None defined yet.
Learning from the Self-future: On-policy Self-distillation for dLLMs
FutureSim: Replaying World Events to Evaluate Adaptive Agents