Class 3 - How ChatGPT and Perplexity Work
| FREE Advanced SEO Course with Notes
Watch the Full Video: ChatGPT and Perplexity Work
LLM = Large Language Model (Large Numbers Model)
LLM = Large Language Model (Large Numbers Model)
Pre-Training: It predicts which words are most likely to come next based on the user’s query.
Supervised Fine Tuning(SFT): It retrieves the relevant information based on the user’s query and generates an answer for the user.
Reinforced Learning From Human Feedback(RLHF): The model follows safety rules. It is designed to avoid answering harmful questions or generating unsafe content, such as instructions for making a bomb or creating illegal or dangerous images, etc.
Vector: It converts the properties and dimensions of an entity into mathematical representations and assigns them a score.
Retrieval-Augmented Generation(RAG): It works on embeddings, vector databases, and live search.