Learning Curricula In Open-ended Worlds

Abstract

Deep reinforcement learning (RL) provides powerful methods for training optimal sequential decision-making agents. As collecting real-world interactions can entail additional costs and safety risks, the common paradigm of sim2real conducts training in a simulator, followed by real-world deployment. Unfortunately, RL agents easily overfit to the choice of simulated training environments, and worse still, learning ends when the agent masters the specific set of simulated environments. In contrast, the real world is highly open-ended, featuring endlessly evolving environments and challenges, making such RL approaches unsuitable. Simply randomizing over simulated environments is insufficient, as it requires making arbitrary distributional assumptions and can be combinatorially less likely to sample specific environment instances that are useful for learning. An ideal learning process should automatically adapt the training environment to maximize the learning potential of the agent over an

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyjiang2023learning

Related papers

• Generating Automatic Curricula Via Self-supervised Active Domain Randomization (2020)0.00
• Statistical Reinforcement Learning In The Real World: A Survey Of Challenges And Future Directions (2026)0.00
• Curriculum Learning For Reinforcement Learning Domains: A Framework And Survey (2020)0.00
• Discovering Minimal Reinforcement Learning Environments (2024)0.00
• Continuous Coordination As A Realistic Scenario For Lifelong Learning (2021)0.00
• Can Learned Optimization Make Reinforcement Learning Less Difficult? (2024)0.00
• Open-ended Learning Leads To Generally Capable Agents (2021)0.00
• An Empirical Investigation Of The Challenges Of Real-world Reinforcement Learning (2020)0.00