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

    Neuronpedia

    INDEX
    Explanations

    non-standard non-alphanumeric characters, specifically those representing actions or feelings

    oai_token-act-pair · gpt-3.5-turbo

    repeated instances of a specific character sequence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Skydragon
    -0.58
     Samar
    -0.57
     Mash
    -0.56
     Gaul
    -0.56
     Mobil
    -0.55
     fragmentation
    -0.54
     Azerb
    -0.53
     Wagner
    -0.53
     Rhodes
    -0.52
     mathemat
    -0.52
    POSITIVE LOGITS
    s
    0.98
    sure
    0.86
    t
    0.86
    ï¸ı
    0.79
    ll
    0.78
    ski
    0.78
    tis
    0.78
    sent
    0.77
    ve
    0.76
    scl
    0.75
    Act Density 0.293%

    No Known Activations