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

    Neuronpedia

    INDEX
    Explanations

    but in many languages

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     prolific
    0.45
     tout
    0.44
     asexual
    0.43
     specifically
    0.43
     use
    0.42
    id
    0.42
     style
    0.42
     todo
    0.42
     vertical
    0.42
     T
    0.40
    POSITIVE LOGITS
     nhưng
    0.97
     ancak
    0.84
     ngunit
    0.84
     لكن
    0.82
     ولكن
    0.79
     lakini
    0.75
     ஆனால்
    0.75
     Ancak
    0.75
     però
    0.74
     אך
    0.72
    Act Density 0.001%

    No Known Activations