What is llemma?
Llemma stands for *Large Language models for Education, Motivation and Assessment*.
This is a proof of concept to see if large language models (LLMs) like ChatGPT can provide useful formative feedback to students on essays and other coursework. The goal is that students can upload essays and request packages of formative feedback, tailored (by teachers and academics) to their specific assignment.
How does it work? Students upload work to the system, and the text of their assignment is extracted. The full or partial text is then sent to an LLM (currently OpenAI's ChatGPT), as part of specially-crafted to provided tailored feedback to students. Packages of these prompts are written by teachers and academics to make sure the automatically generated feedback is targetted towards the learning outcomes of their course.
