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

    Neuronpedia

    INDEX
    Explanations

    personal statements or declarations of opinions

    oai_token-act-pair · gpt-3.5-turbo

    first-person and third-person pronouns and their associated actions or states

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     newcom
    -0.68
    £ı
    -0.68
     Mechan
    -0.64
     Flavoring
    -0.62
    aic
    -0.62
     Grail
    -0.62
    ablishment
    -0.62
     conclud
    -0.60
    summary
    -0.58
     Various
    -0.57
    POSITIVE LOGITS
     shouldn
    1.50
     should
    1.24
     couldn
    1.23
     ought
    1.19
     needed
    1.16
     deserved
    1.15
     wouldn
    1.12
     SHOULD
    1.11
    'd
    1.11
     must
    1.09
    Act Density 0.160%

    No Known Activations