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

    Neuronpedia

    INDEX
    Explanations

    words related to searching, extracting, or gathering something specific from a source

    oai_token-act-pair · gpt-3.5-turbo

    actions related to extracting or gathering information and resources

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     aggress
    -0.67
    ogun
    -0.63
    govern
    -0.63
    eworld
    -0.61
    paced
    -0.61
    inem
    -0.61
    alone
    -0.59
     warr
    -0.59
     withd
    -0.59
    legged
    -0.58
    POSITIVE LOGITS
    ãĤ©
    0.99
    ãĥķãĤ¡
    0.91
     glean
    0.81
     clues
    0.80
    ources
    0.76
    ilts
    0.76
    è£
    0.72
    ãĥĥãĥĪ
    0.71
     extracts
    0.71
     scraps
    0.71
    Act Density 0.137%

    No Known Activations