Undergrad student, Imperial College London
1 paper at NeurIPS 2025
We introduce a model of prediction with limited selectivity, and prove instance-dependent bounds on the optimal error rate.
