Practice · Praxis

Multiple choice Easy dyna-qmodel-based-rl ~50s

A Dyna-Q agent with planning_steps = 10 takes one real step. Compared with plain Q-learning, what extra work does it do?

A It trains a neural network on 10 transitions sampled from a replay buffer, against a target network B It plans an A* path to the goal using its model, then follows that path for the next 10 steps C It runs 10 Monte Carlo tree-search simulations from its current state before choosing its next action D It records the transition in a model, then replays 10 transitions from that model through the same Q-update

Hints are free to take and they are counted. A correct answer after two hints moves your mastery less than one without — which is the honest way to score it.