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

    Neuronpedia

    INDEX
    Explanations

    phrases related to pairs or sets of items

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "pair."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ashtra
    -0.84
    ubs
    -0.75
    hes
    -0.73
    emetery
    -0.73
    ration
    -0.72
    ablishment
    -0.71
    âĸ¬
    -0.68
    ombie
    -0.68
    aband
    -0.67
    ructure
    -0.66
    POSITIVE LOGITS
     thirds
    0.81
     scissors
    0.80
     halves
    0.76
     shoes
    0.75
     twins
    0.72
    footed
    0.72
     glasses
    0.70
     sneakers
    0.69
     legs
    0.69
     jeans
    0.66
    Act Density 0.049%

    No Known Activations