<aside> <img src="/icons/list_gray.svg" alt="/icons/list_gray.svg" width="40px" /> Contents

</aside>

What to learn

In this task, you will get a sense of how LLMs deal with longer text. For instance, 5000 words or longer.

You will discover that different tools have different ‘context windows’ (how much text they can see) and vary in suitability. This is why the main tool we are using is Claude.ai from Anthropic which has a context window of up to about 100,000 words even in the free version.

Other tools, such as ChatGPT only offer this in the paid version (up to about 15,000 words). The longest context window available is in Google Gemini Advanced (up to about 800,000 words) but only in the paid version.

Note: The context window includes everything in one chat including what you ask and what the model generates in response. If your conversations are too long, the model may stop being able to ‘see’ the text you are asking about and start making things up.

What to do

<aside> <img src="/icons/cursor-click_gray.svg" alt="/icons/cursor-click_gray.svg" width="40px" />

Click on the triangle next to each step to see more details and/or get resources necessary for the task.

Read through all steps first.

</aside>

What to focus on

When reviewing the model outputs, focus on:

  1. Choices the model makes about what is relevant
  2. Impact of prompts
  3. Clarity of explanations
  4. Accuracy of quotes