Dinghao Pan

Prototype Tuning: A Meta-Learning Approach for Few-Shot Document-Level Relation Extraction with Large Language Models

Few-Shot Document-Level Relation Extraction (FSDLRE) aims to develop models capable of generalizing to new categories with minimal support examples. Although Large Language Models (LLMs) demonstrate exceptional In-Context Learning (ICL) capabilities on many few-shot tasks, their performance on FSDLR…