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

    Neuronpedia

    INDEX
    Explanations

    1-800-799-SAFE hotline

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     proportionate
    0.43
     permeated
    0.40
    ثم
    0.38
     bathtub
    0.38
     resampling
    0.38
     speedometer
    0.37
     convolutional
    0.37
     permeates
    0.36
     முதலிய
    0.36
     infiltrating
    0.36
    POSITIVE LOGITS
     SAFE
    0.70
    SAFE
    0.68
     SAF
    0.54
     STOP
    0.51
     Safe
    0.50
     HELP
    0.50
    STOP
    0.48
    Safe
    0.47
    HELP
    0.47
     Saf
    0.44
    Act Density 0.029%

    No Known Activations