# Continual Learning on CIFAR-100 Split 10-task (test)

Source: [Continual Learning on CIFAR-100 Split 10-task (test)](<https://www.sota2.com/research/sota/continual-learning-on-cifar-100-split-10-task-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-27T06:29:25.543Z

Evaluation of Continual Learning performance on the test split of the CIFAR-100 dataset configured across 10 sequential tasks, measured using AA (Average Accuracy) and AF (Average Forgetting).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Continual Learning |
| Dataset (including any stated split) | CIFAR-100 Split 10-task (test) |
| Metric | Average Accuracy (AA) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Average Accuracy (AA)](<https://www.sota2.com/research/sota/continual-learning-on-cifar-100-split-10-task-test.md?metric=Average%20Accuracy%20(AA)>): Higher is better; unit: PERCENT.

- [Average Forgetting (AF)](<https://www.sota2.com/research/sota/continual-learning-on-cifar-100-split-10-task-test.md?metric=Average%20Forgetting%20(AF)>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Average Accuracy (AA) | Result date | Report paper |
| --- | --- | --- | --- |
| Non CL Baseline (FiLM) | 91.31 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Non CL Baseline (FiLM) | 90.68 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Non CL Baseline (last layer) | 88.5 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Non CL Baseline (last layer) | 88.4 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Non CL Baseline (FiLM) | 86.06 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (FiLM) | 85.26 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Non CL Baseline (last layer) | 85.1 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (last layer) | 82.6 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (last layer) | 82.48 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (FiLM) | 79.79 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (FiLM) | 79.39 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Cosine classifier | 79.02 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Cosine classifier | 78.93 | Not recorded | 68fa78bb1e209131cea7a1ca |
| PEFT Ensemble (last layer) | 78.81 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Cosine classifier | 72.78 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Naive | 9.63 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Naive | 9.55 | Not recorded | 68fa78bb1e209131cea7a1ca |
| Naive | 9.35 | Not recorded | 68fa78bb1e209131cea7a1ca |

### Non CL Baseline (FiLM)

Average Accuracy (AA): 91.31. Result ID: 69ef008847f34ce924d470ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### Non CL Baseline (FiLM)

Average Accuracy (AA): 90.68. Result ID: 69ef008847f34ce924d470cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### Non CL Baseline (last layer)

Average Accuracy (AA): 88.5. Result ID: 69ef008847f34ce924d470d4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### Non CL Baseline (last layer)

Average Accuracy (AA): 88.4. Result ID: 69ef008847f34ce924d470d5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### Non CL Baseline (FiLM)

Average Accuracy (AA): 86.06. Result ID: 69ef008847f34ce924d470cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

### PEFT Ensemble (FiLM)

Average Accuracy (AA): 85.26. Result ID: 69ef008847f34ce924d470cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### Non CL Baseline (last layer)

Average Accuracy (AA): 85.1. Result ID: 69ef008847f34ce924d470d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

### PEFT Ensemble (last layer)

Average Accuracy (AA): 82.6. Result ID: 69ef008847f34ce924d470d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### PEFT Ensemble (last layer)

Average Accuracy (AA): 82.48. Result ID: 69ef008847f34ce924d470d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### PEFT Ensemble (FiLM)

Average Accuracy (AA): 79.79. Result ID: 69ef008847f34ce924d470ca.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

### PEFT Ensemble (FiLM)

Average Accuracy (AA): 79.39. Result ID: 69ef008847f34ce924d470cc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### Cosine classifier

Average Accuracy (AA): 79.02. Result ID: 69ef008847f34ce924d470c9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### Cosine classifier

Average Accuracy (AA): 78.93. Result ID: 69ef008847f34ce924d470c8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### PEFT Ensemble (last layer)

Average Accuracy (AA): 78.81. Result ID: 69ef008847f34ce924d470d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

### Cosine classifier

Average Accuracy (AA): 72.78. Result ID: 69ef008847f34ce924d470c7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

### Naive

Average Accuracy (AA): 9.63. Result ID: 69ef008847f34ce924d470c6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"privacy":"non-DP"}. This record does not assert independent reproduction.

### Naive

Average Accuracy (AA): 9.55. Result ID: 69ef008847f34ce924d470c5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"8","delta":"1e-5"}. This record does not assert independent reproduction.

### Naive

Average Accuracy (AA): 9.35. Result ID: 69ef008847f34ce924d470c4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78bb1e209131cea7a1ca.md>)

Recorded evaluation configuration: {"epsilon":"1","delta":"1e-5"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
