arXiv subjects

Ying Yang

Publications and source records attributed to Ying Yang.

At least 55 records · Page 3Linked to original sources

arXiv2025

Smooth Total variation Regularization for Interference Detection and Elimination (STRIDE) for MRI

MRI is increasingly desired to function near electronic devices that emit potentially dynamic electromagnetic interference (EMI). To accommodate for this, we propose the STRIDE method, which improves on previous external-sensor-based EMI removal methods by exploiting inherent MR image smoothness in its total variation. STRIDE measures data from both EMI detectors and primary MR imaging coils, transforms this data into the image domain, and for each column of the resulting image array, combines and subtracts data from the EMI detectors in a way that optimizes for total-variation smoothness. Performance was tested on phantom and in-vivo datasets with a 0.5T scanner. STRIDE resulted in visually better EMI removal, higher temporal SNR, larger EMI removal percentage, and lower RMSE than standard implementations. STRIDE is a robust technique that leverages inherent MR image properties to provide improved EMI removal performance over standard algorithms, particularly for time-varying noise sources.

eess.IV↗

arXiv2025

Spatially Randomized Designs Can Enhance Policy Evaluation

This article studies the benefits of using spatially randomized experimental designs which partition the experimental area into distinct, non-overlapping units with treatments assigned randomly. Such designs offer improved policy evaluation in online experiments by providing more precise policy value estimators and more effective A/B testing algorithms than traditional global designs, which apply the same treatment across all units simultaneously. We examine both parametric and nonparametric methods for estimating and inferring policy values based on this randomized approach. Our analysis includes evaluating the mean squared error of the treatment effect estimator and the statistical power of the associated tests. Additionally, we extend our findings to experiments with spatio-temporal dependencies, where treatments are allocated sequentially over time, and account for potential temporal carryover effects. Our theoretical insights are supported by comprehensive numerical experiments.

math.ST↗

arXiv2025

OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search

Traditional e-commerce search systems employ multi-stage cascading architectures (MCA) that progressively filter items through recall, pre-ranking, and ranking stages. While effective at balancing computational efficiency with business conversion, these systems suffer from fragmented computation and optimization objective collisions across stages, which ultimately limit their performance ceiling. To address these, we propose \textbf{OneSearch}, the first industrial-deployed end-to-end generative framework for e-commerce search. This framework introduces three key innovations: (1) a Keyword-enhanced Hierarchical Quantization Encoding (KHQE) module, to preserve both hierarchical semantics and distinctive item attributes while maintaining strong query-item relevance constraints; (2) a multi-view user behavior sequence injection strategy that constructs behavior-driven user IDs and incorporates both explicit short-term and implicit long-term sequences to model user preferences comprehensively; and (3) a Preference-Aware Reward System (PARS) featuring multi-stage supervised fine-tuning and adaptive reward-weighted ranking to capture fine-grained user preferences. Extensive offline evaluations on large-scale industry datasets demonstrate OneSearch's superior performance for high-quality recall and ranking. The rigorous online A/B tests confirm its ability to enhance relevance in the same exposure position, achieving statistically significant improvements: +1.67% item CTR, +2.40% buyer, and +3.22% order volume. Furthermore, OneSearch reduces operational expenditure by 75.40% and improves Model FLOPs Utilization from 3.26% to 27.32%. The system has been successfully deployed across multiple search scenarios in Kuaishou, serving millions of users, generating tens of millions of PVs daily.

cs.IR↗

arXiv2025

Scalable Bayesian inference for heat kernel Gaussian processes on manifolds

We establish a scalable manifold learning method and theory, motivated by the problem of estimating fMRI activation manifolds in the Human Connectome Project (HCP). Our primary contribution is the development of an efficient estimation technique for heat kernel Gaussian processes in the exponential family model. This approach handles large sample sizes $n$, preserves the intrinsic geometry of data, and significantly reduces computational complexity from $\mathcal{O}(n^3)$ to $\mathcal{O}(n)$ via a novel reduced-rank approximation of the graph Laplacian's transition matrix and a Truncated Singular Value Decomposition for the eigenpair computation. The numerical experiments demonstrate the scalability and improved accuracy of our method for manifold learning tasks involving complex large-scale data.

