Full Professor, King's College London
1 paper at NeurIPS 2025
a new autoevaluation method that is provably better (at least no worse) than evaluation irrespective of the quality of the autoevaluator
