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

    Neuronpedia

    INDEX
    Explanations

    phrases containing the word "out" with some preceding context

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating an exit or departure

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     arsen
    -0.72
     tyr
    -0.71
     Brach
    -0.67
     0004
    -0.63
     compr
    -0.62
    avorite
    -0.61
    EStream
    -0.60
    inical
    -0.60
     Khe
    -0.59
     destro
    -0.58
    POSITIVE LOGITS
    stretched
    1.35
    lander
    1.21
    fitted
    1.15
    dated
    1.08
    flows
    1.07
    numbered
    1.06
    lier
    1.05
    smart
    1.05
    doors
    1.05
    landish
    1.03
    Act Density 0.052%

    No Known Activations