Kediska Hababka Feejignaanta

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. What does token attention combine when forming an output representation?
   • Permanent copies of the entire conversation on disk
   • Value information from allowed positions using computed weights
   • Only manually chosen keywords with no learned scores
   • A translation dictionary that removes every other token
2. Warqadee calaamadee 2017 soo bandhigtay qaab dhismeedka transformer?
   • 'Waxbarashada Dheeraadka ah ee Haraaga ah ee Aqoonsiga Sawirka'
   • 'Ku ciyaarista Atari Waxbarasho Xoojin qoto dheer'
   • 'Fiiro gaar ah ayaa ah waxa aad u baahan tahay'
   • 'Shabakadaha Adversarial Generative'
3. Maxay yihiin saddexda qaybood ee loo xisaabiyay calaamad kasta?
   • Weydiinta, furaha, iyo qiimaha
   • Gelida, soo saarista, iyo eexda
   • Dhexdhexaad, dhexdhexaad, iyo qaab
   • Codeeyaha, furaha, iyo khasaaraha
4. Habka softmax (QK^T / √d_k) V, maxaa loo qaybiyaa √d_k?
   • Si natiijada loo beddelo tirooyin
   • Si meesha looga saaro tirooyinka taban
   • Si aad moodelku ugu shaqeeyo GPU
   • Si loo ilaaliyo dhibcooyinka badeecada inay noqdaan kuwo aad u weyn softmax ka hor
5. Why does dense self-attention involve growing pairwise work for very long token sequences?
   • Sababtoo ah waa inay soo dejiso xog cusub internetka
   • Sababtoo ah qoraalka dheer ayaa had iyo jeer ka kooban kelmado dhif ah
   • Sababtoo ah qiimuhu wuxuu ku koraa labajibbaaran dhererka isku xigxiga (O(n²))
   • Sababtoo ah waxay u beddeshaa luuqad barnaamijeed oo gaabis ah
6. What capacity does multi-head attention provide through separate learned projections?
   • A guarantee that every head has one unique named meaning
   • Several attention representations that can capture different relationships in parallel
   • Removal of every token embedding
   • A reduction of vocabulary size by definition
7. In a scalar attention example, values two and six receive weights three-quarters and one-quarter. What is their weighted output?
   • Sideed
   • Afar
   • Laba
   • Saddex
8. What does a causal attention mask prevent in an autoregressive text decoder?
   • Using later token positions when forming the current position's representation
   • Using any previous token
   • Computing any query or key vectors
   • Learning all model parameters