With parameter-free option-A shortcuts, increasing depth from 18 to 34 layers reverses the plain-network degradation: ResNet-34 has lower training and validation error than ResNet-18, while matched ResNet-34 substantially outperforms plain-34. ResNet-18 and plain-18 have comparable final accuracy, though ResNet-18 converges faster early in training.

With parameter-free option-A shortcuts, increasing depth from 18 to 34 layers reverses the plain-network degradation: ResNet-34 has lower training and validation error than ResNet-18, while matched ResNet-34 substantially outperforms plain-34. ResNet-18 and plain-18 have comparable final accuracy, though ResNet-18 converges faster early in training. · CiteArk

Measurement results

0/4 assessed

0 supported0 disputed or conflicting4 not assessed

The complete result group stays with its claim. Assessed rows come first; their judgments do not replace the whole-claim conclusion. Expand a row for conditions, sources, and evidence history.

Each row is a measurement with its reported value, latest available observation, and cumulative evidence state.
Reported	Latest observation	Evidence judgment
27.88%	—	Not assessed
25.03%	—	Not assessed
2.8 percentage points	—	Not assessed
3.5 percentage points	—	Not assessed

Reported and observed measurements

Assessments (0)

Runs (0)

Measurement results