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

    Neuronpedia

    INDEX
    Explanations

    references to concepts of correctness and suitability

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    InjectAttribute
    -0.70
    numerusform
    -0.64
    BeginContext
    -0.64
    cheibe
    -0.62
    MLLoader
    -0.59
    bitField
    -0.57
    blockList
    -0.57
     adjourn
    -0.57
     发表于
    -0.57
    ]){
    -0.56
    POSITIVE LOGITS
     appropriate
    1.14
     proper
    1.10
     correct
    1.01
     wrong
    0.98
    Wrong
    0.98
    Appropriate
    0.98
     richtigen
    0.98
     Wrong
    0.95
     juiste
    0.93
     odpowied
    0.93
    Act Density 0.107%

    No Known Activations