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

    Neuronpedia

    INDEX
    Explanations

    blocking and restricting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     или
    1.55
     or
    1.44
     hoặc
    1.37
     eller
    1.33
     أو
    1.31
    或
    1.30
    หรือ
    1.28
     หรือ
    1.22
     अथवा
    1.22
    或者
    1.21
    POSITIVE LOGITS
     only
    0.89
     "~
    0.88
    あくまで
    0.85
     yalnızca
    0.83
     stderr
    0.81
     ONLY
    0.80
     basically
    0.80
     suicides
    0.78
     something
    0.78
     lowercase
    0.78
    Act Density 0.819%

    No Known Activations