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

    Neuronpedia

    INDEX
    Explanations

    sentences indicating a strong opinion or criticism

    oai_token-act-pair · gpt-3.5-turbo

    sentences that end with a period, indicating completed thoughts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ishable
    -0.83
    isphere
    -0.83
     jugg
    -0.77
     disemb
    -0.76
     vag
    -0.74
     sleeper
    -0.73
     purse
    -0.72
     immortal
    -0.70
     gaze
    -0.70
     tyr
    -0.69
    POSITIVE LOGITS
     However
    1.32
     Instead
    1.27
     Nevertheless
    1.25
     Moreover
    1.23
     Nonetheless
    1.23
     Additionally
    1.22
     Though
    1.20
     Asked
    1.20
     Furthermore
    1.20
     Specifically
    1.19
    Act Density 0.726%

    No Known Activations