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

    Neuronpedia

    INDEX
    Explanations

    past thoughts and possibilities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    indeed
    0.61
    Indeed
    0.58
    确实
    0.55
     indeed
    0.55
     nejen
    0.53
     Indeed
    0.49
    確實
    0.46
    往往
    0.42
    并非
    0.41
    嫩
    0.41
    POSITIVE LOGITS
     বুঝি
    0.61
     might
    0.60
    Might
    0.50
    might
    0.48
     Might
    0.48
     too
    0.46
     could
    0.46
     poderia
    0.45
     podría
    0.43
     pourrait
    0.43
    Act Density 0.008%

    No Known Activations