Skip to content

Commit

Permalink
Merge pull request #38 from gardarjuto/master
Browse files Browse the repository at this point in the history
Add Mix-ME paper from ALOE
  • Loading branch information
Aneoshun authored Feb 1, 2024
2 parents fb9c650 + 8e768ac commit 40ac01b
Showing 1 changed file with 20 additions and 0 deletions.
20 changes: 20 additions & 0 deletions _data/paperlist.yml
Original file line number Diff line number Diff line change
@@ -1,5 +1,25 @@
papers:

- title: "Mix-ME: Quality-Diversity for Multi-Agent Learning"
authors:
- Garðar Ingvarsson
- Mikayel Samvelyan
- Bryan Lim
- Manon Flageat
- Antoine Cully
- Tim Rocktäschel
year: 2023
pdfurl: "https://arxiv.org/pdf/2311.01829.pdf"
bibtex: |
"@article{ingvarsson2023mixme,
title={Mix-ME: Quality-Diversity for Multi-Agent Learning},
author={Gar{\dh}ar Ingvarsson and Mikayel Samvelyan and Bryan Lim and Manon Flageat and Antoine Cully and Tim Rockt{\"a}schel},
journal={arXiv preprint arXiv:2311.01829},
year={2023} }"
tags:
- robotics
abstract: "In many real-world systems, such as adaptive robotics, achieving a single, optimised solution may be insufficient. Instead, a diverse set of high-performing solutions is often required to adapt to varying contexts and requirements. This is the realm of Quality-Diversity (QD), which aims to discover a collection of high-performing solutions, each with their own unique characteristics. QD methods have recently seen success in many domains, including robotics, where they have been used to discover damage-adaptive locomotion controllers. However, most existing work has focused on single-agent settings, despite many tasks of interest being multi-agent. To this end, we introduce Mix-ME, a novel multi-agent variant of the popular MAP-Elites algorithm that forms new solutions using a crossover-like operator by mixing together agents from different teams. We evaluate the proposed methods on a variety of partially observable continuous control tasks. Our evaluation shows that these multi-agent variants obtained by Mix-ME not only compete with single-agent baselines but also often outperform them in multi-agent settings under partial observability."

- title: "Diversity Preservation in Minimal Criterion Coevolution through Resource Limitation."
authors:
- Jonathan C. Brant
Expand Down

0 comments on commit 40ac01b

Please sign in to comment.