None defined yet.
Latent On-Policy Self-Distillation
ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning