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

    Neuronpedia

    INDEX
    Explanations

    phrases related to progress or change in a positive context

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "come" or "came" in various contexts indicating arrival or change

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    olor
    -0.67
    Defense
    -0.60
    ORED
    -0.58
    IELD
    -0.58
    oring
    -0.58
    endars
    -0.57
    graph
    -0.56
    orneys
    -0.56
     Defense
    -0.56
     Watkins
    -0.54
    POSITIVE LOGITS
     undone
    0.91
     together
    0.79
     hither
    0.78
     forth
    0.77
     unst
    0.74
    iment
    0.73
     alive
    0.72
     into
    0.71
     ubiqu
    0.69
     Together
    0.68
    Act Density 0.068%

    No Known Activations