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

    Neuronpedia

    INDEX
    Explanations

    mention of contrasts or comparisons between different entities or concepts

    oai_token-act-pair · gpt-3.5-turbo

    repeated conjunctions and phrases indicating abundance or accumulation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    scrib
    -0.76
    lear
    -0.70
    ounced
    -0.69
    fell
    -0.69
    cedes
    -0.66
    habi
    -0.65
    ounces
    -0.64
    ction
    -0.62
    itute
    -0.61
    byter
    -0.61
    POSITIVE LOGITS
     finally
    1.65
     ultimately
    1.31
     assorted
    1.09
     eventually
    1.08
     frankly
    1.07
     perhaps
    1.06
     consequently
    1.05
     occasionally
    1.01
     generally
    1.01
     even
    1.01
    Act Density 0.265%

    No Known Activations