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

    Neuronpedia

    INDEX
    Explanations

    rough estimates

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     correctamente
    0.54
     assuming
    0.53
    Assuming
    0.51
     correctly
    0.49
     corretamente
    0.49
     accurately
    0.47
    正确的
    0.46
     Assuming
    0.46
     정보를
    0.46
    assuming
    0.46
    POSITIVE LOGITS
     rough
    1.26
     Rough
    1.06
    rough
    1.02
    粗
    0.91
    Rough
    0.89
     roug
    0.80
     roughness
    0.74
     crude
    0.73
     educated
    0.67
     coarse
    0.66
    Act Density 0.032%

    No Known Activations