stat.ME↗

arXiv2025

A Contextual Bandits Approach for Personalization of Hand Gesture Recognition

In human-computer interaction applications like hand gesture recognition, supervised learning models are often trained on a large population of users to achieve high task accuracy. However, due to individual variability in sensor signals and user behavior, static models may not provide optimal performance for all users. Personalizing pretrained models via calibration--collecting labeled data from each user--can improve performance but introduces user friction and struggles with limited data. To overcome these issues, we propose a calibrationless longitudinal personalization method: a contextual multi-arm bandit (MAB) algorithm combined with a pretrained neural network for gesture recognition. This reinforcement-learning-style approach enables personalization using binary reward signals, either user-provided or inferred by the system. We validated this method in a user study. Participants wore a surface electromyography (sEMG) device and played multiple rounds of a 2-D navigation game using six hand gestures. In the session, they completed a baseline round and then a round with our algorithm; in the second session, they played another round with our algorithm. Our approach led to a significant reduction in users' average false negative rate by 0.113 from the initial to the final round, with further decreases between sessions. Average precision also trended upward (by 0.139) from the start to end of a round, continuing in the next session. Notably, some users who could not complete the game with the baseline model succeeded with our contextual MAB model. In summary, our

cs.HC↗

arXiv2025

Unraveling the Interplay between Carryover Effects and Reward Autocorrelations in Switchback Experiments

A/B testing has become the gold standard for policy evaluation in modern technological industries. Motivated by the widespread use of switchback experiments in A/B testing, this paper conducts a comprehensive comparative analysis of various switchback designs in Markovian environments. Unlike many existing works which derive the optimal design based on specific and relatively simple estimators, our analysis covers a range of state-of-the-art estimators developed in the reinforcement learning (RL) literature. It reveals that the effectiveness of different switchback designs depends crucially on (i) the size of the carryover effect and (ii) the auto-correlations among reward errors over time. Meanwhile, these findings are estimator-agnostic, i.e., they apply to most RL estimators. Based on these insights, we provide a workflow to offer guidelines for practitioners on designing switchback experiments in A/B testing.

stat.ML↗

arXiv2025

High-dimensional Asymptotics of Generalization Performance in Continual Ridge Regression

Continual learning is motivated by the need to adapt to real-world dynamics in tasks and data distribution while mitigating catastrophic forgetting. Despite significant advances in continual learning techniques, the theoretical understanding of their generalization performance lags behind. This paper examines the theoretical properties of continual ridge regression in high-dimensional linear models, where the dimension is proportional to the sample size in each task. Using random matrix theory, we derive exact expressions of the asymptotic prediction risk, thereby enabling the characterization of three evaluation metrics of generalization performance in continual learning: average risk, backward transfer, and forward transfer. Furthermore, we present the theoretical risk curves to illustrate the trends in these evaluation metrics throughout the continual learning process. Our analysis reveals several intriguing phenomena in the risk curves, demonstrating how model specifications influence the generalization performance. Simulation studies are conducted to validate our theoretical findings.

stat.ME↗

arXiv2025

OneSug: The Unified End-to-End Generative Framework for E-commerce Query Suggestion

