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

    Neuronpedia

    INDEX
    Explanations

    signed in, to overwrite

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     recognises
    0.46
    ived
    0.42
    opera
    0.40
    rational
    0.40
     recognising
    0.39
    欢迎
    0.38
    rolo
    0.38
    豁
    0.37
    icono
    0.37
    vered
    0.36
    POSITIVE LOGITS
    closePath
    0.44
     poderia
    0.40
     könnte
    0.37
    candy
    0.37
    司
    0.37
     podría
    0.36
     ਕਿ
    0.35
    Candy
    0.35
     candy
    0.35
     erzählen
    0.35
    Act Density 0.000%

    No Known Activations