Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    phrases indicating goals or objectives

    oai_token-act-pair · gpt-3.5-turbo

    expressions that address the reader directly using second-person pronouns

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Ô
    -0.75
     Highlights
    -0.72
     Reports
    -0.70
    pedia
    -0.68
    Parameters
    -0.67
     Amendments
    -0.66
    izable
    -0.66
     resumes
    -0.65
     merits
    -0.63
    Comments
    -0.63
    POSITIVE LOGITS
     lucky
    1.21
     unsure
    1.11
     able
    1.11
     fortunate
    1.10
     interested
    1.08
     unlucky
    1.06
     willing
    1.02
     unaware
    1.02
     inclined
    1.02
     unable
    0.98
    Act Density 0.101%

    No Known Activations