Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem
CoRR(2024)
摘要
Fine-tuning is a widespread technique that allows practitioners to transfer
pre-trained capabilities, as recently showcased by the successful applications
of foundation models. However, fine-tuning reinforcement learning (RL) models
remains a challenge. This work conceptualizes one specific cause of poor
transfer, accentuated in the RL setting by the interplay between actions and
observations: forgetting of pre-trained capabilities. Namely, a model
deteriorates on the state subspace of the downstream task not visited in the
initial phase of fine-tuning, on which the model behaved well due to
pre-training. This way, we lose the anticipated transfer benefits. We identify
conditions when this problem occurs, showing that it is common and, in many
cases, catastrophic. Through a detailed empirical analysis of the challenging
NetHack and Montezuma's Revenge environments, we show that standard knowledge
retention techniques mitigate the problem and thus allow us to take full
advantage of the pre-trained capabilities. In particular, in NetHack, we
achieve a new state-of-the-art for neural models, improving the previous best
score from 5K to over 10K points in the Human Monk scenario.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要