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

    Neuronpedia

    INDEX
    Explanations

    phrases related to various fears, such as fear of speaking out, fear of retribution, fear of legal issues, and more

    oai_token-act-pair · gpt-3.5-turbo

    expressions of fear and anxiety related to potential negative consequences

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Registered
    -0.75
    Film
    -0.73
    minus
    -0.72
    atto
    -0.71
     congratulations
    -0.71
     congr
    -0.71
    Oak
    -0.69
     precincts
    -0.68
     precinct
    -0.68
     Parables
    -0.68
    POSITIVE LOGITS
     retribution
    0.90
     repercussions
    0.86
    angering
    0.86
     repr
    0.85
     unwanted
    0.84
     uncontroll
    0.80
     misunderstood
    0.80
     harm
    0.79
     contam
    0.79
     undue
    0.79
    Act Density 0.456%

    No Known Activations