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

    Neuronpedia

    INDEX
    Explanations

    remains a challenge

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     remained
    0.82
    Remain
    0.82
     Remain
    0.78
     remain
    0.75
    remain
    0.71
     remains
    0.68
     răm
    0.68
     Remains
    0.64
     remaining
    0.63
     permanec
    0.62
    POSITIVE LOGITS
     elusive
    0.45
    霖
    0.45
     unimplemented
    0.42
     stubbornly
    0.42
     unfinished
    0.41
    っていない
    0.41
    šli
    0.41
     unfilled
    0.39
     stubborn
    0.38
    bridge
    0.38
    Act Density 0.007%

    No Known Activations