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

    Neuronpedia

    INDEX
    Explanations

    prevention of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    は
    1.83
    는
    1.73
    是
    1.55
    ুল
    1.30
    い
    1.23
     in
    1.20
     as
    1.16
    ਰ
    1.11
    กัน
    1.10
    な
    1.06
    POSITIVE LOGITS
    ad
    1.15
    us
    1.06
    лло
    1.02
    на
    0.96
    ہ
    0.95
    тна
    0.93
    ت
    0.93
    но
    0.92
    ोत्तर
    0.91
    л
    0.91
    Act Density 0.022%

    No Known Activations