Query suggestion plays a crucial role in enhancing user experience in e-commerce search systems by providing relevant query recommendations that align with users' initial input. This module helps users navigate towards personalized preference needs and reduces typing effort, thereby improving search experience. Traditional query suggestion modules usually adopt multi-stage cascading architectures, for making a well trade-off between system response time and business conversion. But they often suffer from inefficiencies and suboptimal performance due to inconsistent optimization objectives across stages. To address these, we propose OneSug, the first end-to-end generative framework for e-commerce query suggestion. OneSug incorporates a prefix2query representation enhancement module to enrich prefixes using semantically and interactively related queries to bridge content and business characteristics, an encoder-decoder generative model that unifies the query suggestion process, and a reward-weighted ranking strategy with behavior-level weights to capture fine-grained user preferences. Extensive evaluations on large-scale industry datasets demonstrate OneSug's ability for effective and efficient query suggestion. Furthermore, OneSug has been successfully deployed for the entire traffic on the e-commerce search engine in Kuaishou platform for over 1 month, with statistically significant improvements in user top click position (-9.33%), CTR (+2.01%), Order (+2.04%), and Revenue (+1.69%) over the online multi-stage strategy, showing great potential in e-commercial conversion.

cs.IR↗

arXiv2025

Kimi k1.5: Scaling Reinforcement Learning with LLMs

Language model pretraining with next token prediction has proved effective for scaling compute but is limited to the amount of available training data. Scaling reinforcement learning (RL) unlocks a new axis for the continued improvement of artificial intelligence, with the promise that large language models (LLMs) can scale their training data by learning to explore with rewards. However, prior published work has not produced competitive results. In light of this, we report on the training practice of Kimi k1.5, our latest multi-modal LLM trained with RL, including its RL training techniques, multi-modal data recipes, and infrastructure optimization. Long context scaling and improved policy optimization methods are key ingredients of our approach, which establishes a simplistic, effective RL framework without relying on more complex techniques such as Monte Carlo tree search, value functions, and process reward models. Notably, our system achieves state-of-the-art reasoning performance across multiple benchmarks and modalities -- e.g., 77.5 on AIME, 96.2 on MATH 500, 94-th percentile on Codeforces, 74.9 on MathVista -- matching OpenAI's o1. Moreover, we present effective long2short methods that use long-CoT techniques to improve short-CoT models, yielding state-of-the-art short-CoT reasoning results -- e.g., 60.8 on AIME, 94.6 on MATH500, 47.3 on LiveCodeBench -- outperforming existing short-CoT models such as GPT-4o and Claude Sonnet 3.5 by a large margin (up to +550%).

cs.AI↗

arXiv2025

Light as Deception: GPT-driven Natural Relighting Against Vision-Language Pre-training Models

While adversarial attacks on vision-and-language pretraining (VLP) models have been explored, generating natural adversarial samples crafted through realistic and semantically meaningful perturbations remains an open challenge. Existing methods, primarily designed for classification tasks, struggle when adapted to VLP models due to their restricted optimization spaces, leading to ineffective attacks or unnatural artifacts. To address this, we propose \textbf{LightD}, a novel framework that generates natural adversarial samples for VLP models via semantically guided relighting. Specifically, LightD leverages ChatGPT to propose context-aware initial lighting parameters and integrates a pretrained relighting model (IC-light) to enable diverse lighting adjustments. LightD expands the optimization space while ensuring perturbations align with scene semantics. Additionally, gradient-based optimization is applied to the reference lighting image to further enhance attack effectiveness while maintaining visual naturalness. The effectiveness and superiority of the proposed LightD have been demonstrated across various VLP models in tasks such as image captioning and visual question answering.

cs.CV↗

arXiv2025

Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation

This paper studies off-policy evaluation (OPE) in reinforcement learning with a focus on behavior policy estimation for importance sampling. Prior work has shown empirically that estimating a history-dependent behavior policy can lead to lower mean squared error (MSE) even when the true behavior policy is Markovian. However, the question of why the use of history should lower MSE remains open. In this paper, we theoretically demystify this paradox by deriving a bias-variance decomposition of the MSE of ordinary importance sampling (IS) estimators, demonstrating that history-dependent behavior policy estimation decreases their asymptotic variances while increasing their finite-sample biases. Additionally, as the estimated behavior policy conditions on a longer history, we show a consistent decrease in variance. We extend these findings to a range of other OPE estimators, including the sequential IS estimator, the doubly robust estimator and the marginalized IS estimator, with the behavior policy estimated either parametrically or non-parametrically.

