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

    Neuronpedia

    INDEX
    Explanations

    words related to groups or collections

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "ions."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    alogue
    -0.67
     rapport
    -0.63
     Tanz
    -0.61
    foundland
    -0.60
     referen
    -0.60
     req
    -0.60
     cutoff
    -0.60
     witnesses
    -0.59
     interven
    -0.58
     citiz
    -0.58
    POSITIVE LOGITS
    hower
    0.84
    bats
    0.81
    pit
    0.81
    undle
    0.81
    hoe
    0.80
    ike
    0.77
    eca
    0.77
    ova
    0.76
    IUM
    0.74
    ocation
    0.73
    Act Density 0.022%

    No Known Activations