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

    Neuronpedia

    INDEX
    Explanations

    capacity

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     Submitted
    -0.09
    üd
    -0.09
     bezig
    -0.08
    Coronavirus
    -0.08
     Mish
    -0.08
    urias
    -0.08
     приключ
    -0.08
    Arrangement
    -0.07
     successfully
    -0.07
     warmly
    -0.07
    POSITIVE LOGITS
     degrade
    0.13
     degradation
    0.13
     deterioration
    0.12
     deterior
    0.12
     degraded
    0.12
     ухуд
    0.11
    性能
    0.11
     degrading
    0.11
     performance
    0.10
     degrad
    0.10
    Act Density 0.035%

    No Known Activations