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

    Neuronpedia

    INDEX
    Explanations

    word count requests

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     way
    0.67
     makeup
    0.60
     home
    0.59
     make
    0.57
     (
    0.57
     s
    0.55
     organization
    0.54
     other
    0.54
     men
    0.54
     is
    0.54
    POSITIVE LOGITS
    ?”.
    0.84
    ？”
    0.83
    Comments
    0.82
    ➘
    0.80
    čiti
    0.80
    ?”
    0.77
    ógł
    0.76
    !”.
    0.76
     politiques
    0.76
     паліты
    0.75
    Act Density 0.014%

    No Known Activations