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

    Neuronpedia

    INDEX
    Explanations

    rel- prefix words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    bursements
    0.42
    気を
    0.40
     লিমি
    0.40
    "".
    0.39
    ]]).
    0.39
    akul
    0.38
     Bedien
    0.38
    ittää
    0.38
    }}_{
    0.37
     большая
    0.37
    POSITIVE LOGITS
     rel
    0.87
     Rel
    0.86
    Rel
    0.83
     reli
    0.82
     REL
    0.75
     Reli
    0.71
     ril
    0.69
     relativistic
    0.67
    rel
    0.66
     relay
    0.65
    Act Density 0.037%

    No Known Activations