# DAG Recovery on 200-node ER4 Graphs (Gaussian Noise, Variance=5.0)

Source: [DAG Recovery on 200-node ER4 Graphs (Gaussian Noise, Variance=5.0)](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0>)

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

Data updated: 2026-05-25T08:47:36.384Z

Evaluation of Directed Acyclic Graph (DAG) recovery algorithms on synthetic graphs generated using the ER4 model structure, featuring 200 nodes, homoscedastic Gaussian noise, and a variance of 5.0. Metrics include SHD, SID, SHD-C, FDR, and TPR.

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 | Directed Acyclic Graph (DAG) recovery |
| Dataset (including any stated split) | 200-node ER4 graphs homoscedastic Gaussian noise, variance = 5.0 |
| Metric | SHD |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 4 of 4 |
| 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

- [SHD](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0.md?metric=SHD>): Lower is better; unit: SCALAR.

- [SID](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0.md?metric=SID>): Lower is better; unit: SCALAR.

- [SHD-C](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0.md?metric=SHD-C>): Lower is better; unit: SCALAR.

- [FDR](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0.md?metric=FDR>): Lower is better; unit: PERCENT.

- [TPR](<https://www.sota2.com/research/sota/dag-recovery-on-200-node-er4-graphs-gaussian-noise-variance-5-0.md?metric=TPR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | SHD | Result date | Report paper |
| --- | --- | --- | --- |
| CoLiDE-EV | 105.6 | Not recorded | 6a13d76b6f71c5ae7dc9e7c4 |
| CoLiDE-NV | 157 | Not recorded | 6a13d76b6f71c5ae7dc9e7c4 |
| DAGMA | 194.4 | Not recorded | 6a13d76b6f71c5ae7dc9e7c4 |
| GOLEM | 336.6 | Not recorded | 6a13d76b6f71c5ae7dc9e7c4 |

### CoLiDE-EV

SHD: 105.6. Result ID: 6a140c47fbc373592c0b4f8d.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

Recorded evaluation configuration: {"noise variance":"5.0"}. This record does not assert independent reproduction.

### CoLiDE-NV

SHD: 157. Result ID: 6a140c47fbc373592c0b4f8c.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

Recorded evaluation configuration: {"noise variance":"5.0"}. This record does not assert independent reproduction.

### DAGMA

SHD: 194.4. Result ID: 6a140c47fbc373592c0b4f8b.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5b1e209131cea3f52d.md>)

[Implementation](<https://github.com/kevinsbello/dagma>)

Recorded evaluation configuration: {"noise variance":"5.0"}. This record does not assert independent reproduction.

### GOLEM

SHD: 336.6. Result ID: 6a140c47fbc373592c0b4f8a.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d76b6f71c5ae7dc9e7c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa690c1e209131cea1fba9.md>)

Recorded evaluation configuration: {"noise variance":"5.0"}. This record does not assert independent reproduction.

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