# Example-count generalization on 1D conservation law 2,000 steps k=5 examples standardized (train)

Source: [Example-count generalization on 1D conservation law 2,000 steps k=5 examples standardized (train)](<https://www.sota2.com/research/sota/example-count-generalization-on-1d-conservation-law-2-000-steps-k-5-examples-standardized-train>)

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

Data updated: 2026-05-12T10:31:14.344Z

Evaluation of Example-count generalization performance on the training split of the 1D conservation law dataset (k=5, 2,000 steps), assessed using Mean Absolute Error (MAE).

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 | Example-count generalization |
| Dataset (including any stated split) | 1D conservation law 2,000 steps k=5 examples standardized (train) |
| Metric | MAE |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| 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

- [MAE](<https://www.sota2.com/research/sota/example-count-generalization-on-1d-conservation-law-2-000-steps-k-5-examples-standardized-train.md?metric=MAE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | MAE | Result date | Report paper |
| --- | --- | --- | --- |
| StructuredFunctionPE | 0.108 | Not recorded | 6a02b410dbc836b87d8246a2 |
| InterpolatedDemoPE | 0.114 | Not recorded | 6a02b410dbc836b87d8246a2 |
| StructuredDemoRolePE | 0.115 | Not recorded | 6a02b410dbc836b87d8246a2 |
| StructuredDemoRolePE | 0.128 | Not recorded | 6a02b410dbc836b87d8246a2 |
| RoleOnlyPE | 0.159 | Not recorded | 6a02b410dbc836b87d8246a2 |
| StructuredInterpolationPE | 0.197 | Not recorded | 6a02b410dbc836b87d8246a2 |
| VanillaICON (nn.Embedding) | 0.485 | Not recorded | 6a02b410dbc836b87d8246a2 |

### StructuredFunctionPE

MAE: 0.108. Result ID: 6a03009147f34ce924e6734a.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"EvE","Run":"2","Key Mechanism":"Interpolated demo PE + local sinusoid + structured RoPE/bias"}. This record does not assert independent reproduction.

### InterpolatedDemoPE

MAE: 0.114. Result ID: 6a03009147f34ce924e67349.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"EvE","Run":"1","Key Mechanism":"Interpolated demo PE + structured demo/role attention bias"}. This record does not assert independent reproduction.

### StructuredDemoRolePE

MAE: 0.115. Result ID: 6a03009147f34ce924e6734b.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"Static-Initial","Run":"1","Key Mechanism":"Power-law demo compression + sinusoid + demo-aware ALiBi/RoPE"}. This record does not assert independent reproduction.

### StructuredDemoRolePE

MAE: 0.128. Result ID: 6a03009147f34ce924e6734e.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"Static-Final","Run":"2","Key Mechanism":"Tanh demo compression + structured RoPE/bias"}. This record does not assert independent reproduction.

### RoleOnlyPE

MAE: 0.159. Result ID: 6a03009147f34ce924e6734c.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"Static-Initial","Run":"2","Key Mechanism":"Role-only PE + compressed-demo rotary attention"}. This record does not assert independent reproduction.

### StructuredInterpolationPE

MAE: 0.197. Result ID: 6a03009147f34ce924e6734d.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

Recorded evaluation configuration: {"Variant":"Static-Final","Run":"1","Key Mechanism":"Linear clamp + sinusoidal example signal"}. This record does not assert independent reproduction.

### VanillaICON (nn.Embedding)

MAE: 0.485. Result ID: 6a03009147f34ce924e6734f.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b410dbc836b87d8246a2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5ae51e209131ce9c3bb3.md>)

Recorded evaluation configuration: {"Variant":"Seed","Run":"n/a","Key Mechanism":"Flat learned lookup (no structure)"}. This record does not assert independent reproduction.

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