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

    Neuronpedia

    INDEX
    Explanations

    phrases or words related to direct connections or relationships between entities or concepts

    oai_token-act-pair · gpt-3.5-turbo

    phrases explicitly stating direct relationships

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    gerald
    -0.81
    glers
    -0.71
    mble
    -0.69
    ittal
    -0.66
     cautiously
    -0.66
     thoroughly
    -0.65
    Daily
    -0.64
     stal
    -0.62
     Daily
    -0.62
    ulton
    -0.61
    POSITIVE LOGITS
     contradicted
    0.96
     contradicts
    0.83
     contradict
    0.80
    ebted
    0.79
     impacted
    0.77
    forward
    0.77
     benefited
    0.77
     attributable
    0.74
     observable
    0.74
     implicated
    0.71
    Act Density 0.029%

    No Known Activations