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

    Neuronpedia

    INDEX
    Explanations

    the word "Well" as an introductory statement or prompt

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "Well," indicating a response or commentary in dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    illary
    -0.76
    İĭ
    -0.68
    adena
    -0.67
    arb
    -0.67
     arom
    -0.65
    Gy
    -0.64
    è¦ļéĨĴ
    -0.64
    âĹ¼
    -0.64
     dash
    -0.60
    mage
    -0.60
    POSITIVE LOGITS
    esley
    0.95
    espie
    0.89
    tenance
    0.83
    ness
    0.82
    come
    0.81
     Enough
    0.81
    NESS
    0.76
    enough
    0.76
    nesses
    0.75
     Alright
    0.72
    Act Density 0.016%

    No Known Activations