Deep Learning · Level 2 of 5
Residual Connection
A shortcut that adds an earlier representation to a transformed version.
The model can learn a change to its input rather than reconstructing everything.
Example
A block computes x plus a learned transformation of x.
Listen to the definition and example
Audio transcript
Residual Connection. A shortcut that adds an earlier representation to a transformed version. The model can learn a change to its input rather than reconstructing everything. For example: A block computes x plus a learned transformation of x.
Explore this concept
Why it matters
This helps explain why a neural model learns well, becomes unstable, or fails to generalize.
Related concepts
Quick recall question
Try answering before looking back at the definition.