2 papers
cs.LG2024
On the benefits of pixel-based hierarchical policies for task generalization
Tudor Cristea-Platon, Bogdan Mazoure, Josh Susskind +1
Reinforcement learning practitioners often avoid hierarchical policies, especially in image-based observation spaces. Typically, the single-task performance improvement over flat-p…
cs.LG2024
Large Language Models as Generalizable Policies for Embodied Tasks
Andrew Szot, Max Schwarzer, Harsh Agrawal +6
We show that large language models (LLMs) can be adapted to be generalizable policies for embodied visual tasks. Our approach, called Large LAnguage model Reinforcement Learning Po…