cs.LG↗

arXiv2025

Kimi-Audio Technical Report

We present Kimi-Audio, an open-source audio foundation model that excels in audio understanding, generation, and conversation. We detail the practices in building Kimi-Audio, including model architecture, data curation, training recipe, inference deployment, and evaluation. Specifically, we leverage a 12.5Hz audio tokenizer, design a novel LLM-based architecture with continuous features as input and discrete tokens as output, and develop a chunk-wise streaming detokenizer based on flow matching. We curate a pre-training dataset that consists of more than 13 million hours of audio data covering a wide range of modalities including speech, sound, and music, and build a pipeline to construct high-quality and diverse post-training data. Initialized from a pre-trained LLM, Kimi-Audio is continual pre-trained on both audio and text data with several carefully designed tasks, and then fine-tuned to support a diverse of audio-related tasks. Extensive evaluation shows that Kimi-Audio achieves state-of-the-art performance on a range of audio benchmarks including speech recognition, audio understanding, audio question answering, and speech conversation. We release the codes, model checkpoints, as well as the evaluation toolkits in https://github.com/MoonshotAI/Kimi-Audio.

eess.AS↗

arXiv2025

Linear-quadratic control for mean-field backward stochastic differential equations with random coefficients

In this paper, we study the linear-quadratic control problem for mean-field backward stochastic differential equations (MF-BSDE) with random coefficients. We first derive a preliminary stochastic maximum principle to analyze the unique solvability of the optimality system for this control problem through the variational method. Subsequently, we reformulate the mean-field linear-quadratic (MF-BSLQ) problem as a constrained BSDE control problem by imposing constraints on the expectation processes, which we solve using the Extended Lagrange multiplier method. Finally, we derive an explicit expression for the optimal control associated with Problem (MF-BSLQ).

math.OC↗

arXiv2025

Testing the cosmic distance duality relation using Type Ia supernovae and BAO observations

In this work, we propose to utilize the observed ratio of spherically-averaged distance to the sound horizon scale from Baryon Acoustic Oscillation (BAO) data to test the cosmic distance duality relation (CDDR) by comparing the luminosity distances (LDs) obtained from Type Ia supernovae (SNIa) observations with angular diameter distances (ADDs) derived from these ratio measurements, using a cosmological-model-independent method. To match the LDs with the ADDs at the identical redshifts, we employ two methods: a compressed form of the Pantheon sample and a hybrid approach that combines the binning method with an artificial neural network (ANN). The Hubble parameter $H(z)$ at any redshift is reconstructed from the observed Hubble parameter data with the ANN to derive the ADD. To avoid potential biases resulted from the specific prior values of the absolute magnitude $M_{\rm B}$ of SNIa and the sound horizon scale $r_{\rm d}$ from BAO measurements, we introduce the fiducial parameter $κ\equiv10^{M_{\rm B} \over 5}\, r_{\rm d}^{3 \over 2} $ and marginalize their impacts by treating them as nuisance parameters with flat prior distributions in our statistical analysis. Subsequently, we update the measurements of ratio of the transverse comoving distance to the sound horizon scale from the latest BAO data released by the Dark Energy Spectroscopic Instrument (DESI) collaboration for CDDR testing. Our results indicate that BAO observation provides a powerful tool for testing the CDDR, independent of both the absolute magnitude $M_{\rm B}$ and sound horizon scale $r_{\rm d}$, as well as any cosmological model.

astro-ph.CO↗

arXiv2025

Randomized Spectral Clustering for Large-Scale Multi-Layer Networks

Large-scale multi-layer networks with large numbers of nodes, edges, and layers arise across various domains, which poses a great computational challenge for the downstream analysis. In this paper, we develop an efficient randomized spectral clustering algorithm for community detection of multi-layer networks. We first utilize the random sampling strategy to sparsify the adjacency matrix of each layer. Then we use the random projection strategy to accelerate the eigen-decomposition of the sum-of-squared sparsified adjacency matrices of all layers. The communities are finally obtained via the k-means of the eigenvectors. The algorithm not only has low time complexity but also saves the storage space. Theoretically, we study the misclassification error rate of the proposed algorithm under the multi-layer stochastic block models, which shows that the randomization does not deteriorate the error bound under certain conditions. Numerical studies on multi-layer networks with millions of nodes show the superior efficiency of the proposed algorithm, which achieves clustering results rapidly. A new R package called MLRclust is developed and made available to the public.

