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

    Neuronpedia

    INDEX
    Explanations

    assumed, leaving, holding office

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     bi
    2.09
    bi
    1.51
     Bi
    1.48
    Bi
    1.48
     би
    1.29
     biometric
    1.09
     offices
    0.99
    办公室
    0.98
     بي
    0.96
     bih
    0.94
    POSITIVE LOGITS
    off
    0.63
    Off
    0.60
     off
    0.55
    オフ
    0.50
    фф
    0.50
     オ
    0.49
     Off
    0.48
    오
    0.47
    offs
    0.47
     Hoff
    0.46
    Act Density 0.007%

    No Known Activations