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

    Neuronpedia

    INDEX
    Explanations

    references to structures or locations named "tower"

    oai_token-act-pair · gpt-3.5-turbo

    references to towers and various types of towers in general

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ãĤ±
    -0.84
    dit
    -0.80
    é»Ĵ
    -0.76
     Lives
    -0.73
     Taste
    -0.70
    matter
    -0.70
    sweet
    -0.69
    furt
    -0.69
    ãĤ¡
    -0.68
    âĸ¬âĸ¬
    -0.67
    POSITIVE LOGITS
     towers
    1.11
     tower
    1.06
     Towers
    0.99
     erected
    0.97
    tower
    0.94
     heights
    0.93
     skysc
    0.93
     crane
    0.89
    raper
    0.85
     blocks
    0.81
    Act Density 0.026%

    No Known Activations