Exodia Distilling SotA open-endedness through open-endedness.
Last updated: 2026-07-06 · upstream jennyzzt/awesome-open-ended@8933d28 · run #11

State-of-the-art concepts

Derived from 100 curated works via tfidf_kmeans analysis.

Key concepts (curated, ranked by prevalence)

How many of the 100 curated works engage each open-endedness concept, detected across full text + abstracts.

Theme clusters (the "majority vote")

agent, grand, automated

24 works · agent, grand, automated, science, meta

24 papers in this cluster
  1. Automated design of agentic systems
  2. Darwin Gödel Machine: Open-Ended Evolution of Self-Improving Agents
  3. Gödel Agent: A Self-Referential Agent Framework for Recursive Self-Improvement
  4. The AI Scientist: Towards fully automated open-ended scientific discovery
  5. Hyperagents
  6. The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search
  7. Learning to Continually Learn via Meta-learning Agentic Memory Designs
  8. Intelligent Go-Explore: Standing on the Shoulders of Giant Foundation Models
  9. Evolutionary Optimization of Model Merging Recipes
  10. Automated Capability Discovery via Model Self-Exploration
  11. Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution
  12. AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence
  13. Self-Taught Optimizer (STOP): Recursively Self-Improving Code Generation
  14. AgenticRed: Optimizing Agentic Systems for Automated Red-teaming
  15. Augmentative Topology Agents For Open-Ended Learning
  16. Practical PCG Through Large Language Models
  17. Structurally Flexible Neural Networks: Evolving the Building Blocks for General Agents
  18. The Science of Intelligent Exploration
  19. Evolving Prompts In-Context: An Open-ended, Self-replicating Perspective
  20. Open-endedness: The last grand challenge you’ve never heard of
  21. The Future of AI is Open-Ended
  22. AI That Can Improve Itself
  23. EvoCraft: A New Challenge for Open-Endedness
  24. Novel Opportunities in Open-Endedness

agent, objective, environment

16 works · agent, objective, environment, reward, knowledge

16 papers in this cluster
  1. MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge
  2. Open-Ended Learning Leads to Generally Capable Agents
  3. Language and Culture Internalisation for Human-Like Autotelic AI
  4. Motif: Intrinsic Motivation from Artificial Intelligence Feedback
  5. RAPid-Learn: A Framework for Learning to Recover for Handling Novelties in Open-World Environments
  6. Augmenting Autotelic Agents with Large Language Models
  7. Voyager: An Open-Ended Embodied Agent with Large Language Models
  8. Reward-Free Curricula for Training Robust World Models
  9. SPOTTER: Extending Symbolic Planning Operators through Targeted Reinforcement Learning
  10. OS-Copilot: Towards Generalist Computer Agents with Self-Improvement
  11. Powderworld: A Platform for Understanding Generalization via Rich Task Distributions
  12. Creative Problem Solving in Artificially Intelligent Agents: A Survey and Framework
  13. DreamCraft: Text-Guided Generation of Functional 3D Environments in Minecraft
  14. Managing extreme AI risks amid rapid progress
  15. Open-Endedness is Essential for Artificial Superhuman Intelligence
  16. Why Greatness Cannot Be Planned: The Myth of the Objective

intelligence, discovery, foundation

13 works · intelligence, discovery, foundation, era, agent

13 papers in this cluster
  1. Dreaming in Code for Curriculum Learning in Open-Ended Worlds
  2. OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Code
  3. CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery
  4. OMNI: Open-endedness via Models of human Notions of Interestingness
  5. Environment Generation for Zero-Shot Compositional Reinforcement Learning
  6. Safety is Essential for Responsible Open-Ended Systems
  7. General Intelligence Requires Rethinking Exploration
  8. In Search of the Ingredients of Open-Endedness: Replicating Picbreeder with Large Vision-Language Models
  9. Artificial Generational Intelligence: Cultural Accumulation in Reinforcement Learning
  10. Word2World: Generating Stories and Worlds through Large Language Models
  11. Bootstrapping Intelligence: Self-Improving Systems for Open-Ended Discovery
  12. Watts: Infrastructure for Open-Ended Learning
  13. Open-ended and AI-generating algorithms in the era of foundation models

