Compare LLM model's essays written on a given topic.
The following demonstration uses Open AI's GPT-3 and Google Cloud's PaLM 2 language models. It attempts to create an essay in one stage by using the following prompt 'Write an essay on chosen topic with the following format: 1. Introduction 2. Content 3. Content 4. Content 5. Conclusion.'
- Enter a topic in the box below, or select from one of the example topics.
- Click the Analyse button.
- View the essays created by OpenAI and Google PaLM by switching between the tabs.
You must be a member of an institution in order to access this demo.
Please sign in below using Jisc Login with your institution credentials.
More information about this demo
This demo utilises two Large Language Models (LLMs); GPT-3 from OpenAI and PaLM 2 from Google. A LLM is a form of Artificial Intelligence (AI), built on a expansive dataset, predominantly scraped from the internet. The main uses for these models are to generate, complete and summarise text; with a multitude of other potential use cases. They work by, based on the initial text input, continually predicting the most likely next token or word to create meaningful responses, in most cases. When working with such models, providing the best prompt to the model is important and this is known as prompt-engineering.
Things to consider
These tools will generate different responses to prompts each time they are run, so if you don't like a response, rerun the same parameters and get a different essay. In the case of OpenAI's GPT-3, the LLM was also trained on data gathered prior to 2021, so may not have knowledge of current affairs past that cut off. Google Clouds PaLM 2 Model offers Multilinguality, Reasoning and Coding abilities, also coming in a variety of different sizes which allows it to be run on a variety of different machines; resulting in a highly versatile model.