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

    Neuronpedia

    INDEX
    Explanations

    a + descriptive word

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    overlap
    0.49
     overlap
    0.47
    рованная
    0.47
     combination
    0.46
     emphasis
    0.44
     акку
    0.43
    elow
    0.42
     kuin
    0.40
     either
    0.39
    чок
    0.38
    POSITIVE LOGITS
     otherwise
    0.65
     very
    0.57
     admittedly
    0.54
    otherwise
    0.54
    行業
    0.53
     sonst
    0.52
     troubled
    0.49
     autrement
    0.47
    very
    0.47
     अन्यथा
    0.47
    Act Density 0.010%

    No Known Activations