In DDPG, exploration is typically achieved by:
Randomly sampling from the replay buffer with higher priority
Injecting noise into the critic network parameters
Adding time-correlated noise (e.g., Ornstein-Uhlenbeck) or Gaussian noise to the deterministic policy output
Correct answer
Adding time-correlated noise (e.g., Ornstein-Uhlenbeck) or Gaussian noise to the deterministic policy output
Question 2 of 21 in the IIT Madras BS Reinforcement Learning (Reinforcement Learning) End Term paper sat on 10 May 2026, in the January 2026 term (Reinforcement Learning 10 May 26 (Session 2)). It carries 1 mark.
