引导式测验 · 中等 等级

AI 幻觉测验

检查流畅的捏造、来源真实性、检索限制、引文验证以及对有关事实可靠性的主张的基于证据的评估。

相关引导路径人工智能幻觉
问题 1 的 8

为什么仅靠下一令牌预测无法验证生成答案中的事实声明?

预览每个问题8 问题

当您参加测验时,就会出现答案和解释。

  1. 为什么仅靠下一令牌预测无法验证生成答案中的事实声明?
    • 每个模型仅连接到不可靠的网站
    • 所有用户都会提出语法错误的问题
    • 合理的延续并不等同于证明该陈述正确的证据
    • 流利的回答自动证明了事实
  2. 为什么幻觉特别危险?
    • 它们总是以不正确的语法编写,因此用户会忽略它们
    • 它们仅出现在错误消息中
    • 它们导致模型崩溃
    • 它们的表达流畅而自信,因此很容易遗漏错误
  3. 检索增强生成(RAG)能否完全消除幻觉?
    • 是的,RAG 保证提供事实答案
    • 不,模型仍然可以矛盾或忽略嘈杂或冲突的检索文本
    • 是的,但仅限周末
    • 否,因为 RAG 从未提供任何文件
  4. 哪个是幻觉的真实例子?
    • 该模型引用了不存在的法院案例
    • 模型拒绝回答问题
    • 模型提出澄清问题
    • 模型返回空白响应
  5. 哪种工作流程可以帮助检查生成的事实答案?
    • Trust every citation because it looks formal
    • Treat a low sampling temperature as proof of correctness
    • Hide the evidence so the answer is easier to read
    • Retrieve relevant material and verify whether its statements and citations support the claims
  6. How should a claim that a new model has eliminated hallucinations be assessed?
    • Accept it solely because the announcement sounds confident
    • Assume one successful example proves every application
    • Use defined tasks, evidence, and measurement conditions rather than assume universal reliability
    • Judge only whether the model produces polished prose
  7. A generated answer cites a real page, but the page does not support the associated factual statement. What remains unverified?
    • The support for that particular claim
    • Whether a URL can be opened
    • Whether the model used fluent sentences
    • Whether the citation contains text
  8. A supplied document says delivery takes five business days, but its generated summary says two. Which check reveals the error?
    • Counting how many sentences the summary contains
    • Comparing the summary's quantity with the source document
    • Checking only whether the prose sounds confident
    • Assuming summaries may replace any source number