examination/topics/qunar-ai-fullstack/ai-memory/fill_blank.json

{
  "topic": "ai-memory",
  "type": "fill_blank",
  "schema_version": "1.0.0",
  "generated": "2026-09-02T00:00:00Z",
  "questions": [
    {
      "id": "fb-001",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "Agent 记忆分层架构的四种记忆类型是：工作记忆、情景记忆、语义记忆和______记忆。",
      "explanation": "记忆分层由工作记忆（短期）、情景记忆（中期）、语义记忆（长期）和程序性记忆组成。",
      "source": null,
      "related": [],
      "answer": [
        "程序性"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-002",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "记忆的正确流转顺序是：工作记忆 → 情景记忆 → 语义记忆 → ______记忆。",
      "explanation": "工作记忆在会话结束后转为情景记忆，提炼为语义记忆，最终指导程序性记忆。",
      "source": null,
      "related": [],
      "answer": [
        "程序性"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-003",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "LRU 缓存使用什么数据结构组合来实现 O(1) 的访问和淘汰？",
      "explanation": "LRU 使用双向链表维护顺序，哈希表实现 O(1) 查找。",
      "source": null,
      "related": [],
      "answer": [
        "双向链表 + 哈希表"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-004",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "混合记忆淘汰策略的评分公式是：score = α × 时间衰减(t) + β × 频率(f) + γ × ______(i)。",
      "explanation": "混合策略综合考虑时间衰减、访问频率和重要性三个维度。",
      "source": null,
      "related": [],
      "answer": [
        "重要性"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-005",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "混合策略中，时间衰减通常使用什么数学函数计算？",
      "explanation": "时间衰减函数如 e^(-λt)，其中 t 是距上次访问的时间。",
      "source": null,
      "related": [],
      "answer": [
        "e^(-λt)"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-006",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "记忆系统中，余弦相似度超过______时触发冲突检测。",
      "explanation": "语义相似度的余弦值 > 0.8 时，记忆对进入冲突检测。",
      "source": null,
      "related": [],
      "answer": [
        "0.8"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-007",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "部分记忆冲突消解中，当评分差值小于______时无法自动消解，需要向用户澄清。",
      "explanation": "当评分差值 < 0.1 时，系统无法自动判定，需主动向用户确认。",
      "source": null,
      "related": [],
      "answer": [
        "0.1"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-008",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "完全记忆冲突的风险等级分为两类：______（危险）和安全。",
      "explanation": "Critical 包括医疗、金融、生产环境配置等；Safe 包括代码风格、文档格式等。",
      "source": null,
      "related": [],
      "answer": [
        "Critical"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-009",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "安全环境的 YAML 配置中，三条优先级规则的权重分别是：时间 0.4、来源 0.4、______ 0.2。",
      "explanation": "安全环境使用时间（0.4）、来源优先级（0.4）和访问次数（0.2）。",
      "source": null,
      "related": [],
      "answer": [
        "访问次数"
      ],
      "answer_rule": "any"
    },
    {
      "id": "fb-010",
      "type": "fill_blank",
      "difficulty": 2,
      "tags": [
        "ai",
        "memory",
        "lru"
      ],
      "question": "记忆冲突解决中，来源可信度的优先级顺序是：用户明确告知 > 从对话推断 > ______。",
      "explanation": "用户明确告知的可信度最高，其次是对话推断，最后是自动生成。",
      "source": null,
      "related": [],
      "answer": [
        "自动生成"
      ],
      "answer_rule": "any"
    }
  ]
}