they behave as high-pass filters, contrasting with Vision Transformers (ViTs) that act as low-pass filters. This frequency domain analysis reveals why certain designs work well for event-based data, which contains valuable high-frequency information but is also sparse and noisy. Based on this observation, we propose SpikePool, which replaces spike-based self-attention with max pooling attention, a low-pass filtering operation, to create a selective band-pass filtering effect. This design preserves meaningful high-frequency content while capturing critical features and suppressing noise, achieving a better balance for event-based data processing. Our approach demonstrates competitive results on event-based datasets for both classification and object detection tasks while significantly reducing training and inference time by up to 42.5% and 32.8%, respectively.

cs.NE↗