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

    Neuronpedia

    INDEX
    Explanations

    statements advocating for particular rights or positions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the verb "be" used in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Inventory
    -0.73
     Lack
    -0.71
     srfAttach
    -0.68
     inconsistency
    -0.66
     Stamina
    -0.65
     Wid
    -0.64
     unfocusedRange
    -0.63
     Anniversary
    -0.63
    stress
    -0.63
     Farn
    -0.63
    POSITIVE LOGITS
     punished
    1.20
     rewarded
    1.19
     consulted
    1.19
     able
    1.13
     allowed
    1.13
     empowered
    1.13
     compensated
    1.09
     persuaded
    1.09
     victimized
    1.09
     afforded
    1.07
    Act Density 0.201%

    No Known Activations