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

    Neuronpedia

    INDEX
    Explanations

    business and financial related phrases

    oai_token-act-pair · gpt-3.5-turbo

    instances of the verb "to be" and related forms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    acca
    -0.95
    arez
    -0.74
    rongh
    -0.74
    éĹĺ
    -0.73
    arettes
    -0.73
    uction
    -0.71
    undrum
    -0.70
    ampa
    -0.68
    uala
    -0.68
    è¦ļéĨĴ
    -0.67
    POSITIVE LOGITS
     nonetheless
    1.52
     nevertheless
    1.35
     alas
    0.96
     retains
    0.90
     retaining
    0.88
    etheless
    0.87
     still
    0.85
     lacks
    0.83
     lacked
    0.82
     beware
    0.81
    Act Density 0.377%

    No Known Activations