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

    Neuronpedia

    INDEX
    Explanations

    Data drift detection

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     mn
    -0.08
     determ
    -0.08
     nid
    -0.08
     milliseconds
    -0.08
     escape
    -0.08
     dành
    -0.08
     communicator
    -0.08
     escapes
    -0.08
     associate
    -0.07
     immed
    -0.07
    POSITIVE LOGITS
     degrading
    0.12
     degradation
    0.11
     degrade
    0.10
     degraded
    0.10
    Mismatch
    0.09
    _ratio
    0.09
    Performance
    0.09
     ухуд
    0.09
     deterioration
    0.09
    Hoe
    0.08
    Act Density 0.004%

    No Known Activations