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

    Neuronpedia

    INDEX
    Explanations

    reworking, recast, rewrite, overworked

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ’
    -3.91
    \
    -3.75
    并未
    -2.34
    几个月
    -2.33
    两年
    -2.33
    "
    -2.19
    I
    -2.05
    ꦭ
    -2.03
     zusätzliche
    -2.00
    _{
    -1.98
    POSITIVE LOGITS
     elegance
    2.48
    劻
    2.31
    ll
    2.17
     jedną
    2.17
    媄
    2.16
     dritten
    2.09
    ÃO
    1.99
     ellos
    1.96
     +'
    1.95
    >);
    1.93
    Act Density 0.001%

    No Known Activations