arXiv subjects

Chenxi Ma

Publications and source records attributed to Chenxi Ma.

At least 19 recordsLinked to original sources

arXiv2026

ILR-SMO: Iterative Latent Refinement for Robust Spatial Multi-Omics Integration

Spatial multi-omics technologies jointly profile diverse molecular modalities with spatial context, providing a comprehensive view of cellular heterogeneity and tissue organization. To integrate spatial multi-omics data and identify spatial domains, a wide range of unsupervised methods has been proposed. However, recent approaches rely on single-step fusion, directly aggregating heterogeneous modalities into a shared representation, making embeddings sensitive to modality imbalance and measurement noise that are inherent to spatial multi-omics data, as well as unstable optimization under weak supervision. Here, we propose ILR-SMO (Iterative Latent Refinement for Spatial Multi-Omics), a unified self-supervised framework that reformulates multimodal integration as a stability-aware refinement process over a shared latent representation. Instead of single-step plain fusion, ILR-SMO progressively integrates modality-specific information through sequential updates, allowing information to be injected in a controlled manner. This incremental refinement enables later updates to correct or compensate for earlier deviations, resulting in more stable and robust representation learning under heterogeneous and noisy modalities. ILR-SMO further incorporates reliability-aware modality gating for adaptive modulation of modality contributions, and employs a joint objective to enforce spatial coherence while preventing representation collapse. Extensive experiments on five spatial multi-omics benchmarks demonstrate that ILR-SMO consistently outperforms seven state-of-the-art methods and exhibits strong robustness across diverse settings.

cs.LG↗

arXiv2026

Cooling-path-selected self-assembly of telecom InAsP quantum dots

Epitaxial quantum dots are leading solid-state sources of quantum light, but reproducible growth on indium phosphide (InP) remains challenging because dot-formation windows are narrow and emission wavelengths are difficult to predict. Here we report a new epitaxial route for InAsP quantum dots on InP(001), in which the cooling path selects the final morphology without intentional deposition of group-III dot material. High-temperature As exposure forms a shallow near-surface InAsP layer through As-P exchange; the cooling path and continued As supply determine whether three-dimensional islands form. Surface-sensitive spectroscopy shows near-surface As enrichment, while cross-sectional STEM/EELS identifies graded, As-rich InAsP islands. The capped emitters span 1.2-1.6 micrometers, with single-photon emission observed up to 100 K. Across 20 wafers, the surface-effective exchange dose D provides a practical calibration for targeting the median emission wavelength. These results establish cooling as an active epitaxial step for telecom InAsP quantum-dot formation and wavelength control.

quant-ph↗

arXiv2026

MicroWorld: Empowering Multimodal Large Language Models to Bridge the Microscopic Domain Gap with Multimodal Attributed Property Graph

Multimodal large language models (MLLMs) show remarkable potential for scientific reasoning, yet their performance in specialized domains such as microscopy remains limited by the scarcity of domain-specific training data and the difficulty of encoding fine-grained expert knowledge into model parameters. To bridge the gap, we introduce MicroWorld, a framework for structured-knowledge augmentation with a complementary visual retrieval pathway, using a multimodal attributed property graph (MAPG) without domain-specific fine-tuning. MicroWorld extracts biomedical entities and relations via scispaCy or LLM-based triplet mining, aligns images and entities in a shared embedding space using Qwen3-VL-Embedding, and assembles a knowledge graph comprising approximately 106K nodes and 2.62M typed edges spanning eight relation categories. At inference time, a graph-augmented retrieval pipeline matches query entities to the MAPG and injects structured knowledge context into the MLLM prompt. On MicroVQA, MicroWorld raises Qwen3-VL-8B-Instruct accuracy from 48.8% to 67.1% (+18.3 percentage points; 37.5% relative), exceeding zero-shot GPT-5 (59.4%). The identical retrieved context raises GPT-5 to 71.5%, and GPT-5's own entity extraction reaches 73.3%. On MicroBench, 8B Single-Pass improves overall accuracy by 4.0 percentage points. The benefits are greatest for knowledge-intensive reasoning; fine-grained visual discrimination and the additional visual-pathway contribution remain limited. Code and accompanying resources: https://github.com/ieellee/MicroWorld.

