Guiding AI reasoning exploration without expensive trial-and-error scaling

Original: Nudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVR

Source: arxiv.org ↗