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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating criticism or advice towards others

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "should" in various contexts related to suggestions and obligations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Ends
    -0.66
     Fra
    -0.66
     Hilbert
    -0.64
    quickShipAvailable
    -0.63
     Dia
    -0.63
     Rox
    -0.62
    vous
    -0.60
    anka
    -0.60
     Oss
    -0.59
    atile
    -0.58
    POSITIVE LOGITS
    nt
    1.11
    ered
    1.09
     beware
    1.08
     ideally
    1.06
     be
    1.03
     aspire
    1.00
     strive
    0.99
     rethink
    0.99
     definitely
    0.98
     reconsider
    0.98
    Act Density 0.107%

    No Known Activations