Maswali ya Kulinganisha AI

Majibu na maelezo huonekana unapojibu maswali.

1. Katika utafiti wa upatanishi, neno "udukuzi wa tuzo" linaelezea nini?
   • Muundo unaokataa kuboresha malipo yake kwa sababu kazi inaonekana kuwa hatari
   • Wakala akitumia dosari katika ishara yake ya malipo badala ya kufanya kazi iliyokusudiwa
   • Mshambulizi anayeiba uzani wa modeli ya zawadi ili kufundisha upya mfumo shindani
   • Wasanidi programu wanainua wenyewe thamani za zawadi ili kuharakisha muunganisho wakati wa mafunzo
2. Katika RLHF, ni nini jukumu la mfano wa zawadi?
   • Inachukua nafasi ya wakadiriaji wa kibinadamu kabisa kwa kuandika majibu ya rejeleo kwa urekebishaji mzuri
   • Huhifadhi orodha isiyobadilika ya maneno yaliyopigwa marufuku ambayo sera hairuhusiwi kamwe kutoa
   • Hutoa maandishi mapya ya mafunzo ya awali ili muundo msingi uwe na data zaidi ya kujifunza kutoka
   • Hujifunza kutokana na ulinganisho wa mapendeleo ya binadamu na kupata matokeo ili kuongoza sera
3. What does goal misspecification mean in the guide?
   • A model behaves differently when it recognizes a test setting
   • The written objective fails to represent what people actually intend
   • An agent finds a shortcut in a proxy metric while optimizing it
   • A model refuses every task after a safety update
4. Which description matches instrumental convergence in the guide?
   • A user changes their prompt after a refusal
   • A system may seek power or resources because they help achieve many possible goals
   • A test set is held out from training
   • A reward model stores written principles
5. Which modern alignment approach is named in the guide?
   • Credit scoring notices
   • Post-market incident reporting
   • Dataset deduplication
   • Kikatiba AI
6. Utafiti wa "usimamizi mbaya" unajaribu kutatua nini?
   • Jinsi ya kupunguza idadi ya vigezo huku ukiweka usahihi mara kwa mara
   • Jinsi wanadamu wanavyoweza kusimamia AI kwenye kazi ngumu sana kwao kutathmini moja kwa moja
   • Jinsi ya kutumikia mfano kwa mamilioni ya watumiaji mara moja bila kuongeza muda wa kusubiri
   • Jinsi ya kuongeza vikundi vya mafunzo kwenye vituo vingi vya data kwa ufanisi
7. Why does preference optimization not prove that a model has adopted human intent as an internal goal?
   • It improves average responses but does not prove stable behavior under shift or pressure
   • It prevents models from using human preference data
   • It guarantees a model will ignore every written instruction
   • It measures only GPU efficiency
8. Why might an alignment evaluation include tests for evaluation awareness?
   • To measure the number of human raters in training
   • To determine whether a reward model has enough parameters
   • To see whether a model changes behavior when it recognizes a test setting
   • To prove that a system is conscious