Guide-yakabatana mibvunzo · Zvakaoma Nhanho

Yakajairika Sampling Quiz

Inotarisa zvakajairwa kushamisika, entropy-yakavakirwa chinzvimbo, iyo yakajairwa_p misa pachikumbaridzo, uye mutsauko kubva kune mukana-chinzvimbo chepamusoro-p sampling.

Nzira dzekutungamira dzinoenderanaTypical Sampling
Mubvunzo 1 ye 8

Ndeupi huwandu hwakakosha hunoitwa sampling inofananidza tokeni yega yega?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Ndeupi huwandu hwakakosha hunoitwa sampling inofananidza tokeni yega yega?
    • Izvo zvinotarisirwa zvemukati zvemukati (entropy) zvekugovera
    • Saizi yemazwi emuenzaniso
    • Nzvimbo yechiratidzo mumutsara
    • Iyo tembiricha parameter chete
  2. Sampling yakajairika inobata sei chiratidzo chimwe chete chinogoneka?
    • Inogara ichisarudza kutanga
    • Inogona kuibvisa kana kushamisika kwayo kuri kure pasi pemutengo unotarisirwa
    • Inowedzera kaviri mukana wayo
    • Inofuratira zvingangoitika zvachose
  3. Muenzaniso wesampling wakakurudzirwa nepfungwa ipi pamusoro pemutauro wevanhu?
    • Vanhu nguva dzose vanosarudza mazwi asingawanzoitiki
    • Vanhu havambodzokorori mazwi
    • Vanhu vanokurukurirana pedyo nemwero wakadzikama wemashoko
    • Vanhu vanotaura nemitsara yakarebesa
  4. Chii chinonzi tau parameter mune yakajairika sampling control?
    • Chiyero chekudzidza panguva yekudzidziswa
    • Nhamba yezvikamu mumuenzaniso
    • Hurefu hwechirevo chepamusoro
    • Iyo cumulative probability mass yakaunganidzwa mune yakajairika set
  5. Compared with greedy decoding, what does typical sampling try to change?
    • It samples from tokens whose surprisal is closer to the distribution’s expected surprisal
    • It always selects the most probable token
    • It guarantees every continuation is factually correct
    • It removes probability scores from decoding
  6. Imhando ipi yesoftware yakajairika sampling inowanzo kuratidzwa kune vanogadzira?
    • Spreadsheet applications
    • Vhura-sosi LLM inference raibhurari seHugging Face Transformers
    • Zvishandiso zvekugadzirisa mifananidzo
    • Database query engines
  7. How does typical sampling rank candidate tokens differently from top-p sampling?
    • By alphabetical order of token strings
    • By token position in the prompt
    • By distance between token surprisal and expected surprisal, rather than probability rank alone
    • By the model’s training loss
  8. Which value does typical_p control?
    • The model’s vocabulary size
    • The number of transformer layers
    • The probability that the answer is factually correct
    • The target cumulative probability mass collected in the typical set