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

    Neuronpedia

    INDEX
    Explanations

    phrases related to counting or ordering

    oai_token-act-pair · gpt-3.5-turbo

    references to quantities related to groups or units of three and two

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     srfAttach
    -0.64
    BIL
    -0.64
    joy
    -0.61
     Changed
    -0.61
    HY
    -0.60
    mbuds
    -0.60
    Scope
    -0.59
    still
    -0.59
    reply
    -0.59
    rist
    -0.59
    POSITIVE LOGITS
     digits
    1.03
     months
    1.03
     paragraphs
    1.01
     weeks
    0.99
     stages
    0.98
    teen
    0.93
     phases
    0.93
     iterations
    0.93
     chapters
    0.92
     thirds
    0.91
    Act Density 0.064%

    No Known Activations