[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fi1sDENQPxmaA8yr0g1nd3ZS3V9WjKVUOKYGKDvBDgJY":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":10,"question":17,"related":18,"source":29,"type":117},[],"2025-03-14 15:42:06",179150964,[8,9],"正确","错误",{"count":11,"courseId":12,"courseImg":13,"courseName":14,"workId":15,"workName":16},50,"e2f50b2f7e53141ca8dc251b2cba11b4","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002Fa096acd23dd8acdfe38f1bd3183c74c3.png","安徽省人工智能通识课（AI+心理学）","67a6e937ad614268a9cd3f37c9a4ce79","第二章作业","Transformer模型的自注意力机制只能处理序列数据,无法处理图像数据",[19,31,41,51,61,71,79,89,98,107],{"answer":20,"createTime":21,"id":22,"options":23,"question":28,"source":29,"type":30},[],"2026-06-02 05:22:24",178401555,[24,25,26,27],"完全相同","相互矛盾","相互促进","毫无关系","认知脑科学与人工神经网络的关系是","v1",0,{"answer":32,"createTime":33,"id":34,"options":35,"question":40,"source":29,"type":30},[],"2026-06-02 06:39:32",178401556,[36,37,38,39],"突触","树突","细胞体","轴突","神经元之间的信息传递是通过什么结构实现的",{"answer":42,"createTime":43,"id":44,"options":45,"question":50,"source":29,"type":30},[],"2026-06-02 05:02:23",178401557,[46,47,48,49],"明斯基","罗森布拉特","Hinton","麦卡洛克和皮茨","单层感知机模型的提出者是",{"answer":52,"createTime":53,"id":54,"options":55,"question":60,"source":29,"type":30},[],"2026-06-01 23:00:59",178401558,[56,57,58,59],"异或问题","非问题","或问题","与问题","单层感知机模型无法解决的问题是",{"answer":62,"createTime":63,"id":64,"options":65,"question":70,"source":29,"type":30},[],"2026-06-01 07:37:02",178401559,[66,67,68,69],"参数更少","计算效率更高","非线性表达能力","线性表达能力","多层感知机模型的核心优势在于",{"answer":72,"createTime":73,"id":74,"options":75,"question":78,"source":29,"type":30},[],"2026-06-02 03:09:09",178401560,[76,48,47,77],"Yann LeCun","Yoshua Bengio","卷积神经网络(CNN)最早是由谁提出的",{"answer":80,"createTime":81,"id":82,"options":83,"question":88,"source":29,"type":30},[],"2026-06-01 20:28:38",178401561,[84,85,86,87],"分类","降维","提取特征","防止过拟合","卷积神经网络中,卷积层的主要作用是",{"answer":90,"createTime":33,"id":91,"options":92,"question":97,"source":29,"type":30},[],178401562,[93,94,95,96],"并行处理","序列处理","自适应学习","非线性映射","循环神经网络(RNN)的核心特性是",{"answer":99,"createTime":33,"id":100,"options":101,"question":106,"source":29,"type":30},[],178401563,[102,103,104,105],"更新门","遗忘门","输入门","输出门","LSTM模型中,用于决定哪些信息需要被保留的门是",{"answer":108,"createTime":109,"id":110,"options":111,"question":116,"source":29,"type":30},[],"2026-06-02 03:47:06",178401564,[112,113,114,115],"层归一化","多头注意力","残差连接","前馈网络","Transformer模型中,用于解决长距离依赖问题的关键技术是",3]