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

    Neuronpedia

    INDEX
    Explanations

    beyond the scope of a

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     AFTER
    0.51
     כמו
    0.51
     مثل
    0.42
    एगा
    0.41
    idane
    0.41
     WITH
    0.40
    AFTER
    0.40
     विस्त
    0.40
     attn
    0.40
     형식
    0.39
    POSITIVE LOGITS
     de
    0.39
    每次
    0.39
     écrite
    0.39
     polite
    0.39
     fácilmente
    0.39
     아니고
    0.38
    如果不
    0.38
     eivät
    0.36
     entre
    0.36
    ឃើញ
    0.36
    Act Density 0.002%

    No Known Activations