arXiv subjects

Yiming Ma

Publications and source records attributed to Yiming Ma.

At least 19 recordsLinked to original sources

arXiv2026

Isomonodromic Deformations for Linear $q$-Difference Systems of Degree One

We construct isomonodromy transformations for linear $q$-difference systems of the form $Y(qz)=A(z)Y(z)$, where $A(z)=A_0+z A_1$ has diagonal leading coefficient. These transformations shift eigenvalues of the leading coefficient $A_1$ together with roots of $\det A(z)$. They lift compatibility to the right eigenpairs of $A(z)$, yielding a discrete local tau function. The resulting deformation equations preserve the Birkhoff connection matrix, and reduce in their $q\to 1$ limit to the isomonodromic deformation of a meromorphic connection on $\mathbb P^1$ with an irregular singularity of Poincar\'e rank one at $\infty$.

nlin.SI↗

arXiv2026

OneShot: Index-in-Ranking with Neural Scoring for Large-Scale Retrieval

In modern recommendation systems, retrieval serves as a primary stage responsible for filtering billions of candidate items down to thousands prior to refined ranking. To make this massive search effective and efficient, the system relies on ranking accuracy and indexing efficiency. However, these two objectives are traditionally misaligned: while the former optimizes for the alignment between ranking predictions and user behavior, the latter optimizes for a structural grouping of item representations which enables fast search among billions of candidates. Thus, despite extensive efforts to scale up interaction modeling for retrieval, they remain fundamentally limited by the structural misalignment between the ranking objectives and the proximity-learned index. In this work, we address this long-standing dichotomy by proposing a new holistic retrieval framework, OneShot. It is an end-to-end, in-model index learning framework that natively aligns index learning with ranking objectives. Using this joint learning as a structural foundation, OneShot pushes the boundaries of retrieval expressiveness by scaling interaction modeling with neural scoring beyond the persistent dot-product bottleneck. OneShot is fully deployed in Instagram's industrial short-video recommendation system, driving significant wins in user daily sessions, engagement, and time-spent. Additionally, OneShot achieves a $20\%$ recall gain at the operational ranking volume and a 10x efficiency improvement at an equivalent recall level.

cs.IR↗

arXiv2026

VAIOM: Continuous-Input, Discrete-Output Decoder-Only Financial Sequence Modeling

Financial observations are continuous, heterogeneous, and noisy, whereas decoder-only next-token models are usually built around discrete symbolic inputs. We introduce Vector-Input Autoregressive Inference for Ordinal-Return Modeling (VAIOM), a decoder-only Transformer for probabilistic next-return modeling on one-hour foreign-exchange bars. VAIOM separates input representation from output likelihood: continuous multivariate financial-event vectors preserve numerical structure at the input, while a categorical distribution over the next volatility-normalized return bucket supports cross-entropy training and likelihood evaluation. The selected 0.9M Hybrid Continuous Input model combines continuous event features with categorical asset metadata, a Mixture-of-Market-States return head, Gap, volatility-regime, and Ordinal auxiliary objectives, and full-sequence supervision. Models and preprocessing are fit using pre-2024 Train data; models are selected on 2024H2 Validation and evaluated without refitting on two 2025 Test periods. Across three independent training seeds, every model outperforms fixed single-bar LightGBM baseline in both Test halves. For the canonical checkpoint, paired gains over LightGBM are 0.029 and 0.043 bits per event. Validation experiments show that continuous input improves over discrete-token input under the same categorical return objective, full-sequence supervision improves over last-position training, and auxiliary representation shaping together with a mixture-structured return head improves return likelihood in controlled comparisons. A supporting capacity study finds that the smallest evaluated complete architecture rung achieves the strongest Validation likelihood on the present corpus.

cs.LG↗

arXiv2026

Two-Sample Homogeneity Test via Entropic Optimal Transport

This paper proposes a two-sample homogeneity test based on entropic optimal transport (EOT) maps from a common reference distribution -- the uniform law on the unit ball. The test statistic is the squared $L^2$-distance between the two empirical EOT maps. For fixed entropic regularization parameter, we prove that the population map discrepancy is identifiable, derive a functional central limit theorem for the empirical map difference under the null, and establish the Gaussian quadratic-form null limit. We also prove consistency against fixed alternatives and characterize local asymptotic power under contiguous alternatives. A weighted multiplier bootstrap is proposed to calibrate the non-pivotal null distribution, and its validity is established. Extensive simulations demonstrate that the proposed EOT-map test has reliable finite-sample size control and exhibits competitive power compared with other existing methods. The method is particularly powerful for location alternatives and, beyond a single scalar discrepancy, it provides additional diagnostic information on how the two distributions differ. Finally, a real data application concludes the paper.

