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

    Neuronpedia

    INDEX
    Explanations

    connecting or emphasizing points

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     или
    0.51
    然而
    0.50
    However
    0.48
     però
    0.48
     HOWEVER
    0.48
    however
    0.45
    但他
    0.45
     however
    0.44
     અથવા
    0.44
     однако
    0.43
    POSITIVE LOGITS
     importantly
    0.77
     sebagainya
    0.54
     whatnot
    0.54
     culminates
    0.54
     juga
    0.53
     crucially
    0.50
     auch
    0.50
     enjoys
    0.50
     provides
    0.49
     myös
    0.49
    Act Density 0.056%

    No Known Activations