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

    Neuronpedia

    INDEX
    Explanations

    distance in blocks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     are
    0.79
    ö
    0.75
    .
    0.70
    ir
    0.68
    mi
    0.66
    .<
    0.63
     A
    0.62
     have
    0.62
     haue
    0.61
     ఆంధ్ర
    0.59
    POSITIVE LOGITS
    ed
    0.79
    اعر
    0.79
    τε
    0.74
    రు
    0.69
    že
    0.69
    čo
    0.68
    戎
    0.68
    efeller
    0.67
     trajet
    0.67
     و
    0.66
    Act Density 0.001%

    No Known Activations