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

    Neuronpedia

    INDEX
    Explanations

    receipts, DuckDuckGo, four

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     hideous
    0.59
     horrific
    0.57
    มะ
    0.56
     horr
    0.54
     prostitute
    0.54
     terrifying
    0.52
     violent
    0.51
     brazen
    0.51
     abhor
    0.50
    レ
    0.50
    POSITIVE LOGITS
     Abbreviations
    0.55
     कार्य
    0.50
    #
    0.50
     nota
    0.50
    Component
    0.48
     Appendix
    0.48
    Inter
    0.48
     Subsidi
    0.48
    kins
    0.47
    Note
    0.47
    Act Density 0.000%

    No Known Activations