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

    Neuronpedia

    INDEX
    Explanations

    phrases expressing equality, fairness, or regardless of a specific condition

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing unconditionality and inclusivity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    itia
    -0.82
    oshenko
    -0.75
    ²
    -0.71
    ãĤ¼ãĤ¦ãĤ¹
    -0.71
    efficients
    -0.70
    ©¶æ¥µ
    -0.70
    jan
    -0.70
    atis
    -0.69
    raq
    -0.69
    ldon
    -0.68
    POSITIVE LOGITS
     whether
    1.07
     circumstance
    0.90
     affiliation
    0.88
     whatsoever
    0.86
     geography
    0.86
     differing
    0.85
     nationality
    0.85
     severity
    0.80
     circumstances
    0.80
     specifics
    0.80
    Act Density 0.048%

    No Known Activations