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

    Neuronpedia

    INDEX
    Explanations

    sensitive topics and requests

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    s
    0.74
     in
    0.64
    在
    0.63
    tn
    0.60
    ání
    0.59
    jší
    0.59
     appetizing
    0.59
    tan
    0.56
    तील
    0.56
    是
    0.56
    POSITIVE LOGITS
    il
    0.85
    ר
    0.83
    us
    0.75
    й
    0.75
    י
    0.72
    ু
    0.72
    Sensitive
    0.69
     sensitive
    0.68
    ג
    0.68
    ol
    0.67
    Act Density 0.012%

    No Known Activations