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

    Neuronpedia

    INDEX
    Explanations

    numerical values

    oai_token-act-pair · gpt-3.5-turbo

    the recurring number "28" in various contexts throughout the document

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    reads
    -0.85
    amaru
    -0.82
    achu
    -0.80
    ĸļ
    -0.77
    allo
    -0.76
    algia
    -0.76
    oms
    -0.75
    ensibly
    -0.74
    rophe
    -0.73
    ãĤ¤ãĥĪ
    -0.72
    POSITIVE LOGITS
    00
    1.02
    nm
    0.86
    th
    0.86
    nd
    0.81
    80
    0.77
    mins
    0.72
    iology
    0.71
    mph
    0.71
    CHAT
    0.70
    50
    0.66
    Act Density 0.033%

    No Known Activations