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

    Neuronpedia

    INDEX
    Explanations

    words related to defending or justification

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "defended."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    seed
    -0.80
    aster
    -0.71
    KE
    -0.70
    oter
    -0.70
    amen
    -0.68
    mad
    -0.66
    ppa
    -0.65
    aida
    -0.64
    akers
    -0.63
    thin
    -0.63
    POSITIVE LOGITS
     defended
    1.00
     criticised
    0.76
     behavi
    0.76
     showc
    0.75
    nesday
    0.74
    folios
    0.73
    indal
    0.73
     defends
    0.73
     critic
    0.72
     conced
    0.71
    Act Density 0.019%

    No Known Activations