HumT DumT：测量和控制LLM中的类人语言
HumT DumT: Measuring and controlling human-like language in LLMs
AI总结：
本文提出HumT和SocioT指标测量LLM输出中的类人语气及社会感知，并发现用户在多数情境下偏好非类人输出；同时提出DumT方法，在不损害模型性能的前提下系统降低类人语气，以缓解拟人化语言带来的欺骗、过度依赖等风险。
摘要：
LLM是否应该生成使它们看起来像人类的语言？类人语言可能会改善用户体验，但也可能导致欺骗、过度依赖和刻板印象。评估这些潜在影响需要一种系统的方法来测量LLM输出中的类人语气。我们引入了HumT和SocioT，这是基于LLM相对概率的类人语气和文本数据中社会感知其他维度的指标。通过在偏好和使用数据集上测量HumT，我们发现在许多情况下，用户更喜欢LLM中不那么像人类的输出。HumT还提供了对拟人化感知和影响的见解：类人LLM输出与温暖、社会亲密度、女性气质和低地位高度相关，这些与上述危害密切相关。我们引入了DumT，这是一种使用HumT系统控制和减少类人语气程度同时保持模型性能的方法。DumT提供了一种实用的方法来减轻与拟人化语言生成相关的风险。
英文摘要：
Should LLMs generate language that makes them seem human? Human-like language might improve user experience, but might also lead to deception, overreliance, and stereotyping. Assessing these potential impacts requires a systematic way to measure human-like tone in LLM outputs. We introduce HumT and SocioT, metrics for human-like tone and other dimensions of social perceptions in text data based on relative probabilities from an LLM. By measuring HumT across preference and usage datasets, we find that users prefer less human-like outputs from LLMs in many contexts. HumT also offers insights into the perceptions and impacts of anthropomorphism: human-like LLM outputs are highly correlated with warmth, social closeness, femininity, and low status, which are closely linked to the aforementioned harms. We introduce DumT, a method using HumT to systematically control and reduce the degree of human-like tone while preserving model performance. DumT offers a practical approach for mitigating risks associated with anthropomorphic language generation.
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。
