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

    Neuronpedia

    INDEX
    Explanations

    instances where tasks or actions have been completed

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to the completion of tasks or challenges

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    maid
    -0.74
    gypt
    -0.67
    lua
    -0.65
    hot
    -0.64
    spr
    -0.63
    brow
    -0.63
    ung
    -0.62
    bour
    -0.59
    hee
    -0.56
    LI
    -0.56
    POSITIVE LOGITS
    teness
    0.97
    issance
    0.89
    ances
    0.85
    TAIN
    0.84
     completion
    0.77
     srf
    0.76
     Completed
    0.72
    taining
    0.72
     feasibility
    0.71
     preparations
    0.70
    Act Density 0.028%

    No Known Activations