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

    Neuronpedia

    INDEX
    Explanations

    describing technical concepts and their use

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Something
    0.92
     something
    0.89
    Something
    0.88
     Someone
    0.85
    someone
    0.84
    something
    0.82
     Somewhere
    0.82
     someone
    0.81
    Somewhere
    0.81
    Someone
    0.80
    POSITIVE LOGITS
     used
    1.09
     utilizada
    0.94
     utilizzato
    0.92
     utilizado
    0.91
     provides
    0.91
     also
    0.85
     também
    0.85
     також
    0.83
     utilisée
    0.83
    এটি
    0.83
    Act Density 0.566%

    No Known Activations