Can large language models reason about medical questions?

Abstract

Although large language models often produce impressive outputs, it remains unclear how they perform in real-world scenarios requiring strong reasoning skills and expert domain knowledge. We set out to investigate whether closed- and open-source models (GPT-3.5, Llama 2, etc.) can be applied to answer and reason about difficult real-world-based questions. We focus on three popular medical benchmarks (MedQA-US Medical Licensing Examination [USMLE], MedMCQA, and PubMedQA) and multiple prompting scenarios: chain of thought (CoT; think step by step), few shot, and retrieval augmentation. Based on an expert annotation of the generated CoTs, we found that InstructGPT can often read, reason, and recall expert knowledge. Last, by leveraging advances in prompt engineering (few-shot and ensemble methods), we demonstrated that GPT-3.5 not only yields calibrated predictive distributions but also reaches the passing score on three datasets: MedQA-USMLE (60.2%), MedMCQA (62.7%), and PubMedQA (78.2%). Open-source models are closing the gap: Llama 2 70B also passed the MedQA-USMLE with 62.5% accuracy.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 63

1. 1.

   Mikolov T., Sutskever I., Chen K., Corrado G.S., Dean J. Distributed representations of words and phrases and their compositionality. Adv. Neural Inf. Process. Syst. 2013;26.

   Not in our library

2. 2.

   Pennington J., Socher R., Manning C. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2014;1532-1543.

   Not in our library10.3115/v1/d14-1162

3. 3.

   Peters M.E., Neumann M., Iyyer M., Gardner M., Clark C., Lee K., Zettlemoyer L. Deep contextualized word representations. arXiv. 2018.

   Not in our library10.48550/arxiv.1802.05365

4. 4.

   Vaswani A., Shazeer N., Parmar N., Uszkoreit J., Jones L., Gomez A.N., Kaiser Ł., Polosukhin I. Attention is all you need. Adv. Neural Inf. Process. Syst. 2017;30.

   Not in our library

5. 5.

   Devlin J., Chang M., Lee K., Toutanova K. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv. 2018.

   Not in our library10.04805/arxiv.1810.04805

6. 6.

   Radford A., Narasimhan K., Salimans T., Sutskever I. 2018.

   Not in our library

7. 7.

   Brown T., Mann B., Ryder N., Subbiah M., Kaplan J.D., Dhariwal P., Neelakantan A., Shyam P., Sastry G., Askell A., Agarwal S., et al. Language models are few-shot learners. Adv. Neural Inf. Process. Syst. 2020;33.

   Not in our library

8. 8.

   Rae J.W., Borgeaud S., Cai T., Millican K., Hoffmann J., Song H.F., Aslanides J., Henderson S., Ring R., Young S., Rutherford E., et al. Scaling Language Models: Methods, Analysis & Insights from Training Gopher. arXiv. 2021.

   Not in our library10.48550/arxiv.2112.11446

9. 9.

   Chowdhery A., Narang S., Devlin J., Bosma M., Mishra G., Roberts A., Barham P., Chung H.W., Sutton C., Gehrmann S., et al. Palm: Scaling language modeling with pathways. J. Mach. Learn. Res. 2023;24:1-113.

   Not in our library

10. 10.

    Thoppilan R., Freitas D.D., Hall J., Shazeer N., Kulshreshtha A., Cheng H., Jin A., Bos T., Baker L., Du Y., et al. LaMDA: Language Models for Dialog Applications. arXiv. 2022.

    Not in our library10.48550/arxiv.2201.08239

11. 11.

    Hoffmann J., Borgeaud S., Mensch A., Buchatskaya E., Cai T., Rutherford E., Casas D.D.L., Hendricks L.A., Welbl J., Clark A., et al. Training Compute-optimal Large Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2203.15556

12. 12.

    Smith S., Patwary M., Norick B., LeGresley P., Rajbhandari S., Casper J., Liu Z., Prabhumoye S., Zerveas G., Korthikanti V., et al. Using DeepSpeed and Megatron to Train Megatron-turing NLG 530B, A Large-scale Generative Language Model. arXiv. 2022.

    Not in our library10.48550/arxiv.2201.11990

13. 13.

    Zhang S., Roller S., Goyal N., Artetxe M., Chen M., Chen S., Dewan C., Diab M., Li X., Lin X.V., et al. OPT: Open Pre-trained Transformer Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2205.01068

14. 14.

