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

    Neuronpedia

    INDEX
    Explanations

    giveaway winner announcements

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens in giveaway or contest winner‐announcement phrases (e.g. “winner will be announced on…”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     experimentation
    -0.06
    Nano
    -0.06
     Lots
    -0.06
    xing
    -0.06
    286
    -0.06
    YGON
    -0.06
    افت
    -0.06
    egas
    -0.06
     円
    -0.06
    ;'
    -0.06
    POSITIVE LOGITS
    //---------------------------------------------------------------------------↵↵
    0.07
    [P
    0.07
     ADD
    0.06
    -то
    0.06
     espect
    0.06
    κρα
    0.06
     дотрим
    0.06
     redu
    0.06
     srov
    0.06
     erm
    0.06
    Act Density 0.008%

    No Known Activations