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

    Neuronpedia

    INDEX
    Explanations

    multilingual explanations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     While
    0.81
    :",
    0.76
    !”
    0.76
     few
    0.76
    :"+
    0.76
     Throughout
    0.76
    !”,
    0.76
     Since
    0.75
    :
    0.75
     diverse
    0.74
    POSITIVE LOGITS
    ↵
    1.08
    ↵↵
    0.96
    कौन
    0.92
    Invalid
    0.88
    Otros
    0.88
    login
    0.86
    \_
    0.85
    <unused1401>
    0.85
    Extra
    0.85
    <unused1317>
    0.85
    Act Density 0.346%

    No Known Activations