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

    Neuronpedia

    INDEX
    Explanations

    decoding

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     Formation
    -0.07
    ávis
    -0.07
    (MPI
    -0.07
     труб
    -0.07
     asserts
    -0.07
     Cumhuriyeti
    -0.07
     marsh
    -0.07
     منت
    -0.07
    arcy
    -0.07
     fights
    -0.07
    POSITIVE LOGITS
     Decoder
    0.12
    Decoder
    0.11
     decoder
    0.11
    decoded
    0.10
    decode
    0.10
     decode
    0.10
     decoding
    0.10
    Decode
    0.09
    _decode
    0.09
    .decode
    0.08
    Act Density 0.006%

    No Known Activations