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

    Neuronpedia

    INDEX
    Explanations

    , based on the activations provided, it seems like this neuron is finding words indicating sequence or ordering, alongside a specific focus on numerical sequences

    oai_token-act-pair · gpt-3.5-turbo

    instances of punctuation, specifically commas

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     enthus
    -0.69
    Attach
    -0.65
    Deal
    -0.65
    Catalog
    -0.64
    MpServer
    -0.64
    Ult
    -0.62
    adden
    -0.62
    ForgeModLoader
    -0.61
    GF
    -0.61
    Cra
    -0.60
    POSITIVE LOGITS
     however
    0.83
     although
    0.78
     namely
    0.76
     though
    0.75
    000
    0.68
     according
    0.64
    hester
    0.64
     moreover
    0.64
     Shank
    0.63
     please
    0.63
    Act Density 0.106%

    No Known Activations