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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating future actions or states

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "be" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    osate
    -0.69
     Genius
    -0.67
     spectrum
    -0.66
     procedure
    -0.64
     residue
    -0.64
     distinction
    -0.64
     relativity
    -0.63
     process
    -0.63
     Outer
    -0.63
     constellation
    -0.62
    POSITIVE LOGITS
     able
    1.35
     rewarded
    1.01
     thankful
    0.99
     glad
    0.98
    friends
    0.97
     grateful
    0.95
     surprised
    0.95
     tempted
    0.94
     pleased
    0.94
    heading
    0.93
    Act Density 0.217%

    No Known Activations