Cod-koobiyeyaasha yaryar ee Autoencoders ee Kediska Fasiraadda

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Waa maxay ujeedada ugu weyn ee tababarka autoencoder yar ee hawlqabadyada moodeelka?
   • Si loo kala furfuro dhaq-dhaqaaqyada qalloocan oo la tarjumi karo, sifooyin inta badan go'doonsan
   • Si loo cadaadiyo qaabka si uu dhakhso ugu socdo
   • Si loo turjumo qoraalka u dhexeeya luqadaha
   • Si loo kordhiyo saxsanaanta moodeelka ee bartilmaameedyada
2. How does an SAE’s sparsity penalty affect hidden-unit activations?
   • It guarantees each unit represents exactly one concept
   • It encourages only a small subset of the wider hidden layer to activate for an input
   • It makes every hidden unit activate equally
   • It removes the reconstruction objective
3. Ku dhawaad imisa astaamood ayaa Anthropic soosaaray markii la cabiray SAEs ilaa Claude 3 Sonnet sanadka 2024?
   • Ilaa 500
   • Ilaa 2,000
   • Ilaa 34 milyan
   • Ilaa 100 bilyan
4. Muxuu tusay 'Albaabka dahabka Claude'?
   • Taasi SAE-yada waxaa lagu tababari karaa oo kaliya sawirada
   • In moodalku aanu soo saari karin qoraal
   • Tilmaamahaasi wax saameyn ah kuma laha wax soo saarka
   • Taas oo isku dhejinta hal sifo oo sareysa ayaa sabab u ah dabeecadda moodeelka
5. Waa maxay sababta lakabka qarsoon ee SAE caadi ahaan aad uga ballaadhan yahay hawlgelinta ay dib u dhisto?
   • Sababtoo ah fikradaha ayaa ka sarreeya cabbirro ka yar inta ay dhab ahaantii u baahan yihiin, sidaas darteed cutubyo badan ayaa ka caawiya kala saaridooda
   • Si aad tababarka uga dhigto mid jaban
   • Sababtoo ah lakabyada ballaaran ayaa had iyo jeer sax ah
   • Si loo yareeyo tirada xuduudaha
6. Waa kuwee xaddidaadda la og yahay ee cod-bixiyeyaasha autoencoders ee hadda jira?
   • Laguma tababari karo farac hoose
   • Tilmaamaha kala qaybsanaanta, halkaas oo hal fikradi u jabto sifooyin badan oo kala duwan
   • Waxay kaliya ka shaqeeyaan shabakadaha convolutional
   • Waxay meesha ka saarayaan baahida calaamad kasta oo aadanaha
7. Which two objectives are commonly combined when training a sparse autoencoder?
   • A policy gradient and value loss
   • A next-token loss and a contrastive loss
   • Reconstruction error and a penalty on hidden activations
   • A calibration error and classification accuracy
8. What does a TopK sparse autoencoder do to encourage sparsity?
   • Keeps the K largest input pixels and discards the decoder
   • Sets all encoder weights to zero
   • Trains K different tokenizers
   • Keeps only the K largest hidden activations for an input