Skip to content

Using Large Language Models in Education​

Learn how to harness Large Language Models in educational design based on evidence

Start Any Time

Work on your pace and you will have instructors available to help you answer any questions.


Approximately 4 weeks, 6-8 hours/week



Certificate Course Description:

Rule-based cognitive models help understand student thinking and problem solving, help guide many aspects of the design of a tutor, and can function as the “smarts” of a system. Cognitive Tutors using rule-based cognitive models have been proven to be successful in improving student learning in a range of learning domain.

In this course, you will learn how to develop rule-based cognitive models for creating Cognitive Tutors using the Cognitive Tutor Authoring Tools (CTAT). You will also learn ACT-R theory and Foundations of Production Systems behind Cognitive Tutors.

Module 1: Leveraging LLMs: Prompt Engineering, from Definition to Evaluation
  • Define Large Language Model (LLM) specifically Autoregressive Large Language Models
  • Apply prompt engineering strategies to optimize the performance of Large Language Models (LLMs) in generating desired educational outputs
  • Differentiate between zero-shot and few-shot prompting, explaining appropriate scenarios for their utilization
  • Demonstrate how to refine prompts through additional instructions or context to shape the output of Large Language Models
  • Critically evaluate the output from Large Language Models for reliability and accuracy, identifying potential instances of hallucination
Module 2: Top-Down Design: Implementing Educational Tasks with LLMs and Learning Science Principles
  • Divide an educational task into subtasks as part of top-down design
  • Employ LLMs to generate tailored learning interactions and enhance existing learning materials in line with learning science principles
Module 3: Integrating LLMs in Educational Tools: Practical Examples
  • Differentiate between LLM APIs and conversational chatbots justifying the use of APIs in educational contexts despite the convenience of chatbots
  • Analyze the implications of hyperparameter tuning in prompt engineering for optimizing the LLM outputs
    • Temperature
    • Top-k
    • Top-p
    • Number of tokens
    • Repetition Penalty
    • Presence Penalty
    • Number of Examples
    • Query Format
    • Output Format
    • Instruction Style
    • Keyword Selection
Module 4: Course Project or Final Exam

At the end of the course, you’ll have an opportunity to do a two-step project from three choices based on area of your interest i.e Educational Tool Design (UX/UI), Educational Tool Development (Technical), or Prompt Engineering Template (Instructional Design). That will provide you with a nice experience to apply the fundamentals you will learn in the modules to a larger, more authentic, context. It will be graded by the instructor and you will receive personalized feedback along with a sample solution.

* Each assignment will be graded by the instructor and you will receive personalized feedback along with sample solution

A computer is required where CTAT can be installed. The instructions are here.

Required prerequisite: Knowledge or example-tracing tutors OR Example-tracing Tutors: Intelligent Tutor Development for Non-programmers

Researchers, developers, product/UX designers, instructional designers, and teachers who want to create intelligent tutors. Anyone interested in edtech.

What you'll learn

This course will help you:

  • Explain ACT-R theory and Foundations of Productions Systems behind cognitive tutor technology
  • Design and linking the interface representation dynamically to the semantic representation of the problem
  • Develop and debug cognitive models
    • Design your own working memory representation
    • Implementing new fact types
    • Using rules to support multiple strategies within a single problem
    • Create a dynamic interface with a Nools tutor
  • Scope problems suitable for creating rule-based intelligent tutors.

Course Instructors

Gautam Yadav

Gautam Yadav is a Learning Engineer in Carnegie Mellon’s Human-Computer Interaction Institute (HCII) specializing in creating effective evidence-based educational technologies and improving data-driven learning experiences for more than 3 years.


Upon successful completion of the program, participants will receive a verified digital certificate of completion from Carnegie Mellon University’s Open Learning Initiative.

In addition to the knowledge and immediately applicable frameworks you will gain by attending your selected courses, you will benefit from:

  • A digital, verified version of your Executive Certificate (Smart Certificate) you can add to your resume and LinkedIn
  • Networking with a global group of your peers and instructors for advancing your career

Register Now

Register and start taking the course in four steps:

1. Enter your email address 

2. Watch this short video for instructions on how to register in OLI.

3. For this course, copy the course key: Coming Soon (If you enter your email address above, we will send you an email as soon as the course is out to sign up!)

4. Click on this link to Carnegie Mellon University’s Open Initiative to register and try out the course for 48 hours before payment is due.

5. (Optional but highly recommended) Set OLI to automatically resume from where you left off in the course.

1) Click on your name in the upper right corner to bring up your OLI profile settings.

2) Change the option on the dropdown to resume automatically.

3) Lastly, select UPDATE.

You are all set!

Set up course to auto resume in OLI