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

    Neuronpedia

    INDEX
    Explanations

    the significance of key concepts and their commonality or importance in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    most important/common thing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    NameInMap
    -0.97
     betweenstory
    -0.96
    principalColumn
    -0.95
    ImageContext
    -0.90
    Autoritní
    -0.90
    tagHelperRunner
    -0.90
    StructEnd
    -0.89
    parsedMessage
    -0.89
    -0.87
     OMITBAD
    -0.87
    POSITIVE LOGITS
     sahibi
    0.29
    mainly
    0.24
     meilleures
    0.24
     lámpara
    0.24
    chossen
    0.23
     Mostly
    0.23
     mainly
    0.23
     yoktur
    0.22
    mostly
    0.22
     usually
    0.21
    Act Density 0.199%

    No Known Activations