Picking training data for AI reasoning without labels or rewards

Original: Single-Rollout Hidden-State Dynamics for Training-Free RLVR Data Selection

Source: arxiv.org ↗