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

    Neuronpedia

    INDEX
    Explanations

    reduces and translates

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    avascript
    0.43
    咩
    0.39
    etele
    0.37
     Scrub
    0.37
     delegate
    0.37
     लीजिएगा
    0.37
    ération
    0.36
     visitation
    0.36
    legraph
    0.36
    ঘা
    0.36
    POSITIVE LOGITS
    uce
    1.44
    UCE
    1.29
    uced
    1.23
    UC
    1.19
    ucing
    1.16
    ucer
    1.16
    duced
    1.16
    uces
    1.15
     duc
    1.15
    uc
    1.13
    Act Density 0.015%

    No Known Activations