4 papers
Demystifying On-Policy Distillation: Where Does the Teacher's Advantage Live?
Yiping Li, Fan Zhao, Zhibo Hou +2
On-policy distillation (OPD) is increasingly studied as a cheaper alternative to reinforcement learning, yet it remains a form of knowledge distillation (KD) that relies on teacher…
From Memorization to Absorption: Mixed-Policy RL for Continual Knowledge Injection
Zhibo Hou, Fan Zhao, Zhiyu An +1
Continual knowledge injection is essential for keeping large language models up-to-date in a fast-evolving world. Existing methods rely on supervised fine-tuning (SFT), which memor…
Beyond Noisy-TVs: Noise-Robust Exploration Via Learning Progress Monitoring
Zhibo Hou, Zhiyu An, Wan Du
When there exists an unlearnable source of randomness (noisy-TV) in the environment, a naively intrinsic reward driven exploring agent gets stuck at that source of randomness and f…
Disentangling Uncertainties by Learning Compressed Data Representation
Zhiyu An, Zhibo Hou, Wan Du
We study aleatoric and epistemic uncertainty estimation in a learned regressive system dynamics model. Disentangling aleatoric uncertainty (the inherent randomness of the system) f…