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

    Neuronpedia

    INDEX
    Explanations

    whitespace characters or separators in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    the beginning of a document or text (specifically looking at the <bos> token which has high activation values).

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    New Auto-Interp
    Negative Logits
    localctx
    -0.66
    RegistryLite
    -0.64
    Попис
    -0.63
    awtextra
    -0.60
     ObjectParameter
    -0.59
    Portale
    -0.59
     "..\..\..\
    -0.57
    ServletConfig
    -0.57
    ))^{
    -0.56
    elemField
    -0.56
    POSITIVE LOGITS
     reaſon
    0.77
     pleaſure
    0.72
     purpoſe
    0.72
     douceur
    0.66
     parlant
    0.65
     étrangères
    0.64
     uſe
    0.64
     prêtres
    0.63
     sintomi
    0.63
     chrétiens
    0.63
    Act Density 0.403%

    No Known Activations