An assortment of tutorials and investigations aimed at establishing optimal practices for leveraging Large Language Models (LLMs) in educational contexts.
- Key Components: A discussion on the key components of a good prompt along with examples that use these components.
- Developing Prompt Templates: A guide to developing a good prompt with examples. This inclodes some tips regarding how to go about crafting prompt templates to make every day tasks more efficient.
- LangChain: Developing a custom AI Tutor chatbot using LangChain.
- Chat Bot: Interactive Astronomy AI tutor using RAG to provide relevant responses and recommend material.
- Report Card Assistant: Interactive chatbot for developing report card comments.
- LLMs as Optimizers: A tutorial on how to use an LLM to optimize the system instructions for a chatbot.
- RAG Moderator: A tutorial on using RAG to moderate LLM outputs.
- ReAct Agents: An exploration of using Reasoning and Action Agents to have an LLM provide fact-based responses.
- Back-and-Forth Interactions: A guide to how teachers can leverage the back-and-forth nature of LLM interactions to improve responses and get more in-depth explanations.
- Evaluating Prompts and Responses: An explorations of strategies to improve user prompts and ChatGPT responses.
- Using ChatGPT as a Tutor: A use-case for having students interact with ChatGPT to improve their understanding.
- LLM mechanisms: An exploration of how the components of a good prompt relate to the the internal mechanisms of an LLM.
- Interacting with the ChatGPT API: For Python programmers who would like to make their interactions with ChatGPT even more efficient.
- Risks and Considerations: A developing list of important risks and considerations when adopting tools like ChatGPT for educational purposes.