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

    Neuronpedia

    INDEX
    Explanations

    proper nouns related to individuals

    oai_token-act-pair · gpt-3.5-turbo

    instances of the verb "was" and its variations, along with past tense forms related to actions or events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Moines
    -0.67
     è£ıè
    -0.62
    WAR
    -0.62
    rones
    -0.61
     Robotics
    -0.60
    ress
    -0.60
    ixties
    -0.59
     Adds
    -0.58
     Rug
    -0.58
     Comes
    -0.58
    POSITIVE LOGITS
     indeed
    1.05
     unfairly
    0.98
     misunderstood
    0.98
     "âĢ¦
    0.94
     misrepresent
    0.92
     "...
    0.92
    nt
    0.92
     "
    0.89
     misled
    0.89
    ©¶æ
    0.87
    Act Density 0.314%

    No Known Activations