stat.ME↗

arXiv2026

N\"ushuVoice: Reviving the Voice of Endangered N\"ushu with Pitch-Aware Text-to-Speech

N\"ushu is an endangered phonetic script historically used by women in Jiangyong County, southern Hunan, China. While existing computational studies of N\"ushu mainly focus on textual digitization and visual recognition, the acoustic reconstruction of its authentic pronunciation remains largely unexplored. Building a N\"ushu text-to-speech (TTS) system is particularly challenging because available recordings are extremely limited and mostly consist of isolated syllable-level pronunciations rather than natural sentence-level utterances. In this work, we introduce N\"ushuVoice, the first TTS benchmark for N\"ushu. We construct a sentence-level N\"ushu text-to-audio dataset that aligns standardized Unicode N\"ushu text, phonetic transcriptions, standard Chinese translations, and archival recordings. To synthesize speech under this extreme low-resource setting, we propose N\"ushu-PitchVITS, an F0-conditioned VITS framework that leverages N\"ushu's five-level pitch notation as an explicit prosodic inductive bias. Experimental results show that N\"ushu-PitchVITS outperforms strong TTS baselines in spectral fidelity, pitch reconstruction, and human-rated intelligibility. We publicly release the dataset and code at: https://anonymous.4open.science/r/Nvshu-TTS-2EB6.

cs.CL↗

arXiv2026

Robust OT-Guided Generative Residual Domain Adaptation for Bike-Sharing Demand Prediction under Temporal Domain Shift

Bike-sharing models trained on historical station-hour data may degrade when deployed in later years because travel patterns change over time. This paper studies March Citi Bike demand prediction from 2021 to 2026 as a temporal domain adaptation problem and proposes Gen-ROTDA, a robust optimal transport-guided residual domain adaptation framework. The method fits a target-domain station-time anchor with a small labeled target subset, transfers residual rather than raw demand, applies a deterministic label-preserving residual feature generator, and trims high-cost transport matches before training the final residual predictor. Experiments compare Gen-ROTDA with anchor-only, source-only, target-only, fine-tuning, MMD adaptation, Sinkhorn OTDA, ROTDA, and Gen-OTDA. Gen-ROTDA achieves the lowest MAE on the main 2025 to 2026 task and is the best OT-family method on average across multi-year tasks, although fine-tuning and MMD adaptation remain strong overall baselines. Under abnormal target-unlabeled records, Gen-ROTDA is much more stable than non-robust OT variants, suggesting that robust transport is useful for noisy temporal transfer in bike-sharing demand prediction.

cs.LG↗

arXiv2026

DeAR: Fine-Grained VLM Adaptation by Decomposing Attention Head Roles

Prompt learning is a dominant paradigm for adapting pre-trained Vision-Language Models (VLMs) to downstream tasks. However, existing methods often rely on a simplistic, layer-centric view, assuming shallow layers capture general features while deep layers handle task-specific knowledge. This assumption results in uncontrolled interactions between learnable tokens and original tokens. Task-specific knowledge could degrades the model's core generalization and creates a trade-off between task adaptation and the preservation of zero-shot generalization. To address this, we challenge the layer-centric view and propose \textbf{DeAR}, a framework that achieves fine-grained VLM adaptation by \textbf{De}composing \textbf{A}ttention head \textbf{R}oles. We posit that the functional specialization within VLMs occurs not between layers, but at the finer-grained level of individual attention heads in the deeper layers. Based on this insight, we introduce a novel metric, Concept Entropy, to systematically classify attention heads into distinct functional roles: \textit{Attribute}, \textit{Generalization}, and \textit{Mixed}. Guided by these roles, we introduce specialized attribute tokens and a Role-Based Attention Mask mechanism to precisely control information flow, ensuring generalization heads remain isolated from task-specific knowledge. We further incorporate a Task-Adaptive Fusion Strategy for inference. Extensive experiments on fifteen datasets show that DeAR achieves a strong balance between task adaptation and generalization, outperforming previous methods across various tasks.

