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

    Neuronpedia

    INDEX
    Explanations

    COUNTABLE UNIONS, URL, FRONT

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     $=$
    0.89
     серьез
    0.89
    ে
    0.88
    ắn
    0.86
     CHCl
    0.86
     చి
    0.83
    هما
    0.82
    i
    0.82
     данный
    0.81
     Однако
    0.81
    POSITIVE LOGITS
    িক
    0.81
    ल
    0.81
    굽
    0.74
    项目中
    0.68
    EN
    0.66
    nee
    0.66
    力
    0.66
    0.65
    م
    0.65
     flange
    0.64
    Act Density 0.001%

    No Known Activations