    Lieber O., Sharir O., Lenz B., Shoham Y. Jurassic-1: Technical details and evaluation. White Paper AI21 Labs. 2021;1:9.

    Not in our library

15. 15.

    Fedus W., Zoph B., Shazeer N. Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity. J. Mach. Learn. Res. 2022;23:5232-5270.

    Not in our library

16. 16.

    Laurençon H., Saulnier L., Wang T., Akiki C., Villanova del Moral A., Le Scao T., Von Werra L., Mou C., González Ponferrada E., Nguyen H., et al. The BigScience ROOTS Corpus: A 1.6TB composite multilingual dataset. Adv. Neural Inf. Process. Syst. 2022;35:31809-31826.

    Not in our library

17. 17.

    Kaplan J., McCandlish S., Henighan T., Brown T.B., Chess B., Child R., Gray S., Radford A., Wu J., et al. Scaling Laws for Neural Language Models. arXiv. 2020.

    Not in our library10.48550/arxiv.2001.08361

18. 18.

    Srivastava A., Rastogi A., Rao A., Shoeb A.A.M., Abid A., Fisch A., Brown A.R., Santoro A., Gupta A., Garriga-Alonso A., et al. Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models. arXiv. 2022.

    Not in our library10.48550/arxiv.2206.04615

19. 19.

    Liu P., Yuan W., Fu J., Jiang Z., Hayashi H., Neubig G. Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. ACM Comput. Surv. 2023;55:1-35.

    Not in our library

20. 20.

    Wei J., Wang X., Schuurmans D., Bosma M., Xia F., Chi E., Le Q.V., Zhou D. Chain-of-thought prompting elicits reasoning in large language models. Adv. Neural Inf. Process. Syst. 2022;35:24824-24837.

    Not in our library

21. 21.

    Zhou D., Schärli N., Hou L., Wei J., Scales N., Wang X., Schuurmans D., Cui C., Bousquet O., Le Q., Chi E. Least-to-most Prompting Enables Complex Reasoning in Large Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2205.10625

22. 22.

    Drozdov A., Schärli N., Akyürek E., Scales N., Song X., Chen X., Bousquet O., Zhou D. Compositional Semantic Parsing with Large Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2209.15003

23. 23.

    Nye M.I., Andreassen A.J., Gur-Ari G., Michalewski H., Austin J., Bieber D., Dohan D., Lewkowycz A., Bosma M., Luan D., et al. Show Your Work: Scratchpads for Intermediate Computation with Language Models. arXiv. 2021.

    Not in our library10.48550/arxiv.2112.00114

24. 24.

    Suzgun M., Scales N., Schärli N., Gehrmann S., Tay Y., Chung H.W., Chowdhery A., Le Q.V., Chi E.H., Zhou D., Wei J. Challenging BIG-bench Tasks and Whether Chain-of-thought Can Solve Them. arXiv. 2022.

    Not in our library10.48550/arxiv.2210.09261

25. 25.

    Kojima T., Gu S.S., Reid M., Matsuo Y., Iwasawa Y. Large language models are zero-shot reasoners. Adv. Neural Inf. Process. Syst. 2022;35:22199-22213.

    Not in our library

26. 26.

    Ouyang L., Wu J., Jiang X., Almeida D., Wainwright C., Mishkin P., Zhang C., Agarwal S., Slama K., Ray A., Schulman J., et al. Training language models to follow instructions with human feedback. Adv. Neural Inf. Process. Syst. 2022;35:27730-27744.

    Not in our library

27. 27.

    Bender E.M., Gebru T., McMillan-Major A., Shmitchell S. Proceedings of the 2021 ACM conference on fairness, accountability, and transparency. 2021;610-623.

    Not in our library

28. 28.

    Wang B., Xu C., Wang S., Gan Z., Cheng Y., Gao J., Awadallah A.H., Li B. Adversarial GLUE: A Multi-task Benchmark for Robustness Evaluation of Language Models. arXiv. 2021.

    Not in our library10.48550/arxiv.2111.02840

29. 29.
    Considering the possibilities and pitfalls of Generative Pre-trained Transformer 3 (GPT-3) in healthcare delivery · 2021

    No notice
30. 30.

