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

    Neuronpedia

    INDEX
    Explanations

    references to drafts, such as draft documents, draft reports, and draft guidelines

    oai_token-act-pair · gpt-3.5-turbo

    references to drafts of documents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    cles
    -0.73
    hips
    -0.72
     destro
    -0.71
     charism
    -0.69
    Sea
    -0.69
    eno
    -0.69
    rian
    -0.68
     fundament
    -0.68
    tics
    -0.68
    cellence
    -0.67
    POSITIVE LOGITS
    smanship
    1.61
     dod
    0.97
    Kings
    0.97
    lings
    0.72
     drafts
    0.71
     draft
    0.70
     dodging
    0.69
     muster
    0.67
    lake
    0.66
    ographical
    0.64
    Act Density 0.027%

    No Known Activations