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

    Neuronpedia

    INDEX
    Explanations

    sentence structures linking different items or events

    oai_token-act-pair · gpt-3.5-turbo

    phrases that suggest conditional situations or alternatives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ertodd
    -0.77
    atform
    -0.71
    rity
    -0.71
    ords
    -0.68
    icable
    -0.66
    amate
    -0.63
    erd
    -0.63
    isconsin
    -0.62
    auga
    -0.61
    SHIP
    -0.59
    POSITIVE LOGITS
     alternatively
    1.32
     equival
    1.10
     worse
    1.01
    thodox
    0.92
    Else
    0.89
     altern
    0.88
     equivalent
    0.88
     Worse
    0.78
     better
    0.76
     equivalents
    0.74
    Act Density 0.049%

    No Known Activations