NYU Langone Health OLAB

We’re interested in human and artificial intelligence. We use artificial intelligence (AI) to better understand and care for the human brain, and in study the human brain to build better AI.

We believe that investigations into the nature of artificial intelligence and human intelligence are fundamentally the same, and that by attempting to build better artificial intelligence we will discover fundamental insights into the key mechanisms behind human intelligence.

These investigations will augment our clinical work in treating neurological disorders and cancer, and lead the transition from ML guided investigations to ML guided interventions.

Papers

Some of our papers:

• Generalist Foundation Models Are Not Clinical Enough for Hospital Operations
• Generalist Large Language Models Outperform Clinical Tools on Medical Benchmarks
• Clinically Grounded Agent-based Report Evaluation: An Interpretable Metric for Radiology Report Generation
• Large-Scale Multi-omic Biosequence Transformers for Modeling Peptide-Nucleotide Interactions
• Medical large language models are easily distracted
• MedMobile: A mobile-sized language model with expert-level clinical capabilities
• CNS-Obsidian: A Neurosurgical Vision-Language Model Built From Scientific Publications

Models

Some of our models:

• LLaVA-Next-Med-OLAB
• MedMobile
• OmniBioTE