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

    Neuronpedia

    INDEX
    Explanations

    adjective-noun phrases suggesting a comparison or categorization of things

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the expression "some sort of" indicating various categories or types of things

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    oons
    -0.84
    omers
    -0.76
    nets
    -0.76
    arers
    -0.74
    markets
    -0.74
     Sands
    -0.73
    ared
    -0.71
    ears
    -0.71
    ires
    -0.71
    riers
    -0.70
    POSITIVE LOGITS
     intermediary
    0.89
     miraculous
    0.76
     arrangement
    0.74
     resemblance
    0.74
     magical
    0.73
     stabilization
    0.73
     helper
    0.73
     internal
    0.72
     coordination
    0.72
     ransom
    0.71
    Act Density 0.055%

    No Known Activations