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

    Neuronpedia

    INDEX
    Explanations

    references to audience reactions and interactions

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     Ide
    -0.86
     Plum
    -0.82
    erald
    -0.81
    empt
    -0.80
     Scot
    -0.78
     Athletics
    -0.76
    phrine
    -0.75
     Franch
    -0.75
     Ski
    -0.74
    grave
    -0.73
    POSITIVE LOGITS
    atics
    1.10
     audience
    0.99
    iences
    0.97
    ience
    0.94
    room
    0.94
    members
    0.92
    atically
    0.90
    ele
    0.89
    IENCE
    0.89
     member
    0.89
    Act Density 0.904%

    No Known Activations