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

    Neuronpedia

    INDEX
    Explanations

    contrasting negation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Ends
    0.44
    等方面
    0.37
    rarr
    0.37
    ers
    0.36
    hre
    0.35
    Basically
    0.34
    Essentially
    0.33
    *=
    0.33
    jasper
    0.33
    swers
    0.33
    POSITIVE LOGITS
     причем
    0.91
    真的
    0.70
     არა
    0.69
    というか
    0.68
     nejen
    0.68
     όχι
    0.67
     いや
    0.66
     Literally
    0.66
     vraiment
    0.64
     Seriously
    0.64
    Act Density 0.029%

    No Known Activations