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

    Neuronpedia

    INDEX
    Explanations

    phrases and words related to repeated actions or exchanges

    oai_token-act-pair · gpt-3.5-turbo

    repetitions of the word "forth."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    cius
    -0.72
     box
    -0.66
     elements
    -0.65
    Quantity
    -0.64
    abase
    -0.62
     Puerto
    -0.61
     boxes
    -0.59
     Chile
    -0.59
     Sanchez
    -0.59
     Beautiful
    -0.59
    POSITIVE LOGITS
    forth
    2.07
    forward
    1.10
    theless
    1.10
    comings
    1.00
    coming
    0.96
    soever
    0.89
    etheless
    0.84
    sburg
    0.83
    rontal
    0.83
    stairs
    0.82
    Act Density 0.004%

    No Known Activations