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

    Neuronpedia

    INDEX
    Explanations

    expressions related to evaluation and performance metrics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.03
    1:0.02
    2:0.38
    3:0.06
    4:0.10
    5:0.03
    6:0.06
    7:0.07
    8:0.05
    9:0.03
    10:0.05
    11:0.06
    Negative Logits
     stip
    -1.81
     deduction
    -1.58
    issions
    -1.56
     administ
    -1.55
     reinstated
    -1.52
     rebate
    -1.48
     exchange
    -1.43
     rule
    -1.40
     sake
    -1.39
     Regist
    -1.39
    POSITIVE LOGITS
    TextColor
    1.68
    jew
    1.68
    lite
    1.66
    achy
    1.59
    tan
    1.47
    blers
    1.46
    iren
    1.45
    Pie
    1.44
    bones
    1.44
    Haunted
    1.43
    Act Density 0.109%

    No Known Activations