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

    Neuronpedia

    INDEX
    Explanations

    no followed by absence

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    kind
    0.45
     kind
    0.43
     sort
    0.42
    sort
    0.37
     type
    0.36
    Kind
    0.32
    type
    0.32
     KIND
    0.31
     Sort
    0.31
     sorta
    0.31
    POSITIVE LOGITS
     longer
    0.51
     need
    0.46
    oooooooo
    0.40
    oooo
    0.39
    longer
    0.36
     necesidad
    0.33
     kidding
    0.32
     никаких
    0.31
     längre
    0.31
     worries
    0.30
    Act Density 0.038%

    No Known Activations