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

    Neuronpedia

    INDEX
    Explanations

    rare characters and artifacts in the text that may not be relevant in the context of the typical text processing tasks

    oai_token-act-pair · gpt-3.5-turbo

    special characters or symbols

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    idges
    -0.88
    iddles
    -0.82
    awar
    -0.76
    apping
    -0.75
     disadvant
    -0.74
    asonic
    -0.74
    asta
    -0.71
    ifax
    -0.69
    icky
    -0.69
    anooga
    -0.68
    POSITIVE LOGITS
    Ĺ
    1.01
    lishing
    0.92
    æĸ¹
    0.88
    ¤
    0.86
    ï¸ı
    0.85
    RAM
    0.84
    lish
    0.83
    é»Ĵ
    0.83
    ULAR
    0.82
    ãĤŃ
    0.81
    Act Density 0.018%

    No Known Activations