Our Research Philosophy

The research activities of our lab focusses on the theoretical foundations of machine intelligence and its applications to a variety of real-world problems.

Our research philosophy is to conduct work that is rigorous and firmly grounded in theory. We place particular emphasis on applying the scientific method in a systematic way, ensuring that research questions, hypotheses, and methodologies are carefully developed and clearly justified. A key aspect of this approach is our commitment to robust evaluation. We prioritise transparent research design and thorough analysis to ensure that the experimental findings are reliable and reproducible.

Our approach is also highly interdisciplinary, and we regularly collaborate with researchers from fields including economics, political science, anthropology, neuroscience, psychology, and other social and behavioural sciences. We actively draw theoretical and methodological inspiration from these disciplines to inform the development and evaluation of machine intelligence algorithms and systems.

Themes

Current areas of interest include:

• Theory and applications of machine learning;
• Multi-agent systems based on foundational models/large language models and/or reinforcement learning;
• AI-based decision-making (in single-agent and multi-agent scenarios);
• AI safety and security;
• AI creativity;
• AI and society;
• Machine learning for mobile, IoT, networked, and cyber-physical systems.

Last updated: 9 May 2026.