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

    Neuronpedia

    INDEX
    Explanations

    superlatives or rankings

    oai_token-act-pair · gpt-3.5-turbo

    superlative adjectives and phrases that denote ranking or popularity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    æ©
    -0.81
    imaru
    -0.77
    heid
    -0.76
    izon
    -0.73
     Rite
    -0.73
    wagen
    -0.72
    ategories
    -0.70
    instead
    -0.69
    halla
    -0.69
    IDS
    -0.68
    POSITIVE LOGITS
     expensive
    1.16
     efficient
    1.07
     important
    1.06
     valuable
    1.06
     populous
    1.05
     exciting
    1.03
     powerful
    1.03
     lucrative
    1.02
     prolific
    1.02
     influential
    1.01
    Act Density 0.082%

    No Known Activations