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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating small or limited quantities

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating small quantities or minorities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    lees
    -0.75
    ographs
    -0.63
    anon
    -0.62
    uploads
    -0.62
    projects
    -0.62
    cv
    -0.61
    attacks
    -0.59
    mos
    -0.58
    sylvania
    -0.58
     assemblies
    -0.58
    POSITIVE LOGITS
     handful
    1.31
     few
    1.20
     fraction
    1.10
     couple
    1.04
    few
    1.04
     glimpse
    0.98
     dozen
    0.95
     fleeting
    0.92
     hint
    0.90
    rouse
    0.89
    Act Density 0.118%

    No Known Activations