arXiv subjects

Jinye Du

Publications and source records attributed to Jinye Du.

3 recordsLinked to original sources

arXiv2025

AKG kernel Agent: A Multi-Agent Framework for Cross-Platform Kernel Synthesis

Modern AI models demand high-performance computation kernels. The growing complexity of LLMs, multimodal architectures, and recommendation systems, combined with techniques like sparsity and quantization, creates significant computational challenges. Moreover, frequent hardware updates and diverse chip architectures further complicate this landscape, requiring tailored kernel implementations for each platform. However, manual optimization cannot keep pace with these demands, creating a critical bottleneck in AI system development. Recent advances in LLM code generation capabilities have opened new possibilities for automating kernel development. In this work, we propose AKG kernel agent (AI-driven Kernel Generator), a multi-agent system that automates kernel generation, migration, and performance tuning. AKG kernel agent is designed to support multiple domain-specific languages (DSLs), including Triton, TileLang, CPP, and CUDA-C, enabling it to target different hardware backends while maintaining correctness and portability. The system's modular design allows rapid integration of new DSLs and hardware targets. When evaluated on KernelBench using Triton DSL across GPU and NPU backends, AKG kernel agent achieves an average speedup of 1.46$\times$ over PyTorch Eager baselines implementations, demonstrating its effectiveness in accelerating kernel development for modern AI workloads.

cs.AI↗

arXiv2024

Distributed Empirical Likelihood Inference With or Without Byzantine Failures

Empirical likelihood is a very important nonparametric approach which is of wide application. However, it is hard and even infeasible to calculate the empirical log-likelihood ratio statistic with massive data. The main challenge is the calculation of the Lagrange multiplier. This motivates us to develop a distributed empirical likelihood method by calculating the Lagrange multiplier in a multi-round distributed manner. It is shown that the distributed empirical log-likelihood ratio statistic is asymptotically standard chi-squared under some mild conditions. The proposed algorithm is communication-efficient and achieves the desired accuracy in a few rounds. Further, the distributed empirical likelihood method is extended to the case of Byzantine failures. A machine selection algorithm is developed to identify the worker machines without Byzantine failures such that the distributed empirical likelihood method can be applied. The proposed methods are evaluated by numerical simulations and illustrated with an analysis of airline on-time performance study and a surface climate analysis of Yangtze River Economic Belt.

stat.ME↗

arXiv2024

Empirical Likelihood Inference over Decentralized Networks

As a nonparametric statistical inference approach, empirical likelihood has been found very useful in numerous occasions. However, it encounters serious computational challenges when applied directly to the modern massive dataset. This article studies empirical likelihood inference over decentralized distributed networks, where the data are locally collected and stored by different nodes. To fully utilize the data, this article fuses Lagrange multipliers calculated in different nodes by employing a penalization technique. The proposed distributed empirical log-likelihood ratio statistic with Lagrange multipliers solved by the penalized function is asymptotically standard chi-squared under regular conditions even for a divergent machine number. Nevertheless, the optimization problem with the fused penalty is still hard to solve in the decentralized distributed network. To address the problem, two alternating direction method of multipliers (ADMM) based algorithms are proposed, which both have simple node-based implementation schemes. Theoretically, this article establishes convergence properties for proposed algorithms, and further proves the linear convergence of the second algorithm in some specific network structures. The proposed methods are evaluated by numerical simulations and illustrated with analyses of census income and Ford gobike datasets.

stat.ME↗