Guide-yakabatana mibvunzo · Zvakaoma Nhanho

Neural Tangent Kernel Theory Quiz

Edza kuti iyo Neural Tangent Kernel inobatanidza sei kusingaperi-hupamhi network kudzidziswa kune kernel nzira uye basa-nzvimbo dynamics mumutsara muganho.

Nzira dzekutungamira dzinoenderanaNeural Tangent Kernel Theory
Mubvunzo 1 ye 8

Mune isingaperi-hupamhi muganho, iyo Neural Tangent Kernel inoita sei panguva yekudzidziswa?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Mune isingaperi-hupamhi muganho, iyo Neural Tangent Kernel inoita sei panguva yekudzidziswa?
    • Zvinogara zvakangofanana
    • Inoshanduka nekukurumidza nhanho imwe neimwe
    • Inoita zero
    • Inokura zvakanyanya
  2. Chii chinonzi NTK nemasvomhu?
    • The Hessian yekurasikirwa
    • Chigadzirwa chemukati cheiyo network's parameter gradients kune maviri ekuisa
    • Iyo softmax yezvakabuda
    • Kusiyana kwezviyero
  3. Ndiani akaunza Neural Tangent Kernel?
    • Hinton uye Salakhutdinov
    • LeCun uye Bengio
    • Jacot, Gabriel, uye Hongler (2018)
    • Vaswani nevamwe vaaishanda navo
  4. Chii chinonzi 'usimbe kudzidzisa' hutongi?
    • Kana kurovedza kwambomira
    • Kana iyo optimizer isingashandisi kumhanya
    • Kana data ichidyiswa zvishoma nezvishoma
    • Kana ma paramita asinga fambe kubva pakutanga uye network ichiita zvine mutsetse
  5. Chinhu chakakosha chinodzikisira cheNTK theory yemanetwork chaiwo ndechekuti inopotsa:
    • Feature kudzidza mune finite-width network
    • Basa re gradient descent
    • Kuvapo kwezviyereso
    • Kushandiswa kwedata rekudzidzisa
  6. Nekuti iyo NTK inoderedza kudzidziswa kune kernel regression, chii chinogoneka?
    • Kubvisa kudiwa kwechero data
    • Kufanotaura kuburitsa kwakadzidziswa pasina kudzidziswa zvizere
    • Kumhanya mamodheru pasina parameter
    • Kuvimbisa zero bvunzo kukanganisa nguva dzose
  7. In the infinite-width limit, what probabilistic model are neural networks equivalent to at initialization?
    • A yakavanzika Markov modhi
    • Muti wechisarudzo ensemble
    • A Gaussian process
    • A finite-state game controller
  8. For least-squares regression in the NTK limit, what mathematical form does the network function's training evolution follow?
    • A discrete rule that randomly changes labels
    • A decision tree that grows one branch per epoch
    • A nonlinear search over audio frames
    • A linear differential equation in function space