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

    Neuronpedia

    INDEX
    Explanations

    sounds like obvious

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Confusion
    0.43
     confusion
    0.40
    Confusion
    0.38
    特色
    0.36
     ясно
    0.36
     Visible
    0.35
     visible
    0.35
    فق
    0.35
     confusión
    0.35
     consistently
    0.34
    POSITIVE LOGITS
     silly
    1.05
     nerdy
    1.03
     cynical
    1.02
     cliché
    1.02
     cliche
    0.99
     cheesy
    0.97
     admittedly
    0.96
     simplistic
    0.95
     unorthodox
    0.93
     controversial
    0.93
    Act Density 0.080%

    No Known Activations