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

    Neuronpedia

    INDEX
    Explanations

    references to the concept of being "broken" or dysfunctional in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    icals
    -0.78
    76561
    -0.76
    minist
    -0.74
    opped
    -0.72
    aeda
    -0.71
    anamo
    -0.69
    izations
    -0.69
    azor
    -0.68
    hedon
    -0.68
    extreme
    -0.67
    POSITIVE LOGITS
    hearted
    1.05
    neck
    1.02
     bones
    0.90
    bones
    0.87
     ribs
    0.84
    broken
    0.76
    adoes
    0.74
    staff
    0.73
     glass
    0.73
    edient
    0.72
    Act Density 0.011%

    No Known Activations