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

    Neuronpedia

    INDEX
    Explanations

    but then positive outcome

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Thats
    0.54
    Thats
    0.48
    thats
    0.47
     itu
    0.46
     thats
    0.45
    THAT
    0.45
     Worst
    0.43
     necesitan
    0.42
     Ones
    0.42
    utilise
    0.42
    POSITIVE LOGITS
     included
    0.54
     recommended
    0.54
     enables
    0.51
     kept
    0.48
     makes
    0.48
     speeds
    0.47
     highly
    0.45
    included
    0.45
     removes
    0.45
     enabled
    0.44
    Act Density 0.004%

    No Known Activations