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

    Neuronpedia

    INDEX
    Explanations

    verbs related to intensive or urgent actions

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating actions or attempts made towards a goal

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     liking
    -0.83
     Philosophy
    -0.73
     Likes
    -0.72
     philosophies
    -0.69
     philos
    -0.69
     richer
    -0.68
     anecdotes
    -0.68
     whining
    -0.67
     graphs
    -0.67
     Improvement
    -0.67
    POSITIVE LOGITS
     apprehend
    1.56
     evacuate
    1.50
     rescue
    1.39
     arrest
    1.36
     evict
    1.34
     protect
    1.34
     prevent
    1.31
     locate
    1.29
     escort
    1.26
     detain
    1.26
    Act Density 0.284%

    No Known Activations