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

    Neuronpedia

    INDEX
    Explanations

    pronouns or names followed by verbs indicating a request or desire

    oai_token-act-pair · gpt-3.5-turbo

    references to people and their desires or expectations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    VIDIA
    -0.73
    everal
    -0.66
    ibe
    -0.65
    berra
    -0.61
    rium
    -0.60
    ggles
    -0.58
     guiActiveUn
    -0.58
    ationally
    -0.57
    Attempts
    -0.57
    riad
    -0.56
    POSITIVE LOGITS
     to
    1.08
     badly
    0.89
     deported
    0.85
     gone
    0.83
    to
    0.82
     cleaned
    0.80
     punished
    0.78
     TO
    0.77
     spoiled
    0.75
     eaten
    0.74
    Act Density 0.117%

    No Known Activations