From Few to Many: Self-Improving Many-Shot Reasoners Through Iterative Optimization and Generation.

Xingchen Wan, Han Zhou, Ruoxi Sun, Sercan . Arik

Browse the full ICLR paper archive.