stat.CO↗

arXiv2024

Spatial Interference Detection in Treatment Effect Model

Modeling the interference effect is an important issue in the field of causal inference. Existing studies rely on explicit and often homogeneous assumptions regarding interference structures. In this paper, we introduce a low-rank and sparse treatment effect model that leverages data-driven techniques to identify the locations of interference effects. A profiling algorithm is proposed to estimate the model coefficients, and based on these estimates, global test and local detection methods are established to detect the existence of interference and the interference neighbor locations for each unit. We derive the non-asymptotic bound of the estimation error, and establish theoretical guarantees for the global test and the accuracy of the detection method in terms of Jaccard index. Simulations and real data examples are provided to demonstrate the usefulness of the proposed method.

stat.ME↗

arXiv2024

MX precipitate behavior in an irradiated advanced Fe-9Cr steel: Helium effects on phase stability

Precipitates are main microstructural features to provide high temperature creep strength and radiation resistance in structural materials for fusion energy systems. However, the mechanisms of precipitate stability under irradiation in candidate structural materials for fusion first-wall and blanket components are poorly understood. In particular, the dual effects of helium transmutation and irradiation-induced damage on precipitate evolution have not been systematically studied in candidate materials, the leading of which are Fe-9Cr reduced activation ferritic/martensitic (RAFM) alloys. To fill this knowledge gap, a fundamental understanding of the single and combined interactions of helium (0-25 appm He/dpa), temperature (300-600C), and atomic displacements (15-100 dpa) on the behavior of MX (M=metal, X=C and/or N) precipitates in an advanced Fe-9Cr RAFM alloy were studied through the use of dual ion irradiation experiments. It was found that helium suppressed the diffusion-mediated mechanisms of precipitate stability (i.e. radiation-enhanced growth) at elevated temperatures and intermediate damage levels but had no effect on precipitate dissolution in the high dose conditions (>50 dpa). A precipitate stability model was used to rationalize the impacts of helium on ballistic dissolution and radiation-enhanced diffusion which are key contributors to overall precipitate stability. This is the second paper in a series of three to provide a systematic evaluation of MX precipitate behavior in RAFM steels under various fusion-relevant ion irradiation conditions.

cond-mat.mtrl-sci↗

arXiv2024

Testing the cosmic distance duality relation using Type Ia supernovae and radio quasars through model-independent methods

In this work, we perform a cosmological-model-independent test on the cosmic distance duality relation (CDDR) by comparing the angular diameter distance (ADD) obtained from the compact radio quasars (QSOs) with the luminosity distance (LD) from the Pantheon Type Ia supernovae (SNIa) sample. The binning method and Artificial Neural Network (ANN) are employed to match ADD data with LD data at the same redshift, and three different parameterizations are adopted to quantify the possible deviations from the CDDR. We initially investigate the impacts of the specific prior values for the absolute magnitude $M_{\rm B}$ from SNIa and the linear size scaling factor $l$ from QSOs on the CDDR test, demonstrating that these prior values introduce significant biases in the CDDR test. To avoid the biases, we propose a method independent of $M_{\rm B}$ and $l$ to test CDDR, which treats the fiducial value of a new variable $κ\equiv10^{M_{\rm B} \over 5}\,l$ as a nuisance parameter and then marginalize its impact with a flat prior in the statistical analysis. The results show that the CDDR is consistent with the observational data, and QSOs can serve as a powerful tool for testing the CDDR independent of cosmological models.

astro-ph.CO↗