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

    Neuronpedia

    INDEX
    Explanations

    keep to, keep a, keep track

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    er
    1.31
    א
    1.18
    र
    1.16
    rato
    1.14
    haha
    1.13
    tip
    1.11
    dehyde
    1.10
    ম
    1.10
     pren
    1.07
    い
    1.07
    POSITIVE LOGITS
     abreast
    1.76
     track
    1.54
     repeating
    1.41
     wickets
    1.39
     tabs
    1.39
     questioning
    1.35
     innovating
    1.35
    iencing
    1.33
     expanding
    1.32
    akes
    1.31
    Act Density 0.020%

    No Known Activations