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

    Neuronpedia

    INDEX
    Explanations

    phrases related to transactions or actions involving the exchange of goods or services

    oai_token-act-pair · gpt-3.5-turbo

    conjunctions used in a context involving multiple actions or scenarios

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    confirmed
    -0.67
     Introduced
    -0.66
    Enhanced
    -0.64
    prim
    -0.64
    insula
    -0.62
    released
    -0.62
    icated
    -0.61
     Corpus
    -0.60
    script
    -0.60
    particularly
    -0.59
    POSITIVE LOGITS
     then
    1.38
     suddenly
    1.29
     THEN
    1.29
     nobody
    1.13
     pretend
    1.13
     forget
    1.05
     magically
    1.04
     wonder
    1.01
     expecting
    1.00
     somehow
    0.97
    Act Density 0.257%

    No Known Activations