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

    Neuronpedia

    INDEX
    Explanations

    Russian words with specific characters

    oai_token-act-pair · gpt-3.5-turbo

    specific Cyrillic characters and letters, particularly those used in certain Slavic languages

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    gdala
    -0.74
    iqueness
    -0.74
    ndra
    -0.72
    arton
    -0.70
    ayer
    -0.69
     reconc
    -0.68
    cause
    -0.68
    reads
    -0.67
    ¿½
    -0.67
    guyen
    -0.66
    POSITIVE LOGITS
    оÐ
    1.29
    и
    1.18
    о
    1.15
    а
    1.13
    ÑĢ
    1.09
    л
    1.09
    ÑĤ
    1.08
    Ñĥ
    1.05
    Ñģ
    0.97
    е
    0.93
    Act Density 0.010%

    No Known Activations