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

    Neuronpedia

    INDEX
    Explanations

    specific keywords indicating a temporal sequence or context transition

    oai_token-act-pair · gpt-3.5-turbo

    the word "Before" indicating prior events or contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    aren
    -0.83
    utter
    -0.79
    pez
    -0.68
    wire
    -0.66
    ILY
    -0.65
    è¦ļéĨĴ
    -0.65
    hyde
    -0.65
    hack
    -0.63
    erry
    -0.63
    amount
    -0.62
    POSITIVE LOGITS
    cluding
    0.75
    rely
    0.72
    pping
    0.72
    irement
    0.71
     realizing
    0.71
     Stats
    0.70
    noon
    0.70
     anyone
    0.66
     concluding
    0.64
    hand
    0.64
    Act Density 0.035%

    No Known Activations