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

    Neuronpedia

    INDEX
    Explanations

    ctl followed by control or commands

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     premieres
    1.20
     for
    1.08
    ↵↵
    1.05
     on
    0.93
     alleys
    0.93
     streaks
    0.93
     underlining
    0.92
     averages
    0.91
    č
    0.91
     over
    0.91
    POSITIVE LOGITS
    in
    1.72
    are
    1.29
    い
    1.11
    na
    1.08
    ม
    1.08
    dır
    1.07
    inac
    1.06
    er
    1.05
    กว่า
    1.05
    or
    1.03
    Act Density 0.001%

    No Known Activations