Projected Task-Specific Layers for Multi-Task Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Roberts, Josselin Somerville, Di, Julia
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910354583322624
author Roberts, Josselin Somerville
Di, Julia
author_facet Roberts, Josselin Somerville
Di, Julia
contents Multi-task reinforcement learning could enable robots to scale across a wide variety of manipulation tasks in homes and workplaces. However, generalizing from one task to another and mitigating negative task interference still remains a challenge. Addressing this challenge by successfully sharing information across tasks will depend on how well the structure underlying the tasks is captured. In this work, we introduce our new architecture, Projected Task-Specific Layers (PTSL), that leverages a common policy with dense task-specific corrections through task-specific layers to better express shared and variable task information. We then show that our model outperforms the state of the art on the MT10 and MT50 benchmarks of Meta-World consisting of 10 and 50 goal-conditioned tasks for a Sawyer arm.
format Preprint
id arxiv_https___arxiv_org_abs_2309_08776
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Projected Task-Specific Layers for Multi-Task Reinforcement Learning
Roberts, Josselin Somerville
Di, Julia
Machine Learning
Artificial Intelligence
Robotics
Multi-task reinforcement learning could enable robots to scale across a wide variety of manipulation tasks in homes and workplaces. However, generalizing from one task to another and mitigating negative task interference still remains a challenge. Addressing this challenge by successfully sharing information across tasks will depend on how well the structure underlying the tasks is captured. In this work, we introduce our new architecture, Projected Task-Specific Layers (PTSL), that leverages a common policy with dense task-specific corrections through task-specific layers to better express shared and variable task information. We then show that our model outperforms the state of the art on the MT10 and MT50 benchmarks of Meta-World consisting of 10 and 50 goal-conditioned tasks for a Sawyer arm.
title Projected Task-Specific Layers for Multi-Task Reinforcement Learning
topic Machine Learning
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2309.08776