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

    Neuronpedia

    INDEX
    Explanations

    names followed by technical terms

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    。
    -1.06
    都是
    -0.96
    “
    -0.90
    突然
    -0.88
    .'
    -0.84
    正式
    -0.84
    獲得
    -0.83
    க்கும்
    -0.83
    .’
    -0.80
    相比
    -0.80
    POSITIVE LOGITS
    ův
    1.19
    ian
    1.10
     theorem
    1.01
    féle
    0.98
     inför
    0.97
    adm
    0.96
     Theorem
    0.90
     posthum
    0.90
    hian
    0.89
    Compli
    0.88
    Act Density 0.151%

    No Known Activations