DATEXIS/DeepICD-R1-zero-32B
TEXT GENERATIONConcurrency Cost:2Model Size:32.8BQuant:FP8Ctx Length:32kPublished:Mar 15, 2026License:otherArchitecture:Transformer Cold

DATEXIS/DeepICD-R1-zero-32B is a 32.8 billion parameter clinical reasoning model from DATEXIS, designed for ICD-10-CM diagnosis outcome prediction from admission notes. It utilizes the DeepICD-R1 framework, treating diagnosis prediction as a reinforcement learning task optimized with structured reward signals. This "R1-Zero" style model was trained primarily through reinforcement learning without supervised fine-tuning, allowing reasoning behaviors to emerge autonomously. It excels at generating reasoning traces and predicting single ICD-10-CM diagnoses from clinical text for research purposes.

Loading preview...