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

    Neuronpedia

    INDEX
    Explanations

    secretive hidden things

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    as
    0.67
    will
    0.57
    our
    0.55
    ecta
    0.54
    are
    0.53
    sche
    0.53
    at
    0.52
    es
    0.52
    strictly
    0.52
    aa
    0.51
    POSITIVE LOGITS
     crush
    0.50
     sustain
    0.48
    Surprisingly
    0.44
    {(-
    0.42
    突
    0.42
    గిన
    0.42
     loan
    0.41
    ходы
    0.40
     secrecy
    0.40
    त्र
    0.40
    Act Density 0.000%

    No Known Activations