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

    Neuronpedia

    INDEX
    Explanations

    phrases related to actions of leaving or exiting

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to exits or departures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ENN
    -0.65
     densely
    -0.63
    anyl
    -0.62
     statically
    -0.62
     pert
    -0.61
     compuls
    -0.61
     raven
    -0.61
     trave
    -0.60
     arsen
    -0.60
    sonian
    -0.60
    POSITIVE LOGITS
    eering
    0.71
     Moreno
    0.71
     Clause
    0.70
    ees
    0.66
    eers
    0.65
    hyde
    0.64
    lycer
    0.64
    ous
    0.63
    ttes
    0.62
     Elim
    0.62
    Act Density 0.060%

    No Known Activations