Learning When to Stop: Adaptive Latent Reasoning via Reinforcement Learning

OffRLReLMLRMRe-assign community