[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f19P0Bjp-LOqq0nMAggghHP4vXMLJU2_fiKDJXIxXkpg":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":13,"related":14,"source":22,"type":43},[],"2025-06-29 08:00:32",1069533043,[8,9,10,11],"K-means 聚类","主成分分析(PCA)","随机森林","层次聚类",{},"以下哪些是常见的无监督学习方法?( )",[15,24,29,34,44,54,63,72,81,90],{"answer":16,"createTime":5,"id":17,"options":18,"question":21,"source":22,"type":23},[],1069532920,[19,20],"对","错","今日头条、抖音等媒体通过自动推荐技术,能够理解用户的喜好并实现精准推送.( )","v2",3,{"answer":25,"createTime":5,"id":26,"options":27,"question":28,"source":22,"type":23},[],1069532924,[19,20],"大语言模型的推理能力与人类相似,因此可以处理所有类型的推理任务.( )",{"answer":30,"createTime":5,"id":31,"options":32,"question":33,"source":22,"type":23},[],1069532932,[19,20],"大语言模型的主要优势是它们能够理解输入文本的语法规则,而不是语言的统计规律.( )",{"answer":35,"createTime":5,"id":36,"options":37,"question":42,"source":22,"type":43},[],1069532934,[38,39,40,41],"透明性","公正性","责任性","可解释性","AI 伦理的核心原则包括?( )",1,{"answer":45,"createTime":5,"id":46,"options":47,"question":52,"source":22,"type":53},[],1069532936,[48,49,50,51],"学习如何处理结构化数据","提供监督信号来调整权重","学习从大量无标注文本数据中捕捉语言规律","提高模型的可解释性","GPT模型中的"预训练"阶段的主要目的是?( )",0,{"answer":55,"createTime":5,"id":56,"options":57,"question":62,"source":22,"type":53},[],1069532938,[58,59,60,61],"视觉","听觉","逻辑思维","触觉","以下哪项不属于人类智能的感知能力?( )",{"answer":64,"createTime":5,"id":65,"options":66,"question":71,"source":22,"type":53},[],1069532940,[67,68,69,70],"让 AI 运行更快","让 AI 学习人类情感","增强 AI 对欺骗性输入的防御能力","让 AI 更容易理解数据","对抗性训练的主要目的是?( )",{"answer":73,"createTime":5,"id":74,"options":75,"question":80,"source":22,"type":53},[],1069532942,[76,77,78,79],"卷积神经网络(CNN)","循环神经网络(RNN)","自注意力网络(Transformer)","生成对抗网络(GAN)","GPT模型属于以下哪种类型的神经网络架构?( )",{"answer":82,"createTime":5,"id":83,"options":84,"question":89,"source":22,"type":43},[],1069532944,[85,86,87,88],"图灵","布尔","罗素","香农","数理逻辑为形式化描述人类的思维提供了坚实的理论基础,成为人工智能大厦的第一块基石.为这一基石做出杰出贡献的科学家有. ( )",{"answer":91,"createTime":5,"id":92,"options":93,"question":98,"source":22,"type":53},[],1069532955,[94,95,96,97],"卷积核","注意力机制","区域候选","反卷积","图像分类中的Vision Transformer的特点是( )"]