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

    Neuronpedia

    INDEX
    Explanations

    When looking through the activations provided, it appears this neuron is consistently finding occurrences of the word "take" in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "take a" followed by various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ndra
    -0.79
    ells
    -0.71
    Ü
    -0.69
     displays
    -0.68
    tions
    -0.66
    tu
    -0.65
    IAS
    -0.65
    IOR
    -0.64
    eller
    -0.64
    α
    -0.63
    POSITIVE LOGITS
     seriously
    0.91
     lightly
    0.84
     plunge
    0.83
     reins
    0.81
     cue
    0.78
     stride
    0.78
     tumble
    0.73
    ãĥīãĥ©
    0.70
     lesson
    0.67
     tack
    0.67
    Act Density 0.147%

    No Known Activations