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

    Neuronpedia

    INDEX
    Explanations

    URLs with identifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     unbiased
    0.44
     overclock
    0.44
     conformal
    0.44
     sloping
    0.43
     crippled
    0.43
     reprogram
    0.42
     ordinal
    0.41
     dormant
    0.40
     heuristics
    0.40
     droplet
    0.40
    POSITIVE LOGITS
    9
    0.79
    7
    0.77
    8
    0.77
    6
    0.76
    4
    0.67
    5
    0.65
    3
    0.60
    2
    0.54
    1
    0.53
    0
    0.52
    Act Density 0.167%

    No Known Activations