James Lee-Jones

I am a PhD student in the FastPL group at Imperial College London supervised by Prof. Alastair Donaldson.

My research aims to improve the reliability of software, particularly GPU programs, via formal analysis, software testing, and verification.

Teaching

In 2026–2027 I will be teaching the practical half of a third year BEng/MEng course on The Theory and Practice of Concurrent Programming with Azalea Raad.

As a graduate teaching assistant, I have helped with the following:

MSc Co-supervision (25–26): Co-supervising an MSc student exploring the use of skeletal program enumeration to automatically test compilers for the WebGPU Shading Language (WGSL).
C Project Mentorship (25–25): Mentored groups of first year students tasked with writing an emulator and assembler for the ARMv8 AArch64 assembly langauge.
Software Reliability (25–26): Delivered tutorials, Coursework Marking, and Group Presentation Marking.
Computing Practical 2 (24–25, 25–26): Performed live code reviews of student operating systems and compilers.

News

Jul 25, 2026	Our tool paper, Dredd: Scalable Source-Level Mutation Testing for Large C/C++ Codebases (Donaldson et al., 2026), has been accepted to the Tool Demonstrations track at ISSTA 2026! Dredd is available on GitHub.
Jul 10, 2026	In 2026–2027 I will be teaching the practical half of a third year BEng/MEng course on The Theory and Practice of Concurrent Programming with Azalea Raad.
Jun 17, 2026	I presented my work on the Uniformity Analysis in the WebGPU Shading Language (Lee-Jones et al., 2026) at PLDI 2026. The live-streamed talk is available on YouTube.
May 29, 2026	I am serving on the Artifact Evaluation Committee for ISSTA 2026.
Apr 04, 2026	My first paper, Uniformity Analysis in the WebGPU Shading Language (Lee-Jones et al., 2026), has been accepted to PLDI 2026!

Latest Posts

Feb 08, 2025	Hello World!

Publications

ISSTA

Dredd: Scalable Source-Level Mutation Testing for Large C/C++ Codebases

Alastair F. Donaldson, James Lee-Jones, and Jonathan Foo

In Proceedings of the 35th ACM SIGSOFT International Symposium on Software Testing and Analysis, ISSTA Companion, 2026

To appear

PDF Code
PLDI

Uniformity Analysis in the WebGPU Shading Language

James Lee-Jones, John Wickerson, and Alastair F. Donaldson

Proc. ACM Program. Lang., Jun 2026

Abs DOI PDF

The WebGPU programming model brings general-purpose GPU programming to the web, allowing untrusted JavaScript to issue parallel workloads to client GPUs. To ensure reliability, WebGPU mandates uniformity analysis—a static check that rejects programs that could cause barrier divergence, a GPU control-flow error that can hang execution and require OS-level recovery. While traditional GPU models treat barrier divergence as undefined behaviour, WebGPU’s need for safety and reliability makes this unacceptable. We present the first comprehensive formal and practical study of uniformity analysis in WebGPU, identifying four key issues and making corresponding contributions: (1) Lack of definition: The analysis currently defines non-uniform programs only as those it rejects, without an independent notion of barrier divergence. We provide an operational semantics for TinyWGSL, a core calculus of the WebGPU Shading Language (WGSL), developed in consultation with WGSL specification editors and implementers. This semantics rigorously defines barrier divergence in the context of WGSL for the first time. (2) Complex specification: The current description of uniformity analysis in WGSL is lengthy and imprecise. We reformulate it via concise formal rules for TinyWGSL and argue their soundness with respect to our semantics. (3) Lack of soundness and precision: Using our semantics, we expose soundness and precision flaws in the analysis as presented in the WGSL specification. In response, we have proposed four significant changes to the specification (all accepted). (4) Testing difficulty: The complex, semi-formal definition of uniformity in the WGSL specification makes it hard to test implementations. We mechanise uniformity analysis in Alloy and use Alloy’s test generation to stress-test the Chromium implementation, revealing specification–implementation discrepancies and a bug in the Chromium implementation. Overall, our work resolves an important GPU programming problem (rigorously defining barrier divergence), brings an interesting new analysis to the attention of the PL community, and demonstrates the impact of applying formal PL techniques to an important industrial language specification.
arXiv

Capture the Flags: Family-Based Evaluation of Agentic LLMs via Semantics-Preserving Transformations

Shahin Honarvar, Amber Gorzynski, James Lee-Jones, and 4 more authors

Jun 2026

arXiv PDF