# Beyond Accuracy: Investigating Error Types in GPT-4 Responses to USMLE Questions

arXiv: https://arxiv.org/abs/2404.13307 · HTML: https://mlsanity.com/paper/2404.13307/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 48 references that are in the map.
- First posted on arXiv 20 Apr 2024.

## Builds on (most cited first, 26 of 26)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Large Language Models Encode Clinical Knowledge](https://mlsanity.com/paper/2212.13138/) (2022)
- [What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams](https://mlsanity.com/paper/2009.13081/) (2020)
- [Graph of Thoughts: Solving Elaborate Problems with Large Language Models](https://mlsanity.com/paper/2308.09687/) (2023)
- [Galactica: A Large Language Model for Science](https://mlsanity.com/paper/2211.09085/) (2022)
- [Fact or Fiction: Verifying Scientific Claims](https://mlsanity.com/paper/2004.14974/) (2020)
- [MedAlpaca -- An Open-Source Collection of Medical Conversational AI Models and Training Data](https://mlsanity.com/paper/2304.08247/) (2023)
- [How is ChatGPT's behavior changing over time?](https://mlsanity.com/paper/2307.09009/) (2023)
- [FacTool: Factuality Detection in Generative AI -- A Tool Augmented Framework for Multi-Task and Multi-Domain Scenarios](https://mlsanity.com/paper/2307.13528/) (2023)
- [Can large language models reason about medical questions?](https://mlsanity.com/paper/2207.08143/) (2022)
- [A Study of Generative Large Language Model for Medical Research and Healthcare](https://mlsanity.com/paper/2305.13523/) (2023)
- [DoctorGLM: Fine-tuning your Chinese Doctor is not a Herculean Task](https://mlsanity.com/paper/2304.01097/) (2023)
- [When the Majority is Wrong: Modeling Annotator Disagreement for Subjective Tasks](https://mlsanity.com/paper/2305.06626/) (2023)
- [PMC-LLaMA: Towards Building Open-source Language Models for Medicine](https://mlsanity.com/paper/2304.14454/) (2023)
- [Diagnostic Reasoning Prompts Reveal the Potential for Large Language Model Interpretability in Medicine](https://mlsanity.com/paper/2308.06834/) (2023)
- [POTATO: The Portable Text Annotation Tool](https://mlsanity.com/paper/2212.08620/) (2022)
- [Everyone's Voice Matters: Quantifying Annotation Disagreement Using Demographic Information](https://mlsanity.com/paper/2301.05036/) (2023)
- [When Do Annotator Demographics Matter? Measuring the Influence of Annotator Demographics with the POPQUORN Dataset](https://mlsanity.com/paper/2306.06826/) (2023)
- [Human Evaluation and Correlation with Automatic Metrics in Consultation Note Generation](https://mlsanity.com/paper/2204.00447/) (2022)
- [Evaluation of GPT-3.5 and GPT-4 for supporting real-world information needs in healthcare delivery](https://mlsanity.com/paper/2304.13714/) (2023)
- [CAVES: A Dataset to facilitate Explainable Classification and Summarization of Concerns towards COVID Vaccines](https://mlsanity.com/paper/2204.13746/) (2022)
- [A Meta-Evaluation of Faithfulness Metrics for Long-Form Hospital-Course Summarization](https://mlsanity.com/paper/2303.03948/) (2023)
- [Discovering novel drug-supplement interactions using a dietary supplements knowledge graph generated from the biomedical literature](https://mlsanity.com/paper/2106.12741/) (2021)

## Built on by (most cited first, 5 of 5)

- [ERVQA: A Dataset to Benchmark the Readiness of Large Vision Language Models in Hospital Environments](https://mlsanity.com/paper/2410.06420/) (2024)
- [Multidimensional Rubric-oriented Reward Model Learning via Geometric Projection Reference Constraints](https://mlsanity.com/paper/2511.16139/) (2025)
- [OrthoDoc: Multimodal Large Language Model for Assisting Diagnosis in Computed Tomography](https://mlsanity.com/paper/2409.09052/) (2024)
- [Red Teaming Large Language Models for Healthcare](https://mlsanity.com/paper/2505.00467/) (2025)
- [Evaluating the capability of large language models to personalize science texts for diverse middle-school-age learners](https://mlsanity.com/paper/2408.05204/) (2024)

## Abstract (arXiv)

GPT-4 demonstrates high accuracy in medical QA tasks, leading with an accuracy of 86.70%, followed by Med-PaLM 2 at 86.50%. However, around 14% of errors remain. Additionally, current works use GPT-4 to only predict the correct option without providing any explanation and thus do not provide any insight into the thinking process and reasoning used by GPT-4 or other LLMs. Therefore, we introduce a new domain-specific error taxonomy derived from collaboration with medical students. Our GPT-4 USMLE Error (G4UE) dataset comprises 4153 GPT-4 correct responses and 919 incorrect responses to the United States Medical Licensing Examination (USMLE) respectively. These responses are quite long (258 words on average), containing detailed explanations from GPT-4 justifying the selected option. We then launch a large-scale annotation study using the Potato annotation platform and recruit 44 medical experts through Prolific, a well-known crowdsourcing platform. We annotated 300 out of these 919 incorrect data points at a granular level for different classes and created a multi-label span to identify the reasons behind the error. In our annotated dataset, a substantial portion of GPT-4's incorrect responses is categorized as a "Reasonable response by GPT-4," by annotators. This sheds light on the challenge of discerning explanations that may lead to incorrect options, even among trained medical professionals. We also provide medical concepts and medical semantic predications extracted using the SemRep tool for every data point. We believe that it will aid in evaluating the ability of LLMs to answer complex medical questions. We make the resources available at https://github.com/roysoumya/usmle-gpt4-error-taxonomy .