Mutambi-Kutsoropodza Nzira dzemibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

1. Munzira yeActor-Critic, basa re 'mutsoropodzi' nderei?
   • Kufungidzira kukosha kwematunhu (kana zviito) uye kuongorora sarudzo dzemutambi
   • Kusarudza danho rekutora
   • Kuchengeta replay buffer
   • Kuongorora zvakatipoteredza zvisina tsarukano
2. Ko iyo 'advantage' A(s,a) = Q(s,a) - V(s) inoyera chii?
   • Mubairo wechikamu chose
   • Chiito chiri nani zvakadii pane avhareji yehukoshi hwehurumende
   • Chiyero chekudzidza
   • Nhamba yezviito zviripo
3. Nei kuwedzera mutsoropodzi kuchideredza mutsauko uchienzaniswa neakachena mutemo-gradient nzira seREINFORCE?
   • Inobvisa mibairo kubva kuequation
   • Inoita iyo policy deterministic
   • Mutsoropodzi anopa hwaro, saka zvigadziriso zvinoshandisa isina ruzha mukana wechiratidzo
   • Inowedzera the discount factor
4. Ndeupi hunyanzvi hunoitwa neActor-Critic nzira dzine nzira dzakajeka-dzakavakirwa seQ-Kudzidza kubata zvinonetsa?
   • Zviito zvakasarudzika
   • Kushandisa mibayiro
   • Kubvisa ramangwana
   • Kuenderera mberi kwekuita nzvimbo
5. Ndeipi pane zvinotevera inonzi Actor-Critic-based algorithm inoshandiswa zvakanyanya kuRLHF pamhando dzemitauro?
   • PPO (Proximal Policy Optimization)
   • K-zvinoreva
   • Linear regression
   • Q-Kudzidza
6. Ndechipi chiratidzo chinowanzo shandiswa nemutambi kugadzirisa mutemo wake?
   • Iko kupinza kwepixel mbishi
   • Iyo mukana kana TD kukanganisa kwakapihwa nemutsoropodzi
   • Iyo replay buffer size
   • A fixed random number
7. An action has a positive advantage estimate. How does a policy-gradient actor typically use that signal?
   • Decrease its probability in every state
   • Set the critic’s value estimate to zero
   • Increase the probability of choosing that action in the state
   • Replace the policy with a Q-table
8. An action’s estimated advantage is negative. What direction should the actor update favor?
   • Increase it because the critic was surprised
   • Delete the state from the environment
   • Raise the reward discount factor automatically
   • Reduce the likelihood of that action relative to the policy baseline