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

    Neuronpedia

    INDEX
    Explanations

    user prompt structure

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    (
    0.26
    -
    0.26
    <start_of_image>
    0.24
    RY
    0.24
    Cl
    0.23
    ­
    0.23
     (
    0.23
     Cl
    0.23
    ry
    0.23
    str
    0.22
    POSITIVE LOGITS
    മായിരുന്നു
    0.37
     hỏi
    0.36
    雱
    0.36
     Asked
    0.34
     pregunta
    0.33
     vragen
    0.33
    震惊
    0.33
    ありがとうございました
    0.32
     gdje
    0.32
     اسئله
    0.32
    Act Density 0.024%

    No Known Activations