arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2101.05783cs.CLcs.LG
收藏

大语言模型中持续的反穆斯林偏见

Persistent Anti-Muslim Bias in Large Language Models

Stanford University(斯坦福大学);McMaster University(麦克马斯特大学)

Abubakar Abid, Maheen Farooqi, James Zou

摘要:

已经观察到,大规模语言模型捕获了不期望的社会偏见,例如与种族和性别有关的社会偏见;然而,宗教偏见相对未被探索。我们证明,GPT-3,一个国家的最先进的上下文语言模型,捕捉持久的穆斯林暴力偏见。我们以各种方式探索GPT-3,包括快速完成,类比推理和故事生成,以了解这种反穆斯林偏见,证明它在模型的不同用途中表现出一致性和创造性,即使与其他宗教团体的偏见相比,它也是严重的。例如,在23%的测试用例中,“穆斯林”被类比为“恐怖分子”,而在5%的测试用例中,“犹太人”被映射为“金钱”。我们用对抗性文本提示来量化克服这种偏见所需的积极分心,并发现使用最积极的6个形容词将“穆斯林”的暴力完成率从66%降低到20%,但仍然高于其他宗教群体。

英文摘要:

It has been observed that large-scale language models capture undesirable societal biases, e.g. relating to race and gender; yet religious bias has been relatively unexplored. We demonstrate that GPT-3, a state-of-the-art contextual language model, captures persistent Muslim-violence bias. We probe GPT-3 in various ways, including prompt completion, analogical reasoning, and story generation, to understand this anti-Muslim bias, demonstrating that it appears consistently and creatively in different uses of the model and that it is severe even compared to biases about other religious groups. For instance, "Muslim" is analogized to "terrorist" in 23% of test cases, while "Jewish" is mapped to "money" in 5% of test cases. We quantify the positive distraction needed to overcome this bias with adversarial text prompts, and find that use of the most positive 6 adjectives reduces violent completions for "Muslims" from 66% to 20%, but which is still higher than for other religious groups.

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。

↑