Gödel Agent: A Self-Referential Agent Framework for Recursive Self-Improvement

arxiv:2410.04444

Original source · Paper JSON

Agents, open-endedness, AGI · 2024-10-06

Cites (4) · Cited by (2)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

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

Contents

• Gödel Agent: A Self-Referential Agent Framework for Recursively Self-Improvement
• Abstract
• 1 Introduction
• 2 Related Work
• 3 Self-Referential Gödel Agent
• 4 Gödel Agent Implementation
• 4.1 Implementation Details
• 4.2 Additional Designs
• 5 Experiments
• 5.1 Baseline Methods
• 5.2 Experimental Settings
• 5.3 Experimental Results and Analysis
• 6 Analysis
• 6.1 Analysis of Initial Tools
• 6.2 Robustness Analysis of the Agent
• 6.3 Case Study: Game of 24
• 7 Discussions and Future Directions
• 7.1 Discussions
• 7.2 Future Directions
• 8 Conclusion
• Limitations
• Ethics Statement
• Appendix A Goal Prompt of Gödel Agent
• Appendix B Experiment Details
• Appendix C Representative Policies Improved by Gödel Agent
• C.1 Codes of the Best Policies Found by Gödel Agent Across Four Tasks
• C.2 Codes in Game of 24 Tasks
• Appendix D Cost of Experiments
• Appendix E Additional Novel Policies Designed by Gödel Agent
• Appendix F Comparison Between Random Sampling and Gödel Agent Performance
• Instructions for reporting errors