Mohammad Mahmoud Computer Science · UTM

Projects

Two kinds of work live here: lessons I built that you can try right now, and tools my student teams built under my supervision. On the second kind, the label on each card says who did what.

Interactive lessons

Every exercise asks for your answer before it shows you anything, and a wrong answer gets feedback on the mistake it most likely came from. Plain pages, no accounts; progress stays in your browser. All four lessons.

Built by student teams I supervised

Computing education2 papers

Varphi

Built by Hassan El-Sheikha, undergraduate, under my supervision

A description language and debugger for Turing machines. My CSC363 students use it to write and step through their machines. Published at WCCCE ’25 and ’26.

Quantum computingQCE ’26

Graph isomorphism on a quantum annealer

M. Akhtar, D. Wrona, R. Chowdhury, one undergraduate and two graduate students; I derived the formulations and directed the project

Three graph-isomorphism formulations, each refining the last, run on D-Wave’s Advantage2. The paper measures how much of the work classical pruning does before the annealer starts.

Theorem proving

Tiny Proof

Seven-student team, January–April 2025; I defined the problem and what counted as a working result

An AI-assisted prover for Lean 4 that extends DeepSeek-Prover-V1.5 with a tree search over swappable heuristics.

Interpretability

Transformer Visualizer

Five-student team, under my mentorship

A teaching tool for looking inside a language model, stage by stage, across the layers and attention heads of GPT-2 and BLOOM-560M.

Databases

Querying graph databases

Four-student team, CSC492, Winter 2026

A survey of how graph database systems answer multi-hop queries, with reproducible benchmarks of two published techniques and their limitations written down.

Model theoryNSERC USRA

Vaught’s conjecture

Summer research student, NSERC Undergraduate Student Research Award, 2026

Organizing and streamlining the recent results on a classic open problem in model theory, which also has a computability-theoretic equivalent.