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

    Neuronpedia

    INDEX
    Explanations

    apologies and expressions of regret

    oai_token-act-pair · gpt-3.5-turbo

    expressions of apology and regret

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     mosqu
    -0.85
    population
    -0.79
    unin
    -0.75
     monop
    -0.73
     density
    -0.70
    rones
    -0.70
     Locations
    -0.69
     Sear
    -0.69
    nergy
    -0.68
    Residents
    -0.68
    POSITIVE LOGITS
     apology
    2.06
     apologise
    2.05
     apologized
    2.00
     apologised
    1.99
     apologize
    1.94
     apologies
    1.90
     regrets
    1.85
     remorse
    1.81
     apologizing
    1.79
     regret
    1.78
    Act Density 0.631%

    No Known Activations