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

    Neuronpedia

    INDEX
    Explanations

    short phrases that involve specific actions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the punctuation mark ',' (comma)

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Reward
    -0.64
    osc
    -0.63
    MAX
    -0.62
    Switch
    -0.62
    Stock
    -0.60
    int
    -0.60
    num
    -0.60
    grain
    -0.57
    untarily
    -0.57
    Availability
    -0.57
    POSITIVE LOGITS
     meanwhile
    1.35
     however
    1.35
     huh
    1.08
     moreover
    1.00
     unsurprisingly
    0.91
     albeit
    0.88
     alas
    0.87
     though
    0.85
     therefore
    0.82
     according
    0.82
    Act Density 0.569%

    No Known Activations