A Definition of Open-Ended Learning Problems for Goal-Conditioned Agents

arxiv:2311.00344

Original source · Paper JSON

Agents, open-endedness, AGI · 2023-11-01

Cites (1) · Cited by (1)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2
3. Part 3
4. Part 4

Contents

• A Definition of Open-Ended Learning for Goal-Conditioned Agents
• Abstract
• 1 Introduction
• 2 Defining OEL: Elements from the literature
• 2.1 Genealogy of the notion
• 2.2 Current perspectives
• 3 Open-ended learning: a definition
• 3.1 Open-ended process
• 3.2 Open-ended RL
• 3.3 Open-ended GCRL
• 3.4 First-order and second-order open-ended GCRL
• 4 Combining open-ended learning with other properties
• 4.1 Goal-conditioned continual reinforcement learning
• 4.2 Lifelong open-ended learning
• 4.3 The origin of goals: extrinsic, autotelic, and teachable open-ended learning agents
• 5 Evaluating open-ended learning agents
• 5.1 Assessing the open-ended learning property
• 5.2 Comparing open-ended learning agents
• 6 Discussion and Conclusion
• Acknowledgements