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

    Neuronpedia

    INDEX
    Explanations

    instances where someone is pointing to something

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "to"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    soever
    -0.93
    mare
    -0.81
    lance
    -0.78
    bit
    -0.74
    worth
    -0.74
    stress
    -0.74
    tails
    -0.71
    itialized
    -0.71
    MAS
    -0.71
    mask
    -0.70
    POSITIVE LOGITS
     similarities
    0.76
     inconsistencies
    0.69
     anecdotal
    0.69
     finger
    0.65
     ample
    0.63
     clues
    0.63
     preced
    0.62
     warnings
    0.62
     successes
    0.62
     Genocide
    0.62
    Act Density 0.063%

    No Known Activations