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

    Neuronpedia

    INDEX
    Explanations

    geometric shapes, particularly circles, and related terms like 'circles' or 'circled'

    oai_token-act-pair · gpt-3.5-turbo

    references to circles and circular shapes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ufact
    -0.85
    enf
    -0.76
    eele
    -0.74
    ingham
    -0.72
    audi
    -0.71
    ono
    -0.70
    ERAL
    -0.70
    EG
    -0.69
    essee
    -0.69
    owell
    -0.68
    POSITIVE LOGITS
     shaped
    1.08
    jer
    1.02
     shape
    0.97
     circle
    0.94
     jerk
    0.91
     circling
    0.84
     circles
    0.81
    SHIP
    0.80
     circled
    0.78
    shaped
    0.77
    Act Density 0.056%

    No Known Activations