Using LLMs for safe low-level programming
Episode 5
Duration:
Talk
Speakers:
.png)
.png)
Talk
Details
This talk covered two technical results from ICSE'2025 on using Large Language Models (LLMs) for safe low-level programming. The results demonstrate LLMs inferring machine-checkable memory safety invariants in legacy C code, and how LLMs assist in fixing compilation errors in Rust codebases.
Session Code: E5H