arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.02428cs.CLcs.AI
收藏

受教于互联网:探究 OpenAI GPT-3 中的偏见

Taught by the Internet, Exploring Bias in OpenAIs GPT3

FACULTY OF ECONOMICS AND BUSINESS(经济与商学院)

Ali Ayaz, Aditya Nawalgaria, Ruilian Yin

AI总结:

本文通过构建基于 GPT-3 的求职者追踪系统测试性别偏见,并评估现有偏见缓解技术在超大规模 NLP 模型上的有效性。

摘要:

本研究深入梳理了关于自然语言处理模型中偏见的现有文献,以及为缓解偏见问题而提出的技术,包括首先解决偏见为何重要。此外,研究还结合规模远超以往版本的新开发模型,对这些技术作了进一步分析。为实现这些目标,本文作者针对 OpenAI 的 GPT-3 开展研究;GPT-3 是目前面向消费者可用的最大 NLP 模型。其拥有 1750 亿个参数,而 BERT 仅有 3.4 亿个参数,因此 GPT-3 是测试 NLP 模型常见缺陷的理想模型。测试通过开发一个使用 GPT-3 的求职者追踪系统来进行。出于可行性和时间限制,测试主要聚焦性别偏见,而非所有或多种类型的偏见。最后,研究考察并测试了当前的缓解技术,以衡量其有效程度。

英文摘要:

This research delves into the current literature on bias in Natural Language Processing Models and the techniques proposed to mitigate the problem of bias, including why it is important to tackle bias in the first place. Additionally, these techniques are further analysed in the light of newly developed models that tower in size over past editions. To achieve those aims, the authors of this paper conducted their research on GPT3 by OpenAI, the largest NLP model available to consumers today. With 175 billion parameters in contrast to BERTs 340 million, GPT3 is the perfect model to test the common pitfalls of NLP models. Tests were conducted through the development of an Applicant Tracking System using GPT3. For the sake of feasibility and time constraints, the tests primarily focused on gender bias, rather than all or multiple types of bias. Finally, current mitigation techniques are considered and tested to measure their degree of functionality.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