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

    Neuronpedia

    INDEX
    Explanations

    use best practice

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    astrous
    0.50
     अद्भुत
    0.50
     엄청
    0.47
     allerlei
    0.46
     disastrous
    0.46
     berbahaya
    0.45
     impresionante
    0.45
     особы
    0.45
     niezwy
    0.44
     ناقص
    0.44
    POSITIVE LOGITS
     möglichst
    1.12
     preferably
    1.08
    preferably
    0.98
    尽量
    0.96
    尽可能
    0.91
     reputable
    0.88
     almeno
    0.87
     Preferably
    0.84
    至少
    0.81
     최대한
    0.81
    Act Density 0.052%

    No Known Activations