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

    Neuronpedia

    INDEX
    Explanations

    phrases involving contrast or balance in actions or situations

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing continuity or persistence in actions or states

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    uala
    -0.68
    eah
    -0.63
     Bre
    -0.63
    uca
    -0.62
    hack
    -0.62
     Roh
    -0.62
    resa
    -0.61
    onica
    -0.61
    éĹĺ
    -0.60
    rament
    -0.59
    POSITIVE LOGITS
     otherwise
    0.84
     contrasted
    0.82
     simultaneously
    0.79
     ignoring
    0.75
     nonetheless
    0.75
     thereby
    0.73
     forgetting
    0.72
     nevertheless
    0.72
     whereas
    0.72
     unlike
    0.70
    Act Density 0.540%

    No Known Activations