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

    Neuronpedia

    INDEX
    Explanations

    words related to the concept of "somewhat" or incremental change

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "somewhat."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    iens
    -0.80
    aring
    -0.77
    yrs
    -0.76
    ults
    -0.76
    abad
    -0.74
     Fighters
    -0.73
    mberg
    -0.72
    elsen
    -0.71
    ULTS
    -0.69
    ocaust
    -0.69
    POSITIVE LOGITS
     unusual
    0.83
     independ
    0.82
     obscure
    0.82
     amusing
    0.81
     inaccurate
    0.80
     tang
    0.80
     insensitive
    0.80
     surprising
    0.80
     coinc
    0.78
     constrained
    0.78
    Act Density 0.009%

    No Known Activations