Guide-yakabatana mibvunzo · Pakati Nhanho

Variational Autoencoders Quiz

Review VAE distributions, reparameterization, KL regularization, ELBO training, sampling, and latent-diffusion use. Check evidence and scope.

Nzira dzekutungamira dzinoenderanaVariational Autoencoders
Mubvunzo 1 ye 8

Chii chinoitwa ne encoder mune yeVAE inoburitsa kune yakapihwa yekupinza?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Chii chinoitwa ne encoder mune yeVAE inoburitsa kune yakapihwa yekupinza?
    • Maparamendi ekugovera kungangoita, sekureva uye musiyano
    • Nzvimbo imwechete yakagadziriswa munzvimbo yakadzikama
    • Chiratidzo chekirasi
    • Ikopi yakagadziridzwa yezvakaiswa
  2. Chii chinangwa cheiyo reparameterization trick?
    • Kuita kuti decoder ive yakakura
    • Kubvumira magradients kuyerera kuburikidza neasina kurongeka sampling nhanho
    • Kubvisa kudiwa kwedecoder
    • Kumanikidza nzvimbo yakadzikama zvakare
  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. Basa rechinangwa iro VAEs rinowedzera panguva yekudzidziswa rinonzi:
    • Cross-entropy ceiling
    • Kurasikirwa nemuvengi
    • Kureva squared margin
    • Evidence Lower Bound (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. Mune yakadzikama yekuparadzira modhi seStable Diffusion, ibasa rei rinoitwa neVAE?
    • Inonyora mifananidzo yekudzidzira
    • Iyo inomanikidza mifananidzo kune yakavanzika nzvimbo uko kupararira kunomhanya, yobva yadhirodha kudzokera kumapixels
    • Inotsiva iyo encoder yemavara
    • Inoronga mazinga eruzha
  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)