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

    Neuronpedia

    INDEX
    Explanations

    top of high places

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ణు
    0.36
    bcrypt
    0.36
    θο
    0.36
    uffle
    0.35
     ডিপ
    0.35
     backdoor
    0.35
    saus
    0.35
     downg
    0.35
    蒻
    0.34
    isArray
    0.34
    POSITIVE LOGITS
     atop
    2.38
     perched
    1.79
     hilltop
    1.75
     summit
    1.59
     rooftop
    1.54
     vantage
    1.54
    atop
    1.53
     sommet
    1.42
     cima
    1.42
    頂
    1.41
    Act Density 0.041%

    No Known Activations