The field of natural language processing is witnessing significant advancements in the application of large language models (LLMs) to biomedical domains. Recent developments indicate a growing trend towards leveraging LLMs to improve the accuracy and reliability of biomedical information extraction, ontology alignment, and named entity recognition. The integration of retrieval-augmented generation (RAG) frameworks and dynamic prompting strategies has been shown to substantially enhance the performance of LLMs in these tasks. Furthermore, the use of LLMs as oracles for ontology alignment and the development of efficient evaluation methodologies, such as EffiEval, are notable innovations. Noteworthy papers include: Retrieval Augmented Large Language Model System for Comprehensive Drug Contraindications, which achieved significant improvements in model accuracy for drug contraindication information. Another notable paper is ARCE: Augmented Roberta with Contextualized Elucidations for Ner in Automated Rule Checking, which established a new state-of-the-art on a benchmark AEC dataset. Overall, these advancements demonstrate the potential of LLMs to transform the field of biomedical research and applications.
Advances in Large Language Models for Biomedical Applications
Sources
Retrieval Augmented Large Language Model System for Comprehensive Drug Contraindications
Retrieval augmented generation based dynamic prompting for few-shot biomedical named entity recognition using large language models
Arce: Augmented Roberta with Contextualized Elucidations for Ner in Automated Rule Checking
Prompt Tuning for Few-Shot Continual Learning Named Entity Recognition
Data-Efficient Biomedical In-Context Learning: A Diversity-Enhanced Submodular Perspective
Large Language Models as Oracles for Ontology Alignment
Scaling Up Active Testing to Large Language Models
Leveraging Large Language Models for Rare Disease Named Entity Recognition
EffiEval: Efficient and Generalizable Model Evaluation via Capability Coverage Maximization
Neural Bandit Based Optimal LLM Selection for a Pipeline of Tasks
eDIF: A European Deep Inference Fabric for Remote Interpretability of LLM
GenOM: Ontology Matching with Description Generation and Large Language Model