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

    Neuronpedia

    INDEX
    Explanations

    phrases related to actions resulting in impact or revelation

    oai_token-act-pair · gpt-3.5-turbo

    variations of the word "blow" in different contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ournal
    -0.71
    heid
    -0.70
    aho
    -0.65
    angelo
    -0.65
    cer
    -0.65
    ript
    -0.64
    Gameplay
    -0.64
    gent
    -0.62
    Adapt
    -0.62
    rete
    -0.60
    POSITIVE LOGITS
     apart
    0.87
    job
    0.86
     blown
    0.84
     away
    0.83
     holes
    0.83
    pipe
    0.83
     blew
    0.83
     whist
    0.81
    blown
    0.79
    guns
    0.77
    Act Density 0.038%

    No Known Activations