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

    Neuronpedia

    INDEX
    Explanations

    phrases related to instructing or providing guidance on various topics or tasks

    oai_token-act-pair · gpt-3.5-turbo

    phrases that start with "How to."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    lights
    -0.82
    enance
    -0.78
     Coff
    -0.78
     Appears
    -0.73
    court
    -0.67
     Nightmares
    -0.66
     Boone
    -0.65
    mails
    -0.64
    urance
    -0.64
    comments
    -0.63
    POSITIVE LOGITS
     navigate
    1.24
     prepare
    1.11
     evade
    1.11
     proceed
    1.11
     achieve
    1.11
     avoid
    1.09
     overcome
    1.08
     circumvent
    1.08
     organize
    1.07
     manage
    1.07
    Act Density 0.078%

    No Known Activations