가이드 연계 퀴즈 · 중간 레벨

Microsoft AI 퀴즈

Microsoft의 Copilot 제품, Foundry 및 모델 카탈로그를 비교하고 테넌트 권한, 데이터 조건 및 구성 확인이 필요한 이유를 알아보세요.

관련 가이드 경로Microsoft AI
질문 1 의 8

사용자는 Microsoft 제품 전반에 걸쳐 공유된 Copilot 브랜드에서 무엇을 추론해야 합니까?

모든 질문 미리보기8 질문

퀴즈를 풀면 정답과 설명이 나타납니다.

  1. 사용자는 Microsoft 제품 전반에 걸쳐 공유된 Copilot 브랜드에서 무엇을 추론해야 합니까?
    • Products can have different workflows, licensing, and data controls
    • Every product has identical features and limits
    • All products use the same model configuration
    • A brand name guarantees access to every feature
  2. Microsoft Foundry는 가이드에서 어떤 역할을 합니까?
    • A consumer photo-album service
    • A single fixed model with identical settings
    • A platform for building AI applications and agents
    • A replacement for every Microsoft 365 app
  3. 모델이 카탈로그에 나타납니다. 모든 계정의 기능에 대해 무엇을 설정합니까?
    • All features are enabled worldwide
    • Nothing beyond catalog listing; region and account access still need checking
    • Every model can use every tool
    • The model is available in every product
  4. Enterprise Assistant는 문서 권한을 어떻게 존중해야 합니까?
    • Use a service account to search every file regardless of user access
    • Show restricted document titles but hide the text
    • Ask the model to decide which records are private
    • Use records the requesting user is authorized to access
  5. 엔터프라이즈 검색을 활성화하기 전에 문서 권한을 검토하는 이유는 무엇입니까?
    • Search can make existing overshared information easier to discover
    • Retrieval automatically fixes access rules
    • A model prevents oversharing by default
    • Permissions matter only for public documents
  6. What should teams check before processing data with a Microsoft service?
    • Assume one Microsoft agreement covers every service
    • Only the model’s training date
    • The relevant product’s data, licensing, and retention terms
    • A generic answer from a chatbot
  7. How should a team evaluate a custom AI workflow?
    • Use a model benchmark as the only measure
    • Test the complete workflow, including citations, tools, errors, and updates
    • Check only whether the app launches
    • Assume connectors cannot change results
  8. What should be recorded when comparing a Microsoft AI deployment?
    • Only the assistant’s display name
    • Only the vendor brand
    • The user’s personal email password
    • Product, tenant, account, model, and connectors involved