在思维链中引出不确定性以缓解对预测有害用户行为的偏见
Eliciting Uncertainty in Chain-of-Thought to Mitigate Bias against Forecasting Harmful User Behaviors
AI总结：
本文研究如何利用思维链中的不确定性表示来缓解 LLM 在社交媒体对话预测中对有害用户行为的预测偏见，并在 5 个开源模型和 2 个数据集上验证其效果。
摘要：
对话预测任务要求模型预测一段正在进行的对话的结果。例如，它可以应用于社交媒体内容审核中，在有害用户行为发生之前进行预测，从而实现预防性干预。虽然大型语言模型（LLMs）最近被提出作为对话预测的有效工具，但尚不清楚它们可能存在哪些偏见，尤其是在预测我们在审核过程中要求它们预测的（可能有害的）结果时。本文探讨了模型不确定性在多大程度上可以作为缓解潜在偏见的工具。具体而言，我们提出三个主要研究问题：1）当我们要求模型表示其不确定性时，LLM 的预测准确率如何变化；2）当我们要求模型表示其不确定性时，LLM 的偏见如何变化；3）我们如何利用不确定性表示来减少或完全缓解偏见，而无需大量训练数据点。我们针对 5 个开源语言模型回答了这些问题，并在 2 个专门用于评估社交媒体审核对话预测的数据集上进行了测试。
英文摘要：
Conversation forecasting tasks a model with predicting the outcome of an unfolding conversation. For instance, it can be applied in social media moderation to predict harmful user behaviors before they occur, allowing for preventative interventions. While large language models (LLMs) have recently been proposed as an effective tool for conversation forecasting, it's unclear what biases they may have, especially against forecasting the (potentially harmful) outcomes we request them to predict during moderation. This paper explores to what extent model uncertainty can be used as a tool to mitigate potential biases. Specifically, we ask three primary research questions: 1) how does LLM forecasting accuracy change when we ask models to represent their uncertainty; 2) how does LLM bias change when we ask models to represent their uncertainty; 3) how can we use uncertainty representations to reduce or completely mitigate biases without many training data points. We address these questions for 5 open-source language models tested on 2 datasets designed to evaluate conversation forecasting for social media moderation.
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。
