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

    Neuronpedia

    INDEX
    Explanations

    validation and Purdue OWL

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    has
    0.59
    Deck
    0.56
    افة
    0.55
    á
    0.55
    اب
    0.54
    ik
    0.54
    aka
    0.54
    hy
    0.54
    ال
    0.53
    ©
    0.52
    POSITIVE LOGITS
     к
    0.88
     К
    0.73
     FreeBSD
    0.73
    Ｂ
    0.67
     Б
    0.67
     varic
    0.67
    ก
    0.67
     photod
    0.66
     क
    0.64
     С
    0.64
    Act Density 0.001%

    No Known Activations