cs.CV↗

arXiv2026

How Implicit Bias Accumulates and Propagates in LLM Long-term Memory

Long-term memory mechanisms enable Large Language Models (LLMs) to maintain continuity and personalization across extended interaction lifecycles, but they also introduce new and underexplored risks related to fairness. In this work, we study how implicit bias, defined as subtle statistical prejudice, accumulates and propagates within LLMs equipped with long-term memory. To support systematic analysis, we introduce the Decision-based Implicit Bias (DIB) Benchmark, a large-scale dataset comprising 3,776 decision-making scenarios across nine social domains, designed to quantify implicit bias in long-term decision processes. Using a realistic long-horizon simulation framework, we evaluate six state-of-the-art LLMs integrated with three representative memory architectures on DIB and demonstrate that LLMs' implicit bias does not remain static but intensifies over time and propagates across unrelated domains. We further analyze mitigation strategies and show that a static system-level prompting baseline provides limited and short-lived debiasing effects. To address this limitation, we propose Dynamic Memory Tagging (DMT), an agentic intervention that enforces fairness constraints at memory write time. Extensive experimental results show that DMT substantially reduces bias accumulation and effectively curtails cross-domain bias propagation.

cs.LG↗

arXiv2026

Integrated Photonic Quantum Computing: From Silicon to Lithium Niobate

Quantum technologies have surpassed classical systems by leveraging the unique properties of superposition and entanglement in photons and matter. Recent advancements in integrated quantum photonics, especially in silicon-based and lithium niobate platforms, are pushing the technology toward greater scalability and functionality. Silicon circuits have progressed from centimeter-scale, dual-photon systems to millimeter-scale, high-density devices that integrate thousands of components, enabling sophisticated programmable manipulation of multi-photon states. Meanwhile, lithium niobate, thanks to its wide optical transmission window, outstanding nonlinear and electro-optic coefficients, and chemical stability, has emerged as an optimal substrate for fully integrated photonic quantum chips. Devices made from this material exhibit high efficiency in in generating, manipulating, converting, storing, and detecting photon states, thereby establishing a basis for deterministic multi-photon generation and single-photon quantum interactions, as well as comprehensive frequency-state control. This review explores the development of integrated photonic quantum technologies based on both silicon and lithium niobate, highlighting invaluable insights gained from silicon-based systems that can assist the scaling of lithium niobate technologies. It examines the functional integration mechanisms of lithium niobate in electro-optic tuning and nonlinear energy conversion, showcasing its transformative impact throughout the photonic quantum computing process. Looking ahead, we speculate on the developmental pathways for lithium niobate platforms and their potential to revolutionize areas such as quantum communication, complex system simulation, quantum sampling, and optical quantum computing paradigms.

physics.optics↗

arXiv2025

SeedProteo: Accurate De Novo All-Atom Design of Protein Binders

We present SeedProteo, a diffusion-based model for de novo all-atom protein design. We demonstrate how to repurpose a cutting-edge folding architecture into a powerful generative design framework by effectively integrating self-conditioning features. Extensive benchmarks highlight the model's capabilities across two distinct tasks: in unconditional generation, SeedProteo exhibits superior length generalization and structural diversity, maintaining robustness for long sequences and complex topologies; in binder design, it achieves state-of-the-art performance among open-source methods, attaining the highest in-silico design success rates, structural diversity and novelty. Finally, we validate SeedProteo through wet-lab assays on two therapeutic targets, achieving hit rates of 70%-80% and picomolar-level binding affinities, establishing leading results. To facilitate community adoption, we provide public access to SeedProteo via a webserver (https://seedfold.io/proteinDesign).

q-bio.BM↗

arXiv2025

SeedFold: Scaling Biomolecular Structure Prediction

Highly accurate biomolecular structure prediction is a key component of developing biomolecular foundation models, and one of the most critical aspects of building foundation models is identifying the recipes for scaling the model. In this work, we present SeedFold, a folding model that successfully scales up the model capacity. Our contributions are threefold: first, we identify an effective width-scaling strategy for the Pairformer to increase representation capacity; second, we introduce a novel linear triangular attention that reduces computational complexity to enable efficient scaling; finally, we construct a large-scale distillation dataset to substantially enlarge the training set. Experiments on FoldBench show that SeedFold outperforms AlphaFold3 on most protein-related tasks.

