[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fXhZa3ghNPJW9EQgOkhxTTnXkUFO_WwY7BKAvxE-XVXM":3},{"answer":4,"createTime":5,"id":6,"options":7,"origin":12,"question":13,"related":14,"source":25,"type":36},[],"2026-05-14 09:37:27",1083069922,[8,9,10,11],"数据污染","过度优化或"应试"","评估维度的局限性","测试题目过于简单",{},"评估大模型能力的常用基准测试面临的挑战和争议主要有哪些?( )",[15,27,37,46,56,64,73,76,85,90],{"answer":16,"createTime":17,"id":18,"options":19,"question":24,"source":25,"type":26},[],"2026-05-14 09:37:26",1083068943,[20,21,22,23],"提供新奇视角","生成多样化示例","进行概念组合与重构","提供最终的、无需修改的成品方案","在认知策略层面,AI辅助进行"创意启发"的主要作用不包括以下哪项?( )","v2",0,{"answer":28,"createTime":17,"id":29,"options":30,"question":35,"source":25,"type":36},[],1083069347,[31,32,33,34],"参数规模巨大","训练数据海量","可能展现出涌现能力","物理体积庞大","大语言模型的"大"主要体现在哪些方面?( )",1,{"answer":38,"createTime":17,"id":39,"options":40,"question":45,"source":25,"type":26},[],1083069491,[41,42,43,44],"内容梳理","概念澄清","自我评估","目标规划","将复杂、抽象的专业术语用更通俗易懂的语言进行转述,或者通过生动、贴切的类比和隐喻来帮助理解,这属于AI在哪项认知策略上的应用?( )",{"answer":47,"createTime":5,"id":48,"options":49,"question":55,"source":25,"type":36},[],1083069800,[50,51,52,53,54],"文本合成","图像合成","音频合成","视频合成","三维模型合成","AIGC(人工智能生成内容)的主要应用领域\u002F内容类型有哪些?( )",{"answer":57,"createTime":5,"id":58,"options":59,"question":62,"source":25,"type":63},[],1083069836,[60,61],"对","错","斯坦福小镇(Generative Agents)案例的核心贡献是展示了如何让智能体通过观察和模仿来学习操作智能手机上的各种应用程序.( )",3,{"answer":65,"createTime":17,"id":66,"options":67,"question":72,"source":25,"type":26},[],1083069873,[68,69,70,71],"情感投入","快速响应","批判性参与","编程技巧",""关键洞察"指出,与AI在认知策略层面进行协同,本质上是一场需要高度____的互动过程.( )",{"answer":74,"createTime":5,"id":6,"options":75,"question":13,"source":25,"type":36},[],[8,9,10,11],{"answer":77,"createTime":5,"id":78,"options":79,"question":84,"source":25,"type":36},[],1083069951,[80,81,82,83],"清晰性","目标导向与迭代优化","任务分解","上下文提供","构建有效提示词的四大核心原则是什么?( )",{"answer":86,"createTime":17,"id":87,"options":88,"question":89,"source":25,"type":63},[],1083070034,[60,61],"大语言模型的"幻觉"现象是指模型生成的内容在语言上不流畅、语法错误百出,但信息内容却是完全准确的.( )",{"answer":91,"createTime":17,"id":92,"options":93,"question":94,"source":25,"type":63},[],1083070058,[60,61],"无论是哪种模态的AIGC,其生成能力均以海量、高质量且模态间对齐良好的训练数据为基础.( )"]