How should I practice Multi-Head Attention Forward Pass?

Start by reading the full statement, solve the visible examples, then compare the pattern against related Meta phone screen problems in FastPrep.

Is Multi-Head Attention Forward Pass associated with Meta?

FastPrep tags this as a Meta phone screen coding practice problem when the problem catalog has that company and stage context.

Multi-Head Attention Forward Pass

Meta● HardFULLTIMEPHONE SCREEN

Learn

Problem statement

Implement a projection-free multi-head scaled dot-product attention forward pass. query, key, and value are row-major integer matrices with the same feature width d; key and value have the same number of rows. heads divides d.

Split each row into contiguous head slices. For each query row and head, compute dot products against every key slice, divide by sqrt(d / heads), apply a numerically stable softmax, and use the weights to combine the matching value slices. Concatenate the head outputs.

Return one comma-separated string per query row with every coordinate formatted to exactly six decimal places. There are no learned projections or masks.

Function

multiHeadAttention(query: int[][], key: int[][], value: int[][], heads: int) → String[]

Examples

Example 1

query = [[1]]key = [[0],[1]]value = [[10],[20]]heads = 1return = ["17.310586"]

Softmax over scores 0 and 1 weights the second value more heavily.

Example 2

query = [[1,0]]key = [[1,0],[0,1]]value = [[4,8],[6,2]]heads = 2return = ["4.537883,5.000000"]

Each one-dimensional head computes its own attention distribution.

Constraints

1 <= query rows, key rows, d <= 40.
All rows have width d; 1 <= heads <= d and divides d.
Entries are between -100 and 100.