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

    Neuronpedia

    INDEX
    Explanations

    core connection unique dialect legitimate

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    gia
    0.43
     nrows
    0.42
     verden
    0.41
     sequins
    0.41
    ليز
    0.41
    wendungs
    0.41
     PI
    0.40
    nelle
    0.40
    fice
    0.39
     incision
    0.39
    POSITIVE LOGITS
    atributo
    0.42
     丿
    0.42
    ศัก
    0.41
    ┤
    0.40
    Seperti
    0.39
    څ
    0.39
     superhero
    0.38
    აბ
    0.38
    茘
    0.38
    وک
    0.38
    Act Density 6.284%

    No Known Activations