Scholar
Max Peeperkorn
Google Scholar ID: usCCE-gAAAAJ
PhD Candidate / Research Associate, University of Kent
Citations & Impact
Citations
301
H-index
6
i10-index
4
Publications
15
Co-authors
9
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Paper “Is Temperature the Creativity Parameter of Large Language Models?” featured on IBM Think and awarded the Artificial Intelligence Journal Best Student Paper Award at ICCC'24. Other publications include 'Mind the Gap: Conformative Decoding to Improve Output Diversity of Instruction-Tuned Large Language Models' (preprint), 'On characterizations of large language models and creativity evaluation' (14th International Conference on Computational Creativity), etc.
Research Experience
- Current research projects: creative application of Large Language Models; Neural Network Verification, working on learning branching heuristics as part of Elena Botoeva's BRIGHT project. Previously worked in the creative industries as a software engineer and graphic designer.
Education
- PhD candidate in Computer Science at the University of Kent, advised by Anna Jordanous and Dan Brown (University of Waterloo, Canada). MSc (cum laude) from the Media Technology program at Leiden University, Netherlands. Bachelor's degree in Graphic Design from the Royal Academy of Art, The Hague, Netherlands.
Background
- Research interests: creative application of Large Language Models, effects of hyperparameter choices, fine-tuning, and decoding strategies on their creative abilities. Specializes in Computer Science, particularly at the intersection of Natural Language Processing, Creativity, and Neural Network Verification.
Miscellany
- Personal interests: previously worked in the creative industries, specializing in unique (web-based) projects and art installations.
