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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences within brackets

    oai_token-act-pair · gpt-3.5-turbo

    instances of brackets or symbols indicating segmentation or grouping

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     cones
    -0.71
     intellig
    -0.71
     equivalents
    -0.71
     seams
    -0.70
     poisoning
    -0.69
     stagger
    -0.69
     values
    -0.68
    ores
    -0.67
     piping
    -0.67
     transported
    -0.67
    POSITIVE LOGITS
    ...]
    1.55
    âĢ¦]
    1.54
    Pg
    1.29
    !]
    1.22
    ](
    1.07
    AUT
    1.06
    actionDate
    1.05
    ?]
    1.04
    paragraph
    1.03
    REDACTED
    0.98
    Act Density 0.027%

    No Known Activations