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

    Neuronpedia

    INDEX
    Explanations

    statements or points of emphasis that end in a period

    oai_token-act-pair · gpt-3.5-turbo

    sentences that conclude a thought or statement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     ogre
    -0.73
     mate
    -0.72
     footing
    -0.69
     aven
    -0.69
    ishment
    -0.68
     deity
    -0.68
     reversible
    -0.68
    bolt
    -0.67
    anus
    -0.66
    kee
    -0.64
    POSITIVE LOGITS
     Including
    1.24
     Examples
    1.18
     Universities
    1.11
     Their
    1.10
     These
    1.10
     Along
    1.09
     Those
    1.09
     Often
    1.08
     However
    1.08
     They
    1.07
    Act Density 0.744%

    No Known Activations