Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions

arxiv:1901.01753

Original source · Paper JSON

Agents, open-endedness, AGI · 2019-01-07

Cites (1) · Cited by (6)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2
3. Part 3
4. Part 4
5. Part 5
6. Part 6

Contents

• Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
• Abstract
• 1 Introduction
• 2 Background
• 2.1 Behavioral Diversity and Stepping Stones
• 2.2 Open-Ended Search via Minimal Criterion Coevolution (MCC)
• 2.3 Evolution Strategies (ES)
• 3 The Paired Open-Ended Trailblazer (POET) Algorithm
• 4 Experiment Setup And Results
• 4.1 Environment and Experiment Setup
• 4.2 Environment Encoding and Mutation
• 4.3 Results
• 5 Discussion, Future Work, and Conclusion
• Acknowledgments
• 6 Supplemental Information
• 6.1 Producing New Environments
• 6.2 ES-Based Transfer
• Instructions for reporting errors