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

    Neuronpedia

    INDEX
    Explanations

    continuing or completing thoughts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ard
    0.45
    ip
    0.41
    rain
    0.40
    inline
    0.39
    illa
    0.39
    an
    0.38
    ear
    0.38
    raul
    0.38
    ush
    0.38
    static
    0.38
    POSITIVE LOGITS
     CONTINUE
    0.43
    始まる
    0.42
     “…
    0.41
     ANYTHING
    0.41
    埧
    0.40
     allerlei
    0.40
    塡
    0.40
     THEIR
    0.39
     शुरु
    0.39
    各种
    0.39
    Act Density 0.075%

    No Known Activations