StudyBrixFree worksheets, study guides and courses

University · Artificial intelligence

University Artificial intelligence practice

15 questions on deep learning, ai engineering, generative ai, machine learning. The answer key is a separate download.

Open and downloadGet the answer keyChange the questions

What is on the worksheet

  1. A neuron computes ReLU(wx+b) with w=9, x=0, b=9. Find its output. Round the final result to 4 decimal places if needed.

  2. A strictly sequential AI workflow spends 270 ms retrieving, 130 ms generating and 60 ms validating. Find total latency, ignoring retries. Round the final result to 4 decimal places if needed. Answer in ms.

  3. A labeled query has 39 relevant documents. A retriever finds 20 of them. Calculate retrieval recall. Round the final result to 4 decimal places if needed.

  4. A neuron computes ReLU(wx+b) with w=6, x=-8, b=7. Find its output. Round the final result to 4 decimal places if needed.

  5. A fitted linear model has w=3, b=28. Find its prediction at x=5. Round the final result to 4 decimal places if needed.

  6. A neuron computes ReLU(wx+b) with w=-9, x=1, b=-3. Find its output. Round the final result to 4 decimal places if needed.

  7. PCA covariance eigenvalues are 22, 12, 3. Find the fraction of total variance explained by the first component. Round the final result to 4 decimal places if needed.

  8. A fully connected layer maps 21 inputs to 10 outputs with one bias per output. How many trainable parameters does it have? Round the final result to 4 decimal places if needed.

  9. A strictly sequential AI workflow spends 300 ms retrieving, 760 ms generating and 20 ms validating. Find total latency, ignoring retries. Round the final result to 4 decimal places if needed. Answer in ms.

  10. A neuron computes ReLU(wx+b) with w=1, x=-5, b=4. Find its output. Round the final result to 4 decimal places if needed.

  11. A fitted linear model has w=11, b=-8. Find its prediction at x=1. Round the final result to 4 decimal places if needed.

  12. For TP=48, FP=59, FN=47, compute positive-class F1 score. Round the final result to 4 decimal places if needed.

  13. For TP=42, FP=19, FN=2, compute positive-class F1 score. Round the final result to 4 decimal places if needed.

  14. Three prediction errors (prediction minus target) are -3, 10, 10. Find mean squared error. Round the final result to 4 decimal places if needed.

  15. A fully connected layer maps 42 inputs to 17 outputs with one bias per output. How many trainable parameters does it have? Round the final result to 4 decimal places if needed.