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

    Neuronpedia

    INDEX
    Explanations

    take risks, control, or advantage

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     triumphant
    0.38
    データを
    0.37
     bot
    0.36
     forth
    0.36
     dermat
    0.36
    マットレス
    0.36
    stadt
    0.35
    up
    0.35
    組み
    0.35
     passport
    0.34
    POSITIVE LOGITS
     Taken
    0.63
    Taken
    0.59
     taken
    0.57
    taken
    0.57
    Take
    0.54
     TAKE
    0.53
     Take
    0.52
     TAK
    0.50
     genomen
    0.47
     seriously
    0.47
    Act Density 0.014%

    No Known Activations