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

    Neuronpedia

    INDEX
    Explanations

    differences or conflicting situations

    oai_token-act-pair · gpt-3.5-turbo

    the use of the word "while" in contrasting statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ahime
    -0.95
    iotic
    -0.83
    iggurat
    -0.82
    illet
    -0.80
    wow
    -0.78
    helm
    -0.78
    orb
    -0.77
    red
    -0.76
    hack
    -0.76
    Register
    -0.75
    POSITIVE LOGITS
     retaining
    1.12
     maintaining
    1.11
     others
    1.06
     simultaneously
    1.01
     preserving
    0.95
     ignoring
    0.92
     acknowledging
    0.92
     insisting
    0.86
     keeping
    0.85
     avoiding
    0.84
    Act Density 0.051%

    No Known Activations