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

    Neuronpedia

    INDEX
    Explanations

    endings of sentences

    oai_token-act-pair · gpt-3.5-turbo

    sentences or phrases that emphasize the conclusion or final point of a discussion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ascus
    -0.79
     satisf
    -0.74
    thur
    -0.73
     deity
    -0.73
     messed
    -0.73
     derog
    -0.72
    bul
    -0.71
     fucked
    -0.70
    ishment
    -0.68
     maniac
    -0.68
    POSITIVE LOGITS
     Whereas
    1.38
     Firstly
    1.25
     According
    1.21
     Specifically
    1.17
     Though
    1.16
     While
    1.15
     Earlier
    1.14
     Particularly
    1.14
     Unlike
    1.12
     Consider
    1.12
    Act Density 0.514%

    No Known Activations