Yanli Wang

Yanli Wang

I am a master's student at SYSU

About Me

Introduction

I am a master’s student at the School of Software Engineering, Sun Yat-sen University. My supervisor is Prof. Yanlin Wang and thank her for her multi-faceted supports. Before going to Sun Yat-sen University, I received the B.S. degree in E-Commerce from South China University of Technology in June 2023.

My research mainly focuses on: (1) Large Language Model; (2) Code Intelligence.

Educations

  • Sun Yat-sen University
    M.S. in Software Engineering, August 2023 - Present.

  • South China University of Technology
    B.S. in E-Commerce, August 2019 - June 2023.

Publications

SabreCoder: From Logical to Computational Sparsity: Structure-Aware Block-Sparse Attention for Long Code Completion
ACL 2026 Paper | Code
Yanli Wang, Yanlin Wang, Bowen Zhang, Yiwei Zhang, Daya Guo, Jiachi Chen, Hongyu Zhang, Zibin Zheng
TLDR: SabreCoder is a training-free structure-aware block-sparse attention mechanism that bridges logical and computational sparsity for long code completion. By parsing code into semantic chunks, constructing chunk-level sparse patterns via dependency analysis and similarity matching, and mapping them to GPU-friendly block-sparse formats, SabreCoder reduces Time-To-First-Token (TTFT) by 45-55% while maintaining accuracy within 3% of dense attention.

RepoTransBench: A Real-World Multilingual Benchmark for Repository-Level Code Translation
TSE 2025 Paper | Code
Yanli Wang, Yanlin Wang, Suiquan Wang, Daya Guo, Jiachi Chen, John Grundy, Xilin Liu, Yuchi Ma, Mingzhi Mao, Hongyu Zhang, Zibin Zheng
TLDR: RepoTransBench is a comprehensive repository-level code translation benchmark featuring 1,897 real-world repository samples across 13 language pairs with automatically executable test suites. Unlike previous fine-grained benchmarks that focus on snippets, functions, or files, RepoTransBench addresses real-world demands where entire repositories need translation.

AlignCoder: Aligning Retrieval with Target Intent for Repository-Level Code Completion
ASE 2025 Paper | Code
Tianyue Jiang, Yanli Wang, Yanlin Wang, Daya Guo, Ensheng Shi, Yuchi Ma, Jiachi Chen, Zibin Zheng
TLDR: AlignCoder generates multiple candidate completions to create enhanced queries that bridge the semantic gap between initial queries and target code, using reinforcement learning to train a retriever for more accurate repository-level code completion.

RLCoder: Reinforcement Learning for Repository-Level Code Completion
ICSE 2025 Paper | Code
Yanlin Wang, Yanli Wang, Daya Guo, Jiachi Chen, Ruikai Zhang, Yuchi Ma, and Zibin Zheng
TLDR: RLCoder is a novel reinforcement learning framework, which can enable the retriever to learn to retrieve useful content for code completion without the need for labeled data, and introduces a stop signal mechanism.

Awards

  • National Scholarship, 2025
  • 1st Class Graduate Scholarship at SYSU, 2024 & 2025
  • 2nd Class Graduate Scholarship at SYSU, 2023
  • Zhaolian Finance Scholarship, 2024
  • 1st Prize, Baidu Business AI Technology Innovation Competition (South China Region, ¥3000), 2024
  • 1st Class Scholarship at SCUT, 2022
  • 2nd Class Scholarship at SCUT, 2020
  • Merit Student at SCUT, 2020 & 2022
  • 2nd Prize, International Mathematical Contest in Modeling, 2022
  • 1st Prize, National College Student Statistical Modeling Competition (Guangdong), 2022
  • 2nd Prize, Youxin Cloud Cup: Big Data Management and Application Innovation Competition – Rural Revitalization Track (¥5000), 2022
  • 2nd Prize, National College Student Mathematical Modeling Competition (Guangdong), 2021
  • 2nd Prize, National College Student Mathematical Competition (Guangdong), 2021
  • 2nd Prize, National Middle School Computer Production Competition (Henan), 2018