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

    Neuronpedia

    INDEX
    Explanations

    double hyphens that separate conversations between two or more speakers

    oai_token-act-pair · gpt-3.5-turbo

    instances of hesitation or uncertainty in dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     LIA
    -0.72
    æ³
    -0.70
    agher
    -0.66
     Slug
    -0.66
    osal
    -0.66
    å¼
    -0.64
    éĩ
    -0.64
     QC
    -0.63
    æĪ¦
    -0.62
     Kard
    -0.61
    POSITIVE LOGITS
    certain
    0.87
     uh
    0.86
     gasp
    0.79
    wait
    0.77
    natureconservancy
    0.73
    different
    0.73
    [[
    0.70
    ________
    0.69
    argo
    0.66
    ––
    0.66
    Act Density 0.187%

    No Known Activations