about

I am an MSc Computer Science candidate at the University of the Western Cape and the Centre for Artificial Intelligence Research (CAIR), supervised by Thomas Meyer (University of Cape Town) and Louise Leenen (University of the Western Cape). I work in formal non-monotonic reasoning: logic that draws sensible default conclusions and then revises them when a more specific case overrides the default. My thesis builds on the KLM framework, a standard mathematical account of how that revision should behave, and uses it to give learned rule sets a proper logical reading. The explanation lives in the rules themselves, not in a story fitted around a black box after the fact.

Moving onto a PhD in neurosymbolic AI, I want to focus on the inner workings rather than post-hoc explanations: how logical structure can be embedded into a neural network during training, and how a network might learn its own symbolic representations directly from data.

education

MSc Computer Science
University of the Western Cape
submitting 2026

BSc (Hons) Computer Science, cum laude
University of the Western Cape
2024

BSc Computer Science, cum laude
University of the Western Cape
2023

research

2026
Toward Defeasible Machine Learning: When Learned Rule Sets Become Defeasible Theories
Ruvarashe Madzime, Thomas Meyer
24th International Workshop on Nonmonotonic Reasoning (NMR 2026), Federated Logic Conference (FLoC), Lisbon, Portugal

2026
Toward a Defeasible Semantics for Symbolic Classifiers
Ruvarashe Madzime
Doctoral Consortium, 23rd International Conference on Principles of Knowledge Representation and Reasoning (KR 2026 DC), Federated Logic Conference (FLoC), Lisbon, Portugal

2026
First Steps Towards Human-AI Ranking Aggregation
Jonas Karge, Roy Ferguson, Daniel Grimaldi, Jonas Haldimann, Ruvarashe Madzime, Thomas Meyer
Joint Workshop on Statistics and Knowledge Integration for Logic, Learning, Ethical Decisions, and LLMs (SKILLED-LLMs 2026), Federated Logic Conference (FLoC), Lisbon, Portugal

2025
An Override-Aware Classifier for Transparent AI
Ruvarashe Madzime, Thomas Meyer, Louise Leenen
6th Southern African Conference on Artificial Intelligence Research (SACAIR 2025), Cape Town, South Africa

2024
Enhanced Electronic Health Record Text Summarisation Using Large Language Models
Ruvarashe Madzime, Clement Nyirenda
5th Southern African Conference on Artificial Intelligence Research (SACAIR 2024), Bloemfontein, South Africa