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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences where something is described as the "right" choice, decision, thing, or fit

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating the concept of "right" or correctness in varying contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ĸļ
    -0.75
    ADRA
    -0.73
    cit
    -0.70
     conclud
    -0.70
    mat
    -0.66
     Bei
    -0.64
    aryl
    -0.64
    bryce
    -0.62
    è¦ļéĨĴ
    -0.62
     srfAttach
    -0.61
    POSITIVE LOGITS
    eous
    1.17
     wing
    0.86
     fit
    0.72
     hemisphere
    0.71
     flank
    0.70
    lander
    0.69
     thing
    0.69
     amount
    0.68
    wing
    0.68
     fielder
    0.67
    Act Density 0.031%

    No Known Activations