arXiv subjects

Xiaoyu Ma

Publications and source records attributed to Xiaoyu Ma.

At least 37 records · Page 2Linked to original sources

arXiv2026

Challenges and Research Directions for Large Language Model Inference Hardware

Large Language Model (LLM) inference is hard. The autoregressive Decode phase of the underlying Transformer model makes LLM inference fundamentally different from training. Exacerbated by recent AI trends, the primary challenges are memory and interconnect rather than compute. To address these challenges, we highlight four architecture research opportunities: High Bandwidth Flash for 10X memory capacity with HBM-like bandwidth; Processing-Near-Memory and 3D memory-logic stacking for high memory bandwidth; and low-latency interconnect to speedup communication. While our focus is datacenter AI, we also review their applicability for mobile devices.

cs.AR↗

arXiv2026

Advances and Innovations in the Multi-Agent Robotic System (MARS) Challenge

Recent advancements in multimodal large language models and vision-languageaction models have significantly driven progress in Embodied AI. As the field transitions toward more complex task scenarios, multi-agent system frameworks are becoming essential for achieving scalable, efficient, and collaborative solutions. This shift is fueled by three primary factors: increasing agent capabilities, enhancing system efficiency through task delegation, and enabling advanced human-agent interactions. To address the challenges posed by multi-agent collaboration, we propose the Multi-Agent Robotic System (MARS) Challenge, held at the NeurIPS 2025 Workshop on SpaVLE. The competition focuses on two critical areas: planning and control, where participants explore multi-agent embodied planning using vision-language models (VLMs) to coordinate tasks and policy execution to perform robotic manipulation in dynamic environments. By evaluating solutions submitted by participants, the challenge provides valuable insights into the design and coordination of embodied multi-agent systems, contributing to the future development of advanced collaborative AI systems.

cs.RO↗

arXiv2025

The $\infty$-S test via regression quantile affine LASSO

A novel test in the linear $\ell_1$ (LAD) and quantile regressions is proposed, based on the scores provided by the dual variables (signs) arising in the calculation of the (so-called) affine-lasso estimate--a Rao-type, Lagrange multiplier test using the thresholding, towards the null hypothesis of the test, function of the latter estimate.

stat.ME↗

arXiv2025

BLURR: A Boosted Low-Resource Inference for Vision-Language-Action Models

Vision-language-action (VLA) models enable impressive zero shot manipulation, but their inference stacks are often too heavy for responsive web demos or high frequency robot control on commodity GPUs. We present BLURR, a lightweight inference wrapper that can be plugged into existing VLA controllers without retraining or changing model checkpoints. Instantiated on the pi-zero VLA controller, BLURR keeps the original observation interfaces and accelerates control by combining an instruction prefix key value cache, mixed precision execution, and a single step rollout schedule that reduces per step computation. In our SimplerEnv based evaluation, BLURR maintains task success rates comparable to the original controller while significantly lowering effective FLOPs and wall clock latency. We also build an interactive web demo that allows users to switch between controllers and toggle inference options in real time while watching manipulation episodes. This highlights BLURR as a practical approach for deploying modern VLA policies under tight compute budgets.

cs.RO↗

arXiv2025

Modeling Spatio-temporal Extremes via Conditional Variational Autoencoders

Extreme weather events are widely studied in fields such as agriculture, ecology, and meteorology. The spatio-temporal co-occurrence of extreme events can strengthen or weaken under changing climate conditions. In this paper, we propose a novel approach to model spatio-temporal extremes by integrating climate indices via a conditional variational autoencoder (cXVAE). A convolutional neural network (CNN) is embedded in the decoder to convolve climatological indices with the spatial dependence within the latent space, thereby allowing the decoder to be dependent on the climate variables. There are three main contributions here. First, we demonstrate through extensive simulations that the proposed conditional XVAE accurately emulates spatial fields and recovers spatially and temporally varying extremal dependence with very low computational cost post training. Second, we provide a simple, scalable approach to detecting condition-driven shifts and whether the dependence structure is invariant to the conditioning variable. Third, when dependence is found to be condition-sensitive, the conditional XVAE supports counterfactual experiments allowing intervention on the climate covariate and propagating the associated change through the learned decoder to quantify differences in joint tail risk, co-occurrence ranges, and return metrics. To demonstrate the practical utility and performance of the model in real-world scenarios, we apply our method to analyze the monthly maximum Fire Weather Index (FWI) over eastern Australia from 2014 to 2024 conditioned on the El Niño/Southern Oscillation (ENSO) index.

