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

    Neuronpedia

    INDEX
    Explanations

    expressions of gratitude and requests

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express intent or requests

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    furt
    -0.76
     Nash
    -0.66
    grounds
    -0.66
    wise
    -0.63
     Abel
    -0.63
    bars
    -0.62
    raw
    -0.62
     Nile
    -0.62
    metadata
    -0.60
    cases
    -0.59
    POSITIVE LOGITS
     emulate
    1.06
     propose
    1.03
     encourage
    1.01
     recreate
    1.00
     participate
    0.97
     congratulate
    0.96
     clarify
    0.93
     capitalize
    0.92
     assure
    0.92
     postpone
    0.91
    Act Density 0.065%

    No Known Activations