q-bio.BM↗

arXiv2025

Multidimensional Stochastic Dominance Test Based on Center-outward Quantiles

Stochastic dominance (SD) provides a quantile-based partial ordering of random variables and has broad applications. Its extension to multivariate settings, however, is challenging due to the lack of a canonical ordering in $\mathbb{R}^d$ ($d \ge 2$) and the set-valued character of multivariate quantiles. Based on the multivariate center-outward quantile function in Hallin et al. (2021), this paper proposes new first- and second-order multivariate stochastic dominance (MSD) concepts through comparing contribution functions defined over quantile contours and regions. To address computational and inferential challenges, we incorporate entropy-regularized optimal transport, which ensures faster convergence rate and tractable estimation. We further develop consistent Kolmogorov-Smirnov and Cram\'er- von Mises type test statistics for MSD, establish bootstrap validity, and demonstrate through extensive simulations good finite-sample performance of the tests. Our approach offers a theoretically rigorous, and computationally feasible solution for comparing multivariate distributions.

stat.ME↗

arXiv2025

UniMPR: A Unified Framework for Multimodal Place Recognition with Heterogeneous Sensor Configurations

Place recognition is a critical component of autonomous vehicles and robotics, enabling global localization in GPS-denied environments. Recent advances have spurred significant interest in multimodal place recognition (MPR), which leverages complementary strengths of multiple modalities. Despite its potential, most existing MPR methods still face three key challenges: (1) dynamically adapting to various modality inputs within a unified framework, (2) maintaining robustness with missing or degraded modalities, and (3) generalizing across diverse sensor configurations and setups. In this paper, we propose UniMPR, a unified framework for multimodal place recognition. Using only one trained model, it can seamlessly adapt to any combination of common perceptual modalities (e.g., camera, LiDAR, radar). To tackle the data heterogeneity, we unify all inputs within a polar BEV feature space. Subsequently, the polar BEVs are fed into a multi-branch network to exploit discriminative intra-model and inter-modal features from any modality combinations. To fully exploit the network's generalization capability and robustness, we construct a large-scale training set from multiple datasets and introduce an adaptive label assignment strategy for extensive pre-training. Experiments on seven datasets demonstrate that UniMPR achieves state-of-the-art performance under varying sensor configurations, modality combinations, and environmental conditions. Our code will be released at https://github.com/QiZS-BIT/UniMPR.

cs.CV↗

arXiv2025

CTransformer: Deep-transformer-based 3D cell membrane tracking with subcellular-resolved molecular quantification

Deep learning segmentation and fluorescence imaging techniques allow the cellular morphology of living embryos to be constructed spatiotemporally. These development processes involve numerous molecules distributed at the subcellular scale, such as cell adhesion (E-cadherin), which accumulate at cell-cell interfaces to regulate intercellular connection. However, quantifying molecular distributions within specific subcellular regions across the entire embryo, where cell movement and molecular redistribution occur rapidly, is challenging due to the need for simultaneous cell morphology reconstruction and lineage tracing due to photobleaching and phototoxicity. We report a transformer-based pipeline, CTransformer, that establishes a 4D cellular morphology map before the 550-cell (late) stage. CTransformer constructed 4D cellular morphology atlases, reaching 80% accuracy at the 550-cell stage. Through this advanced architecture, we use only one channel to reconstruct cell morphology and achieve cell tracing. With each cell's morphology as a reference, the distribution of specific molecules throughout the cell body and at cell interfaces can be quantitatively measured in another fluorescence channel. We apply this methodology to track E-cadherin during embryonic development of the worm Caenorhabditis elegans, from fertilization to gastrulation. Our results reveal that E-cadherin is tightly regulated across individual embryos, both within single cells and at cell-cell interfaces, displaying an anterior-posterior gradient and cell- and lineage-specific patterns. Furthermore, its spatiotemporal heterogeneity influences cell mechanics and embryonic morphogenesis, helping explain how C. elegans achieves stereotypical developmental patterns at cellular resolution.

physics.bio-ph↗

arXiv2025

SilverTorch: A Unified Model-based System to Democratize Large-Scale Recommendation on GPUs

