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

    Neuronpedia

    INDEX
    Explanations

    names with a specific pattern, where "oti" followed by a single digit

    oai_token-act-pair · gpt-3.5-turbo

    references to negotiations and related terms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    spin
    -0.75
    mom
    -0.70
    die
    -0.70
    desktop
    -0.66
    open
    -0.65
    eele
    -0.64
    cor
    -0.62
    spot
    -0.61
    riad
    -0.61
     sockets
    -0.61
    POSITIVE LOGITS
    oti
    1.21
     Ng
    0.90
    otiation
    0.88
    ations
    0.88
     renegoti
    0.87
    terness
    0.84
    ovan
    0.76
    ouk
    0.76
    zzo
    0.74
     negotiated
    0.73
    Act Density 0.006%

    No Known Activations