Open WorldDiscovery GitHub
← Research library
2026arXiv

RSI-Anything: Unlocking Recursive Self-Improvement with World Models

What this work does

Uses predictive world models to guide action and learning-data selection, while observed feedback controls policy or workflow improvement and conditional world-model adaptation.

THE RESEARCH CONVERSATION

Discussion & reactions

Share a question, an insight, or a replication experience. Sign in with GitHub to join the conversation.

Open GitHub discussions ↗