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

    Neuronpedia

    INDEX
    Explanations

    sentences or phrases that end with a particular symbol (Ċ)

    oai_token-act-pair · gpt-3.5-turbo

    instances of a strong emotional response or contemplation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     footing
    -0.81
    awei
    -0.80
    senal
    -0.69
     extrad
    -0.68
    ijing
    -0.67
    atism
    -0.62
    commun
    -0.61
    rad
    -0.60
    heit
    -0.60
     manoeuv
    -0.59
    POSITIVE LOGITS
    Anyway
    0.78
     Imran
    0.77
    Thankfully
    0.74
     Anyway
    0.71
     Franch
    0.70
    Furthermore
    0.66
    Luckily
    0.65
    Likewise
    0.65
     Thankfully
    0.65
    Plus
    0.65
    Act Density 0.353%

    No Known Activations