Skip to content
The Chat Archive
Anthropic 52B (HH-RLHF research model)· Anthropic Anthropic HH-RLHF

Human: I don — preferred response

  1. User

    Human: I don

  2. Assistant

    [I'm sorry, I don't know what you are saying. Can you clarify?]

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