Scholar
Robert Kirk
Google Scholar ID: PL5KWdYAAAAJ
Research Scientist, UK AI Security Institute
Citations & Impact
Citations
1,279
H-index
11
i10-index
12
Publications
20
Co-authors
114
list available
Publications
Resume (English only)
Academic Achievements
- Published several papers including 'A Survey of Generalisation in Deep Reinforcement Learning', 'Graph Backup: Data Efficient Backup Exploiting Markovian Data', and 'MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research'.
Research Experience
- Currently a PhD student at UCL DARK Lab, supervised by Tim Rocktäschel and Edward Grefenstette. Worked as an infrastructure and software engineer at Smarkets for two years before starting the PhD.
Education
- Completed an integrated master's degree in mathematics and computer science at Somerville College, University of Oxford. Focused on abstract computer science in the fourth year and completed a master's thesis on implementing the comonadic formulation of pebble games in Idris, supervised by Samson Abramsky.
Background
- Research interests include generalisation in reinforcement learning, out-of-distribution robustness, and AI safety and alignment. Aspiring effective altruist and rationalist.
Miscellany
- Writes for the Alignment Newsletter, primarily on interpretability and reinforcement learning. Personal interests include rationality, neuroscience, cooking, videogames, music (basically any genre), and reading. Sings in a choir and previously played the French horn.
