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

    Neuronpedia

    INDEX
    Explanations

    specific phrases indicating coincidence or circumstance

    oai_token-act-pair · gpt-3.5-turbo

    phrases containing the structure "happened to be."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ãĤ¦ãĤ¹
    -0.78
    intensive
    -0.69
    Log
    -0.67
    urized
    -0.67
     Limits
    -0.66
    chlor
    -0.66
    rar
    -0.65
    lime
    -0.65
    Easy
    -0.65
    simple
    -0.64
    POSITIVE LOGITS
     coincide
    1.21
     stumble
    1.19
     reside
    1.00
     befriend
    0.94
     belong
    0.91
     resemble
    0.89
     meet
    0.88
     be
    0.86
     stumbled
    0.85
     inherit
    0.84
    Act Density 0.101%

    No Known Activations