Serving deep learning based recommendation models (DLRM) at scale is challenging. Existing approaches rely on dedicated ANN indexing and filtering services on CPUs, suffering from non-negligible costs and missing co-design opportunities. Such inefficiency makes them difficult to support complex model architectures, such as learned similarities and multi-task retrieval. In this paper, we present SilverTorch, a model-based serving system that brings all components into one unified model. It unifies model serving by replacing standalone indexing and filtering services with model layers. We propose a model-based GPU Bloom index for feature filtering and a fused Int8 ANN kernel for nearest neighbor search. Through co-design of the ANN search and feature filtering, we reduce GPU memory usage and eliminate computation. Benefiting from this design, we scale up retrieval by introducing an OverArch scoring layer and a multi-task retrieval with a Value Model to aggregate scores. These advancements improve the retrieval accuracy and enable future studies for serving more complex models. Our evaluation on industry-scale datasets show that SilverTorch achieves up to 23.7\times higher throughput compared to the state-of-the-art approaches. We also demonstrate that SilverTorch solution is 13.35\times more cost-efficient than CPU-based solution while improving accuracy via serving more complex models. SilverTorch is deployed at scale, serving hundreds of models online and supporting recommendation for diverse applications.

cs.IR↗

arXiv2025

CellINR: Implicitly Overcoming Photo-induced Artifacts in 4D Live Fluorescence Microscopy

4D live fluorescence microscopy is often compromised by prolonged high intensity illumination which induces photobleaching and phototoxic effects that generate photo-induced artifacts and severely impair image continuity and detail recovery. To address this challenge, we propose the CellINR framework, a case-specific optimization approach based on implicit neural representation. The method employs blind convolution and structure amplification strategies to map 3D spatial coordinates into the high frequency domain, enabling precise modeling and high-accuracy reconstruction of cellular structures while effectively distinguishing true signals from artifacts. Experimental results demonstrate that CellINR significantly outperforms existing techniques in artifact removal and restoration of structural continuity, and for the first time, a paired 4D live cell imaging dataset is provided for evaluating reconstruction performance, thereby offering a solid foundation for subsequent quantitative analyses and biological research. The code and dataset will be public.

eess.IV↗

arXiv2025

SLENet: A Novel Multiscale CNN-Based Network for Detecting the Rats Estrous Cycle

In clinical medicine, rats are commonly used as experimental subjects. However, their estrous cycle significantly impacts their biological responses, leading to differences in experimental results. Therefore, accurately determining the estrous cycle is crucial for minimizing interference. Manually identifying the estrous cycle in rats presents several challenges, including high costs, long training periods, and subjectivity. To address these issues, this paper proposes a classification network-Spatial Long-distance EfficientNet (SLENet). This network is designed based on EfficientNet, specifically modifying the Mobile Inverted Bottleneck Convolution (MBConv) module by introducing a novel Spatial Efficient Channel Attention (SECA) mechanism to replace the original Squeeze Excitation (SE) module. Additionally, a Non-local attention mechanism is incorporated after the last convolutional layer to enhance the network's ability to capture long-range dependencies. The dataset used 2,655 microscopic images of rat vaginal epithelial cells, with 531 images in the test set. Experimental results indicate that SLENet achieved an accuracy of 96.31%, outperforming baseline EfficientNet model (94.2%). This finding provide practical value for optimizing experimental design in rat-based studies such as reproductive and pharmacological research, but this study is limited to microscopy image data, without considering other factors like temporal patterns, thus, incorporating multi-modal input is necessary for future application.

eess.IV↗

arXiv2025

MHD simulation of tilt instability during the dynamic FRC magnetic compression process

The nonlinear evolution of the tilt instability in a field reversed configuration (FRC) during the dynamic magnetic compression process has been investigated using magnetohydrodynamic (MHD) simulations with the NIMROD code [C. R. Sovinec \textit{et al.}, J. Comput. Phys. \textbf{195}, 355 (2004)]. The tilt mode induces significant deformations in the linear growth phase and results in complete confinement loss of the FRC in the nonlinear phase, with no evidence of dynamic nonlinear stabilization. The growth rate of the tilt mode increases with the compression field ramping rate and approaches an asymptotic value. Toroidal flow can reduce both the growth rate and the nonlinear saturation amplitude of the tilt mode. The stabilizing effect of the toroidal rotation is enhanced with higher compression field ramping rates due to the spontaneous toroidal field generation and increased flow shear during compression. Although the tilt mode remains unstable with a toroidal rotation Mach number close to 0.5, the onset of tilt distortion can be delayed, allowing a magnetic compression ratio up to 5.3 before the compressional heating terminates.

physics.plasm-ph↗