Language Models are Surprisingly Fragile to Drug Names in Biomedical Benchmarks
Medical knowledge is context-dependent and requires consistent reasoning across various natural language expressions of semantically equivalent phrases. This is particularly crucial for drug names, where patients often use brand names like Advil or Tylenol instead of their generic equivalents. To study this, we create a new robustness dataset, RABBITS, to evaluate performance differences on medical benchmarks after swapping brand and generic drug names using physician expert annotations. We assess both open-source and API-based LLMs on MedQA and MedMCQA, revealing a consistent performance drop ranging from 1-10\%. Furthermore, we identify a potential source of this fragility as the contamination of test data in widely used pre-training datasets. All code is accessible at https://github.com/BittermanLab/RABBITS, and a HuggingFace leaderboard is available at https://huggingface.co/spaces/AIM-Harvard/rabbits-leaderboard.
Code (1)
Tasks
MedQASimilar Papers 제목 키워드 기반
A Lightweight Neural Model for Biomedical Entity Linking
Biomedical entity linking aims to map biomedical mentions, such as diseases and drugs, to standard entities in a given knowledge base. The specific challenge in this context is that the same biomedical entity can have a …
Entity LinkingmodelUsing Snomed to recognize and index chemical and drug mentions.
In this paper we describe a new named entity extraction system. Our work proposes a system for the identification and annotation of drug names in Spanish biomedical texts based on machine learning and deep learning model…
BIG-bench Machine LearningEntity Extraction using GANFrom Knowledge Generation to Knowledge Verification: Examining the BioMedical Generative Capabilities of ChatGPT
The generative capabilities of LLM models offer opportunities for accelerating tasks but raise concerns about the authenticity of the knowledge they produce. To address these concerns, we present a computational approach…
Prompt EngineeringDrug and Disease Interpretation Learning with Biomedical Entity Representation Transformer
Concept normalization in free-form texts is a crucial step in every text-mining pipeline. Neural architectures based on Bidirectional Encoder Representations from Transformers (BERT) have achieved state-of-the-art result…
Drug DiscoveryMetric LearningTransfer LearningTripletBioinformatic Scaling of Allosteric Interactions in Biomedical Isozymes
Allosteric (long-range) interactions can be surprisingly strong in proteins of biomedical interest. Here we use bioinformatic scaling to connect prior results on nonsteroidal anti-inflammatory drugs to promising new drug…