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

    Neuronpedia

    INDEX
    Explanations

    phrases expressing comfort and discomfort in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.02
    1:0.00
    2:0.07
    3:0.05
    4:0.13
    5:0.03
    6:0.03
    7:0.45
    8:0.01
    9:0.03
    10:0.07
    11:0.06
    Negative Logits
    ennial
    -1.64
    acio
    -1.60
    aceutical
    -1.55
    lass
    -1.53
    phabet
    -1.51
    govtrack
    -1.50
    ials
    -1.47
    idal
    -1.46
    ritic
    -1.45
    ocus
    -1.43
    POSITIVE LOGITS
     admitting
    1.78
     knowing
    1.67
     comfortable
    1.55
    Cub
    1.51
     sharing
    1.47
     snug
    1.45
     trusting
    1.42
     admit
    1.39
     comfortably
    1.38
     improv
    1.38
    Act Density 0.008%

    No Known Activations