Previous question · 1081
Reverse Nodes In K Group
hardData Structures
Next question · 1083
Decode Ways
mediumAlgorithms
Random coding question
1082. Masked attention scores
medium
Generalstaff
Function contract
Inputs
- K
- np.ndarray
- Q
- np.ndarray
- mask
- np.ndarray
Returns
- value
- np.ndarray
Examples
Input
{
"K": [
[
1,
0
],
[
0,
1
]
],
"Q": [
[
1,
0
],
[
0,
1
]
],
"mask": [
[
1,
0
],
[
1,
1
]
]
}Output
[
[
1,
0
],
[
0.3302384506733431,
0.6697615493266569
]
]Input Signature
| Argument | Type |
|---|---|
| K | np.ndarray |
| Q | np.ndarray |
| mask | np.ndarray |
Output Signature
| Return Name | Type |
|---|---|
| value | np.ndarray |
Constraints
Hint 1
Hint 2
Hint 3
Roles
ML Engineer
AI Engineer
Data Scientist
Research Scientist
Companies
GeneralLevels
staff
senior
entry
Tags
scaled-dot-product-attention
masking
numerically-stable-softmax
numpy-matrix-multiplication
Similar Questions
Linear Regressionmedium
ML CodingActivation function relueasy
ML CodingText normalizationmedium
ML Coding48 people are solving this problem
Loading editor for the selected language
Ln 1, Col 1
Program Output
DATAINTERVIEW
Interactive coding terminal
Python / NumPy environment ready.
Type a command below or run code from the editor.
run scriptExecute the current editor contents
submitGrade the current editor contents against all solution test cases
python ("print hello world")Run an isolated Python command
helpShow all available commands