# Prompt Selection on Perturbation-100 (test)

Source: [Prompt Selection on Perturbation-100 (test)](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test>)

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

Data updated: 2026-03-23T11:14:19.248Z

Evaluation of model performance on the Prompt Selection task using the Perturbation-100 test set, measured by metrics such as Top-1 Hit, Mean Regret, and AUROC Score.

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 | Prompt Selection |
| Dataset (including any stated split) | Perturbation-100 (test) |
| Metric | Top-1 Hit |
| Unit | PERCENT |
| Direction | Higher 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

- [Top-1 Hit](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=Top-1%20Hit>): Higher is better; unit: PERCENT.

- [Mean Regret](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=Mean%20Regret>): Lower is better; unit: PERCENT.

- [AUROC Score](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=AUROC%20Score>): Higher is better; unit: SCALAR.

- [95% CI Lower Bound](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=95%25%20CI%20Lower%20Bound>): Higher is better; unit: SCALAR.

- [Delta vs 0.5 (%)](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=Delta%20vs%200.5%20(%25)>): Higher is better; unit: PERCENT.

- [Spearman ρ](<https://www.sota2.com/research/sota/prompt-selection-on-perturbation-100-test.md?metric=Spearman%20%CF%81>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Top-1 Hit | Result date | Report paper |
| --- | --- | --- | --- |
| Risk-Cos | 76 | Not recorded | 69c0c90e0636e5bf4afc3c63 |
| Risk-Margin | 76 | Not recorded | 69c0c90e0636e5bf4afc3c63 |
| Risk-Entropy | 57 | Not recorded | 69c0c90e0636e5bf4afc3c63 |
| Risk-Loss (NLL) | 52 | Not recorded | 69c0c90e0636e5bf4afc3c63 |

### Risk-Cos

Top-1 Hit: 76. Result ID: 69c11f5447f34ce924a97880.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c90e0636e5bf4afc3c63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c0c90e0636e5bf4afc3c63.md>)

Recorded evaluation configuration: {"Mode":"Prefill-only"}. This record does not assert independent reproduction.

### Risk-Margin

Top-1 Hit: 76. Result ID: 69c11f5447f34ce924a97881.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c90e0636e5bf4afc3c63.md>)

Recorded evaluation configuration: {"Mode":"Prefill-only"}. This record does not assert independent reproduction.

### Risk-Entropy

Top-1 Hit: 57. Result ID: 69c11f5447f34ce924a97882.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c90e0636e5bf4afc3c63.md>)

Recorded evaluation configuration: {"Mode":"Prefill-only"}. This record does not assert independent reproduction.

### Risk-Loss (NLL)

Top-1 Hit: 52. Result ID: 69c11f5447f34ce924a97883.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c90e0636e5bf4afc3c63.md>)

Recorded evaluation configuration: {"Mode":"Prefill-only"}. This record does not assert independent reproduction.

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