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

    Neuronpedia

    INDEX
    Explanations

    words and phrases indicating outcomes or results

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.02
    1:0.02
    2:0.06
    3:0.06
    4:0.17
    5:0.02
    6:0.06
    7:0.35
    8:0.03
    9:0.03
    10:0.07
    11:0.04
    Negative Logits
     Consent
    -1.75
    notes
    -1.61
     Pledge
    -1.60
     remembrance
    -1.60
     AES
    -1.58
     Hash
    -1.56
    mem
    -1.54
    ネ
    -1.53
     Memories
    -1.53
     Password
    -1.52
    POSITIVE LOGITS
     unfair
    1.73
     cheaper
    1.70
    ophobic
    1.68
     gloom
    1.67
     smoother
    1.64
    ONSORED
    1.58
     absurdity
    1.56
     worse
    1.56
     unprepared
    1.52
     inefficient
    1.50
    Act Density 0.001%

    No Known Activations