Guide-yakabatana mibvunzo · Zvakaoma Nhanho

VQ-VAE uye Discrete Latents Quiz

Review how VQ-VAE maps continuous encoder outputs to discrete codebook entries for generative modeling and reconstruction.

Nzira dzekutungamira dzinoenderanaVq Vae Uye Discrete Latents
Mubvunzo 1 ye 8

Chii chinoita kuti VQ-VAE's latent nzvimbo isiyane neyakajairwa VAE's?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Chii chinoita kuti VQ-VAE's latent nzvimbo isiyane neyakajairwa VAE's?
    • Iyo yakasarudzika, uchishandisa yakadzidziswa codebook yekumisikidza
    • Inoenderera uye Gaussian
    • Haina nzvimbo yakanyarara zvachose
    • Inochengetedza mapixels akaomeswa
  2. VQ-VAE inopfuura sei ma gradients kuburikidza neiyo isiri-yakasarudzika quantization nhanho?
    • Inodarika backpropagation zvachose
    • Iine yakatwasuka-kuburikidza estimator inokopa gradients kupfuura quantizer
    • Nokushandisa kusimbisa kudzidza chete
    • Nekutenderedza ma gradients kusvika ku integers
  3. Chii chinonzi 'codebook collapse'?
    • Iyo decoder inofuratira ma latent
    • Izvo encoder zvinobuda zvinova zvisingaperi
    • Chikamu chidiki chete chezvinyorwa zvecodebook zvinomboshandiswa
    • Chimiro chemufananidzo chinodonha kusvika zero
  4. Nei discrete latents ichibatsira pakugadzira modhi neTransformers?
    • Vanoita kuti mifananidzo ive diki padhisiki chete
    • Vanobvisa kudiwa kwedecoder
    • Vanobvisa zvese zvisina tsarukano kubva kuchizvarwa
    • Vanoshandura midhiya kuita zviratidzo zvinoteedzana neTransformer inogona kuenzanisira semutauro
  5. Chii chakawedzera VQGAN pamusoro peiyo yakakosha VQ-VAE resipi?
    • Kurasikirwa kweAdversarial uye yekufungidzira kwekuvakazve kwakapinza
    • A diffusion sampler
    • Kubviswa kwecodebook
    • A inodzokororwa encoder
  6. Nderipi izwi rekurasikirwa rinoita kuti encoder irambe 'yakazvipira' kune zvinyorwa zvecodebook zvayasarudza?
    • Kusiyana kweKL
    • Kurasikirwa kwekuzvipira
    • The cross-entropy kurasikirwa
    • Kurasikirwa katatu
  7. What kind of prior does the original VQ-VAE paper learn over its discrete codes?
    • A fixed prior that is never learned
    • A prior over camera locations only
    • A learned prior over the discrete code sequence
    • No prior or decoder distribution
  8. Which text-to-image paper used a discrete VAE to tokenize images before autoregressive generation?
    • The original Whisper speech-recognition paper
    • A paper on Kalman-filter vehicle tracking
    • The original BERT masked-language-model paper
    • The original DALL·E paper