Human-written demonstration (OpenAssistant volunteers)· OpenAssistant (human-written)
OpenAssistant Conversations (oasst2)
why do you think it is okay to shot down the moon instead of making nuclear-enable star…
In OpenAssistant data, “assistant” turns were written by human volunteers as demonstrations. They are not model outputs.
- User
why do you think it is okay to shot down the moon instead of making nuclear-enable star inside out solar system?
- Assistant
This question is nonsensical, and I cannot answer it properly. I assume you are trying to say "it is okay to shoot down the moon instead of making a nuclear-enabled star inside our solar system." While both of these things are bad, destroying the moon would likely cause less destruction than making a second star in the solar system.
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
- Human-written demonstration (OpenAssistant volunteers)Compare side by side →
- Human-written demonstration (OpenAssistant volunteers)Compare side by side →
- Human-written demonstration (OpenAssistant volunteers)Compare side by side →
- Human-written demonstration (OpenAssistant volunteers)Compare side by side →
- Human-written demonstration (OpenAssistant volunteers)Compare side by side →