TY - RPRT TI - From Few to Many: Self-Improving Many-Shot Reasoners Through Iterative Optimization and Generation AU - Xingchen Wan AU - Han Zhou AU - Ruoxi Sun AU - Hootan Nakhost AU - Ke Jiang AU - Sercan Ö. Arık PY - 2025 UR - https://arxiv.org/abs/2502.00330 ID - 2502.00330 ER -