- formatting
- images
- links
- math
- code
- blockquotes
•
•
•
•
•
推理时对齐方法综述
Inference Time Alignment, Controlled Decoding, Guided Sequence Genalization
大模型 Scaling Law
LLM pretraining, LLM scaling law
大模型与人类价值对齐问题
LLM alignment, Value Compass, from Microsoft Research Asia.