RSI-Anything uses world models for recursive self-improvement, lifting lab yields to 96%
jiqizhixin · x · 2026-10-12
A new arXiv paper introduces RSI-Anything, a domain-general recursive self-improvement (RSI) framework built on JEPA-Anything that works across coding, GUI, planning and science tasks. It uses predictive world models to guide action and training-data selection, while component and compatibility checks decide whether updated policy-model pairs are retained.
Key results:
- Code repair: with the coder fixed, success rises from 26.3% (static verifier) to 29.2% (predictive selection) and 31.2% after world-model adaptation, at similar test usage.
- GUI tasks: higher completion rates with 39.4% fewer average actions.
- Lab planning: scores more than double from 21.80% to 46.01% across five policy versions.
- Reaction optimization: 96.15% mean best yield vs 82.16% for Bayesian optimization under the same 40-experiment budget.
An end-to-end immune-science application ties workflow evolution to therapeutic hypothesis generation. Paper and code are public.
Related event: RSI-Anything: World models enable recursive self-improvement across domains(4 posts)→
More from coding & agent
- 3,300+ open-source AI agent tools in one auto-updating, categorized list — Strict-Quarter7214 · 2026-10-12
- yoagent: a Rust agent runtime that reuses pi extensions and DeepSeek Harness plugins — liyuanhao · 2026-10-12
- xses: a unified CLI to sync and resume AI coding agent sessions across 5 tools — Choice_Concern_5387 · 2026-10-12
- 15 role-based prompts claim to replace a $114,500 ML education, from theory to MLOps — bigaiguy · 2026-10-12
- Use small models for grunt work, save SOTA for orchestration, says agentic engineering practitioner — petergyang · 2026-10-12
- Screego: open-source self-hosted screen sharing for devs with 10.5k GitHub stars — thisguyknowsai · 2026-10-12