Quiz bu xóotaayu njaxas

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

1. Lan mooy wuute ci njaxasu xóotaayu jeton yi?
   • Ñaata xayma (ban diisaay) token bu nekk jot
   • Baatu-kàddu yiñ jëfandikoo
   • Done yiñ tàggat
   • Làkku génne
2. naka la ko MoD di def ngir mëna xam bu baax budget bi?
   • Dañu wàññi genn-wàllu poid model bi ci anam wu mucc
   • Ni ñuy jëfandikoo benn fraksioŋ top-k bu token ci blok bu nekk
   • Jëfandikoo xeetu jëfandikoo bu wuute ci tegtal bu nekk
   • Ni ñuy dindi bépp bàyyi xel
3. Lan mooy dall token yi routeur bi tannul ci benn block buñ jox?
   • Dañu leen di efaase ci limu toppalante
   • Dañu leen di yónnee ci beneen model
   • Dañuy romb blok bi jaaraleko ci lëkkaloo bi des, jaar ci te duñu soppiku
   • Dañu leen di soppi def ay nataal
4. Ban labo moo siiwal pexem njaxasu xóotaayu?
   • 2019
   • 2015
   • 2012
   • 2024
5. Lan mooy jur boole MoD ak njaxasu-ekspert yi?
   • MoDE, biy jaar ci xóotaayu tànneef ak kàngam yépp
   • Modèle bu amul benn parametre
   • Reseau convolutionnel bu sell
   • Sistem tokeniseer kese
6. Lan moo waral ñu taamu top-k bu takku ci kaw threshold bu token bu nekk ci MoD?
   • Dafay tax model bi génne mbind mu bari
   • Dafay tëye jëmmu tensor yi, te loolu lu baax la ci hardware bi
   • Dina dindi soxla tàggat yaram
   • Dafay yokk vocabulaire bi
7. What does the learned router do at a Mixture-of-Depths block?
   • Selects a new model family for every prompt
   • Predicts labels for the next training batch
   • Scores tokens so the block can select which ones receive full attention and MLP computation
   • Changes the vocabulary size at inference
8. Why does causal generation in MoD use auxiliary predictors for routing?
   • To let a model read the entire answer before generating
   • To remove the fixed top-k compute budget
   • To convert the sequence into an image
   • To make routing decisions without looking at future tokens