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

    Neuronpedia

    INDEX
    Explanations

    references to cultural or geographical terms and concepts

    oai_token-act-pair · gpt-3.5-turbo

    references to various ethnic groups, particularly those with "arian" in their names

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     touching
    -0.63
     Multi
    -0.62
    Wo
    -0.61
     perf
    -0.61
     foul
    -0.60
     lasting
    -0.59
     handles
    -0.59
     wond
    -0.59
     rout
    -0.58
     disabled
    -0.58
    POSITIVE LOGITS
    arian
    4.96
    arians
    3.84
    aria
    1.94
    arius
    1.74
    ary
    1.55
    arial
    1.52
    arist
    1.51
    ari
    1.49
    aries
    1.49
    ariat
    1.45
    Act Density 0.009%

    No Known Activations