arXiv subjects

Yixin Han

Publications and source records attributed to Yixin Han.

5 recordsLinked to original sources

arXiv2025

RLRC: Reinforcement Learning-based Recovery for Compressed Vision-Language-Action Models

Vision-Language-Action models (VLA) have demonstrated remarkable capabilities and strong potential in complex robotic manipulation. However, their large parameter sizes and high inference latency hinder real-world deployment, especially on resource-constrained platforms. To address this, we conduct a systematic empirical study of model compression for VLAs. Building on these insights, we present \textit{RLRC}, a three-stage compression and recovery pipeline consisting of structured pruning, performance recovery via SFT and RL, and subsequent quantization. The RL stage incorporates a critic warm-up strategy and BC loss regularization to stabilize training and preserve policy behavior. RLRC achieves up to an 8 times memory reduction and 2.3 times inference speedup while maintaining the original task success rate. Extensive experiments across multiple VLA backbones show that RLRC consistently outperforms existing compression baselines, highlighting its effectiveness for on-device deployment. Project website: https://rlrc-vla.github.io

cs.RO↗

arXiv2024

FASTNav: Fine-tuned Adaptive Small-language-models Trained for Multi-point Robot Navigation

With the rapid development of large language models (LLM), robots are starting to enjoy the benefits of new interaction methods that large language models bring. Because edge computing fulfills the needs for rapid response, privacy, and network autonomy, we believe it facilitates the extensive deployment of large models for robot navigation across various industries. To enable local deployment of language models on edge devices, we adopt some model boosting methods. In this paper, we propose FASTNav - a method for boosting lightweight LLMs, also known as small language models (SLMs), for robot navigation. The proposed method contains three modules: fine-tuning, teacher-student iteration, and language-based multi-point robot navigation. We train and evaluate models with FASTNav in both simulation and real robots, proving that we can deploy them with low cost, high accuracy and low response time. Compared to other model compression methods, FASTNav shows potential in the local deployment of language models and tends to be a promising solution for language-guided robot navigation on edge devices.

cs.RO↗

arXiv2023

Leverage classifier: Another look at support vector machine

Support vector machine (SVM) is a popular classifier known for accuracy, flexibility, and robustness. However, its intensive computation has hindered its application to large-scale datasets. In this paper, we propose a new optimal leverage classifier based on linear SVM under a nonseparable setting. Our classifier aims to select an informative subset of the training sample to reduce data size, enabling efficient computation while maintaining high accuracy. We take a novel view of SVM under the general subsampling framework and rigorously investigate the statistical properties. We propose a two-step subsampling procedure consisting of a pilot estimation of the optimal subsampling probabilities and a subsampling step to construct the classifier. We develop a new Bahadur representation of the SVM coefficients and derive unconditional asymptotic distribution and optimal subsampling probabilities without giving the full sample. Numerical results demonstrate that our classifiers outperform the existing methods in terms of estimation, computation, and prediction.

stat.ME↗

arXiv2022

Model-free controlled variable selection via data splitting

Addressing the simultaneous identification of contributory variables while controlling the false discovery rate (FDR) in high-dimensional data is a crucial statistical challenge. In this paper, we propose a novel model-free variable selection procedure in sufficient dimension reduction framework via a data splitting technique. The variable selection problem is first converted to a least squares procedure with several response transformations. We construct a series of statistics with global symmetry property and leverage the symmetry to derive a data-driven threshold aimed at error rate control. Our approach demonstrates the capability for achieving finite-sample and asymptotic FDR control under mild theoretical conditions. Numerical experiments confirm that our procedure has satisfactory FDR control and higher power compared with existing methods.

stat.ME↗

arXiv2022

High-dimensional Inference and FDR Control for Simulated Markov Random Fields

Identifying important features linked to a response variable is a fundamental task in various scientific domains. This article explores statistical inference for simulated Markov random fields in high-dimensional settings. We introduce a methodology based on Markov Chain Monte Carlo Maximum Likelihood Estimation (MCMC-MLE) with Elastic-net regularization. Under mild conditions on the MCMC method, our penalized MCMC-MLE method achieves $\ell_{1}$-consistency. We propose a decorrelated score test, establishing both its asymptotic normality and that of a one-step estimator, along with the associated confidence interval. Furthermore, we construct two false discovery rate control procedures via the asymptotic behaviors for both p-values and e-values. Comprehensive numerical simulations confirm the theoretical validity of the proposed methods.

stat.ML↗