AI Architecture & Systems
1
Microarchitecture
2
Kernel
3
Compiler
4
Architecture
5
System
6
Algorithms
Algorithms
/
§6.14
Alignment and Preference Learning
6.14.1
Reward Model, Preference Data, and Pairwise Ranking
#
6.14.2
RLHF and RLAIF
#
6.14.3
PPO, KL Regularization, and Reference Policy
#
6.14.4
DPO, IPO, KTO, and Direct Preference Optimization
#
6.14.5
Rejection Sampling and Best-of-N Data Selection
#
6.14.6
Safety Alignment, Refusal, and Preference Generalization
#
← Previous
6.13 SFT and Parameter-Efficient Fine-Tuning
→ Next
6.15 Reasoning and RL Post-Training