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

    Neuronpedia

    INDEX
    Explanations

    dates specified as "Mar" followed by a numerical day indicating the 9th or 10th day

    oai_token-act-pair · gpt-3.5-turbo

    date references

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ħĭ
    -0.80
    anwhile
    -0.73
    culosis
    -0.70
    mble
    -0.70
     tray
    -0.66
     ours
    -0.66
     Chosen
    -0.65
    ¥ŀ
    -0.64
    代
    -0.63
     mole
    -0.61
    POSITIVE LOGITS
    Mar
    1.27
    riage
    1.14
    ried
    1.08
    athon
    1.01
    itime
    0.95
    ston
    0.93
    ital
    0.87
    cipled
    0.87
    onso
    0.86
    ette
    0.86
    Act Density 0.007%

    No Known Activations