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

    Neuronpedia

    INDEX
    Explanations

    introduces specifications and details

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    𝐞
    0.86
    tedir
    0.86
    س
    0.81
    €™
    0.80
    NSDictionary
    0.77
    𝐚
    0.75
    ারের
    0.75
    cdots
    0.74
     ollut
    0.73
    য়ে
    0.72
    POSITIVE LOGITS
    ০
    1.50
    i
    1.30
    𝑛
    1.18
    ي
    1.16
    ی
    1.12
    ಿ
    1.11
    reuse
    1.09
    𝑡
    1.08
     upfront
    1.06
    ি
    1.05
    Act Density 0.398%

    No Known Activations