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

    Neuronpedia

    INDEX
    Explanations

    phrases related to explaining how something operates or works

    oai_token-act-pair · gpt-3.5-turbo

    phrases that describe functionality or the workings of various systems or mechanisms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    gart
    -0.81
    voy
    -0.69
    imar
    -0.68
    TERN
    -0.68
    Bir
    -0.67
    miss
    -0.65
    moon
    -0.64
    ark
    -0.61
    intage
    -0.60
    ãģķ
    -0.60
    POSITIVE LOGITS
     proced
    0.72
     unfolded
    0.68
    :[
    0.67
     nutshell
    0.65
     differs
    0.65
     unfold
    0.65
    iquette
    0.62
     Different
    0.62
     unfolding
    0.62
     shaping
    0.62
    Act Density 0.097%

    No Known Activations