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

    Neuronpedia

    INDEX
    Explanations

    adjectives that describe surprising events or actions

    oai_token-act-pair · gpt-3.5-turbo

    intensifiers that convey strong emotional responses or reactions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ateurs
    -0.99
    hops
    -0.93
    lees
    -0.88
    agents
    -0.87
    assies
    -0.87
    ometers
    -0.87
    hers
    -0.85
    rants
    -0.84
    units
    -0.84
    asters
    -0.84
    POSITIVE LOGITS
     combination
    1.00
     assortment
    0.97
     outbreak
    0.95
     array
    0.94
     tale
    0.91
     mixture
    0.91
     piece
    0.91
     swath
    0.91
     chunk
    0.89
     flurry
    0.84
    Act Density 0.346%

    No Known Activations