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

    Neuronpedia

    INDEX
    Explanations

    words related to safety and fairness

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating safety, fairness, and reasonable assumptions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    rez
    -0.58
    Introduced
    -0.55
    unia
    -0.55
    otte
    -0.55
    reen
    -0.55
     ensu
    -0.53
    resp
    -0.49
     peacefully
    -0.49
    uko
    -0.48
     toile
    -0.48
    POSITIVE LOGITS
     conjecture
    0.86
     to
    0.84
     enough
    0.80
     speculation
    0.77
     theor
    0.74
     inference
    0.72
     conject
    0.71
     misconception
    0.70
     speculate
    0.69
    Reviewer
    0.67
    Act Density 0.117%

    No Known Activations