Computer Science · Formal Methods

Jacqueline Mitchell

Reasoning rigorously about software — proving systems correct rather than merely testing them.

About

Jacqueline Mitchell

PhD Student @ USC

I'm a third-year Computer Science PhD student studying formal methods at the University of Southern California (USC), advised by Prof. Chao Wang. My research interests are in abstract interpretation, specifically in automating the challenging aspects of software verification and the design of verification tools. I am also interested in using formal techniques to improve the quality and security of LLM-generated code.

I am also an Operations Team Member at SIGPLAN-M, a wonderful group which pairs up mentors and mentees in the PL community. Check it out!

Prior to USC, I received my B.S. in Computer Science with a minor in Mathematics from UC Davis, where I was advised by Prof. Aditya V. Thakur.

Here's my CV!

Recent News

The proof continues.

Paper accepted at POPL 2027
Principles of Programming Languages · Denver, CO
Distinguished Paper Award, CAV 2026
Computer Aided Verification · Montréal
Research internship, Microsoft Research
RiSE group · Cambridge, UK
Invited talk: mechanized concurrent separation logic
Iris Workshop
Preprint: a verified register allocator
arXiv:2602.04417

Publications

Results, established.

2027

Every derivable triple holds in the operational semantics — verified programs are memory-safe and data-race free.

2026

The compiled program behaves exactly like its source.

Well-typed probabilistic programs never get stuck.

2025

Entailment in this fragment can be decided automatically.

Local guarantees compose into a global one.