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

    Neuronpedia

    INDEX
    Explanations

    cost

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     birds
    -0.07
    ())).
    -0.07
     Serialization
    -0.06
    (cn
    -0.06
     nal
    -0.06
    .literal
    -0.06
    "]).
    -0.06
     racist
    -0.06
    -spin
    -0.06
     readOnly
    -0.06
    POSITIVE LOGITS
    .eu
    0.07
    IM
    0.06
    takes
    0.06
     LAP
    0.06
     Quarry
    0.06
     mí
    0.06
    Que
    0.06
     SEL
    0.06
     nothing
    0.06
    _Generic
    0.06
    Act Density 0.014%

    No Known Activations