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

    Neuronpedia

    INDEX
    Explanations

    Bacon and specific published works

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     most
    -0.84
     Washington
    -0.83
     envio
    -0.80
     despite
    -0.80
     and
    -0.80
     when
    -0.78
     intensely
    -0.77
     people
    -0.77
     at
    -0.77
     as
    -0.75
    POSITIVE LOGITS
    Bacon
    1.16
     Bacon
    1.12
    Nov
    0.93
    ᕙ
    0.91
     эксперимента
    0.90
    踯
    0.90
     induk
    0.89
    Intel
    0.85
     kapture
    0.85
     réguli
    0.84
    Act Density 0.015%

    No Known Activations