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

    Neuronpedia

    INDEX
    Explanations

    permanent loss or damage

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    。
    0.96
    d
    0.88
    a
    0.87
    1
    0.84
     as
    0.79
    t
    0.76
    ።
    0.74
    ).
    0.73
    ।
    0.73
    ü
    0.73
    POSITIVE LOGITS
     everlasting
    0.93
     Permanent
    0.88
     Perman
    0.78
    ни
    0.78
     permanently
    0.77
    permanent
    0.75
    Perman
    0.75
    永久
    0.75
     permanent
    0.74
    น
    0.73
    Act Density 0.006%

    No Known Activations