Compositional RL Agents That Follow Language Commands in Temporal Logic
Name
CBMM-Memo-127.pdf
Size
2.12 MB
Format
Adobe PDF
Checksum (MD5)
bfdc03fb04696869ba2c3a331f43557d
Author(s)
Kuo, Yen-Ling
Barbu, Andrei
Katz, Boris
Date Issued
July 19, 2021
Publisher
Center for Brains, Minds and Machines (CBMM), Frontiers in Robotics and AI
Series/Report no.
CBMM Memo;127
Abstract
We demonstrate how a reinforcement learning agent can use compositional recurrent neural net- works to learn to carry out commands specified in linear temporal logic (LTL). Our approach takes as input an LTL formula, structures a deep network according to the parse of the formula, and determines satisfying actions. This compositional structure of the network enables zero-shot generalization to sig- nificantly more complex unseen formulas. We demonstrate this ability in multiple problem domains with both discrete and continuous state-action spaces. In a symbolic domain, the agent finds a sequence of letters that satisfy a specification. In a Minecraft-like environment, the agent finds a sequence of actions that conform to a formula. In the Fetch environment, the robot finds a sequence of arm config- urations that move blocks on a table to fulfill the commands. While most prior work can learn to execute one formula reliably, we develop a novel form of multi-task learning for RL agents that allows them to learn from a diverse set of tasks and generalize to a new set of diverse tasks without any additional training. The compositional structures presented here are not specific to LTL, thus opening the path to RL agents that perform zero-shot generalization in other compositional domains.
Persistent DSpace Link
