GPT-4 vs. GPT-3.5 AS CODING ASSISTANTS

Author:

Moussiades Lefteris1,Zografos George1,Papakostas George1

Affiliation:

1. International Hellenic University

Abstract

Abstract Large Language Models are not limited to understanding and producing natural language. Instead, they do equally well in understanding and generating source code in various programming languages. At the time of writing, GPT-4 was considered the most potent large language model from OpenAI. This paper compares GPT-4 with its immediate ancestor, GPT-3.5, as coding assistants. More specifically, we have constructed appropriate tests to check whether the two systems can a) answer typical questions that can arise during code development, b) generate reliable code in response to user requirements, and c) contribute to code debugging. The test results for both models are impressive. However, the performance of GPT-4 is outstanding and signals an increase in the productivity of programmers and the reorganization of software development procedures based on these new tools.

Publisher

Research Square Platform LLC

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3