Skip to content
ml-recall

#instructgpt

3 questions from Alignment & Reasoning

Timed test

3 questions · 3 min · score and explanations at the end

Take a test
Q 1 / 3Score 0 / 0

Question 1: Short answer

easyAI-checkedAlignment & Reasoning

Complete InstructGPT's three-step RLHF recipe. First, labelers write demonstrations of the desired behavior, and a pretrained GPT-3 is fine-tuned on them with learning. Second, labelers compare several outputs for the same prompt, and a model is trained to predict which output they prefer. Third, the fine-tuned policy is optimized against that model's score with the RL algorithm .

Fill in every blank (hints in the boxes). Enter moves to the next blank.