quiz bu lëkkale ak tegtal · Dëgër Niveau

InfoNCE ak SimCLR

Saytu InfoNCE mébet bu baax ak bu baaxul, xalaati SimCLR ak boppu projection, ak cër yi tàngoor ak dayo lots.

Yooni njiit yu jëm ciInfonce ak Simclr Mébet
Laaj 1 de 8

Lan la mébetu InfoNCE di ñaax ab model mu def?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. Lan la mébetu InfoNCE di ñaax ab model mu def?
    • Defal ab laajte bu gëna méngoo ak njariñam dëgg moo gën ñaawtéef yu bari
    • Dangay fàttaliku valeur pixel yi ci nataal bu nekk
    • Wàññil limu paramet model yi
    • Xalaatal baat bi ci topp ci santaane
  2. Ci SimCLR, naka lañuy defaree ñaar ñu baax?
    • Soo boole ab nataal ak etiketu nit
    • Ni ñuy jëfandikoo ñaari yokk yu wuute ci benn nataal
    • Soo tànnee ñaari nataal yu wuute lool
    • Sooy wëlbati done yépp
  3. Ban cër la tàngoor τ def ci NT-Xent / InfoNCE?
    • Dafay tëral ni ñuy jàngee
    • Dafay saytu dayo lote bi ci saasi
    • Dafay xayma poñ yi ñuy nuru; τ bu wàññeeku dafay ñaw séddale bi ba noppi fësal negatif yu dëgër yi
    • Dafay wane limu couche yi ci encoder bi
  4. Lan mooy dall boppu projection bu SimCLR ginaaw biñu ko njëkkee tàggat?
    • Mooy nekk klassifikatër bu mujj
    • Dañu koy gelé ba noppi jëfandikoo ko ni encoder
    • Dañu koy ñaari yoon ngir mëna jàpp
    • Dañu koy sànni; représentation bi laata bopp bi lañuy jëfandikoo ngir liggéey yi ci suuf
  5. Lu tax SimCLR di yéem ci dayo yu bari?
    • Joxe misaal yu bari yu baaxul ci jéego bu nekk ci tàggat yaram
    • Ngir wàññi jëfandikoo mémoire
    • Ngir moytu jëfandikoo yokkute
    • Ngir dindi soxla parametru tàngoor
  6. In SimCLR, which examples usually act as negatives for one anchor view?
    • The matching view of the same image
    • The other augmented views in the batch that come from different images
    • Only examples with the same class label
    • Views produced by a second encoder trained on labels
  7. Why does SimCLR use a nonlinear projection head before computing its contrastive loss?
    • It replaces the image encoder
    • It supplies class labels during pretraining
    • The learned projection helps the contrastive objective while the representation before the head can transfer downstream
    • It removes the need for data augmentation
  8. For an InfoNCE anchor, what should happen to the positive similarity relative to sampled negatives?
    • Every negative should have a higher score than the positive
    • All scores must be identical
    • The positive is excluded from the objective
    • The positive should receive a higher score than the negatives