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

    Neuronpedia

    INDEX
    Explanations

    low risk, unlikely, rare

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     nejen
    0.54
     måste
    0.49
     leider
    0.48
     teilweise
    0.47
    一定要
    0.47
     BOTH
    0.46
    必须要
    0.46
     właśnie
    0.44
     нередко
    0.42
     반드시
    0.42
    POSITIVE LOGITS
     fortunately
    1.15
     thankfully
    1.12
     unlikely
    1.11
     luckily
    1.01
     negligible
    1.00
     harmless
    0.98
    Thankfully
    0.98
    Fortunately
    0.97
     Fortunately
    0.96
     Thankfully
    0.91
    Act Density 0.053%

    No Known Activations