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

    Neuronpedia

    INDEX
    Explanations

    drop-in replacements

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    お
    0.55
    ל
    0.53
    htmlspecialchars
    0.52
    은
    0.51
    따
    0.51
    ३
    0.50
    嗲
    0.50
    htra
    0.50
    ری
    0.50
    ના
    0.49
    POSITIVE LOGITS
    ando
    0.55
     появления
    0.52
    ».
    0.51
     zehn
    0.50
    andering
    0.49
     условиях
    0.49
    expiration
    0.48
     dernière
    0.48
     побы
    0.47
    隐含规则
    0.47
    Act Density 0.000%

    No Known Activations