cs.CV↗

arXiv2026

Generation of continuous-wave laser light at 148.4 nm using cavity-enhanced second harmonic generation in $BaMgF_4$

The development of a nuclear optical clock based on \textsuperscript{229}Th is expected to influence the next generation of time and frequency standards. At its heart is the need for a continuous wave (CW) at laser at 148.4~nm to drive the clock transition between the ground and excited isomeric state. In this work we experimentally investigate the potential of BaMgF\textsubscript{4} crystals to create a CW solid state laser at the vacuum ultraviolet (VUV) wavelength of 148.4~nm via cavity-enhanced second harmonic generation. For this purpose, a BaMgF\textsubscript{4} crystal was grown, optically polished and periodically poled. The crystal was inserted into a power-enhancement cavity, resonant at the fundamental wavelength of 296.8~nm and the generated laser light at 148.4~nm was characterized. Within this proof-of-concept investigation, a VUV output power of typically ($16\pm1$)~pW is obtained. This marks the first time that this type of crystal is used to generate VUV laser light. The experimental findings are compared to theoretical expectations and provide a clear path for future improvements, which could have implications not only for compact laser sources in the VUV but also solid-state as well as single-ion based \textsuperscript{229}Th nuclear clocks.

physics.optics↗

arXiv2026

Liutex Core Lines as a New Framework for Tropical Cyclone Vortex Analysis

Liutex core lines are systematically applied to Hurricane Dorian to investigate the three-dimensional vortical organization of an intense tropical cyclone. Unlike conventional analysis based on wind speed or vorticity, Liutex core lines identify the local vortex axes and provide a compact geometrical representation of rotational structures. The results show that Dorian is not characterized by a single vortex core, but by a large population of vortices embedded within a coherent hurricane-scale circulation. Stronger and more vertically organized vortices are concentrated in the inner-core region, whereas numerous weaker, inclined, and curved vortical structures extend outward and collectively retain the spiral organization of the storm. The complete core line field therefore reveals a three-dimensional, multiscale vortical skeleton of the hurricane. These results suggest that a tropical cyclone can be viewed as an organized multiscale vortex system rather than a single monolithic vortex. Beyond visualization, Liutex core lines provide a potential framework for quantifying vortex geometry, spatial organization, and temporal evolution, offering a new approach for studying structural changes and vortex dynamics in tropical cyclones.

physics.ao-ph↗

arXiv2026

Calibration of systematic distortions in quantum emitter localization microscopy for deterministic nanophotonic fabrication

Quantum photonic technologies greatly benefit from quantum light emitters with high brightness, indistinguishability, and reliable polarization characteristics. Achieving optimal performance relies on the accurate localization of emitters and their deterministic integration into tailored photonic structures with nanometer-scale accuracy. Although marker-based photoluminescence imaging techniques can achieve statistical fitting uncertainties below 10 nm, the ultimate integration yield is often limited by uncorrected systematic distortions in custom cryo-optical setups that compromise metrological accuracy. Here, we present an in situ calibration protocol that uses lithographically defined gold nanodisk arrays as references to calibrate optical distortions with a Zernike vector-field model. On held-out validation patterns beyond the calibration dataset, this correction reduces the residual systematic bias to 5.3 nm with a 2D scatter of 24.6 nm across the analyzed field of view. Furthermore, we demonstrate that applying this correction to the deterministic fabrication of circular mesa structures around semiconductor quantum dots reduces the variance in emission polarization by 49%, indicating improved registration accuracy. This calibration strategy offers a practical route to high-yield deterministic integration of quantum emitters into scalable quantum photonic circuits.

