Identity Mappings in ResNet: Stacking Blocks
Intermediate
The problem, from first principles
A deep residual network is a stack of blocks. Each block adds its residual to the running signal, so the output is the input plus a sum of block contributions. Stacking is the step that produces the paper's deep models.
From theory to code
Implement stack_residuals(x, fs), which applies each residual block in order, adding each block's output to its input.
Constraints
- An empty list returns
xunchanged.
Hints
Hint 1
Loop over the functions, replacing the running output with out + f(out) each time.
Keywords
research-papersarchitectureresnet
Python
Run your code to see results
No results yet
Run your code to check it against the examples, or submit to run every test.