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

    Neuronpedia

    INDEX
    Explanations

    phrases related to secrets, hidden information, or concealed objects

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "hidden" or "hiddenness."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    annis
    -0.86
    nia
    -0.84
    xual
    -0.81
     Pwr
    -0.79
    ctive
    -0.77
    SHIP
    -0.76
    ocations
    -0.74
    udeb
    -0.73
    ufact
    -0.73
    urses
    -0.73
    POSITIVE LOGITS
     hidden
    0.91
     Hidden
    0.84
     behind
    0.83
     recess
    0.82
    door
    0.78
     stash
    0.77
     hatch
    0.77
     gem
    0.76
     tunnel
    0.76
     treasure
    0.75
    Act Density 0.014%

    No Known Activations