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

    Neuronpedia

    INDEX
    Explanations

    asking for specifics

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     genoemd
    0.83
     comentario
    0.79
     genannt
    0.78
    Người
    0.75
    ayant
    0.74
    Tidak
    0.73
    Seperti
    0.70
    étranger
    0.70
     tekint
    0.70
    unico
    0.69
    POSITIVE LOGITS
     what
    1.67
     exactly
    1.67
     why
    1.59
     where
    1.54
     how
    1.46
     which
    1.46
     whether
    1.46
     precisely
    1.34
    exactly
    1.22
     who
    1.21
    Act Density 1.039%

    No Known Activations