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

    Neuronpedia

    INDEX
    Explanations

    building, transportation, war industry

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     daß
    0.68
     Owing
    0.68
     owing
    0.60
     endeavour
    0.59
     muß
    0.59
     practise
    0.58
     Notwithstanding
    0.58
     endeavoured
    0.58
    Owing
    0.57
    Notwithstanding
    0.57
    POSITIVE LOGITS
     shitty
    0.50
     দিলো
    0.50
     medications
    0.48
     গেলো
    0.48
    啊
    0.46
    ?!
    0.46
     ✨
    0.44
     चीज़
    0.44
     healthcare
    0.43
     啊
    0.43
    Act Density 0.001%

    No Known Activations