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

    Neuronpedia

    INDEX
    Explanations

    planted upside down

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    د
    0.59
    م
    0.57
    ف
    0.55
    مة
    0.51
    נ
    0.51
    tolerant
    0.50
    י
    0.49
    convert
    0.49
    ي
    0.47
    מ
    0.47
    POSITIVE LOGITS
     masks
    0.45
     flavours
    0.45
     launchers
    0.45
     CHAS
    0.45
     kết
    0.45
     embarked
    0.44
     famílias
    0.44
     setelah
    0.44
     shoppers
    0.43
     usuários
    0.43
    Act Density 0.004%

    No Known Activations