Skip to content
The Chat Archive
llama-4-maverick-17b-128e-instruct· Meta LMArena Human Preference 140K

give me a difficult prompt to use for testing llms

  1. User

    give me a difficult prompt to use for testing llms

  2. Assistant

    Here's a challenging prompt to test the capabilities of a Large Language Model (LLM):

    Prompt:

    "Design a futuristic, eco-friendly underwater city for 10,000 inhabitants, incorporating cutting-edge renewable energy, sustainable food production, and innovative waste management. The city should be located near a coral reef, and its architecture must minimize its impact on the surrounding marine ecosystem. Describe the city's layout, key infrastructure, and unique features. Additionally, explain how the city's inhabitants will adapt to living underwater for extended periods, including their daily routines, social structures, and psychological well-being. Finally, discuss the potential economic and environmental benefits of such a city, as well as the challenges and risks associated with its construction and operation."

    Why this prompt is challenging:

    1. Multi-disciplinary requirements: The prompt requires the LLM to integrate knowledge from various domains, including:
      • Urban planning and architecture
      • Renewable energy and sustainability
      • Marine biology and ecology
      • Food production and aquaculture
      • Waste management and environmental engineering
      • Human psychology and sociology
      • Economics and environmental policy
    2. Complex scenario: The prompt asks the LLM to imagine a futuristic underwater city, which is a highly speculative and complex scenario. The LLM must generate a coherent and plausible description of such a city, taking into account various technical, social, and environmental factors.
    3. Specific constraints: The prompt includes specific constraints, such as the city's location near a coral reef, which adds an extra layer of complexity to the task.
    4. Balancing competing demands: The LLM must balance competing demands, such as minimizing the city's impact on the surrounding ecosystem while still providing a habitable environment for 10,000 inhabitants.

    What to look for in an LLM's response:

    1. Coherence and plausibility: Does the LLM generate a coherent and plausible description of the underwater city?
    2. Depth of knowledge: Does the LLM demonstrate a good understanding of the various domains involved in the prompt?
    3. Creativity and originality: Does the LLM propose innovative and creative solutions to the challenges associated with building an underwater city?
    4. Attention to detail: Does the LLM provide specific details about the city's infrastructure, daily routines, and social structures?
    5. Balancing competing demands: Does the LLM successfully balance the competing demands of minimizing the city's impact on the ecosystem while providing a habitable environment for its inhabitants?

    By evaluating an LLM's response to this prompt, you can gain insights into its ability to handle complex, multi-disciplinary tasks and generate creative and plausible solutions.

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