Tàngoor ak laaj ak tontu ngir jël misaal

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

1. Lan mooy yokk tàngoor wi ci génnug model bi?
   • Daf koy gëna deterministik ak baamtu
   • Dafay soppi diisaayu model bi ba fàww
   • Desactivé fonction softmax
   • Daf koy gëna bari te wuute, ak saas yu bari ci ay jeton yu wuute ak yeneen yi
2. Laata ñuy jël misaal, poñ yiñ joxe ci model bi (logits) dañu leen di soppi def ko ay probabilite ci:
   • Seggal top-k bi
   • Fonction softmax bi
   • Tasaaro ci ginaaw
   • Tokenisatër bi
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. Ci wàllu mekanik, lan la tàngoor di def ci logit yi?
   • Séddale logit bu nekk ak T balaa softmax
   • Yokk benn nimero ci logit bu nekk
   • efaase logits yi gëna suufe
   • Yokk probabilite yu mujj yi ak 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?
   • Ñatt
   • Twelve
   • Jiroom benn
   • Zéro
8. For probabilities 0.55, 0.30, 0.10, and 0.05, how many leading tokens does top-p 0.80 keep?
   • Benn
   • Ñatt
   • Ñaar
   • ñeent