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

    Neuronpedia

    INDEX
    Explanations

    phrases related to evaluation, assessment, or examination

    oai_token-act-pair · gpt-3.5-turbo

    processes and actions related to evaluation, approval, and utilization

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    habi
    -0.58
    olated
    -0.58
     penal
    -0.56
    mare
    -0.55
    lete
    -0.55
     nutshell
    -0.54
    cept
    -0.53
    orious
    -0.53
    pat
    -0.53
    Sequ
    -0.53
    POSITIVE LOGITS
     purposes
    1.84
     sake
    1.34
     reasons
    1.14
    ummies
    0.96
     purpose
    0.90
    EngineDebug
    0.89
    farious
    0.75
    okers
    0.71
     seekers
    0.71
     sometime
    0.69
    Act Density 0.369%

    No Known Activations