stat.ML↗

arXiv2025

Measuring the magnetic fields in the chromospheres of low-mass stars

Magnetic fields in the upper atmospheres of solar-like stars are believed to provide an enormous amount of energy to power the hot coronae and drive large-scale eruptions that could impact the habitability of planetary systems around these stars. However, these magnetic fields have never been routinely measured on stars beyond the solar system. Through decade-long spectropolarimetric observations, we have now achieved the measurements of magnetic fields in the lower and middle chromospheres of three M-dwarfs. Our results indicate that the line-of-sight component of the chromospheric magnetic fields can reach up to hundreds of Gauss, whose sign frequently opposes that of the photospheric field. The measurements highlight the magnetic field complexity and the variation with height close to the surface of these M-dwarfs. They provide critical constraints on the energy budget responsible for heating and eruptions of stellar upper atmospheres, and enable assessments of how stellar magnetic activity may affect exoplanet environments.

astro-ph.SR↗

arXiv2025

Revisit Modality Imbalance at the Decision Layer

Multimodal learning integrates information from different modalities to enhance model performance, yet it often suffers from modality imbalance, where dominant modalities overshadow weaker ones during joint optimization. This paper reveals that such an imbalance not only occurs during representation learning but also manifests significantly at the decision layer. Experiments on audio-visual datasets (CREMAD and Kinetic-Sounds) show that even after extensive pretraining and balanced optimization, models still exhibit systematic bias toward certain modalities, such as audio. Further analysis demonstrates that this bias originates from intrinsic disparities in feature-space and decision-weight distributions rather than from optimization dynamics alone. We argue that aggregating uncalibrated modality outputs at the fusion stage leads to biased decision-layer weighting, hindering weaker modalities from contributing effectively. To address this, we propose that future multimodal systems should focus more on incorporate adaptive weight allocation mechanisms at the decision layer, enabling relative balanced according to the capabilities of each modality.

cs.LG↗

arXiv2025

Mixup Helps Understanding Multimodal Video Better

Multimodal video understanding plays a crucial role in tasks such as action recognition and emotion classification by combining information from different modalities. However, multimodal models are prone to overfitting strong modalities, which can dominate learning and suppress the contributions of weaker ones. To address this challenge, we first propose Multimodal Mixup (MM), which applies the Mixup strategy at the aggregated multimodal feature level to mitigate overfitting by generating virtual feature-label pairs. While MM effectively improves generalization, it treats all modalities uniformly and does not account for modality imbalance during training. Building on MM, we further introduce Balanced Multimodal Mixup (B-MM), which dynamically adjusts the mixing ratios for each modality based on their relative contributions to the learning objective. Extensive experiments on several datasets demonstrate the effectiveness of our methods in improving generalization and multimodal robustness.

cs.CV↗

arXiv2025

Validation-Free Sparse Learning: A Phase Transition Approach to Feature Selection

