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

    Neuronpedia

    INDEX
    Explanations

    location or being

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Though
    -1.70
     Though
    -1.45
    though
    -1.41
     though
    -1.34
     yearly
    -1.30
     anybody
    -1.29
     somebody
    -1.23
     everybody
    -1.18
    Anybody
    -1.16
     maneras
    -1.16
    POSITIVE LOGITS
    calaureate
    1.24
     insbesondere
    1.07
     velmi
    1.02
     ​​
    0.98
     sublime
    0.98
     XNUMX
    0.96
     Ditto
    0.96
     FINALLY
    0.95
     donuts
    0.92
     затем
    0.92
    Act Density 0.056%

    No Known Activations