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

    Neuronpedia

    INDEX
    Explanations

    instances of failing to do something

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "fail" and its variations as well as failures to meet expectations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    soDeliveryDate
    -0.81
    EStream
    -0.74
    Sold
    -0.70
    Textures
    -0.69
    Flo
    -0.69
    Reader
    -0.65
    Fra
    -0.64
     Cutter
    -0.63
     Canary
    -0.63
     Vec
    -0.62
    POSITIVE LOGITS
     comply
    1.27
     adequately
    1.26
     capitalize
    1.24
     disclose
    1.24
     anticipate
    1.21
     recognize
    1.21
     adhere
    1.21
     heed
    1.20
     fulfill
    1.20
     grasp
    1.18
    Act Density 0.065%

    No Known Activations