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

    Neuronpedia

    INDEX
    Explanations

    development and improvement

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     can
    0.46
     procedente
    0.45
     case
    0.44
     hotel
    0.44
     reproduced
    0.44
     explained
    0.43
     differentiated
    0.42
     republished
    0.42
     resides
    0.41
     computed
    0.41
    POSITIVE LOGITS
    Drain
    0.54
    Пу
    0.52
    П
    0.51
    𝔬
    0.51
    Описание
    0.49
    必ず
    0.49
    И
    0.49
    Б
    0.47
    顔
    0.46
    С
    0.44
    Act Density 0.001%

    No Known Activations