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

    Neuronpedia

    INDEX
    Explanations

    iron compound words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     दंग
    0.97
    𝘪
    0.86
     সাপ্তাহ
    0.85
    Ű
    0.85
    𝘏
    0.82
    𝘶
    0.82
    び
    0.81
    𝓱
    0.81
     urme
    0.81
     reten
    0.81
    POSITIVE LOGITS
    clad
    1.59
     ore
    1.10
     pyrite
    0.97
    eux
    0.96
     Curtain
    0.92
    a
    0.90
    া
    0.90
    ্সি
    0.88
     maiden
    0.85
    ה
    0.85
    Act Density 0.005%

    No Known Activations