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

Key points

• We introduce RSI-Anything, a domain-general framework built on JEPA-Anything that uses predictive world models to guide action and learning-data selection.1
• Across diverse computational and scientific domains, RSI-Anything demonstrates substantial gains.1
• An end-to-end immune-science application further connects workflow evolution and predictive modeling with therapeutic hypothesis generation.1
• These results demonstrate how predictive world models and retained system updates enable recursive improvement across diverse environments.1

Sources (2)

• [1]RSI-Anything: Unlocking Recursive Self-Improvement with World Models
  Hugging Face Daily Papers · Oct 9, 12:00 AM

  Recursive self-improvement (RSI) aims to transform interaction experience into persistent improvements, yet external feedback is costly and world-model predictions may become unreliable as policies evolve.

  We introduce RSI-Anything, a domain-general framework built on JEPA-Anything that uses predictive world models to guide action and learning-data selection.

• [2]RSI-Anything: Unlocking Recursive Self-Improvement with World Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 05:50 PM · same content

Extractive summary: sentences quoted from the sources.