# Generalization on Synthetic Dataset

Source: [Generalization on Synthetic Dataset](<https://www.sota2.com/research/sota/generalization-on-synthetic-dataset>)

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

Data updated: 2026-02-26T06:00:55.279Z

Evaluation of model generalization performance on a synthetic dataset, measured using R^2 scores calculated across different data subsets (10%, 20%, 50%, 70%).

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 | Generalization |
| Dataset (including any stated split) | Synthetic Dataset |
| Metric | R^2 (10%) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [R^2 (10%)](<https://www.sota2.com/research/sota/generalization-on-synthetic-dataset.md?metric=R%5E2%20(10%25)>): Higher is better; unit: SCALAR.

- [R^2 (20%)](<https://www.sota2.com/research/sota/generalization-on-synthetic-dataset.md?metric=R%5E2%20(20%25)>): Higher is better; unit: SCALAR.

- [R^2 (50%)](<https://www.sota2.com/research/sota/generalization-on-synthetic-dataset.md?metric=R%5E2%20(50%25)>): Higher is better; unit: SCALAR.

- [R^2 (70%)](<https://www.sota2.com/research/sota/generalization-on-synthetic-dataset.md?metric=R%5E2%20(70%25)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | R^2 (10%) | Result date | Report paper |
| --- | --- | --- | --- |
| AdverISF (Two-Stage) | 0.251 | Not recorded | 698969eac97a0b3e3b5dc3e5 |
| MLP | 0.201 | Not recorded | 698969eac97a0b3e3b5dc3e5 |
| AdverISF (Joint) | 0.179 | Not recorded | 698969eac97a0b3e3b5dc3e5 |
| infoR–LSF | 0.165 | Not recorded | 698969eac97a0b3e3b5dc3e5 |
| VIB | 0.159 | Not recorded | 698969eac97a0b3e3b5dc3e5 |

### AdverISF (Two-Stage)

R^2 (10%): 0.251. Result ID: 6998e7875698665ec44bed39.

[Reporting paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

Recorded evaluation configuration: {"Training Mode":"Two-Stage"}. This record does not assert independent reproduction.

### MLP

R^2 (10%): 0.201. Result ID: 6998e7875698665ec44bed35.

[Reporting paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### AdverISF (Joint)

R^2 (10%): 0.179. Result ID: 6998e7875698665ec44bed38.

[Reporting paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

Recorded evaluation configuration: {"Training Mode":"Joint"}. This record does not assert independent reproduction.

### infoR–LSF

R^2 (10%): 0.165. Result ID: 6998e7875698665ec44bed37.

[Reporting paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### VIB

R^2 (10%): 0.159. Result ID: 6998e7875698665ec44bed36.

[Reporting paper](<https://www.sota2.com/research/paper/698969eac97a0b3e3b5dc3e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa56501e209131ce99b93c.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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