The growing environmental footprint of artificial intelligence (AI), especially in terms of storage and computation, calls for more frugal and interpretable models. Sparse models (e.g., linear, neural networks) offer a promising solution by selecting only the most relevant features, reducing complexity, preventing over-fitting and enabling interpretation-marking a step towards truly intelligent AI. The concept of a right amount of sparsity (without too many false positive or too few true positive) is subjective. So we propose a new paradigm previously only observed and mathematically studied for compressed sensing (noiseless linear models): obtaining a phase transition in the probability of retrieving the relevant features. We show in practice how to obtain this phase transition for a class of sparse learners. Our approach is flexible and applicable to complex models ranging from linear to shallow and deep artificial neural networks while supporting various loss functions and sparsity-promoting penalties. It does not rely on cross-validation or on a validation set to select its single regularization parameter. For real-world data, it provides a good balance between predictive accuracy and feature sparsity. A Python package is available at https://github.com/VcMaxouuu/HarderLASSO containing all the simulations and ready-to-use models.

stat.ML↗

arXiv2025

Entanglement certification from moments of positive maps

Entanglement certification is crucial in physical experiments, particularly when only partial knowledge of the quantum state is available. In this context, we present an entanglement criterion based on positive but not completely positive maps, which eliminates the need to identify eigenvalues of the output state. Notably, the Faddeev-LeVerrier algorithm establishes a relationship between the coefficients of characteristic polynomials and the moments of a matrix. This enables the existence of negative eigenvalues through the moments of the output state. The effectiveness of our criterion relies on the selection of positive maps, similar to the original positive maps criterion.

quant-ph↗

arXiv2025

Airalogy: AI-empowered universal data digitization for research automation

Research data are the foundation of Artificial Intelligence (AI)-driven science, yet current AI applications remain limited to a few fields with readily available, well-structured, digitized datasets. Achieving comprehensive AI empowerment across multiple disciplines is still out of reach. Present-day research data collection is often fragmented, lacking unified standards, inefficiently managed, and difficult to share. Creating a single platform for standardized data digitization needs to overcome the inherent challenge of balancing between universality (supporting the diverse, ever-evolving needs of various disciplines) and standardization (enforcing consistent formats to fully enable AI). No existing platform accommodates both facets. Building a truly multidisciplinary platform requires integrating scientific domain knowledge with sophisticated computing skills. Researchers often lack the computational expertise to design customized and standardized data recording methods, whereas platform developers rarely grasp the intricate needs of multiple scientific domains. These gaps impede research data standardization and hamper AI-driven progress. In this study, we address these challenges by developing Airalogy (https://airalogy.com), the world's first AI- and community-driven platform that balances universality and standardization for digitizing research data across multiple disciplines. Airalogy represents entire research workflows using customizable, standardized data records and offers an advanced AI research copilot for intelligent Q&A, automated data entry, analysis, and research automation. Already deployed in laboratories across all four schools of Westlake University, Airalogy has the potential to accelerate and automate scientific innovation in universities, industry, and the global research community-ultimately benefiting humanity as a whole.

cs.AI↗

arXiv2025

Improving Multimodal Learning Balance and Sufficiency through Data Remixing

Different modalities hold considerable gaps in optimization trajectories, including speeds and paths, which lead to modality laziness and modality clash when jointly training multimodal models, resulting in insufficient and imbalanced multimodal learning. Existing methods focus on enforcing the weak modality by adding modality-specific optimization objectives, aligning their optimization speeds, or decomposing multimodal learning to enhance unimodal learning. These methods fail to achieve both unimodal sufficiency and multimodal balance. In this paper, we, for the first time, address both concerns by proposing multimodal Data Remixing, including decoupling multimodal data and filtering hard samples for each modality to mitigate modality imbalance; and then batch-level reassembling to align the gradient directions and avoid cross-modal interference, thus enhancing unimodal learning sufficiency. Experimental results demonstrate that our method can be seamlessly integrated with existing approaches, improving accuracy by approximately 6.50%$\uparrow$ on CREMAD and 3.41%$\uparrow$ on Kinetic-Sounds, without training set expansion or additional computational overhead during inference. The source code is available at https://github.com/MatthewMaxy/Remix_ICML2025.

cs.LG↗

arXiv2025

Dynamic Authentication and Granularized Authorization with a Cross-Domain Zero Trust Architecture for Federated Learning in Large-Scale IoT Networks

