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

    Neuronpedia

    INDEX
    Explanations

    in inner

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     excluding
    0.82
     randomly
    0.73
     relaxed
    0.72
     double
    0.72
     narrow
    0.71
     before
    0.71
     across
    0.71
     direction
    0.71
     reduced
    0.70
     compatible
    0.70
    POSITIVE LOGITS
    此事
    0.96
    dyž
    0.93
    mér
    0.92
     často
    0.92
    ững
    0.92
    quefois
    0.91
     اين
    0.90
    ţă
    0.90
    endidikan
    0.88
    unehm
    0.88
    Act Density 0.071%

    No Known Activations