Salvato in:
Dettagli Bibliografici
Autori principali: Le, Chenqian, Gong, Ziheng, Wang, Chihang, Ni, Haowei, Li, Panfeng, Chen, Xupeng
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2506.12182
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908642733719552
author Le, Chenqian
Gong, Ziheng
Wang, Chihang
Ni, Haowei
Li, Panfeng
Chen, Xupeng
author_facet Le, Chenqian
Gong, Ziheng
Wang, Chihang
Ni, Haowei
Li, Panfeng
Chen, Xupeng
contents Large language models (LLMs) have shown great potential in medical question answering (MedQA), yet adapting them to biomedical reasoning remains challenging due to domain-specific complexity and limited supervision. In this work, we study how prompt design and lightweight fine-tuning affect the performance of open-source LLMs on PubMedQA, a benchmark for multiple-choice biomedical questions. We focus on two widely used prompting strategies - standard instruction prompts and Chain-of-Thought (CoT) prompts - and apply QLoRA for parameter-efficient instruction tuning. Across multiple model families and sizes, our experiments show that CoT prompting alone can improve reasoning in zero-shot settings, while instruction tuning significantly boosts accuracy. However, fine-tuning on CoT prompts does not universally enhance performance and may even degrade it for certain larger models. These findings suggest that reasoning-aware prompts are useful, but their benefits are model- and scale-dependent. Our study offers practical insights into combining prompt engineering with efficient finetuning for medical QA applications.
format Preprint
id arxiv_https___arxiv_org_abs_2506_12182
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Instruction Tuning and CoT Prompting for Contextual Medical QA with LLMs
Le, Chenqian
Gong, Ziheng
Wang, Chihang
Ni, Haowei
Li, Panfeng
Chen, Xupeng
Computation and Language
Large language models (LLMs) have shown great potential in medical question answering (MedQA), yet adapting them to biomedical reasoning remains challenging due to domain-specific complexity and limited supervision. In this work, we study how prompt design and lightweight fine-tuning affect the performance of open-source LLMs on PubMedQA, a benchmark for multiple-choice biomedical questions. We focus on two widely used prompting strategies - standard instruction prompts and Chain-of-Thought (CoT) prompts - and apply QLoRA for parameter-efficient instruction tuning. Across multiple model families and sizes, our experiments show that CoT prompting alone can improve reasoning in zero-shot settings, while instruction tuning significantly boosts accuracy. However, fine-tuning on CoT prompts does not universally enhance performance and may even degrade it for certain larger models. These findings suggest that reasoning-aware prompts are useful, but their benefits are model- and scale-dependent. Our study offers practical insights into combining prompt engineering with efficient finetuning for medical QA applications.
title Instruction Tuning and CoT Prompting for Contextual Medical QA with LLMs
topic Computation and Language
url https://arxiv.org/abs/2506.12182