With the increasing number of connected devices and complex networks involved, current domain-specific security techniques become inadequate for diverse large-scale Internet of Things (IoT) systems applications. While cross-domain authentication and authorization brings lots of security improvement, it creates new challenges of efficiency and security. Zero trust architecture (ZTA), an emerging network security architecture, offers a more granular and robust security environment for IoT systems. However, extensive cross-domain data exchange in ZTA can cause reduced authentication and authorization efficiency and data privacy concerns. Therefore, in this paper, we propose a dynamic authentication and granularized authorization scheme based on ZTA integrated with decentralized federated learning (DFL) for cross-domain IoT networks. Specifically, device requests in the cross-domain process are continuously monitored and evaluated, and only necessary access permissions are granted. To protect user data privacy and reduce latency, we integrate DFL with ZTA to securely and efficiently share device data across different domains. Particularly, the DFL model is compressed to reduce the network transmission load. Meanwhile, a dynamic adaptive weight adjustment mechanism is proposed to enable the DFL model to adapt to data characteristics from different domains. We analyze the performance of the proposed scheme in terms of security proof, including confidentiality, integrity and availability. Simulation results demonstrate the superior performance of the proposed scheme in terms of lower latency and higher throughput compared to other existing representative schemes.

cs.NI↗

arXiv2024

Flexible and efficient emulation of spatial extremes processes via variational autoencoders

Many real-world processes have complex tail dependence structures that cannot be characterized using classical Gaussian processes. More flexible spatial extremes models exhibit appealing extremal dependence properties but are often exceedingly prohibitive to fit and simulate from in high dimensions. In this paper, we aim to push the boundaries on computation and modeling of high-dimensional spatial extremes via integrating a new spatial extremes model that has flexible and non-stationary dependence properties in the encoding-decoding structure of a variational autoencoder called the XVAE. The XVAE can emulate spatial observations and produce outputs that have the same statistical properties as the inputs, especially in the tail. Our approach also provides a novel way of making fast inference with complex extreme-value processes. Through extensive simulation studies, we show that our XVAE is substantially more time-efficient than traditional Bayesian inference while outperforming many spatial extremes models with a stationary dependence structure. Lastly, we analyze a high-resolution satellite-derived dataset of sea surface temperature in the Red Sea, which includes 30 years of daily measurements at 16703 grid cells. We demonstrate how to use XVAE to identify regions susceptible to marine heatwaves under climate change and examine the spatial and temporal variability of the extremal dependence structure.

stat.ML↗

arXiv2024

Multi-class within-day dynamic traffic equilibrium with strategic travel time information

Most research on within-day dynamic traffic equilibrium with information provision implicitly considers travel time information, often assuming information to be perfect or imperfect based on travelers' perception error. However, lacking explicit formulation of information limits insightful analysis of information impact on dynamic traffic equilibrium and the potential benefits of leveraging information provision to improve system-level performance. To address this gap, this paper proposes a within-day dynamic traffic equilibrium model that explicitly formulates strategic information provision as an endogenous element. In the proposed framework, two classes of travelers receive different types of travel time information: one class receives instantaneous travel time reflecting the prevailing traffic conditions, while the other class receives strategic forecasts of travel times, generated by accounting for travelers' reactions to instantaneous information based on strategic thinking from behavioral game theory. The resulting multi-class within-day dynamic equilibrium differs from existing models by explicitly modeling information provision and consideration of information consistency. The inherent dynamics of real-time updated traffic information, traffic conditions, and travelers' choice behaviors are analytically modeled, with the resulting dynamic equilibrium formulated as a fixed-point problem. The theoretical propositions and numerical findings offer rich insights into the impact of information on the traffic network, strategic forecast information penetration, the relationship between the proposed equilibrium and traditional dynamic traffic equilibria, and information accuracy. This research contributes to a deeper understanding of the interplay between information and traffic dynamics, paving the way for more effective traffic management strategies.

