{"description":"Test your understanding of position embeddings and their role in enabling Transformers to interpret word order.","questions":[{"answer":"To give the model information about word order","number":1,"options":["To reduce model parameters","To improve output formatting","To give the model information about word order","To replace token embeddings"],"question":"Why are position embeddings necessary in Transformer models?"},{"answer":"Sine and cosine","number":2,"options":["Square and root","Tangent and cotangent","Sine and cosine","Exponential and logarithmic"],"question":"Which of the following functions are used in standard position embeddings?"},{"answer":"They are periodic and allow unique encoding of positions","number":3,"options":["They are easy to store in memory","They are periodic and allow unique encoding of positions","They are bounded between -100 and 100","They match token vocabulary length"],"question":"What is the main property of sine and cosine that makes them useful for position encoding?"},{"answer":"By adding them element-wise","number":4,"options":["By multiplying them","By concatenating them","By adding them element-wise","By sorting them"],"question":"How are position embeddings combined with word embeddings in Transformer models?"},{"answer":"Word order and structural relationships","number":5,"options":["Vocabulary frequency","Sentence complexity","Word order and structural relationships","Character-level spelling"],"question":"What aspect of language understanding do position embeddings help capture?"},{"answer":"Same as word embedding dimension","number":6,"options":["Same as word embedding dimension","Always 3","Dependent on the number of tokens in vocabulary","Double the number of attention heads"],"question":"What is the typical dimension of a position embedding in practice?"},{"answer":"At the input stage with token embeddings","number":7,"options":["Final classification layer","Before the feed-forward network","At the input stage with token embeddings","At the output projection layer"],"question":"In which layer are position embeddings applied in Transformer architectures?"},{"answer":"They allow models to generalize better across sequences of different lengths","number":8,"options":["They are easier to visualize in 3D","They reduce model size","They allow models to generalize better across sequences of different lengths","They don\u2019t require sine/cosine functions"],"question":"What is a key benefit of relative position embeddings compared to absolute ones?"},{"answer":"Machine translation","number":9,"options":["Named entity recognition","Machine translation","Character counting","Stop word removal"],"question":"Which NLP task especially benefits from capturing the relative order of tokens?"},{"answer":"3D scatter plot","number":10,"options":["Bar chart","Pie chart","3D scatter plot","Confusion matrix"],"question":"What kind of plot can help visualize how position embeddings shift with word order?"}],"title":"Position Embedding"}