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

    Neuronpedia

    INDEX
    Explanations

    thinking, negation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     важ
    0.39
     मलया
    0.39
     }^{+}$
    0.37
     વિદ્ય
    0.37
    蔗
    0.36
     Manly
    0.36
     Valuable
    0.35
    兕
    0.35
     اختصاص
    0.35
     মনোনীত
    0.35
    POSITIVE LOGITS
     誕生日
    0.47
    thinking
    0.43
     weak
    0.42
    hud
    0.42
     ELSE
    0.42
     thinking
    0.41
    hurt
    0.41
    guru
    0.41
    implicitly
    0.40
    thin
    0.40
    Act Density 0.009%

    No Known Activations