Pitt Clinical NLP and AI Innovation Laboratory

llms-vote

Code for the Paper "Large Language Models Vote: Prompting for Rare Disease Identification"

icl-minima

Code for the Paper "In-Context Learning Functions with Varying Number of Minima"

snn-for-fsl

Code for the Paper "Few-Shot Learning for Clinical Natural Language Processing Using Siamese Neural Networks: Algorithm Development and Validation Study"

HealthPrompt

Prompt Learning is a new paradigm in Natural Language Processing(NLP), which uses pre-trained language models to perform downstream tasks, avoiding fine tuning of a model. This zero-shot learning approach can be applied to healthcare applications, where there is limited availability of data. This projects aims to design prompt methodologies for downstream applications using PLMs, without the need for model training. This approach can change "Deep Learning" to "Deep Thinking" in Healthcare sector.

LLM4CDS

Synthetic patient datasets for the paper entitled "Enhancing Large Language Models for Clinical Decision Support by Incorporating Clinical Practice Guidelines"

pittnail.github.io

The official website for PittNAIL

AMIANLP2022

website for AMIA NLP Pre-symposium

AMIANLP2023

EBAIC

Website for The 1st International Workshop on Ethics and Bias of Artificial Intelligence in Clinical Applications (EBAIC)

Rare-disease-identification

Rare disease identification from free-text clinical notes with ontologies and weak supervision

CLEARdashboard