MedRSI: Stanford and Oxford unveil first recursive self-improvement framework for medical agents
rohanpaul_ai · x · 2026-09-23
Stanford+Oxford researchers (Junde Wu et al.) published MedRSI on arXiv, the first recursive self-improvement framework for medicine: agents continuously transform diagnostic failures into new clinical capabilities via tool composition and task-specific model training. To handle safety in clinical settings, it introduces two mechanisms — clinical-cost-aware failure prioritization (improving errors by potential clinical consequence, not frequency) and fast discovery with slow registration (new tools enter the persistent agent only after demonstrating sustained benefit across subsequent patient cohorts). On public glaucoma and heart disease benchmarks plus two private clinical tasks, MedRSI progressively develops segmentation, measurement, prediction, multimodal reasoning, and generative capabilities, surpassing manually engineered agents.
More from Research
- Autonomous agent edits just 1,024 weights to lift Qwen3.5 0.8B from 7.4% to 37.4% on MATH-500 — VictorSanh · 2026-11-12
- Gait dataset labeled by encounter, spanning pre-op to 3 months post-TKA — zawthet · 2026-11-12
- Walking alignment measured 2.5-3.1° more neutral than X-ray, iPad gait study finds — zawthet · 2026-11-12
- MIT's Tegmark coins 'formal-washing': Signal protocol proof is far from complete — tegmark · 2026-11-12
- Signal formal verification hits 80%, Tegmark says Wired overstated completion — tegmark · 2026-11-12
- Category-theoretic framework asks: when does a physical system actually compute? — neurovium · 2026-11-12