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

    Neuronpedia

    INDEX
    Explanations

    hidden origins or true nature

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     seem
    0.80
    据说
    0.77
    にならない
    0.76
    íž
    0.76
     seemingly
    0.76
     seems
    0.75
    seems
    0.75
     seemed
    0.72
     rappelle
    0.72
    fails
    0.72
    POSITIVE LOGITS
     hidden
    1.25
     genuine
    1.18
     ulterior
    1.13
     fraude
    1.11
     hoax
    1.10
     involvement
    1.10
     ocult
    1.09
    hidden
    1.09
     origin
    1.09
     origen
    1.08
    Act Density 0.252%

    No Known Activations