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

    Neuronpedia

    INDEX
    Explanations

    proper names, specifically the name "Thomas" and its variations

    oai_token-act-pair · gpt-3.5-turbo

    references to the name "Thomas."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    raq
    -0.73
    âĶĢâĶĢâĶĢâĶĢ
    -0.64
    bull
    -0.64
     arena
    -0.62
     forums
    -0.61
     eye
    -0.60
     grill
    -0.59
    rooms
    -0.59
     shaft
    -0.59
     inbox
    -0.59
    POSITIVE LOGITS
    Thomas
    3.33
     Thomas
    1.78
     Willis
    1.09
     Francois
    1.06
    Tom
    1.05
    Thompson
    1.01
    Thom
    1.00
     Ludwig
    0.97
     Amos
    0.96
    William
    0.95
    Act Density 0.016%

    No Known Activations