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

    Neuronpedia

    INDEX
    Explanations

    the word segment "unw" followed by a high-value activation word component

    oai_token-act-pair · gpt-3.5-turbo

    references to unwritten rules or social norms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    phrine
    -0.94
    ãĤ¼
    -0.84
    hyde
    -0.81
    uyomi
    -0.81
    senal
    -0.79
    pmwiki
    -0.77
     Defenders
    -0.73
    anwhile
    -0.72
     hemor
    -0.72
    姫
    -0.72
    POSITIVE LOGITS
    arranted
    1.05
    ield
    1.04
    ritten
    1.04
    ashed
    0.99
    irth
    0.99
    inding
    0.96
    ashington
    0.94
    atcher
    0.93
    itt
    0.93
    avering
    0.92
    Act Density 0.009%

    No Known Activations