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

    Neuronpedia

    INDEX
    Explanations

    phrases related to compensating for something or filling a gap

    oai_token-act-pair · gpt-3.5-turbo

    phrases involving the concept of "making up" or compensating for something

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    luaj
    -0.72
    Corp
    -0.72
    atis
    -0.71
    ridor
    -0.68
    akings
    -0.68
    iw
    -0.68
     DOI
    -0.66
    erb
    -0.65
    ér
    -0.64
    hens
    -0.63
    POSITIVE LOGITS
     excuses
    0.68
     ground
    0.66
     minds
    0.64
    ulates
    0.64
     thirds
    0.64
    clusions
    0.64
    swing
    0.63
     sided
    0.62
    stories
    0.62
    ulate
    0.59
    Act Density 0.023%

    No Known Activations