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

    Neuronpedia

    INDEX
    Explanations

    assurances, guarantees, and promises given by a speaker to the audience

    oai_token-act-pair · gpt-3.5-turbo

    expressions of promises and assurances

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     awkwardly
    -0.80
     puzzling
    -0.72
     curious
    -0.66
    NetMessage
    -0.65
     contrasting
    -0.65
    pmwiki
    -0.62
     confused
    -0.61
     cumbers
    -0.60
     desperation
    -0.59
    arkin
    -0.59
    POSITIVE LOGITS
     wont
    0.89
     nothing
    0.88
     won
    0.86
    nothing
    0.86
     WILL
    0.85
     NEVER
    0.82
     never
    0.81
     none
    0.81
     anytime
    0.81
     confidentiality
    0.80
    Act Density 0.401%

    No Known Activations