Learning Individual Scientific Taste with Privileged Hindsight Distillation

Hanwei Fan, Yanqiao Wang, Yiqi Zhu, Xuanyu Lei, Peng Li, Yang Liu · arXiv preprint

Learns individual scientific taste from sparse pairwise feedback through shared criterion assessments, a population preference mapping, and researcher-specific adaptation. Privileged Hindsight Distillation uses later scientific evidence during training while preserving proposal-visible inference.