    Sezgin E., Sirrianni J., Linwood S.L. Operationalizing and Implementing Pretrained, Large Artificial Intelligence Linguistic Models in the US Health Care System: Outlook of Generative Pretrained Transformer 3 (GPT-3) as a Service Model. JMIR medical informatics. 2022;10.

    Not in our library10.2196/32875

31. 31.

    Hendrycks D., Burns C., Basart S., Zou A., Mazeika M., Song D., Steinhardt J. Measuring Massive Multitask Language Understanding. arXiv. 2020.

    Not in our library10.48550/arxiv.2009.03300

32. 32.

    Jin D., Pan E., Oufattole N., Weng W.H., Fang H., Szolovits P. What disease does this patient have? a large-scale open domain question answering dataset from medical exams. Appl. Sci. 2021;11:6421.

    Not in our library

33. 33.
    Large language models encode clinical knowledge · 2023

    Corrected
34. 34.

    Singhal K., Tu T., Gottweis J., Sayres R., Wulczyn E., Hou L., Clark K., Pfohl S., Cole-Lewis H., Neal D., et al. Towards Expert-level Medical Question Answering with Large Language Models. arXiv. 2023.

    Not in our library10.48550/arxiv.2305.09617

35. 35.

    Nori H., King N., McKinney S.M., Carignan D., Horvitz E. Capabilities of GPT-4 on Medical Challenge Problems. arXiv. 2023.

    Not in our library10.48550/arxiv.2303.13375

36. 36.
    How Does ChatGPT Perform on the United States Medical Licensing Examination (USMLE)? The Implications of Large Language Models for Medical Education and Knowledge Assessment · 2023

    No notice
37. 37.
    Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models · 2023

    No notice
38. 38.

    Touvron H., Martin L., Stone K., Albert P., Almahairi A., Babaei Y., Bashlykov N., Batra S., Bhargava P., Bhosale S., et al. Llama 2: Open Foundation and Fine-tuned Chat Models. arXiv. 2023.

    Not in our library10.48550/arxiv.2307.09288

39. 39.

    Chen Z., Hernández-Cano A., Romanou A., Bonnet A., Matoba K., Salvi F., Pagliardini M., Fan S., Köpf A., Mohtashami A., et al. MEDITRON-70B: Scaling Medical Pretraining for Large Language Models. arXiv. 2023.

    Not in our library10.48550/arxiv.2311.16079

40. 40.

    Wang Y., Ma X., Chen W. Augmenting Black-box LLMs with Medical Textbooks for Clinical Question Answering. arXiv. 2023.

    Not in our library10.48550/arxiv.2309.02233

41. 41.

    Liévin V., Motzfeldt A.G., Jensen I.R., Winther O. International Conference on Machine Learning. 2023;20950-20977.

    Not in our library

42. 42.

    Wang X., Wei J., Schuurmans D., Le Q.V., Chi E.H., Zhou D. Self-consistency Improves Chain of Thought Reasoning in Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2203.11171

43. 43.

    Li Y., Lin Z., Zhang S., Fu Q., Chen B., Lou J., Chen W. On the Advance of Making Language Models Better Reasoners. arXiv. 2022.

    Not in our library10.48550/arxiv.2206.02336

44. 44.

    Pal A., Umapathi L.K., Sankarasubbu M. Medmcqa: A large-scale multi-subject multi-choice dataset for medical domain question answering. PMLR. 2022;174:248-260.

    Not in our library

45. 45.

    Jin Q., Dhingra B., Liu Z., Cohen W.W., Lu X. PubMedQA: A Dataset for Biomedical Research Question Answering. arXiv. 2019.

    Not in our library10.48550/arxiv.1909.06146

46. 46.

    Chung H.W., Hou L., Longpre S., Zoph B., Tay Y., Fedus W., et al. Scaling Instruction-finetuned Language Models. arXiv. 2022.

    Not in our library10.48550/arxiv.2210.11416

47. 47.

    Lewis P., Perez E., Piktus A., Petroni F., Karpukhin V., Goyal N., Küttler H., Lewis M., Yih W.T., Rocktäschel T., et al. Retrieval-augmented generation for knowledge-intensive nlp tasks. Adv. Neural Inf. Process. Syst. 2020;33:9459-9474.

    Not in our library

48. 48.

    Borgeaud S., Mensch A., Hoffmann J., Cai T., Rutherford E., Millican K., Van Den Driessche G.B., Lespiau J.B., Damoc B., Clark A., et al. Improving language models by retrieving from trillions of tokens. PMLR. 2022;162:2206-2240.

