collaborators

5 papers

cs.SE2026

WarmTuner: Program-Specific Warm Starts for Compiler Autotuning via Offline-to-Online Reinforcement Learning

Tianlu Qiao, Mingxuan Zhu, Zeyu Sun +1

Compilers are fundamental software tools that translate high-level programs into machine code. Modern compilers expose hundreds of optimizations, each turned on or off through an o…

cs.SE2026

GapForge: Directed Compiler Fuzzing via Coverage-Gap Analysis

Mingxuan Zhu, Qingyuan Liang, Junjie Chen +2

Modern compiler codebases (e.g., GCC and LLVM) are large and complex, making comprehensive coverage across diverse code regions highly challenging. Most existing test generation te…

cs.CL2026

The Detection-Extraction Gap: Models Know the Answer Before They Can Say It

Hanyang Wang, Mingxuan Zhu

Modern reasoning models continue generating long after the answer is already determined. Across five model configurations, two families, and three benchmarks, we find that 52--88%…

cs.SE2025

Directional Diffusion-Style Code Editing Pre-training

Qingyuan Liang, Zeyu Sun, Qihao Zhu +6

Code pre-trained models have shown promising effectiveness in various software engineering tasks. Among these tasks, many tasks are related to software evolution and/or code editin…

cs.SE2025

Isolating Compiler Bugs through Compilation Steps Analysis

Yujie Liu, Mingxuan Zhu, Shengyu Cheng +1

Compilers are essential to software systems, and their bugs can propagate to dependent software. Ensuring compiler correctness is critical. However, isolating compiler bugs remains…