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

    Neuronpedia

    INDEX
    Explanations

    verbs related to actions or decisions being taken

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
    Cong
    -0.71
     accompanies
    -0.70
    ese
    -0.67
     dissatisf
    -0.64
     livest
    -0.62
    bie
    -0.61
    cape
    -0.59
    Friend
    -0.59
     Smile
    -0.58
    Bey
    -0.58
    POSITIVE LOGITS
     advantage
    1.23
    aways
    1.08
     heed
    0.97
     precautions
    0.95
     aback
    0.95
     care
    0.92
     refuge
    0.89
     responsibility
    0.86
     precaution
    0.84
     aim
    0.83
    Act Density 0.099%

    No Known Activations