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

    Neuronpedia

    INDEX
    Explanations

    how to construct

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     по
    0.74
    0.70
    J
    0.66
     Fortunately
    0.63
    B
    0.62
     อย่าง
    0.61
     But
    0.60
     However
    0.60
     Plus
    0.59
     Luckily
    0.59
    POSITIVE LOGITS
     someone
    1.05
     they
    1.00
    <unused329>
    0.97
    <unused946>
    0.96
    <unused758>
    0.95
    <unused1804>
    0.95
    <unused322>
    0.94
     समास
    0.94
    <unused339>
    0.93
    LOTREntity
    0.93
    Act Density 2.809%

    No Known Activations