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

    Neuronpedia

    INDEX
    Explanations

    negations or phrases indicating difficulties and complaints

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    sometimes
    -0.22
     sometimes
    -0.22
     Sometimes
    -0.18
    adesh
    -0.17
     bazen
    -0.17
     иногда
    -0.17
    Sometimes
    -0.16
    ubl
    -0.16
     occasionally
    -0.15
    даÑĤ
    -0.14
    POSITIVE LOGITS
     anything
    0.26
     any
    0.26
     TOO
    0.24
     too
    0.23
     necessarily
    0.23
     much
    0.22
     major
    0.22
     overly
    0.22
     materially
    0.21
     terribly
    0.21
    Act Density 0.221%

    No Known Activations