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

    Neuronpedia

    INDEX
    Explanations

    holds, caught, hanging, or trailing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    불
    0.39
     Fat
    0.36
     تمر
    0.36
     burnt
    0.35
     같아요
    0.35
     Stre
    0.35
     Lutheran
    0.35
    Burns
    0.35
     불
    0.34
     Revol
    0.34
    POSITIVE LOGITS
     off
    0.50
    𒅖
    0.44
     onwards
    0.39
    segn
    0.39
     Джу
    0.39
    स्थ्य
    0.38
     plazas
    0.38
    rdquo
    0.38
     транспорта
    0.38
    िस्थित
    0.37
    Act Density 0.073%

    No Known Activations