Fundamentals of testing AI-based systems · Question 22 of 60
Why can even locked AI systems with large neural networks exhibit non-deterministic behavior in practice?
2 answers are correct. Medium
Show answer
Answer
- C Because of variations in hardware execution, particularly with parallel computation or GPUs
- D Because of floating-point precision limits
Explanation
A locked system should really always answer the same way for the same input. With large neural networks, this does not always work in practice: floating-point numbers have limited precision, and depending on the hardware, especially with parallel computation on GPUs, rounding turns out slightly differently.