Automatic Data Augmentation for Generalization in Reinforcement...

Deep reinforcement learning (RL) agents often fail to generalize beyond their training environments. To alleviate this problem, recent work has proposed the use of data augmentation. However...

Verifying your browser | OpenReview Verifying your browser Complete the check below to continue to OpenReview Please complete the verification above. Have an OpenReview account? Sign in to skip this check.

related reading

• Just Ask for Generalization | Eric Jangevjang.com
• Quantifying Generalization in Reinforcement Learningarxiv.org
• [2103.06326] S4RL: Surprisingly Simple Self-Supervision for Offline Reinforcement Learningarxiv.org
• Sporks of AGIsergeylevine.substack.com
• The upcoming GPT-3 moment for RL | Mechanize, Inc.mechanize.work
• A Taxonomy of RL Environments for LLM Agentsleehanchung.github.io
• [2606.12016] Generalization Hacking: Models Can Game Reinforcement Learning by Preventing Behavioral Generalizationarxiv.org
• Reinforcement Learning With Verifiable Rewards: How Data and Verifiers Shape RLVRsnorkel.ai
• Introducing OpenReward | General Reasoninggr.inc
• How DeepMind's Generally Capable Agents Were Trained — LessWronglesswrong.com
• Noisy Data Breaks RLVRddkang.substack.com
• Deep Reinforcement Learning Doesn't Work Yetalexirpan.com