Zenaique
Topics
Practice
Study
Browse
Reference
Pricing
Search…
⌘K
Premium
/
RLHF & Alignment
Premium
RLHF & Alignment
Premium questions for this topic. Unlock with Pro to study the full set.
18 questions
Unlock with Pro
01
Premium
Why must SFT happen…
Short Answer
Medium
02
Premium
Your preference labelers agree…
Short Answer
Medium
06
Premium
Explain DPO's implicit reward…
Short Answer
Hard
13
Premium
Compare best of N…
Short Answer
Medium
14
Premium
KL blew past target…
Short Answer
Hard
15
Premium
Live thumbs up training…
Short Answer
Hard
16
Premium
Refusals jumped from 2%…
Short Answer
Medium
17
Premium
What is the alignment…
Short Answer
Medium
18
Premium
Sketch the reward overoptimization…
Short Answer
Hard
03
Premium
Pick the real tradeoff…
Multiple Choice
Medium
04
Premium
What happens when the…
Multiple Choice
Hard
05
Premium
What capability does DPO…
Multiple Choice
Medium
08
Premium
Pick why reward model…
Multiple Choice
Medium
11
Premium
Your only feedback is…
Multiple Choice
Medium
12
Premium
When is RLAIF preferable…
Multiple Choice
Medium
07
Premium
Spot the off policy…
Spot the Error
Hard
09
Premium
Spot the flaw in…
Spot the Error
Medium
10
Premium
Spot the unsafe assumption…
Spot the Error
Hard