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

    Neuronpedia

    INDEX
    Explanations

    question mark followed by a word

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ->
    0.45
     פּ
    0.43
     يُ
    0.43
    當然
    0.41
    STRUCTION
    0.40
     できる
    0.39
     TF
    0.39
     dépenses
    0.39
     పొంద
    0.39
     legitimacy
    0.38
    POSITIVE LOGITS
     maroon
    0.42
    ř
    0.42
    nobody
    0.41
    iknya
    0.41
    рок
    0.39
    rische
    0.38
    kati
    0.38
    ared
    0.38
    unk
    0.37
     crumpled
    0.37
    Act Density 0.003%

    No Known Activations