Tile: L8: Large Language Models | Artificial Intelligence Across Domains (PUBLIC) | MI-Moodle

L8: Large Language Models
In this module, learn how computers “understand” human language and generate their responses.

Ever since OpenAI released its ChatGPT chatbot in November 2022, artificial intelligence has been a hot topic among the general public. For some, AI and ChatGPT are almost synonymous (but we hope it's already clear that AI encompasses much more than that). In this module, we will not only look at the possible applications of large language models, but also consider the technical and sociocultural perspectives: What technical concepts underlie the capabilities of AI chatbots? What limitations, problems, and potential impacts can be expected, and how can these be addressed?

After this module, you will be able to
Name the potentials of language models.
Explain the basic ideas and principles of how language models generate texts and are trained with available internet data.
Use technical background knowledge to prompt language AIs more effectively.
Discuss challenges of language models such as biases, prejudices, and unwanted content in outputs, and explain the lack of transparency of large language models.
Discuss changes and impacts in your own field.
We strongly recommend that you work through Module 08 on a computer in order to make the best use of the learning resources.
- Select activity Establishing the connection
  
  Establishing the connection
  
  Students must
  
  Mark as done
- Select activity 1. INTRODUCTION (Kopie)
  1. Introduction
- Select activity I. Real or AI-generated?
  
  I. Real or AI-generated?
  
  Students must
  
  Mark as done
- Select activity 2. UNDERSTANDING AND PRODUCING LANGUAGE
  2. Understanding and producing language
- Select activity I. Language models
  
  I. Language models
  
  Students must
  
  Mark as done
- Select activity II. Large language models
  
  II. Large language models
  
  Students must
  
  Mark as done
- Select activity III. Large language models: Pretraining
  
  III. Large language models: Pretraining
  
  Students must
  
  Mark as done
- Select activity IV. Key technical concepts of GPTs
  
  IV. Key technical concepts of GPTs
  
  Students must
  
  Mark as done
- Select activity V. Large language models: Finetuning
  
  V. Large language models: Finetuning
  
  Students must
  
  Mark as done
- Select activity 3. ETHICAL CONCERNS
  3. Ethical concerns
- Select activity I. Understanding the problem of hallucinations and their effects
  
  I. Understanding the problem of hallucinations and their effects
  
  Students must
  
  Mark as done
- Select activity II. Other limitations of LLMs
  
  II. Other limitations of LLMs
  
  Students must
  
  Mark as done
- Select activity 4. PROMPTING
  4. Prompting
- Select activity I. Prompt Engineering and Prompt Strategies
  
  I. Prompt Engineering and Prompt Strategies
  
  Students must
  
  Mark as done
- Select activity II. Creating a good prompt
  
  II. Creating a good prompt
  
  Students must
  
  Mark as done
- Select activity III. Practicing
  
  III. Practicing
  
  Students must
  
  Mark as done
- Select activity IV. Discussion
  
  IV. Discussion
  
  Students must
  
  Mark as done
- Select activity 5. LARGE LANGUAGE MODELS “ACROSS DOMAINS”
  5. Large language models “across domains”
- Select activity I. Problem solving in various areas with large language models
  
  I. Problem solving in various areas with large language models
  
  Students must
  
  Mark as done
- Select activity Test your acquired knowledge
  
  Test your acquired knowledge