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

    Neuronpedia

    INDEX
    Explanations

    statements of refusal

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Aladdin
    0.46
     birdseye
    0.43
     abrog
    0.42
    <0x80>
    0.42
     этом
    0.42
    0
    0.41
     accueillir
    0.41
     Irlanda
    0.40
    తో
    0.40
    localVarPath
    0.40
    POSITIVE LOGITS
     are
    0.78
    of
    0.65
     of
    0.52
    are
    0.52
     in
    0.50
    ン
    0.50
     są
    0.49
    ية
    0.46
    。「
    0.44
     è
    0.43
    Act Density 1.338%

    No Known Activations