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

    Neuronpedia

    INDEX
    Explanations

    fractions and numbers in forms like "½" and "9"

    oai_token-act-pair · gpt-3.5-turbo

    fractions or values represented in a specific numeric format

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Pag
    -0.71
     validation
    -0.70
     logic
    -0.65
     Riot
    -0.64
     cloning
    -0.64
     logging
    -0.63
     Trees
    -0.61
     MEM
    -0.60
     Vis
    -0.60
     referenced
    -0.59
    POSITIVE LOGITS
    ½
    3.79
     ½
    2.52
    âĢ¢âĢ¢
    1.29
    âĻ¦
    1.08
    ¢
    1.04
    âģ
    1.04
    Ö¼
    1.00
    £
    0.98
    ±
    0.93
    ++++
    0.89
    Act Density 0.016%

    No Known Activations