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

    Neuronpedia

    INDEX
    Explanations

    relative vs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    什么
    2.28
    م
    1.75
    かわいい
    1.70
     chick
    1.69
    Стра
    1.69
    वर्ड
    1.61
     roll
    1.60
     Somos
    1.59
     Soma
    1.59
     soulful
    1.59
    POSITIVE LOGITS
    mt
    2.19
    eng
    2.14
    models
    2.08
    passwd
    2.05
    matrices
    2.03
    md
    2.02
    missions
    1.98
    maps
    1.97
     ilişk
    1.95
    leq
    1.95
    Act Density 0.136%

    No Known Activations