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

    Neuronpedia

    INDEX
    Explanations

    sentences describing situations or events where things are going well or fine

    oai_token-act-pair · gpt-3.5-turbo

    positive sentiments and well-being indicators

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     incest
    -0.74
     Sacrifice
    -0.71
     jealousy
    -0.70
     scorn
    -0.70
     betrayal
    -0.70
     hypocrisy
    -0.69
     stakes
    -0.68
     Spons
    -0.66
     Vide
    -0.65
     Insp
    -0.65
    POSITIVE LOGITS
     alright
    1.23
     satisfactory
    1.16
     stabilized
    1.13
     calmed
    1.11
     safely
    1.07
     okay
    1.04
     healed
    1.03
     ok
    1.02
     satisf
    1.02
    stable
    1.01
    Act Density 0.846%

    No Known Activations