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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "admit" and its various forms, indicating themes of acknowledgment or confession

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    blo
    -0.16
    abouts
    -0.16
    /Gate
    -0.16
    erness
    -0.15
    ìĶ
    -0.14
    olding
    -0.14
    lei
    -0.14
    ¶Į
    -0.14
    ÑĢÑĥÑĩ
    -0.13
    ackson
    -0.13
    POSITIVE LOGITS
     defeat
    0.30
     freely
    0.21
    ting
    0.21
     responsibility
    0.20
     defeats
    0.18
     feeling
    0.18
    ably
    0.18
     readily
    0.17
     defeated
    0.17
    ance
    0.17
    Act Density 0.029%

    No Known Activations