Xujie Si

This paper presents an efficient tree diffing method using SAT solving.

Research focuses on improving neural network verification with minimal NAP specifications.

A new algorithm improves code refinement using LLMs more efficiently.

APPL streamlines development with large language models using an intuitive, Python-like syntax.

Tyro enhances type error identification for programmers, especially beginners.

Researchers enhance image and motion generation based on user instructions.