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

    Neuronpedia

    INDEX
    Explanations

    want me to

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Ç
    0.37
    ط
    0.37
    ว่า
    0.36
    гляд
    0.36
    وال
    0.35
    wym
    0.35
    ม
    0.35
    that
    0.35
    isms
    0.35
     kwamba
    0.35
    POSITIVE LOGITS
     want
    0.57
     desperately
    0.47
     quiero
    0.46
     partake
    0.46
     WANT
    0.44
     revenge
    0.43
     wants
    0.42
     quería
    0.42
     unbedingt
    0.41
    want
    0.40
    Act Density 0.023%

    No Known Activations