Understanding language models

1. Pre-training Alignment vs Fine-tuning Alignment (Prompt Engineering Secrets)