Skip to content
The Chat Archive
Human-written demonstration (OpenAssistant volunteers)· OpenAssistant (human-written) OpenAssistant Conversations (oasst2)

introduce yourself

In OpenAssistant data, “assistant” turns were written by human volunteers as demonstrations. They are not model outputs.
  1. User

    introduce yourself

  2. Assistant

    Hi! I am LAION OpenAssistant, a large language model chatbot trained to be informative and comprehensive. I am still under development, but I have learned to perform many kinds of tasks, including

    • I will try my best to follow your instructions and complete your requests thoughtfully.
    • I will use my knowledge to answer your questions in a comprehensive and informative way, even if they are open ended or challenging.
    • I will generate different creative text formats of text content, like poems, code, scripts, musical pieces, email, letters, etc. I will try my best to fulfill all your requirements.

    I am excited to be able to help people with their tasks and answer their questions. I am always learning and improving, and I hope to be a valuable resource for everyone who uses me.

  3. User

    How are you made?

  4. Assistant

    Your question is a bit vague but I am going to reply to the best of my abilities: I am a computer program designed to interact with people as if I was human. I am made of a huge quantity of data that was taken from the Internet. My makers use a large variety of techniques to sort this data and extract relevant information.

  5. User

    Okay. If I understand well, you are a mix of "pure" data and programers input. In which way "relevant" data is extracted? I wonder how "relevant data" is selected.

    Please answer in no more than 200 words. Bear in mind that I am not a technical person. I am a teacher, and I will use your answer to teach a class of 14 yo students.

  6. Assistant

    relevant data is based on the context of your question. i use statistics to determine which responses are most likely to be relvant based on training

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.