评估GPT-3生成的仇恨内容审核解释
Evaluating GPT-3 Generated Explanations for Hateful Content Moderation
AI总结：
本研究提出分析框架，通过2,400名受访者调查评估GPT-3生成的仇恨言论解释，发现其虽高质量但说服力可能导致内容审核错误判断，强调需谨慎应用。
摘要：
近期研究聚焦于通过微调或提示使用大型语言模型（LLM）为仇恨言论生成解释。尽管该领域兴趣日益增长，但这些生成解释的有效性和潜在局限性仍鲜为人知。一个关键担忧是，这些由LLM生成的解释可能导致用户和内容审核员对被标记内容的性质做出错误判断。例如，LLM生成的解释可能不准确地使内容审核员相信一段良性内容是仇恨性的。鉴于此，我们提出了一个用于审查仇恨言论解释的分析框架，并开展了一项评估此类解释的广泛调查。具体而言，我们提示GPT-3为仇恨和非仇恨内容生成解释，并开展了一项涉及2,400名独特受访者的调查来评估生成的解释。我们的研究结果显示：（1）人类评估者认为GPT生成的解释在语言流畅性、信息量、说服力和逻辑合理性方面质量较高，（2）然而，这些解释的说服力因采用的提示策略而异，（3）这种说服力可能导致对内容仇恨性质的错误判断。我们的研究强调了在内容审核中应用LLM生成解释时需谨慎的必要性。代码和结果可在https://github.com/Social-AI-Studio/GPT3-HateEval获取。
英文摘要：
Recent research has focused on using large language models (LLMs) to generate explanations for hate speech through fine-tuning or prompting. Despite the growing interest in this area, these generated explanations' effectiveness and potential limitations remain poorly understood. A key concern is that these explanations, generated by LLMs, may lead to erroneous judgments about the nature of flagged content by both users and content moderators. For instance, an LLM-generated explanation might inaccurately convince a content moderator that a benign piece of content is hateful. In light of this, we propose an analytical framework for examining hate speech explanations and conducted an extensive survey on evaluating such explanations. Specifically, we prompted GPT-3 to generate explanations for both hateful and non-hateful content, and a survey was conducted with 2,400 unique respondents to evaluate the generated explanations. Our findings reveal that (1) human evaluators rated the GPT-generated explanations as high quality in terms of linguistic fluency, informativeness, persuasiveness, and logical soundness, (2) the persuasive nature of these explanations, however, varied depending on the prompting strategy employed, and (3) this persuasiveness may result in incorrect judgments about the hatefulness of the content. Our study underscores the need for caution in applying LLM-generated explanations for content moderation. Code and results are available at https://github.com/Social-AI-Studio/GPT3-HateEval.
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。
