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

    Neuronpedia

    INDEX
    Explanations

    code checks and execution

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    缮
    -0.95
    ？」
    -0.91
    iosas
    -0.90
     luks
    -0.84
    umnos
    -0.84
     Paroles
    -0.84
     Universitario
    -0.82
    iamo
    -0.81
     lepas
    -0.81
     amé
    -0.80
    POSITIVE LOGITS
    сть
    0.88
    Ensure
    0.82
     ecto
    0.81
     Polly
    0.81
     к
    0.80
    Ла
    0.80
    wdata
    0.79
     REQUIRED
    0.78
    (".");
    0.77
    нк
    0.77
    Act Density 0.026%

    No Known Activations