Text Understanding in GPT-4 vs Humans

AmazUtah_NLP at SemEval-2024 Task 9: A MultiChoice Question Answering System for Commonsense Defying Reasoning


View a PDF of the paper titled Text Understanding in GPT-4 vs Humans, by Thomas R. Shultz and 2 other authors

View PDF

Abstract:We examine whether a leading AI system GPT4 understands text as well as humans do, first using a well-established standardized test of discourse comprehension. On this test, GPT4 performs slightly, but not statistically significantly, better than humans given the very high level of human performance. Both GPT4 and humans make correct inferences about information that is not explicitly stated in the text, a critical test of understanding. Next, we use more difficult passages to determine whether that could allow larger differences between GPT4 and humans. GPT4 does considerably better on this more difficult text than do the high school and university students for whom these the text passages are designed, as admission tests of student reading comprehension. Deeper exploration of GPT4 performance on material from one of these admission tests reveals generally accepted signatures of genuine understanding, namely generalization and inference.

Submission history

From: Thomas Shultz [view email]
[v1]
Mon, 25 Mar 2024 21:17:14 UTC (313 KB)
[v2]
Sun, 20 Oct 2024 14:11:18 UTC (430 KB)
[v3]
Fri, 20 Dec 2024 17:50:38 UTC (441 KB)



Source link
lol

By stp2y

Leave a Reply

Your email address will not be published. Required fields are marked *

No widgets found. Go to Widget page and add the widget in Offcanvas Sidebar Widget Area.