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

    Neuronpedia

    INDEX
    Explanations

    buttermilk", "but contains", "but not", "butch"

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    0.47
    0.45
    டன்
    0.43
    0.41
     Neste
    0.40
     따라서
    0.40
    வே
    0.40
    0.39
    0.38
    0.38
    POSITIVE LOGITS
    termilk
    0.74
    ternut
    0.71
     dennoch
    0.67
    thole
    0.63
     nevertheless
    0.60
    rition
    0.57
    tery
    0.54
     nonetheless
    0.54
     তবুও
    0.54
    ters
    0.54
    Act Density 0.007%

    No Known Activations