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

    Neuronpedia

    INDEX
    Explanations

    instances where actions or responsibilities are being taken seriously

    oai_token-act-pair · gpt-3.5-turbo

    references to responsibilities, concerns, and issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    etheless
    -0.80
    entimes
    -0.65
     accompanies
    -0.64
    furt
    -0.61
    los
    -0.60
    ITNESS
    -0.60
    izabeth
    -0.59
    whose
    -0.57
    DA
    -0.55
    nor
    -0.55
    POSITIVE LOGITS
     squarely
    0.91
     aside
    0.86
     down
    0.82
     into
    0.81
     overboard
    0.81
     seriously
    0.75
     intact
    0.74
     onto
    0.73
     firmly
    0.72
     forward
    0.72
    Act Density 0.204%

    No Known Activations