Files and links (1)
Preprint (Author's original)arXiv.org - Non-exclusive license to distribute, Open
Preprint
MedHallu: A Comprehensive Benchmark for Detecting Medical Hallucinations in Large Language Models
20 Feb 2025
Abstract
Advancements in Large Language Models (LLMs) and their increasing use in
medical question-answering necessitate rigorous evaluation of their
reliability. A critical challenge lies in hallucination, where models generate
plausible yet factually incorrect outputs. In the medical domain, this poses
serious risks to patient safety and clinical decision-making. To address this,
we introduce MedHallu, the first benchmark specifically designed for medical
hallucination detection. MedHallu comprises 10,000 high-quality question-answer
pairs derived from PubMedQA, with hallucinated answers systematically generated
through a controlled pipeline. Our experiments show that state-of-the-art LLMs,
including GPT-4o, Llama-3.1, and the medically fine-tuned UltraMedical,
struggle with this binary hallucination detection task, with the best model
achieving an F1 score as low as 0.625 for detecting "hard" category
hallucinations. Using bidirectional entailment clustering, we show that
harder-to-detect hallucinations are semantically closer to ground truth.
Through experiments, we also show incorporating domain-specific knowledge and
introducing a "not sure" category as one of the answer categories improves the
precision and F1 scores by up to 38% relative to baselines.
Metrics
18 Record Views
Details
- Title
- MedHallu: A Comprehensive Benchmark for Detecting Medical Hallucinations in Large Language Models
- Creators
- Shrey PanditJiawei XuJunyuan HongZhangyang WangTianlong ChenKaidi XuYing Ding
- Resource Type
- Preprint
- Language
- English
- Academic Unit
- Computer Science (Computing)
- Other Identifier
- 991022030047204721