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

    Neuronpedia

    INDEX
    Explanations

    references to the word "with" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.02
    1:0.01
    2:0.09
    3:0.06
    4:0.16
    5:0.03
    6:0.04
    7:0.30
    8:0.03
    9:0.04
    10:0.06
    11:0.09
    Negative Logits
    ibaba
    -2.15
    fast
    -1.58
    arching
    -1.47
    facebook
    -1.45
    qu
    -1.45
    arming
    -1.44
    coming
    -1.43
    asive
    -1.42
    isance
    -1.42
    qual
    -1.42
    POSITIVE LOGITS
     Tsukuyomi
    1.76
    actionDate
    1.72
     withd
    1.63
     taxp
    1.53
     linem
    1.52
     beforehand
    1.51
     burner
    1.51
     Abedin
    1.48
     politely
    1.48
     Hou
    1.46
    Act Density 0.002%

    No Known Activations