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

    Neuronpedia

    INDEX
    Explanations

    reading subtle signals

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Aunque
    0.97
     aunque
    0.84
     ignoring
    0.82
     ayant
    0.81
     Embora
    0.81
     Aufgrund
    0.79
     Apesar
    0.79
     Anche
    0.78
     även
    0.78
     Además
    0.78
    POSITIVE LOGITS
     वगैरह
    0.84
    0.81
    noons
    0.81
    ogra
    0.78
    Favourite
    0.77
    0.77
    Margins
    0.77
    কৃষ্ণ
    0.77
     बांट
    0.75
    listen
    0.73
    Act Density 0.058%

    No Known Activations