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

    Neuronpedia

    INDEX
    Explanations

    negative contractions indicating denial or negation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    er
    -0.74
     Thales
    -0.74
    tır
    -0.73
     _('
    -0.73
     PACE
    -0.71
     alá
    -0.70
    Meyer
    -0.69
     Gibbs
    -0.68
     Weiss
    -0.67
    eze
    -0.67
    POSITIVE LOGITS
     shouldn
    1.05
     isn
    1.04
     wasn
    1.01
     mustn
    1.01
     couldn
    1.00
     Shouldn
    0.99
     hadn
    0.98
     didn
    0.97
     aren
    0.97
    __":
    0.95
    Act Density 0.075%

    No Known Activations