physics.optics↗

arXiv2026

TraceLab: Characterizing Coding Agent Workloads for LLM Serving

Coding agents are rapidly becoming a major application of agentic LLMs, but serving them efficiently remains challenging. Progress on this challenge requires understanding real workload patterns, yet the data needed for such analysis is largely absent. Existing public traces and benchmarks do not capture real, day-to-day coding-agent usage across multiple agents and model families for serving-system analysis. To help fill this gap, we collect and release a trace of roughly 4,300 coding-agent sessions, containing about 350,000 LLM steps and 430,000 tool calls from our own day-to-day use of Claude Code and Codex. Our analysis shows that coding-agent workloads feature long autonomous loops, long contexts with short outputs, diverse and heavily-tailed tool calls, and high but imperfect prefix cache hit rates. These findings point to concrete opportunities for optimizing serving, including lower-overhead tool calling, append-length-aware prefill, semantic-aware tool-latency prediction, and improved KV-cache management around human-paced gaps. We release the dataset, trace collection pipeline, and analysis code at https://github.com/uw-syfi/TraceLab.git the project website is https://tracelab.cs.washington.edu.

cs.LG↗

arXiv2025

Non-Uniform Class-Wise Coreset Selection for Vision Model Fine-tuning

Coreset selection aims to identify a small yet highly informative subset of data, thereby enabling more efficient model training while reducing storage overhead. Recently, this capability has been leveraged to tackle the challenges of fine-tuning large foundation models, offering a direct pathway to their efficient and practical deployment. However, most existing methods are class-agnostic, causing them to overlook significant difficulty variations among classes. This leads them to disproportionately prune samples from either overly easy or hard classes, resulting in a suboptimal allocation of the data budget that ultimately degrades the final coreset performance. To address this limitation, we propose Non-Uniform Class-Wise Coreset Selection (NUCS), a novel framework that both integrates class-level and sample-level difficulty. We propose a robust metric for global class difficulty, quantified as the winsorized average of per-sample difficulty scores. Guided by this metric, our method performs a theoretically-grounded, non-uniform allocation of data selection budgets inter-class, while adaptively selecting samples intra-class with optimal difficulty ranges. Extensive experiments on a wide range of visual classification tasks demonstrate that NUCS consistently outperforms state-of-the-art methods across 10 diverse datasets and pre-trained models, achieving both superior accuracy and computational efficiency, highlighting the promise of non-uniform class-wise selection strategy for advancing the efficient fine-tuning of large foundation models.

cs.LG↗

arXiv2025

Unifying Segment Anything in Microscopy with Vision-Language Knowledge

Accurate segmentation of regions of interest in biomedical images holds substantial value in image analysis. Although several foundation models for biomedical segmentation have currently achieved excellent performance on certain datasets, they typically demonstrate sub-optimal performance on unseen domain data. We owe the deficiency to lack of vision-language knowledge before segmentation. Multimodal Large Language Models (MLLMs) bring outstanding understanding and reasoning capabilities to multimodal tasks, which inspires us to leverage MLLMs to inject Vision-Language Knowledge (VLK), thereby enabling vision models to demonstrate superior generalization capabilities on cross-domain datasets. In this paper, we propose a novel framework that seamlessly uses MLLMs to guide SAM in learning microscopy cross-domain data, unifying Segment Anything in Microscopy, named uLLSAM. Specifically, we propose the Vision-Language Semantic Alignment (VLSA) module, which injects VLK into Segment Anything Model (SAM). We find that after SAM receives global VLK prompts, its performance improves significantly, but there are deficiencies in boundary contour perception. Therefore, we further propose Semantic Boundary Regularization (SBR) to regularize SAM. Our method achieves performance improvements of 11.8% in SA across 9 in-domain microscopy datasets, achieving state-of-the-art performance. Our method also demonstrates improvements of 9.2% in SA across 10 out-of-domain datasets, exhibiting strong generalization capabilities. Code is available at https://github.com/ieellee/uLLSAM.

