[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fw2u8CZm-Q3UKEU-l7K6xhKBs9ArB5BzF6FU9haIe99A":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":19,"related":20,"source":30,"type":31},[],"2026-04-10 09:36:43",336342302,[8,9,10,11],"双向编码、自回归解码","动态掩码、整词掩码","文本填充、文档旋转","词向量因式分解、跨层参数共享",{"count":13,"courseId":14,"courseImg":15,"courseName":16,"workId":17,"workName":18},53,"53e1d2ef4961cca8eea3e23969ad2cb9","https:\u002F\u002Ftihai-oss-cloud.itihey.com\u002Fimg\u002F03a579384a6dc297c89809b582fcc767.png","默认课程","work_52079728","第二次作业","ALBERT 为了减少参数量,主要采用了哪两项技术",[21,32,41,50,59,68,71,80,91,101],{"answer":22,"createTime":5,"id":23,"options":24,"question":29,"source":30,"type":31},[],336342297,[25,26,27,28],"词袋模型","规则翻译模型","去噪自编码器","单向语言模型","BART 本质上属于哪一类模型","v1",0,{"answer":33,"createTime":5,"id":34,"options":35,"question":40,"source":30,"type":31},[],336342298,[36,37,38,39],"RoBERTa","T5","GPT","BERT","下列哪一类模型属于 Decoder-only 模型",{"answer":42,"createTime":5,"id":43,"options":44,"question":49,"source":30,"type":31},[],336342299,[45,46,47,48],"SOP 和 NSP","文本生成和机器翻译","MLM 和 NSP","MLM 和 SOP","BERT 的两个核心预训练任务是哪个组合",{"answer":51,"createTime":5,"id":52,"options":53,"question":58,"source":30,"type":31},[],336342300,[54,55,56,57],"对句子做情感分类","预测下一个句子","判断输入中的单词是否被替换过","生成整段文本","在 ELECTRA 中,判别器(Discriminator)的主要任务是什么",{"answer":60,"createTime":5,"id":61,"options":62,"question":67,"source":30,"type":31},[],336342301,[63,64,65,66],"15%","25%","10%","20%","在 BERT 的 Masked Language Model 任务中,输入文本中大约有多少比例的 token 会被 mask",{"answer":69,"createTime":5,"id":6,"options":70,"question":19,"source":30,"type":31},[],[8,9,10,11],{"answer":72,"createTime":5,"id":73,"options":74,"question":79,"source":30,"type":31},[],336342303,[75,76,77,78],"保留 NSP,并减少训练数据","用 LSTM 替代 Transformer","使用动态掩码,并舍弃 NSP","只使用静态掩码,并缩小批次","下列哪一项更符合 RoBERTa 的改进特点",{"answer":81,"createTime":82,"id":83,"options":84,"question":90,"source":30,"type":31},[],"2026-05-23 00:07:33",382646132,[85,86,87,88,89],"一向","朝向","剛才","以前","接近","向來道邊有賣餅家.向來",{"answer":92,"createTime":82,"id":93,"options":94,"question":100,"source":30,"type":31},[],382646133,[95,96,97,98,99],"將近","并且","大約","尚且","超過","時人以為年且百歲.且",{"answer":102,"createTime":82,"id":103,"options":104,"question":110,"source":30,"type":31},[],382646134,[105,106,107,108,109],"旋即","探索","尋找","古代長度單位,約合一丈","思索","所患尋差.尋"]