Quiz buy gëna xéewale taamu direct

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

1. Lan la DPO dindi buñu ko méngale ak RLHF cosaan?
   • Royuwaayu neexal bu wuute ak bouclage jàng buy dooleel
   • Soxla bépp done bu nit ñi taamu
   • Royuwaayu làkk biñ njëkka tàggat
   • Jëfandikoo probabilite log
2. Ban done la benn misaalu tàggat DPO am?
   • ab laaj ak ab poñ neexal bu nimero
   • ab laaj ak ab tontu buñu tann ak ab tontu buñu bañ
   • Ab tontu bu gaaw ak benn tontu bu jaar yoon
   • Xeetu neexal bi kese mooy génne
3. Ban cër la xeetu royuwaay bi di def ci DPO?
   • Dafay defar tontu yiñ bañ
   • Dafay joxe poñ ci tontu yu am neexal yu leer
   • Mooy liñuy jëfandikoo ngir natt politik bi ngir wàññi dem bi
   • Dafay wecci done yiñ tàggat
4. How does the DPO loss use a labeled preference pair?
   • It regresses both responses to the same scalar target
   • It trains a separate value critic to predict token-level returns
   • It maximizes the rejected response without a reference
   • It increases the chosen response’s policy-versus-reference log-ratio relative to the rejected response
5. Lan mooy parametru beta (tamperatiir) bi ci DPO bi?
   • Ak naka lañuy tëyee politik bi ci xeetu royuwaay bi
   • Tolluwaayu jàng bu optimisatër bi
   • Limu ñaari tànneef yi
   • Dayo vokabileer bi
6. Which alignment objective can learn from binary feedback that an output is desirable or undesirable?
   • PPO, which requires a value critic for every label
   • KTO, a prospect-theoretic objective for binary desirability feedback
   • DARTS, which searches differentiable network cells
   • Beam search, which changes model weights from rankings
7. In a standard DPO preference example, which response is preferred?
   • The response with more tokens by definition
   • The response with the lowest reference likelihood
   • The response selected by the preference labeler for the given prompt
   • A random response unrelated to the prompt
8. In DPO, what expression serves as the implicit reward for a response, up to an input-dependent constant?
   • A separately trained scalar reward-model head
   • A human-written point score for each token
   • The average sequence length of preference data
   • Beta times the policy/reference log-probability ratio