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

    Neuronpedia

    INDEX
    Explanations

    keep it short, keep track

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    <0x0D>
    1.60
    し
    1.18
    ка
    1.17
     identific
    1.08
     assegn
    1.08
    2
    1.08
     invoc
    1.05
    1.02
     coloc
    0.96
                        
    0.91
    POSITIVE LOGITS
     on
    1.48
     keep
    1.19
    AS
    1.13
    AV
    1.13
    on
    1.06
    та
    1.06
    υ
    1.05
    ur
    1.04
    OR
    1.04
    Keep
    1.03
    Act Density 0.069%

    No Known Activations