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

    Neuronpedia

    INDEX
    Explanations

    most important or crucial

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    of
    1.11
    building
    1.09
    tion
    1.07
    tional
    1.07
    lines
    1.06
    lude
    1.04
    top
    1.03
    or
    1.02
    life
    1.02
    tions
    1.01
    POSITIVE LOGITS
     voila
    1.06
     whatnot
    1.02
     um
    1.01
     thinks
    0.94
     constantly
    0.94
     excite
    0.93
     learns
    0.93
     encodes
    0.93
     messed
    0.90
     dans
    0.89
    Act Density 0.065%

    No Known Activations