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

    Neuronpedia

    INDEX
    Explanations

    sadly, comfort, masked, today

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     gradi
    0.53
    假如
    0.46
     eset
    0.45
     cached
    0.44
     infil
    0.43
     spacers
    0.42
     ඔබට
    0.40
     equivalently
    0.40
     decode
    0.39
     proguard
    0.39
    POSITIVE LOGITS
    小组
    0.51
    ਯ
    0.47
    AND
    0.46
    CHREIB
    0.46
    !।
    0.46
    AN
    0.46
    communication
    0.46
    revise
    0.45
    UR
    0.45
    Ꮲ
    0.44
    Act Density 0.001%

    No Known Activations