# Parameter Reconstruction on Physics-based parameter reconstruction (test)

Source: [Parameter Reconstruction on Physics-based parameter reconstruction (test)](<https://www.sota2.com/research/sota/parameter-reconstruction-on-physics-based-parameter-reconstruction-test>)

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

Data updated: 2026-02-26T05:48:45.871Z

Evaluation of model performance on Parameter Reconstruction using the test split of the Physics-based parameter reconstruction dataset. Key metrics include Loss, Parameter RMSE, and Runtime per epoch.

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 | Parameter Reconstruction |
| Dataset (including any stated split) | Physics-based parameter reconstruction (test) |
| Metric | Loss |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 3 of 3 |
| 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

- [Loss](<https://www.sota2.com/research/sota/parameter-reconstruction-on-physics-based-parameter-reconstruction-test.md?metric=Loss>): Lower is better; unit: SCALAR.

- [Parameter RMSE](<https://www.sota2.com/research/sota/parameter-reconstruction-on-physics-based-parameter-reconstruction-test.md?metric=Parameter%20RMSE>): Lower is better; unit: SCALAR.

- [Runtime per epoch](<https://www.sota2.com/research/sota/parameter-reconstruction-on-physics-based-parameter-reconstruction-test.md?metric=Runtime%20per%20epoch>): Lower is better; unit: TIME.

## Recorded results

| Method | Loss | Result date | Report paper |
| --- | --- | --- | --- |
| Tsit5 | 0.0059 | Not recorded | 68fa5c841e209131ce9d1a3b |
| Arnoldi | 0.0061 | Not recorded | 68fa5c841e209131ce9d1a3b |
| Dopri5 | 0.0063 | Not recorded | 68fa5c841e209131ce9d1a3b |

### Tsit5

Loss: 0.0059. Result ID: 698bb5857a967eddb770c6d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa5c841e209131ce9d1a3b.md>)

Recorded evaluation configuration: {"Numerical Solver":"Tsit5","Differentiation strategy":"disc. → diff.","Adjoint method type":"RecursiveCheckpointAdjoint"}. This record does not assert independent reproduction.

### Arnoldi

Loss: 0.0061. Result ID: 698bb5857a967eddb770c6d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa5c841e209131ce9d1a3b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5c841e209131ce9d1a3b.md>)

[Implementation](<https://github.com/pnkraemer/experiments-lanczos-adjoints>)

Recorded evaluation configuration: {"Numerical Solver":"Arnoldi","Differentiation strategy":"differentiate-then-discretise","Adjoint method type":"Arnoldi's adjoints"}. This record does not assert independent reproduction.

### Dopri5

Loss: 0.0063. Result ID: 698bb5857a967eddb770c6d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa5c841e209131ce9d1a3b.md>)

Recorded evaluation configuration: {"Numerical Solver":"Dopri5","Differentiation strategy":"diff. → disc.","Adjoint method type":"BacksolveAdjoint"}. This record does not assert independent reproduction.

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