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

    Neuronpedia

    INDEX
    Explanations

    words related to making a task more difficult or easier, depending on the context

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating difficulty or ease regarding tasks or situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    chn
    -0.71
     Flags
    -0.63
     SUM
    -0.62
     chang
    -0.62
    iche
    -0.61
    WER
    -0.60
    ODY
    -0.59
     LOVE
    -0.59
    aret
    -0.58
    Introduced
    -0.58
    POSITIVE LOGITS
    enged
    0.74
    imaru
    0.69
     enforce
    0.67
    aneously
    0.67
     prey
    0.67
     unwanted
    0.66
    itary
    0.66
    anced
    0.63
    forced
    0.62
     rout
    0.62
    Act Density 0.063%

    No Known Activations