RSI-Anything: Unlocking Recursive Self-Improvement with World Models
- Published
- Source
- arXiv
- Paper number
- 1192
- Field
- LLMs / NLP
- arXiv ID
- 2610.13912
Key points
- Continuously checks how well the world model's predictions match actual observations, automatically deciding whether to update the policy or the world model.
- In code repair, adding world-model-guided selection and adaptation raised the success rate step by step from 26.3% to 29.2% to 31.2%.
- In chemical reaction optimization, it achieved a mean best observed yield of 96.15% within a 40-experiment budget, far ahead of Bayesian optimization at 82.16%.
- Laboratory planning scores more than doubled from 21.80% to 46.01% across five policy versions, showing that improvements accumulate.
- An immunology application connected workflow evolution with predictive modeling in an end-to-end demonstration of therapeutic hypothesis generation.
Paper links
External research summaries. These are not HDATF publications or measured product results.