Implement Scaled Dot-Product Attention

Question: Implement Scaled Dot-Product Attention Implement single-head scaled dot-product attention. You are given three matrices: Query matrix Q with shape n × d Key matrix K with shape n × d Value matrix V with shape n × d Compute: [ Attention(Q,K,V)=softmax\left(\frac{QK^T}{\sqrt d}\right)V ] The softmax is applied independently to each row. Input Format n d n rows of Q, each containing d floating-point numbers n rows of K, each containing d floating-point numbers n rows of V, each containing d floating-point numbers Output Format Output the attention result matrix with shape n × d. Print d floating-point numbers per row, rounded to 6 decimal places. Constraints 1 <= n <= 100 1 <= d <= 64 Input values are in [-10, 10] Requirements Correctly implement scaled dot-product attention. Use a numerically stable softmax. Explain the time and space complexity. Example Input: 2 2 1 0 0 1 1 0 0 1 1 2 3 4 Output: 1.660477 2.660477 2.339523 3.339523 Example Input 1 1 2 3 4 Output 4.000000