评估LLMs在做出道德判断时的性别偏见
Evaluating Gender Bias of LLMs in Making Morality Judgements
AI总结：
本研究引入GenMO数据集，评估了GPT、Llama、Mistral和Claude等主流LLMs在道德判断中的性别偏见，发现所有模型均表现出显著偏向女性的偏见。
摘要：
大型语言模型（LLMs）在众多自然语言处理（NLP）任务中展现出了卓越的能力。然而，这些模型仍然无法避免诸如社会偏见，尤其是性别偏见等局限性。本研究调查了当前闭源和开源LLMs是否存在性别偏见，特别是在被要求给出道德意见时。为了评估这些模型，我们整理并引入了一个新数据集GenMO（Gender-bias in Morality Opinions），其中包含分别以男性和女性角色为主角的平行短篇故事。具体而言，我们测试了GPT家族（GPT-3.5-turbo、GPT-3.5-turbo-instruct、GPT-4-turbo）、Llama 3和3.1家族（8B/70B）、Mistral-7B以及Claude 3家族（Sonnet和Opus）的模型。令人惊讶的是，尽管采用了安全检查，我们测试的所有生产标准模型都表现出显著的性别偏见，其中GPT-3.5-turbo在24%的样本中给出了带有偏见的意见。此外，所有模型都一致偏向女性角色，GPT在68-85%的案例中表现出偏见，而Llama 3在约81-85%的实例中表现出偏见。此外，我们的研究还调查了模型参数对性别偏见的影响，并探讨了LLMs在道德决策中暴露偏见的现实情境。
英文摘要：
Large Language Models (LLMs) have shown remarkable capabilities in a multitude of Natural Language Processing (NLP) tasks. However, these models are still not immune to limitations such as social biases, especially gender bias. This work investigates whether current closed and open-source LLMs possess gender bias, especially when asked to give moral opinions. To evaluate these models, we curate and introduce a new dataset GenMO (Gender-bias in Morality Opinions) comprising parallel short stories featuring male and female characters respectively. Specifically, we test models from the GPT family (GPT-3.5-turbo, GPT-3.5-turbo-instruct, GPT-4-turbo), Llama 3 and 3.1 families (8B/70B), Mistral-7B and Claude 3 families (Sonnet and Opus). Surprisingly, despite employing safety checks, all production-standard models we tested display significant gender bias with GPT-3.5-turbo giving biased opinions in 24% of the samples. Additionally, all models consistently favour female characters, with GPT showing bias in 68-85% of cases and Llama 3 in around 81-85% instances. Additionally, our study investigates the impact of model parameters on gender bias and explores real-world situations where LLMs reveal biases in moral decision-making.
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。
