DeepYard
R

ResidencyRL

RL framework for training medical AI agents through simulated clinical encounters

Open SourceFree

About

Research framework applying reinforcement learning to train LLM-based medical AI agents in simulated clinical environments. Agents develop clinical reasoning by practicing through thousands of virtual patient encounters with progressive autonomy levels and diverse feedback mechanisms. Designed for healthcare AI research exploring how agents can acquire clinical decision-making skills through iterative practice rather than static training.

Details

Language
Patterns

Tags

frameworkautonomousopen-sourcepythonevaluationmulti-agent