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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating preparation or anticipation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.01
    1:0.02
    2:0.10
    3:0.10
    4:0.09
    5:0.01
    6:0.07
    7:0.33
    8:0.03
    9:0.02
    10:0.07
    11:0.08
    Negative Logits
    rers
    -1.76
    cause
    -1.71
    cro
    -1.66
    matter
    -1.61
    hack
    -1.60
    ationally
    -1.55
    alties
    -1.48
    iencies
    -1.45
    link
    -1.45
     DOI
    -1.44
    POSITIVE LOGITS
     rehears
    1.69
     farewell
    1.65
    paren
    1.61
     preparations
    1.56
    �
    1.53
     fuss
    1.50
    estern
    1.46
     prom
    1.45
     anticipation
    1.36
     Prepar
    1.36
    Act Density 0.005%

    No Known Activations