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

    Neuronpedia

    INDEX
    Explanations

    steps or actions in a set of instructions, likely related to technical or procedural tasks

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    lycer
    -0.91
     Courts
    -0.88
     motto
    -0.86
     indo
    -0.86
     Stall
    -0.86
    sun
    -0.86
     Levant
    -0.85
     Finals
    -0.84
    league
    -0.84
     ejac
    -0.82
    POSITIVE LOGITS
    ignty
    1.23
    ername
    1.10
    lish
    1.08
    artisan
    1.07
    lication
    1.04
    ering
    1.01
    lished
    0.99
    nir
    0.98
    geoning
    0.97
    arthed
    0.96
    Act Density 0.300%

    No Known Activations