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

    Neuronpedia

    INDEX
    Explanations

    exploring dark themes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ).
    0.46
    .)
    0.36
    ):
    0.31
    ].
    0.30
     judicious
    0.29
     salud
    0.29
     legend
    0.29
     trist
    0.28
     fairly
    0.28
     trim
    0.28
    POSITIVE LOGITS
    Dieser
    0.40
     කිරීමට
    0.38
     ಮಾಡಲು
    0.37
    Questo
    0.36
    Lorsque
    0.35
    ꒷
    0.35
    Xiaomi
    0.35
    この
    0.35
    Ბ
    0.34
    Yeni
    0.33
    Act Density 0.101%

    No Known Activations