Generalization to New Sequential Decision Making Tasks with In-Context Learning
CoRR(2023)
摘要
Training autonomous agents that can learn new tasks from only a handful of
demonstrations is a long-standing problem in machine learning. Recently,
transformers have been shown to learn new language or vision tasks without any
weight updates from only a few examples, also referred to as in-context
learning. However, the sequential decision making setting poses additional
challenges having a lower tolerance for errors since the environment's
stochasticity or the agent's actions can lead to unseen, and sometimes
unrecoverable, states. In this paper, we use an illustrative example to show
that naively applying transformers to sequential decision making problems does
not enable in-context learning of new tasks. We then demonstrate how training
on sequences of trajectories with certain distributional properties leads to
in-context learning of new sequential decision making tasks. We investigate
different design choices and find that larger model and dataset sizes, as well
as more task diversity, environment stochasticity, and trajectory burstiness,
all result in better in-context learning of new out-of-distribution tasks. By
training on large diverse offline datasets, our model is able to learn new
MiniHack and Procgen tasks without any weight updates from just a handful of
demonstrations.
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