Hello
- User
Hello
- Assistant
Hello! How can I assist you today?
- User
What's your model name?
- Assistant
I’m based on the GPT-4 architecture. How can I help you today?
- User
how big is your data set?
- Assistant
I don’t have access to the exact size or specifics of the dataset used to train me. However, I was trained on a mixture of licensed data, data created by human trainers, and publicly available text from books, websites, and other sources. The dataset is extremely large—on the order of hundreds of billions of words—to help me understand and generate human-like text across a wide range of topics. If you have questions about how training generally works or about AI in general, feel free to ask!
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.
Identical first prompt
- gpt-4-0314Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-3.5-turbo-0301Compare side by side →
- gpt-3.5-turbo-0301Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-3.5-turbo-0301Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-4-0314Compare side by side →
- gpt-3.5-turbo-0301Compare side by side →
claude-instant-1Compare side by side →