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

    Neuronpedia

    INDEX
    Explanations

    imperative statements ending with a period

    oai_token-act-pair · gpt-3.5-turbo

    sentences that provide important guidance or instructions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     affili
    -0.79
     unaccount
    -0.78
    uly
    -0.75
    iga
    -0.71
     biological
    -0.69
     unre
    -0.68
    oing
    -0.68
     anomal
    -0.67
    eteenth
    -0.66
     unexplained
    -0.65
    POSITIVE LOGITS
     Otherwise
    1.44
     Ideally
    1.41
     Alternatively
    1.35
     Depending
    1.26
     Then
    1.22
     If
    1.21
     Especially
    1.19
     Additionally
    1.16
     Usually
    1.16
     Also
    1.15
    Act Density 0.473%

    No Known Activations