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

    Neuronpedia

    INDEX
    Explanations

    closing parenthesis followed by code block

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     conclusão
    -1.27
    particularly
    -1.13
     combinação
    -1.09
     melh
    -1.07
     opinião
    -1.05
     maniere
    -1.02
     missão
    -1.00
     particular
    -0.96
     estratégia
    -0.96
     ligação
    -0.96
    POSITIVE LOGITS
     (
    2.44
     and
    1.18
     {
    1.11
     error
    0.98
     on
    0.96
     {
    0.93
     exquisite
    0.92
    ？（
    0.91
     in
    0.90
     before
    0.90
    Act Density 0.005%

    No Known Activations