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

    Neuronpedia

    INDEX
    Explanations

    unique or specific terms followed by a hyphen that indicate a deliberate or careful selection or action

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to 'hand-picked' selections or items

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Ĥİ
    -0.77
    FORMATION
    -0.73
    ashtra
    -0.72
     bowel
    -0.68
     tremend
    -0.68
    iatus
    -0.65
    ioxide
    -0.64
    omaly
    -0.64
    ibaba
    -0.63
    factor
    -0.63
    POSITIVE LOGITS
    cific
    0.67
    picked
    0.66
    iers
    0.63
    ographed
    0.61
    forward
    0.61
    ricks
    0.60
     centr
    0.60
    cards
    0.59
     Dalai
    0.59
     steering
    0.58
    Act Density 0.129%

    No Known Activations