Maswali ya Halijoto na Sampuli

Majibu na maelezo huonekana unapojibu maswali.

1. Kuongeza joto kunafanya nini kwa pato la mfano?
   • Huifanya iwe ya kubainisha zaidi na kujirudia
   • Inabadilisha kabisa uzito wa mfano
   • Huzima utendakazi wa softmax
   • Huifanya kuwa nasibu zaidi na tofauti, na nafasi zaidi ya tokeni zisizo za kawaida
2. Kabla ya kuchukua sampuli, alama mbichi za ishara (logi) hubadilishwa kuwa uwezekano na:
   • Kichujio cha juu-k
   • Kazi ya softmax
   • Uenezaji wa nyuma
   • Kiashiria
3. What determines the candidate set in top-p sampling, compared with top-k?
   • Top-p always keeps exactly ten tokens
   • Both methods always keep every vocabulary token
   • Top-p uses the smallest leading set reaching a cumulative probability threshold; top-k uses a count
   • Top-k uses a cumulative threshold while top-p uses a fixed count
4. Kimechanic, halijoto hufanya nini kwa kumbukumbu?
   • Hugawanya kila kumbukumbu kwa T kabla ya softmax
   • Huongeza nambari nasibu kwa kila kumbukumbu
   • Hufuta kumbukumbu za chini kabisa
   • Huzidisha uwezekano wa mwisho na T
5. For a structured-data task, what does choosing more predictable token selection fail to establish?
   • That a decoding setting was chosen
   • That the generated content is correct or valid
   • That token scores can be compared
   • That a model can produce an output
6. How does greedy decoding choose the next token?
   • It literally divides every logit by zero
   • It uniformly samples every vocabulary token
   • It permanently retrains the model
   • It selects a highest-scoring eligible token rather than sampling
7. Two token logits differ by six before temperature scaling. With positive temperature two, what is their scaled difference?
   • Tatu
   • Kumi na mbili
   • Sita
   • Sifuri
8. For probabilities 0.55, 0.30, 0.10, and 0.05, how many leading tokens does top-p 0.80 keep?
   • Moja
   • Tatu
   • Mbili
   • Nne