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

    Neuronpedia

    INDEX
    Explanations

    the word "secret."

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "secret."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ©¶æ
    -0.88
    thood
    -0.86
    annis
    -0.86
    ð
    -0.80
    gaard
    -0.79
    adjusted
    -0.78
    ulf
    -0.74
     Pwr
    -0.74
    FG
    -0.73
    older
    -0.73
    POSITIVE LOGITS
    arial
    1.02
    ariat
    1.01
     secret
    0.98
     ingredient
    0.91
    rets
    0.87
     underground
    0.85
     ballot
    0.83
     hidden
    0.82
     secrets
    0.80
     stash
    0.80
    Act Density 0.013%

    No Known Activations