Ikibazo gihuza ibibazo · Hagati Urwego

Ikibazo cya Autoencoders Ikibazo

Ongera usubiremo VAE ikwirakwizwa, reparameterisation, KL isanzwe, amahugurwa ya ELBO, icyitegererezo, hamwe no gukoresha-gukwirakwizwa. Reba ibimenyetso n'urugero.

Inzira zijyanye no kuyoboraImashini zitandukanye
Ikibazo 1 Bya 8

Niki kodegisi mubisohoka VAE kugirango yinjizwe?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Niki kodegisi mubisohoka VAE kugirango yinjizwe?
    • Ibipimo byo gukwirakwiza amahirwe, nkuburyo butandukanye
    • Ingingo imwe ihamye mumwanya wihishe
    • Ikirango cy'ishuri
    • Kopi yubatswe yinyandiko
  2. Niyihe ntego yuburiganya bwo gusubiramo?
    • Gukora decoder nini
    • Kwemerera gradients gutembera muburyo bwo gutoranya intambwe
    • Gukuraho ibikenewe kuri decoder
    • Kugirango ugabanye umwanya wihishe imbere
  3. In the standard Gaussian VAE formulation, what does the KL term encourage?
    • The decoder to predict future input tokens
    • The encoder to remove all randomness
    • The approximate posterior to stay close to the chosen prior, commonly a standard normal
    • Every latent variable to have the same observed value
  4. Igikorwa gifatika VAEs yerekana mugihe cyamahugurwa yitwa:
    • Igisenge cyambukiranya imipaka
    • Igihombo cyo guhangana
    • Hagati ya kare
    • Ibimenyetso Byibanze (ELBO)
  5. In a Gaussian VAE, what enables generation of a new sample?
    • Sample a latent from the prior and decode it with the learned decoder
    • Copy a training image directly without a model
    • Run the encoder backward with no latent sampling
    • Choose a random class label without decoding
  6. Muri moderi ikwirakwizwa rwihishwa nka Stable Diffusion, ni uruhe ruhare VAE igira?
    • Iranga amashusho y'amahugurwa
    • Ihagarika amashusho kumwanya wihishe aho diffusion ikorera, hanyuma igasubira inyuma kuri pigiseli
    • Isimbuza inyandiko kodegisi
    • Iteganya urwego rwurusaku
  7. In variational inference, what does the ELBO provide for the marginal log likelihood log p(x)?
    • An exact value of log p(x) for every model
    • A guarantee that each generated sample is realistic
    • A tractable lower bound that can be maximized instead of directly maximizing log p(x)
    • A posterior distribution that requires no inference model
  8. What distribution does a VAE decoder model given a latent vector?
    • The encoder’s approximate posterior q(z|x)
    • The training set’s class-frequency histogram only
    • The optimizer’s learning-rate schedule
    • The likelihood of an observation conditioned on that latent, p(x|z)