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

    Neuronpedia

    INDEX
    Explanations

    numbers appended to words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ে
    1.22
    orems
    1.21
     nhàng
    1.16
     Fmat
    1.13
    ০
    1.09
    ۾
    1.08
    oretical
    1.08
    oretically
    1.07
    pèce
    1.05
    𒌅
    1.05
    POSITIVE LOGITS
     der
    0.97
     ter
    0.96
     an
    0.94
     a
    0.91
     ber
    0.81
    woods
    0.77
     ben
    0.76
     ay
    0.75
     Woods
    0.75
     kraft
    0.74
    Act Density 0.080%

    No Known Activations