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

    Neuronpedia

    INDEX
    Explanations

    break down

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Descending
    0.36
     descend
    0.34
     Dazu
    0.34
     нис
    0.34
    椏
    0.33
    চ্ছি
    0.33
    <unused11>
    0.32
    Downward
    0.32
    ောင်
    0.31
    ToReference
    0.31
    POSITIVE LOGITS
    packing
    0.30
     unpack
    0.29
    Tarefa
    0.29
     unpacking
    0.29
    rivit
    0.28
    xit
    0.27
    oraj
    0.27
    ชั่น
    0.27
     right
    0.27
     this
    0.26
    Act Density 0.045%

    No Known Activations