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

    Neuronpedia

    INDEX
    Explanations

    instances of characters locking themselves or others in a particular location or restricted space

    oai_token-act-pair · gpt-3.5-turbo

    instances of being confined or restricted

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     suggestions
    -0.66
    Ĥª
    -0.62
    fortune
    -0.62
    ITNESS
    -0.61
     suggestion
    -0.60
     luck
    -0.59
     Sutherland
    -0.58
    sal
    -0.57
     Drawn
    -0.57
     Suggest
    -0.56
    POSITIVE LOGITS
     tightly
    1.22
     shut
    1.08
     doors
    1.07
     securely
    1.00
     tight
    0.99
    tight
    0.98
     behind
    0.97
     indefinitely
    0.96
     tighter
    0.91
     away
    0.90
    Act Density 0.142%

    No Known Activations