가이드 연계 퀴즈 · 중간 레벨

멀티 토큰 예측 교육 퀴즈

DeepSeek-V3의 고유한 순차 설계, 선택적 추론 헤드 및 대상 오프셋을 사용하여 멀티 토큰 손실 및 공유 트렁크 헤드를 검토합니다.

관련 가이드 경로다중 토큰 예측 교육
질문 1 의 8

In the cited Meta study, what architecture predicts several future tokens from a training position?

모든 질문 미리보기8 질문

퀴즈를 풀면 정답과 설명이 나타납니다.

  1. In the cited Meta study, what architecture predicts several future tokens from a training position?
    • A mandatory separate complete language model for every word
    • Independent output heads above a shared transformer trunk
    • One head that can target only the preceding token
    • A larger vocabulary with no additional prediction targets
  2. 사전 학습에서 멀티 토큰 예측 목표를 특히 사용한 모델은 무엇인가요?
    • 원래 Word2Vec
    • 컨볼루셔널 이미지 분류기
    • N-그램 모델
    • DeepSeek-V3
  3. What additional supervision does multi-token training supply at a prefix position?
    • Loss terms for multiple future-token targets
    • Only a larger batch size with the same single target
    • A mandatory image-reconstruction target
    • No prediction loss at any position
  4. How are auxiliary predictions used in the Meta paper’s self-speculative inference experiment?
    • They are accepted unconditionally without verification
    • They automatically retrain the model on every prompt
    • They propose tokens that are then checked by the decoding procedure
    • They encrypt every generated token
  5. 속도 향상이 필요하지 않으면 훈련 후 보조 헤드는 어떻게 되나요?
    • 유지되어야 하며 모델 크기를 두 배로 늘려야 합니다.
    • 폐기할 수 있으므로 배포 크기는 변경되지 않습니다.
    • 주목층으로 변해
    • 그들은 토크나이저가 됩니다
  6. Which evaluated task family showed notable gains in the Meta multi-token study?
    • Code generation benchmarks
    • A guarantee of perfect code compilation in every language
    • Only alphabetizing raw filenames
    • A required quantum-sensor prediction task
  7. In a four-head offset design predicting from prefix position t, which token is the fourth head’s target?
    • The token at t−4
    • The token at t+1 for every head
    • Only the token at t itself
    • The token at t+4
  8. What distinguishes DeepSeek-V3’s stated MTP architecture from the cited independent parallel-head implementation?
    • It predicts only tokens before the prefix
    • It requires every target to be an image pixel
    • Sequential auxiliary modules that preserve a causal chain across prediction depths
    • It removes the language-model output vocabulary