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

    Neuronpedia

    INDEX
    Explanations

    ellipses at the end of a sentence

    oai_token-act-pair · gpt-3.5-turbo

    ellipses or text indicating a pause or trailing off in thought

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     oun
    -0.70
     greens
    -0.69
     attest
    -0.67
     merit
    -0.67
    ãĥ£
    -0.66
     complement
    -0.66
     strength
    -0.64
     cradle
    -0.64
     wielded
    -0.64
     replacement
    -0.63
    POSITIVE LOGITS
    why
    1.01
    BUT
    0.95
    sort
    0.93
    they
    0.93
    oops
    0.91
    wow
    0.89
    maybe
    0.88
    interesting
    0.87
    uh
    0.87
    yeah
    0.87
    Act Density 0.043%

    No Known Activations