To Generate or to Retrieve? On the Effectiveness of Artificial Contexts for Medical Open-Domain Question Answering
This work addresses the challenge of specialized knowledge access in medical QA, offering a practical solution for low-resource hardware settings.
The paper tackles the problem of medical open-domain question answering by introducing MedGENIE, a generate-then-read framework that sets new state-of-the-art results on benchmarks like MedQA-USMLE, MedMCQA, and MMLU, using up to 706 times fewer parameters than large baselines.
Medical open-domain question answering demands substantial access to specialized knowledge. Recent efforts have sought to decouple knowledge from model parameters, counteracting architectural scaling and allowing for training on common low-resource hardware. The retrieve-then-read paradigm has become ubiquitous, with model predictions grounded on relevant knowledge pieces from external repositories such as PubMed, textbooks, and UMLS. An alternative path, still under-explored but made possible by the advent of domain-specific large language models, entails constructing artificial contexts through prompting. As a result, "to generate or to retrieve" is the modern equivalent of Hamlet's dilemma. This paper presents MedGENIE, the first generate-then-read framework for multiple-choice question answering in medicine. We conduct extensive experiments on MedQA-USMLE, MedMCQA, and MMLU, incorporating a practical perspective by assuming a maximum of 24GB VRAM. MedGENIE sets a new state-of-the-art in the open-book setting of each testbed, allowing a small-scale reader to outcompete zero-shot closed-book 175B baselines while using up to 706$\times$ fewer parameters. Our findings reveal that generated passages are more effective than retrieved ones in attaining higher accuracy.