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

    Neuronpedia

    INDEX
    Explanations

    exploring why

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Alternatively
    0.69
     "\\
    0.68
     Vocabulary
    0.68
     י
    0.67
     ק
    0.66
     או
    0.65
     tbsp
    0.64
     Lyrics
    0.63
     În
    0.63
     également
    0.63
    POSITIVE LOGITS
    why
    0.71
    de
    0.65
    real
    0.65
    telling
    0.62
    not
    0.62
    up
    0.61
    des
    0.60
    chocolate
    0.60
    changing
    0.60
    type
    0.59
    Act Density 0.577%

    No Known Activations