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

    Neuronpedia

    INDEX
    Explanations

    sequences of numbers separated by dashes

    oai_token-act-pair · gpt-3.5-turbo

    numerical values and negative symbols

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     contracting
    -0.69
     processes
    -0.62
     incorpor
    -0.62
     Weld
    -0.58
    lly
    -0.58
    onal
    -0.58
     staffing
    -0.57
     streaming
    -0.57
     etiquette
    -0.57
     filtering
    -0.57
    POSITIVE LOGITS
    _-
    1.02
    sama
    0.81
    è¦ļéĨĴ
    0.78
    0
    0.74
    ãĥĺãĥ©
    0.72
    Cs
    0.72
    disc
    0.72
    plus
    0.71
    clad
    0.71
    cell
    0.71
    Act Density 0.046%

    No Known Activations