Skip to content
ml-recall

#reinforce

4 questions from Policy Gradient & Actor-Critic

Timed test

4 questions · 4 min · score and explanations at the end

Take a test
Q 1 / 4Score 0 / 0

Question 1: Single choice

mediumAI-checkedPolicy Gradient & Actor-Critic

REINFORCE uses the log-derivative trick ∇θEy∼πθ[R(y)]=Ey∼πθ[R(y) ∇θlog⁡πθ(y)]\nabla_\theta \mathbb{E}_{y \sim \pi_\theta}[R(y)] = \mathbb{E}_{y \sim \pi_\theta}[R(y)\,\nabla_\theta \log \pi_\theta(y)]. Why is this useful when fine-tuning a language model with a reward?

Choose one option