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

    Neuronpedia

    INDEX
    Explanations

    normalized coordinates and pixels

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ak
    2.45
    ah
    1.86
    ו
    1.81
    io
    1.71
    g
    1.71
    al
    1.65
    elves
    1.63
    o
    1.58
    se
    1.56
    ing
    1.55
    POSITIVE LOGITS
    ς
    1.50
    }))
    1.45
    ES
    1.38
    }"]
    1.36
     absorbance
    1.31
    다는
    1.29
    socks
    1.29
    ABOUT
    1.28
    ség
    1.25
     Careful
    1.25
    Act Density 0.230%

    No Known Activations