This repository provides the official implementation for the paper RePrompT: Recurrent Prompt Tuning for Integrating Structured EHR Encoders with Large Language Models.
In this paper, we introduce RePrompT, a time-aware large language model (LLM) framework that integrates structured EHR signals through soft prompt tuning. The goal of this work is to address issues that arise when applying LLM methods to EHR data; namely, (i) the limitations of free text in representing the temporal structure of medical codes, and (ii) the difficulty of modeling patient-to-patient similarity patterns that underpin the success of traditional EHR models.
To install the required dependencies, run the following command from the project's root directory:
pip install -r requirements.txt