eess.SY↗

arXiv2024

ODFormer: Semantic Fundus Image Segmentation Using Transformer for Optic Nerve Head Detection

Optic nerve head (ONH) detection has been a crucial area of study in ophthalmology for years. However, the significant discrepancy between fundus image datasets, each generated using a single type of fundus camera, poses challenges to the generalizability of ONH detection approaches developed based on semantic segmentation networks. Despite the numerous recent advancements in general-purpose semantic segmentation methods using convolutional neural networks (CNNs) and Transformers, there is currently a lack of benchmarks for these state-of-the-art (SoTA) networks specifically trained for ONH detection. Therefore, in this article, we make contributions from three key aspects: network design, the publication of a dataset, and the establishment of a comprehensive benchmark. Our newly developed ONH detection network, referred to as ODFormer, is based upon the Swin Transformer architecture and incorporates two novel components: a multi-scale context aggregator and a lightweight bidirectional feature recalibrator. Our published large-scale dataset, known as TongjiU-DROD, provides multi-resolution fundus images for each participant, captured using two distinct types of cameras. Our established benchmark involves three datasets: DRIONS-DB, DRISHTI-GS1, and TongjiU-DROD, created by researchers from different countries and containing fundus images captured from participants of diverse races and ages. Extensive experimental results demonstrate that our proposed ODFormer outperforms other state-of-the-art (SoTA) networks in terms of performance and generalizability. Our dataset and source code are publicly available at mias.group/ODFormer.

eess.IV↗

arXiv2024

Unconventional superconducting diode effects via antisymmetry and antisymmetry breaking

Symmetry-breaking plays a pivotal role in unlocking intriguing properties and functionalities in material systems. For example, the breaking of spatial and temporal symmetries leads to a fascinating phenomenon of superconducting diode effect. However, generating and precisely controlling the superconducting diode effect poses significant challenges. Here, we take a novel route with deliberate manipulation of magnetic charge potentials to realize unconventional superconducting flux-quantum diode effects. We achieve this through suitably tailored nanoengineered arrays of nanobar magnets on top of a superconducting thin film. We demonstrate the vital roles of inversion antisymmetry and its breaking in evoking unconventional superconducting effects-a magnetically symmetric diode effect and an odd-parity magnetotransport effect. These effects are non-volatilely controllable through in-situ magnetization switching of the nanobar magnets. Our findings promote the use of antisymmetry (breaking) for initiating unconventional superconducting properties, paving the way for exciting prospects and innovative functionalities in superconducting electronics.

cond-mat.supr-con↗

arXiv2023

TJDR: A High-Quality Diabetic Retinopathy Pixel-Level Annotation Dataset

Diabetic retinopathy (DR), as a debilitating ocular complication, necessitates prompt intervention and treatment. Despite the effectiveness of artificial intelligence in aiding DR grading, the progression of research toward enhancing the interpretability of DR grading through precise lesion segmentation faces a severe hindrance due to the scarcity of pixel-level annotated DR datasets. To mitigate this, this paper presents and delineates TJDR, a high-quality DR pixel-level annotation dataset, which comprises 561 color fundus images sourced from the Tongji Hospital Affiliated to Tongji University. These images are captured using diverse fundus cameras including Topcon's TRC-50DX and Zeiss CLARUS 500, exhibit high resolution. For the sake of adhering strictly to principles of data privacy, the private information of images is meticulously removed while ensuring clarity in displaying anatomical structures such as the optic disc, retinal blood vessels, and macular fovea. The DR lesions are annotated using the Labelme tool, encompassing four prevalent DR lesions: Hard Exudates (EX), Hemorrhages (HE), Microaneurysms (MA), and Soft Exudates (SE), labeled respectively from 1 to 4, with 0 representing the background. Significantly, experienced ophthalmologists conduct the annotation work with rigorous quality assurance, culminating in the construction of this dataset. This dataset has been partitioned into training and testing sets and publicly released to contribute to advancements in the DR lesion segmentation research community.

eess.IV↗