Macluumaadka iyo Kediska Ujeeddooyinka SimCLR

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Maxay ujeeddada InfoNCE ku dhiirigelisaa moodalku inuu sameeyo?
   • Samee su'aal la mid ah sida saxda ah ee dhabta ah marka loo eego kuwa taban oo badan
   • Xusuusnow qiimaha pixel saxda ah ee sawir kasta
   • Yaree tirada cabbiraadaha moodeelka
   • Saadaali kelmada ku xigta jumlad
2. Gudaha SimCLR, sidee loo abuuray lammaane togan?
   • Iyadoo la isku dhejinayo sawir leh calaamadda aadanaha
   • Adigoo adeegsanaya laba kor u qaadis random oo kala duwan isla sawirka
   • Adiga oo dooranaya laba sawir oo gebi ahaanba kala duwan
   • Adiga oo wareejinaya kaydka xogta oo dhan
3. Doorkee ayuu heerkulka τ ka ciyaara NT-Xent/InfoNCE?
   • Waxay dejisaa heerka waxbarashada
   • Waxay si toos ah u xakameysaa cabbirka dufcadda
   • Waxay miisaamaysaa dhibcaha isku midka ah; hoose τ waxay afaysa qaybinta waxayna xooga saartaa xumaanta adag
   • Waxay go'aamisaa tirada lakabyada ku jira encoder-ka
4. Maxaa ku dhacaya madaxa saadaasha SimCLR ka dib tababarka hore?
   • Waxay noqotaa kala-saarkii ugu dambeeyay
   • Waa la qaboojiyey oo loo istcimaali jiray cod-bixiye ahaan
   • Waxa loo koobiyeeyay tilmaamid
   • Waa la iska tuuraa; matalaadda ka hor inta aan madaxa loo isticmaalin hawlaha hoose
5. Maxay SimCLR ugu tiirsan tahay cabbirrada dufcadaha waaweyn?
   • Si aad u bixiso tusaalooyin badan oo taban gudaha tillaabo kasta oo tababar ah
   • Si loo yareeyo isticmaalka xusuusta
   • Si looga fogaado isticmaalka kordhinta
   • Si loo baabi'iyo baahida loo qabo cabbirka heerkulka
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