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

    Neuronpedia

    INDEX
    Explanations

    organizations' names or group titles

    oai_token-act-pair · gpt-3.5-turbo

    references to various organizations and groups

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ÙIJ
    -0.73
    STON
    -0.72
    cially
    -0.72
    entimes
    -0.71
    ¬¼
    -0.71
    ãģł
    -0.68
    antly
    -0.67
    ivery
    -0.66
    worldly
    -0.66
    cffffcc
    -0.65
    POSITIVE LOGITS
     which
    1.11
     whose
    0.99
     who
    0.88
     whom
    0.83
    which
    0.80
    who
    0.75
     wherein
    0.74
     whence
    0.74
     aka
    0.74
     Colo
    0.73
    Act Density 0.259%

    No Known Activations