2024
DOI: 10.1098/rsta.2023.0254
|View full text |Cite
|
Sign up to set email alerts
|

GPT-4 passes the bar exam

Daniel Martin Katz,
Michael James Bommarito,
Shang Gao
et al.

Abstract: In this paper, we experimentally evaluate the zero-shot performance of GPT-4 against prior generations of GPT on the entire uniform bar examination (UBE), including not only the multiple-choice multistate bar examination (MBE), but also the open-ended multistate essay exam (MEE) and multistate performance test (MPT) components. On the MBE, GPT-4 significantly outperforms both human test-takers and prior models, demonstrating a 26% increase over ChatGPT and beating humans in five of seven subject areas. On the … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2024
2024
2024
2024

Publication Types

Select...
5
1
1

Relationship

1
6

Authors

Journals

citations
Cited by 12 publications
(3 citation statements)
references
References 60 publications
0
3
0
Order By: Relevance
“…Much more powerful than GPT-3.5 (available for free), GPT-4 requires a paid subscription. While GPT-4 can pass the multistate bar examination taken by US law students (Katz et al, 2024), even GPT-3.5 can pass the National Board of Medical Examiners test at the level of a third-year medical student (Gilson et al, 2023). ChatGPT has already been integrated into Duolingo Max (which adds AI role play and "explain my answer" features), and will soon be integrated into Microsoft Word, PowerPoint, and Outlook (Warren, 2023).…”
Section: Chatgptmentioning
confidence: 99%
“…Much more powerful than GPT-3.5 (available for free), GPT-4 requires a paid subscription. While GPT-4 can pass the multistate bar examination taken by US law students (Katz et al, 2024), even GPT-3.5 can pass the National Board of Medical Examiners test at the level of a third-year medical student (Gilson et al, 2023). ChatGPT has already been integrated into Duolingo Max (which adds AI role play and "explain my answer" features), and will soon be integrated into Microsoft Word, PowerPoint, and Outlook (Warren, 2023).…”
Section: Chatgptmentioning
confidence: 99%
“…Katz et al . [ 47 ] test GPT-4 and its earlier progenitors on the three components of the bar exam, which in many US jurisdictions must be completed by a legally trained individual to be able to practise law. This paper displays a quite sophisticated technical content and underlying methodology, including for instance a ‘contamination check’ directly assisted by the OpenAI creators to make sure that the exam questions had not been presented before to GPT-4 during its training phase.…”
Section: Legal Practice and Contextmentioning
confidence: 99%
“…Most prominently, generative AI (GenAI) and Large Language Models (LLMs) are revolutionizing education. These technologies (such as ChatGPT) have proven their capabilities by achieving feats like passing the US bar exam [21] and the European Exam in Core Cardiology [22]. LLMs, exemplified by the Generative Pre-trained Transformer (GPT), leverage vast datasets to understand and generate human-like language, enhancing natural language processing and enabling tasks like language translation, computer code development, text generation, and sentiment analysis with exceptional accuracy.…”
Section: Ai and Its Tensions For Sustainabilitymentioning
confidence: 99%