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

    Neuronpedia

    INDEX
    Explanations

    GNU/Linux system information

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     given
    0.64
     work
    0.62
    ,
    0.62
     un
    0.62
     exactly
    0.62
    on
    0.61
     is
    0.60
     only
    0.59
     amp
    0.59
    .
    0.59
    POSITIVE LOGITS
    饉
    0.87
     rctx
    0.86
    ཥ
    0.86
    raseña
    0.80
    ꗥ
    0.78
    <unused1716>
    0.78
    ፔ
    0.78
     cGraph
    0.77
    Benzyloxy
    0.77
     callSettings
    0.76
    Act Density 0.014%

    No Known Activations