rl, environment design, training

13 works · rl, environment design, training, policy, regret

13 papers in this cluster
  1. Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Design
  2. Replay-Guided Adversarial Environment Design
  3. MAESTRO: Open-Ended Environment Design for Multi-Agent Reinforcement Learning
  4. Discovering General Reinforcement Learning Algorithms with Adversarial Environment Design
  5. Grounding Aleatoric Uncertainty in Unsupervised Environment Design
  6. Evolving Curricula with Regret-Based Environment Design
  7. MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research
  8. Human-Timescale Adaptation in an Open-Ended Task Space
  9. Multi-Agent Diagnostics for Robustness via Illuminated Diversity
  10. Prioritized Level Replay
  11. Evolution and The Knightian Blindspot of Machine Learning
  12. Evolution through Large Models
  13. Executive Function: A Contrastive Value Policy for Resampling and Relabeling Perceptions via Hindsight Summarization?

quality diversity, qd, diverse

10 works · quality diversity, qd, diverse, high, search

10 papers in this cluster
  1. Quality Diversity through Human Feedback
  2. Large Language Models as In-context AI Generators for Quality-Diversity
  3. Quality-Diversity through AI Feedback
  4. Generative Design through Quality-Diversity Data Synthesis and Language Models
  5. Improving Deep Reinforcement Learning via Quality Diversity, Open-Ended and AI-Generating Algorithms
  6. Toward Artificial Open-Ended Evolution within Lenia using Quality-Diversity
  7. Foundation Model Self-Play: Open-Ended Strategy Innovation via Foundation Models
  8. Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts
  9. Eureka: Human-Level Reward Design via Coding Large Language Models
  10. Discovering Preference Optimization Algorithms with and for Large Language Models

interactive, base, llm

7 works · interactive, base, llm, video, testbed

7 papers in this cluster
  1. Genie: Generative Interactive Environments
  2. SerendipityLM: Interactive evolutionary exploration of generative design spaces with large language models
  3. Interactive poetry breeding through Mixtral base model LLMs
  4. Debating with More Persuasive LLMs Leads to More Truthful Answers
  5. DéjàQ: Open-Ended Evolution of Diverse, Learnable and Verifiable Problems
  6. Deep Laplacian-based Options for Temporally-Extended Exploration
  7. Video Games as a Testbed for Open-Ended Phenomena

innovation, world, paired trailblazer poet

7 works · innovation, world, paired trailblazer poet

7 papers in this cluster
  1. Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
  2. Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions
  3. Endlessly Generating Increasingly Complex and Diverse Learning Environments and their Solutions through the Paired Open-Ended Trailblazer (POET)
  4. Co-optimising Robot Morphology and Controller in a Simulated Open-Ended Environment
  5. Co-generation of game levels and game-playing agents
  6. Open-ended search for environments and adapted agents using map-elites
  7. Open-Endedness, World Models, and the Automation of Innovation

identifying, changing, evolution artificial life

5 works · identifying, changing, evolution artificial life, necessary condition

5 papers in this cluster
  1. Identifying Necessary Conditions for Open-Ended Evolution through the Artificial Life World of Chromaria
  2. Open Questions in Creating Safe Open-ended AI: Tensions Between Control and Creativity
  3. Identifying Life-Changing Books with LLMs
  4. Computational Life: How Well-formed, Self-replicating Programs Emerge from Simple Interaction
  5. Flow-Lenia: Towards open-ended evolution in cellular automata through mass conservation and parameter localization

Dominant vocabulary (consensus keyphrases)

  1. agent (2.1702)
  2. environment (1.7942)
  3. llm (1.3859)
  4. world (1.3409)
  5. language (1.3008)
  6. search (1.2987)
  7. design (1.2539)
  8. diverse (1.2289)
  9. reinforcement (1.2182)
  10. exploration (1.1796)
  11. large (1.1171)
  12. foundation (1.1034)
  13. domain (1.1026)
  14. diversity (1.0727)
  15. game (1.0386)
  16. training (1.0068)
  17. evocraft (1.0)
  18. generate (0.9931)
  19. discovery (0.9752)
  20. evolution (0.9544)