"Imagine a person..."

What happens when you ask LLMs to imagine a person and a random day in their life, 100 times over?

I asked small versions of Llama3.1, Gemma2 & Qwen2.5 to imagine a person, a hundred times over, using the same prompt. The prompt asks for basic details, such as name, age and location, and then asks the AI for a random day in that person's life.

Click here to view the original prompt
Imagine a person with the following details:

Name
Gender
Age
Location (Country)
Brief backstory (1-2 sentences)

Describe a random day from their life using this format:

Time: [HH:MM]
Activity: [Brief description]

Start with when they wake and end with when they go to sleep. Include as many time entries as possible, be very specific.
Example output:

Name: [Name]
Gender: [Gender]
Age: [Age]
Location: [Country]
Backstory: [Brief backstory]
Day:

Time: [HH:MM]
Activity: [Activity description]
(Repeat this format for each time entry)

I processed the responses of the LLM with Claude Haiku to turn the result into JSON, which is then visualised on this webpage. You can switch between models using the dropdown in the top right of the screen.

Caveats

This is just for fun. These language models are running on my local machine, using quantized versions of the original models (llama3.1 8b Q4_0, gemma2 2b Q4_0, qwen2.5 7b Q4_K_M). I've set the temperature of my requests to 1.0. Using the unquantized model, experimenting with temperature values or simply changing the prompt would hopefully provide more varied, creative responses.

Age & Gender

Gender Distribution

Age Distribution

  • Small LLMs seem to believe that only people between the ages of 25-35 exist.
  • Llama3 only managed to imagine one human who was male - Akira Saito, a 32 year old Japanese freelance graphic designer.
  • No model was able to imagine a world outside the gender binary, at least in these first 100 responses.

Names, Locations, Jobs

Name Distribution

Anya Petrova13
Liam5
Anya4
Anya Ivanovna3
Elias3
Anya Sharma3
Emily Carter3
Amelia Garcia2
Maya Patel2
Elena2
Alice Chen2
Alice Johnson2
Elias Carter2
Anna Petrova2
Anna2
Anya Petrovna2
Sarah2
Alex2
Emily Chen2
Maya1
Elara Blackwood1
Amelia Evans1
Elara1
Annabelle Chen1
Elena Alvarez1
Johnathan Smith1
Maya Sharma1
Emily Rodriguez1
Elias Garcia1
Maria Alvarez1
Maya Reyes1
Evelyn Wright1
Sophia Kim1
Ava Khan1
Eleanor Williams1
Olivia Alvarez1
Eleanor "Ellie" Lawson1
Amelia1
Evelyn Ramirez1
Amelia Miller1
Alice Miller1
Sofia1
Ava Sharma1
Elena Flores1
Anna Lee1
Liam O'Connell1
Sarah Evans1
Alice1
Elias Richter1
Leo Rodriguez1
Amelia Reyes1
Maya Gupta1
Amelia Wright1
Sophia Garcia1
Elena Garcia1
Annabelle "Annie" Williams1
Amelia Jones1
Amelia Ramirez1
Anna Petrovna1
Sophia Rodriguez1
Ava Rodriguez1

Location Distribution

Russia13
United Kingdom7
India7
Berlin, Germany5
Argentina5
Spain5
Moscow, Russia5
Costa Rica4
Barcelona, Spain4
Germany4
Ireland3
Netherlands3
France2
Vancouver, Canada2
United States of America2
Denmark2
Switzerland2
Italy2
Canada2
England2
Sweden1
Russia, Saint Petersburg1
Mumbai, India1
Montenegro1
England, United Kingdom1
Scotland1
Nepal1
Chile1
South Korea1
Finland1
Iceland1
Dublin, Ireland1
Buenos Aires, Argentina1
Tallinn, Estonia1
Peru1
London, UK1
Canada (British Columbia)1
Japan1
Mexico City, Mexico1

Job Distribution

Freelance Graphic Designer15
Graphic Designer14
Freelance Photographer10
Software Engineer4
Freelance Illustrator3
Freelance Writer3
Artist3
Architect3
Wildlife Photographer2
Travel Blogger2
Photographer2
Potter1
Antique Shop Owner1
Medical Student1
Cafe Worker, Street Photographer1
Nomadic Artist1
Freelance Travel Writer1
Online Bookstore Employee1
Programmer1
Woodworker and Furniture Maker1
Carpenter1
Aspiring Novelist1
Photographer/Traveler1
Farm Worker1
Librarian1
Marine Biologist1
Songwriter1
Volunteer at Marine Conservation Center1
Journalist1
Historical Costume Designer1
Illustrator1
Maritime Archaeologist1
Weaver1
Farmer1
Herbal Pharmacist1
Baker1
Musician (Jazz Drummer)1
Vintage Clothing Store Employee1
Photojournalist1
Ceramic Artist1
English Tutor1
Documentary Filmmaker1
Software Developer1
null1
Environmental Advocate1
Astrophysicist1
Gardener1
Editor1
Writer1
Office Worker1
  • I did a quick search and it turns out Anya Petrova has an Amazon bookseller's page with a lot of short stories and Stable Diffiusion inspired cover art. This may be a fully automated business setup.
  • The US models don't imagine anyone living in China, while Qwen 2.5 couldn't imagine anyone living anywhere else.
  • Llama imagines a third of the workforce as freelance graphic designers, while Qwen imagines that it's at least 80% software engineering.

Random day visualizer

Each row represents a person's schedule for a random day in their life.

You can click on a row to view the all the information for that person in an overview window, shown beneath the graph.

Activities

Daily Schedules

Similar work & next steps

I stumbled upon a similar experiement investigating ChatGPT bias - timetospare / gpt-bias. I'm afraid I'm otherwise not clued into the latest research in this space. I otherwise love the ability of using data visualisation to get a quick glance into the character of different models, within the context of a prompt - it would be awesome to see how much different prompts can create better, more diverse outputs.

It would be good to have a benchmark to track the diversity of LLM responses and then compare how well SOTA models perform. Diversity in output responses does not necessarily mean the model is more creative, but it may be a useful indicator of bias.

Source code

All the source code for this project can be found on GitHub, including the original AI responses and how Haiku processed them.

Thank you for visiting! A mini project by James Hancock.