Glossary of Terms for Course on ChatGPT for Teachers
ChatGPT
Code Highlighter
Computational Thinking
Computational thinking is a way of solving problems and analyzing data by breaking it down into smaller parts and using algorithms and other tools to solve the problem.
It is a thought process that is used in many different fields, from engineering to economics, and it is often used in computing and programming. Computational thinking involves breaking down a problem into smaller pieces and understanding how the pieces interact with one another, to create a solution.
Deep Links
The feature of a website that allows any subsection of a document to have its own URL that can be bookmarked and sent via chat.
Fine-Tuning
Generative Text
GPT
Generative Pre-Trained machine learning models. This is the model use for both ChatGPT and GPT-3.
Grammarly
A cloud-based service that helps writers with spelling, grammar and writing style.
Grammarly is also created using large langauge machine learning models.
Markdown
Mermaid Markdown
A text format for encoding diagrams.
In January of 2022 GitHub announced built-in support for Mermaid diagrams into their Markdown variant rendered for README.md
Natural Language Processing
- Also know as: NLP
OpenAI
A small 400-person company (as of 2022) that created the GPT and ChatGPT large-language models.
OpenAI has received over a billion dollars of investment from Microsoft for an exclusivity arrangement that prevents OpenAI from working with the other cloud providers (Google and Amazon)
Prompt
SVG
SVG means Scalable Vector Graphics. It is a standard format for generating vector images (not bitmaps) that can be easily generated by ChatGPT.
Tokens
Common sequences of characters found in text that are assigned numbers and used as inputs to machine learning models.
The number of tokens in both prompts and responses are added together and must be under the token budget for a NLP model.