cs.CV↗

arXiv2025

MicroVQA++: High-Quality Microscopy Reasoning Dataset with Weakly Supervised Graphs for Multimodal Large Language Model

Multimodal Large Language Models are increasingly applied to biomedical imaging, yet scientific reasoning for microscopy remains limited by the scarcity of large-scale, high-quality training data. We introduce MicroVQA++, a three-stage, large-scale and high-quality microscopy VQA corpus derived from the BIOMEDICA archive. Stage one bootstraps supervision from expert-validated figure-caption pairs sourced from peer-reviewed articles. Stage two applies HiCQA-Graph, a novel heterogeneous graph over images, captions, and QAs that fuses NLI-based textual entailment, CLIP-based vision-language alignment, and agent signals to identify and filter inconsistent samples. Stage three uses a MultiModal Large Language Model (MLLM) agent to generate multiple-choice questions (MCQ) followed by human screening. The resulting release comprises a large training split and a human-checked test split whose Bloom's level hard-sample distribution exceeds the MicroVQA benchmark. Our work delivers (i) a quality-controlled dataset that couples expert literature with graph-based filtering and human refinement; (ii) HiCQA-Graph, the first graph that jointly models (image, caption, QA) for cross-modal consistency filtering; (iii) evidence that careful data construction enables 4B-scale MLLMs to reach competitive microscopy reasoning performance (e.g., GPT-5) and achieve state-of-the-art performance among open-source MLLMs. Code and dataset will be released after the review process concludes.

cs.CV↗

arXiv2025

Deterministic Integration of CsPbBr3 Quantum Dots with Plasmonic Ring Microcavities

Perovskite quantum dots hold great promise for quantum information processing as wavelength-tunable single photon sources operable over a broad temperature range. However, their deterministic integration into nanophotonic structures remains a major challenge, limited by their random spatial distribution and non-directional emission. In this work, we employ a two-step electron beam lithography process to deterministically place CsPbBr3 quantum dots within the mode volume of plasmonic ring microcavities. Simulations predict strong field enhancement within the cavity, boosting photon emission rates via the Purcell effect and improving the quantum efficiency of the emitters. Experimentally, coupling ensembles of CsPbBr3 quantum dots to the cavities results in a four-fold enhancement in photoluminescence intensity and a three-fold reduction in fluorescence lifetime at room temperature. Single-emitter coupling is further investigated at cryogenic temperatures, leading to a two-fold reduction in radiative lifetime. These results demonstrate a scalable approach for the integration of perovskite quantum dots into nanophotonic cavities and quantum photonic circuits.

physics.optics↗

arXiv2025

MM-Skin: Enhancing Dermatology Vision-Language Model with an Image-Text Dataset Derived from Textbooks

Medical vision-language models (VLMs) have shown promise as clinical assistants across various medical fields. However, specialized dermatology VLM capable of delivering professional and detailed diagnostic analysis remains underdeveloped, primarily due to less specialized text descriptions in current dermatology multimodal datasets. To address this issue, we propose MM-Skin, the first large-scale multimodal dermatology dataset that encompasses 3 imaging modalities, including clinical, dermoscopic, and pathological and nearly 10k high-quality image-text pairs collected from professional textbooks. In addition, we generate over 27k diverse, instruction-following vision question answering (VQA) samples (9 times the size of current largest dermatology VQA dataset). Leveraging public datasets and MM-Skin, we developed SkinVL, a dermatology-specific VLM designed for precise and nuanced skin disease interpretation. Comprehensive benchmark evaluations of SkinVL on VQA, supervised fine-tuning (SFT) and zero-shot classification tasks across 8 datasets, reveal its exceptional performance for skin diseases in comparison to both general and medical VLM models. The introduction of MM-Skin and SkinVL offers a meaningful contribution to advancing the development of clinical dermatology VLM assistants. MM-Skin is available at https://github.com/ZwQ803/MM-Skin

