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

    Neuronpedia

    INDEX
    Explanations

    sentences ending with a period and potentially containing specific related keywords

    oai_token-act-pair · gpt-3.5-turbo

    punctuation marks, specifically periods and parentheses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     acron
    -0.72
     unbeat
    -0.66
     tyr
    -0.65
     grooming
    -0.64
     oak
    -0.62
     asses
    -0.61
     unstoppable
    -0.61
     touches
    -0.59
     pillar
    -0.59
     barg
    -0.59
    POSITIVE LOGITS
     Instead
    2.60
    Instead
    2.31
     Rather
    2.16
    Rather
    1.83
     Nor
    1.76
     Nonetheless
    1.65
     Nevertheless
    1.62
     instead
    1.53
    instead
    1.52
    Nevertheless
    1.51
    Act Density 0.647%

    No Known Activations