In an era increasingly influenced by artificial intelligence, the detection
of fake news is crucial, especially in contexts like election seasons where
misinformation can have significant societal impacts. This study evaluates the
effectiveness of various LLMs in identifying and filtering fake news content.
Utilizing a comparative analysis approach, we tested four large LLMs -- GPT-4,
Claude 3 Sonnet, Gemini Pro 1.0, and Mistral Large -- and two smaller LLMs --
Gemma 7B and Mistral 7B. By using fake news dataset samples from Kaggle, this
research not only sheds light on the current capabilities and limitations of
LLMs in fake news detection but also discusses the implications for developers
and policymakers in enhancing AI-driven informational integrity.

研究通过对比分析，评估了不同大型和小型 LLMs 在识别和过滤假新闻内容方面的有效性，并利用 Kaggle 的假新闻数据集样本探讨了当前 LLMs 在假新闻检测方面的能力和局限性，同时讨论了提高 AI 驱动信息完整性对开发人员和决策者的影响。

评估大型语言模型在检测虚假新闻中的功效：一项比较分析

Evaluating the Efficacy of Large Language Models in Detecting Fake News:  A Comparative Analysis

Large Language Models (LLMs) are already as persuasive as humans. However, we
know very little about why. This paper investigates the persuasion strategies
of LLMs, comparing them with human-generated arguments. Using a dataset of
1,251 participants in an experiment, we analyze the persuaion strategies of
LLM-generated and human-generated arguments using measures of cognitive effort
(lexical and grammatical complexity) and moral-emotional language (sentiment
and moral analysis). The study reveals that LLMs produce arguments that require
higher cognitive effort, exhibiting more complex grammatical and lexical
structures than human counterparts. Additionally, LLMs demonstrate a
significant propensity to engage more deeply with moral language, utilizing
both positive and negative moral foundations more frequently than humans. In
contrast with previous research, no significant difference was found in the
emotional content produced by LLMs and humans. These findings contribute to the
discourse on AI and persuasion, highlighting the dual potential of LLMs to both
enhance and undermine informational integrity through communication strategies
for digital persuasion.

大型语言模型（LLM）与人类一样具有说服力，但我们对其原因知之甚少。本文研究了 LLM 的说服策略，并将其与人类生成的论点进行比较。通过对 1251 名实验参与者的数据集分析，我们使用认知努力度（词汇和语法复杂性）以及道德情感语言（情感和道德分析）衡量 LLM 生成和人类生成的论点的说服策略。研究表明，LLM 生成的论点需要更多的认知努力，其语法和词汇结构比人类对应论点更为复杂。此外，LLM 表现出更强烈的倾向来运用道德语言，与人类相比，更频繁地使用积极和消极的道德基础。与先前的研究相反，没有发现 LLM 生成的情感内容与人类有显著差异。这些发现有助于关于人工智能和说服力的讨论，突显了 LLM 通过传播策略对信息的真实性产生同时增强和破坏的双重潜力。