cs.CV↗

arXiv2025

Scaling Laws for Data-Efficient Visual Transfer Learning

Current scaling laws for visual AI models focus predominantly on large-scale pretraining, leaving a critical gap in understanding how performance scales for data-constrained downstream tasks. To address this limitation, this paper establishes the first practical framework for data-efficient scaling laws in visual transfer learning, addressing two fundamental questions: 1) How do scaling behaviors shift when downstream tasks operate with limited data? 2) What governs the efficacy of knowledge distillation under such constraints? Through systematic analysis of vision tasks across data regimes (1K-1M samples), we propose the distillation boundary theory, revealing a critical turning point in distillation efficiency: 1) Distillation superiority: In data-scarce conditions, distilled models significantly outperform their non-distillation counterparts, efficiently leveraging inherited knowledge to compensate for limited training samples. 2) Pre-training dominance: As pre-training data increases beyond a critical threshold, non-distilled models gradually surpass distilled versions, suggesting diminishing returns from knowledge inheritance when sufficient task-specific data becomes available. Empirical validation across various model scales (2.5M to 38M parameters) and data volumes demonstrate these performance inflection points, with error difference curves transitioning from positive to negative values at critical data thresholds, confirming our theoretical predictions. This work redefines scaling laws for data-limited regimes, bridging the knowledge gap between large-scale pretraining and practical downstream adaptation, addressing a critical barrier to understanding vision model scaling behaviors and optimizing computational resource allocation.

cs.LG↗

arXiv2024

Circular photonic crystal grating design for charge-tunable quantum light sources in the telecom C-band

Efficient generation of entangled photon pairs at telecom wavelengths is a key ingredient for long-range quantum networks. While embedding semiconductor quantum dots into hybrid circular Bragg gratings has proven effective, it conflicts with $p$-$i$-$n$ diode heterostructures which offer superior coherence. We propose and analyze hybrid circular photonic crystal gratings, incorporating air holes to facilitate charge carrier transport without compromising optical properties. Through numerical simulations, a broad cavity mode with a Purcell factor of 23 enhancing both exciton and biexciton transitions, and exceptional collection efficiency of 92.4% into an objective with numerical aperture of 0.7 are achieved. Furthermore, our design demonstrates direct coupling efficiency over 90% into a single-mode fiber over the entire telecom C-band. The hybrid circular photonic crystal grating thereby emerges as a promising solution for the efficient generation of highly coherent, polarization-entangled photon pairs.

quant-ph↗

arXiv2023

Converse Flexoelectricity of Low-Dimensional Bismuth Selenite (Bi2Se3) Revealed by Piezoresponse Force Microscopy (PFM)

Many kinds of two-dimensional (2D) van der Waals (vdW) have been demonstrated to exhibit electromechanical coupling effects, which makes them promising candidates for next-generation devices, such as piezotronics and nanogenerators. Recently, flexoelectricity was found to account for the out-of-plane electromechanical coupling in many 2D transition metal dichalcogenides (TMDs) who only exhibit in-plane piezoelectricity. However, low dimensional vdW three-dimensional (3D) topological insulators (TIs) have been overlooked regarding their electromechanical properties. In this study, for the first time, we experimentally investigate the electromechanical coupling of low dimensional 3D TIs with a centrosymmetric crystal structure, where a binary compound, bismuth selenite (Bi2Se3), is taken as an example. The results of piezoresponse force microscope (PFM) tests on the Bi2Se3 nanoflakes show that the material exhibits both out-of-plane and in-plane electromechanical responses. The Bi2Se3 nanoflake with a thickness of 37 nm possesses an effective out-of-plane piezoelectric coefficient of ~0.65 pm V-1. With careful analyses, the electromechanical responses are verified to arise from the converse flexoelectricity. The measured effective out-of-plane piezoelectric coefficient is mainly contributed by flexoelectric coefficient, μ_39, which is estimated to be approximately 0.13 nC m-1. However, it is rather difficult to obtain the in-plane component of the flexoelectric tensor from the in-plane PFM measurements since the direction of the in-plane stress is always not normal to the AFM cantilever axis. The results provide useful guidance for understanding the flexoelectric effect of low dimensional vdW materials with centrosymmetric crystal structures. Moreover, the work can pave to way to explore the electromechanical devices based on the flexoelectricity of vdW TIs.

