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

    Neuronpedia

    INDEX
    Explanations

    phrases where someone is expressing themselves or sharing their thoughts

    oai_token-act-pair · gpt-3.5-turbo

    statements or phrases indicating what individuals or groups have to express or communicate

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    mental
    -0.81
     conflic
    -0.72
    pes
    -0.72
    onut
    -0.72
    few
    -0.70
    ortunately
    -0.69
    sbm
    -0.67
    Tai
    -0.66
    tele
    -0.66
    Limited
    -0.64
    POSITIVE LOGITS
     goodbye
    1.32
     aloud
    0.82
     hello
    0.82
    iveness
    0.78
    erous
    0.77
     farewell
    0.76
    eness
    0.74
    anka
    0.73
    escription
    0.68
     Countdown
    0.67
    Act Density 0.032%

    No Known Activations