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

    Neuronpedia

    INDEX
    Explanations

    project details

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     accused
    0.53
     characterised
    0.45
     threshold
    0.45
     heifer
    0.44
     accusing
    0.42
     alder
    0.42
     lounge
    0.42
     boyfriend
    0.42
     guilty
    0.42
     Guild
    0.41
    POSITIVE LOGITS
     глоба
    0.46
    Aplic
    0.45
     аспек
    0.43
     Сасик
    0.43
    aufgaben
    0.43
    𒉡
    0.43
    bben
    0.43
    개
    0.43
    timevals
    0.43
    Glob
    0.42
    Act Density 0.003%

    No Known Activations