cond-mat.mes-hall↗

arXiv2023

Surface quantum dots with pure, coherent, and blinking-free single photon emission

The surface of semiconductor nanostructures has a major impact on their electronic and optical properties. Disorder and defects in the surface layer typically cause degradation of charge carrier transport and radiative recombination dynamics. However, surface vicinity is inevitable for many scalable nano-optical applications. Epitaxially grown quantum dots are the best candidate for high-performance single photon emission and show great potential for quantum technologies. Yet, these emitters only reveal their excellent properties if they are deeply embedded in a semiconductor host. Until today, quantum dots close to surfaces yield weak, broad, and unstable emissions. Here, we show the complete restoration of optical properties from quantum dots grown directly on a semiconductor surface. The vanishing luminescence from the as-grown sample turns into bright, ultra-stable, coherent and blinking-free single photon emission after sulphur passivation. Under quasi-resonant excitation, single photons are generated with 98.8% purity, 77% indistinguishability, linewidths down to 4 $μ$eV and 99.69% persistency across 11 orders of magnitude in time. The emission is stable even after two years and when being subjected to nanomanufacturing processes. Some long-standing stumbling blocks for surface-dominated quantum dots are thereby removed, unveiling new possibilities for hybrid nano-devices and applications in quantum communication or sensing.

cond-mat.mes-hall↗

arXiv2023

Multi-Modality Deep Network for JPEG Artifacts Reduction

In recent years, many convolutional neural network-based models are designed for JPEG artifacts reduction, and have achieved notable progress. However, few methods are suitable for extreme low-bitrate image compression artifacts reduction. The main challenge is that the highly compressed image loses too much information, resulting in reconstructing high-quality image difficultly. To address this issue, we propose a multimodal fusion learning method for text-guided JPEG artifacts reduction, in which the corresponding text description not only provides the potential prior information of the highly compressed image, but also serves as supplementary information to assist in image deblocking. We fuse image features and text semantic features from the global and local perspectives respectively, and design a contrastive loss built upon contrastive learning to produce visually pleasing results. Extensive experiments, including a user study, prove that our method can obtain better deblocking results compared to the state-of-the-art methods.

cs.CV↗

arXiv2022

Geometry-Aware Reference Synthesis for Multi-View Image Super-Resolution

Recent multi-view multimedia applications struggle between high-resolution (HR) visual experience and storage or bandwidth constraints. Therefore, this paper proposes a Multi-View Image Super-Resolution (MVISR) task. It aims to increase the resolution of multi-view images captured from the same scene. One solution is to apply image or video super-resolution (SR) methods to reconstruct HR results from the low-resolution (LR) input view. However, these methods cannot handle large-angle transformations between views and leverage information in all multi-view images. To address these problems, we propose the MVSRnet, which uses geometry information to extract sharp details from all LR multi-view to support the SR of the LR input view. Specifically, the proposed Geometry-Aware Reference Synthesis module in MVSRnet uses geometry information and all multi-view LR images to synthesize pixel-aligned HR reference images. Then, the proposed Dynamic High-Frequency Search network fully exploits the high-frequency textural details in reference images for SR. Extensive experiments on several benchmarks show that our method significantly improves over the state-of-the-art approaches.

cs.CV↗