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

    Neuronpedia

    INDEX
    Explanations

    commands or instructions related to computer programming

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "bat" and its variations in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    sequences containing 'bat', especially in place names or technical file extensions.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @andreas

    The neuron activates strongly for the token "bat" in various contexts.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @andreas
    New Auto-Interp
    Negative Logits
     Matter
    -0.65
     intern
    -0.62
    icer
    -0.61
     Arrest
    -0.58
     juices
    -0.57
     Serve
    -0.57
     Course
    -0.55
     Pace
    -0.55
     garn
    -0.55
     Intercept
    -0.55
    POSITIVE LOGITS
    tenance
    0.89
    DragonMagazine
    0.83
    tered
    0.81
    raits
    0.80
    itude
    0.80
    elligence
    0.79
    zsche
    0.79
    rical
    0.79
    chers
    0.77
    unia
    0.77
    Act Density 0.028%

    No Known Activations