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

    Neuronpedia

    INDEX
    Explanations

    phrases or fragments containing special characters such as symbols and punctuation

    oai_token-act-pair · gpt-3.5-turbo

    repeated instances of specific characters or symbols

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     vom
    -0.70
     maxim
    -0.68
     scattering
    -0.68
     microphone
    -0.67
     blond
    -0.67
     decomp
    -0.66
     pity
    -0.66
     cane
    -0.66
     minim
    -0.66
    romy
    -0.65
    POSITIVE LOGITS
    £
    1.17
    ħ
    1.07
    ¬
    1.06
    ¹
    1.04
    Ĵ
    1.01
    ı
    0.98
    į
    0.96
    º
    0.95
    ¦
    0.93
    ®
    0.92
    Act Density 0.283%

    No Known Activations