In CIFAR-10 label-shuffling experiments, adding exact-SVD EGD increases the area under the training-accuracy curve for Adam, RAdam, and RMSprop at every reported shuffle schedule. Without EGD, RMSprop is strongest in each schedule; the largest printed multiplicative gain is Adam with shuffling at 3 epochs.

Reported and observed measurements

Assessments (0)

No immutable Assessment has been published for this Claim yet.

Runs (0)

No execution has been linked to this Claim yet.

Each row is a measurement with its reported value, latest available observation, and cumulative evidence state.
Metric and conditions	Reported	Latest observation	Evidence judgment
egd: false · optimizer: Adam · shuffle_label: Shuffling at 1 Epoch	43 paper-reported AUC units	—	Not assessed
egd: true · optimizer: Adam · shuffle_label: Shuffling at 1 Epoch	79.2 paper-reported AUC units	—	Not assessed
optimizer: Adam · shuffle_label: Shuffling at 1 Epoch	1.8 fold	—	Not assessed
egd: false · optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch	40.8 paper-reported AUC units	—	Not assessed
egd: true · optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch	70 paper-reported AUC units	—	Not assessed
optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch	1.7 fold	—	Not assessed
egd: false · optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch	60.2 paper-reported AUC units	—	Not assessed
egd: true · optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch	79.5 paper-reported AUC units	—	Not assessed
optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch	1.3 fold	—	Not assessed
egd: false · optimizer: Adam · shuffle_label: Shuffling at 3 Epochs	24.7 paper-reported AUC units	—	Not assessed
egd: true · optimizer: Adam · shuffle_label: Shuffling at 3 Epochs	56 paper-reported AUC units	—	Not assessed
optimizer: Adam · shuffle_label: Shuffling at 3 Epochs	2.3 fold	—	Not assessed

Metric and conditions

Reported

Latest observation

Evidence judgment

egd: false · optimizer: Adam · shuffle_label: Shuffling at 1 Epoch

43 paper-reported AUC units

—

Not assessed

egd: true · optimizer: Adam · shuffle_label: Shuffling at 1 Epoch

79.2 paper-reported AUC units

—

Not assessed

optimizer: Adam · shuffle_label: Shuffling at 1 Epoch

1.8 fold

—

Not assessed

egd: false · optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch

40.8 paper-reported AUC units

—

Not assessed

egd: true · optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch

70 paper-reported AUC units

—

Not assessed

optimizer: RAdam · shuffle_label: Shuffling at 1 Epoch

1.7 fold

—

Not assessed

egd: false · optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch

60.2 paper-reported AUC units

—

Not assessed

egd: true · optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch

79.5 paper-reported AUC units

—

Not assessed

optimizer: RMSProp · shuffle_label: Shuffling at 1 Epoch

1.3 fold

—

Not assessed

egd: false · optimizer: Adam · shuffle_label: Shuffling at 3 Epochs

24.7 paper-reported AUC units

—

Not assessed

egd: true · optimizer: Adam · shuffle_label: Shuffling at 3 Epochs

56 paper-reported AUC units

—

Not assessed

optimizer: Adam · shuffle_label: Shuffling at 3 Epochs

2.3 fold

—

Not assessed

Reported and observed measurements

Assessments (0)

Runs (0)

Measurement results