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

    Neuronpedia

    INDEX
    Explanations

    @implementation

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     withhold
    -0.07
     disabilities
    -0.07
     supervised
    -0.07
     _
    -0.06
     electricity
    -0.06
     mi
    -0.06
     town
    -0.06
     cosmetic
    -0.06
    topics
    -0.06
     checks
    -0.06
    POSITIVE LOGITS
    _AB
    0.07
     EditorGUILayout
    0.07
    ATRIX
    0.07
    (ErrorMessage
    0.07
     Cortex
    0.07
    /model
    0.06
    ameleon
    0.06
     kInstruction
    0.06
    .Room
    0.06
    AMENT
    0.06
    Act Density 0.003%

    No Known Activations