10 Must-Know Terms and Concepts to Better Understand ChatGPT and LLMs
Make the most out of the content online
The new rising star of the data science and AI ecosystem is large language models (LLMs). The rapid increase of the popularity of LLMs started with ChatGPT and soon became prevalent in social media, blogs, and scientific papers.
This article is for those who want to learn about LLMs but find it difficult to understand LLM content because of technical terms or domain-based concepts.
Data science and AI are very broad fields so even if you are working in these fields, LLM-related content might be difficult for you to understand at first. Hence, knowing what these terms and concepts mean beforehand make it easier to digest other LLM-related content.
Transformer
First introduced in Attention is All You Need paper, transformer is a network architecture based entirely on attention mechanism.
It aims to solve the shortcomings of recurrent neural networks, which
- are slow to learn
- suffer from vanishing gradient or long-term dependencies in the case of long sequences
Attention defines which part of the input sequence we need to focus on and the transformer…