    Not in our library

49. 49.

    Lazaridou A., Gribovskaya E., Stokowiec W., Grigorev N. Internet-augmented language models through few-shot prompting for open-domain question answering. arXiv. 2022.

    Not in our library10.48550/arxiv.2203.05115

50. 50.

    Liévin V. 2023.

    Not in our library10.5281/zenodo.10301874

51. 51.

    Ott S., Hebenstreit K., Li’evin V., Hother C.E., Moradi M., Mayrhauser M., Praas R., Winther O., Samwald M. ThoughtSource: A central hub for large language model reasoning data. Sci. Data. 2023;10.

    Not in our library10.1038/s41597-023-02433-3

52. 52.

    Chen M., Tworek J., Jun H., Yuan Q., de Oliveira Pinto H.P., Kaplan J., Edwards H., Burda Y., Joseph N., Brockman G., et al. Evaluating Large Language Models Trained on Code. arXiv. 2021.

    Not in our library10.48550/arxiv.2107.03374

53. 53.

    Zheng L., Chiang W., Sheng Y., Zhuang S., Wu Z., Zhuang Y., Lin Z., Li Z., Li D., Xing E., et al. Judging LLM-as-a-judge with MT-bench and Chatbot Arena. arXiv. 2023.

    Not in our library10.48550/arxiv.2306.05685

54. 54.

    Dettmers T., Pagnoni A., Holtzman A., Zettlemoyer L. QLoRA: Efficient Finetuning of Quantized LLMs. arXiv. 2023.

    Not in our library10.48550/arxiv.2305.14314

55. 55.

    Almazrouei E., Alobeidli H., Alshamsi A., Cappelli A., Cojocaru R., Debbah M., Goffinet E., Heslow D., Launay J., Malartic Q., et al. 2023.

    Not in our library

56. 56.

    Team M.N. 2023.

    Not in our library

57. 57.

    Black S., Biderman S., Hallahan E., Anthony Q., Gao L., Golding L., He H., Leahy C., McDonell K., Phang J., et al. GPT-NeoX-20B: An Open-source Autoregressive Language Model. arXiv. 2022.

    Not in our library10.48550/arxiv.2204.06745

58. 58.

    Yasunaga M., Leskovec J., Liang P. LinkBERT: Pretraining Language Models with Document Links. arXiv. 2022.

    Not in our library10.48550/arxiv.2203.15827

59. 59.

    Karpukhin V., Oguz B., Min S., Wu L., Edunov S., Chen D., Yih W.T. Dense Passage Retrieval for Open-domain Question Answering. arXiv. 2020.

    Not in our library10.48550/arxiv.2004.04906

60. 60.

    Cobbe K., Kosaraju V., Bavarian M., Chen M., Jun H., Kaiser L., Plappert M., Tworek J., Hilton J., Nakano R., et al. Training Verifiers to Solve Math Word Problems. arXiv. 2021.

    Not in our library10.48550/arxiv.2110.14168

61. 61.

    Kadavath S., Conerly T., Askell A., Henighan T., Drain D., Perez E., Plappert M., Tworek J., Hilton J., Nakano R., et al. Language Models (Mostly) Know What They Know. arXiv. 2022.

    Not in our library10.48550/arxiv.2207.05221

62. 62.

    Zhou C., Liu P., Xu P., Iyer S., Sun J., Mao Y., Ma X., Efrat A., Yu P., Yu L., et al. LIMA: Less Is More for Alignment. arXiv. 2023.

    Not in our library10.48550/arxiv.2305.11206

63. 63.

    Singhal K., Tu T., Gottweis J., Sayres R., Wulczyn E., Hou L., Clark K., Pfohl S., Cole-Lewis H., Neal D., et al. Towards Expert-level Medical Question Answering with Large Language Models. arXiv. 2023.

    Not in our library10.48550/arxiv.2305.09617

Cited by papers we hold 5

• Matching patients to clinical trials with large language models · 2024
  No notice
• Large language models encode clinical knowledge · 2023
  Corrected
• Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models · 2023
  No notice
• The shaky foundations of large language models and foundation models for electronic health records · 2023
  No notice
• Evaluating the Feasibility of ChatGPT in Healthcare: An Analysis of Multiple Clinical and Research Scenarios · 2023
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.