SOURCE-LINKED INTELLIGENCE

ThinkPrior: Zero-Rollout Difficulty Priors for Cold-Start Prompt Selection in RLVR

arXiv · AI, language, vision and robotics · article · Sep 8, 2026 · UTC

In reinforcement learning with verifiable rewards (RLVR) trained with group relative policy optimization (GRPO), the KL-free reward-advantage term studied here depends on within-group reward variation. If all rollouts in a group are correct or all are wrong, their group-relative advantages are identically zero; these zero-advantage silent groups provide no reward-advantage gradient, yet uniform sampling spends 39% of a run's rollouts on them. History-based prompt selection must first spend target-policy rollouts to estimate difficulty, creating a cold start with rollout waste; ThinkPrior inste

Read original source ↗ Open in workspace

recordType

paper

region

Global