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

    Neuronpedia

    INDEX
    Explanations

    references to correctness and appropriateness in various contexts

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

    correct or proper version

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    fromnode
    -0.58
     autorytatywna
    -0.55
     Comprometido
    -0.52
    Portály
    -0.51
    帖最后由
    -0.50
    complexContent
    -0.49
     proyectos
    -0.48
    sizeCache
    -0.48
    Искәрмәләр
    -0.44
    Rüyada
    -0.43
    POSITIVE LOGITS
     correct
    0.63
     appropriate
    0.56
     correctly
    0.53
     preferred
    0.48
     corresponding
    0.47
     juiste
    0.47
    Correct
    0.47
    correctly
    0.46
    correct
    0.46
     riktig
    0.46
    Act Density 0.021%

    No Known Activations