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

    Neuronpedia

    INDEX
    Explanations

    training data cutoff

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    始まる
    0.46
    start
    0.41
     ডিগ্রী
    0.40
    しばらく
    0.40
    regex
    0.40
    until
    0.39
     hacked
    0.38
     lysed
    0.38
    Tmp
    0.38
     возника
    0.38
    POSITIVE LOGITS
     الحاله
    0.42
    应用
    0.41
     finalizing
    0.41
    晖
    0.41
    “(
    0.38
    作为
    0.38
     الجديد
    0.38
     رسول
    0.37
     Abbot
    0.37
     zahr
    0.37
    Act Density 0.010%

    No Known Activations