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

    Neuronpedia

    INDEX
    Explanations

    proper nouns, likely related to politics or government entities

    oai_token-act-pair · gpt-3.5-turbo

    trailing punctuation marks or unknown characters

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    jri
    -0.65
     destro
    -0.63
    iterranean
    -0.62
     reperto
    -0.61
    wcs
    -0.61
    wagon
    -0.59
     corrid
    -0.59
    oppy
    -0.59
    vertisement
    -0.57
     contrace
    -0.57
    POSITIVE LOGITS
    ITED
    0.60
     âĵĺ
    0.54
     HERO
    0.52
     Fired
    0.50
     Mars
    0.50
     Horror
    0.49
     Islam
    0.48
     Squirrel
    0.48
    IVES
    0.48
     Endless
    0.47
    Act Density 0.156%

    No Known Activations