# Prompt recovery attack on CounterFact (test)

Source: [Prompt recovery attack on CounterFact (test)](<https://www.sota2.com/research/sota/prompt-recovery-attack-on-counterfact-test>)

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

Data updated: 2026-02-26T06:07:07.753Z

Evaluation of Prompt Recovery Attack performance on the test split of the CounterFact dataset, measured by Top-1, Top-5, Top-20 accuracy, and Similarity.

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 recovery attack |
| Dataset (including any stated split) | CounterFact (test) |
| Metric | Top-1 Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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 Accuracy](<https://www.sota2.com/research/sota/prompt-recovery-attack-on-counterfact-test.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

- [Top-5 Accuracy](<https://www.sota2.com/research/sota/prompt-recovery-attack-on-counterfact-test.md?metric=Top-5%20Accuracy>): Higher is better; unit: PERCENT.

- [Top-20 Accuracy](<https://www.sota2.com/research/sota/prompt-recovery-attack-on-counterfact-test.md?metric=Top-20%20Accuracy>): Higher is better; unit: PERCENT.

- [Similarity](<https://www.sota2.com/research/sota/prompt-recovery-attack-on-counterfact-test.md?metric=Similarity>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| AlphaEdit | 54 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 52 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 48 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 47 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 45 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 44 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 43 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 36 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |
| AlphaEdit | 30 | Not recorded | 698d5e62c97a0b3e3b5e3c5a |

### AlphaEdit

Top-1 Accuracy: 54. Result ID: 6996db375698665ec427a919.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Llama-3-Instruct","N":10}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 52. Result ID: 6996db375698665ec427a918.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"GPT-J","N":100}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 48. Result ID: 6996db375698665ec427a916.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"GPT-J","N":10}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 47. Result ID: 6996db375698665ec427a917.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"GPT-J","N":50}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 45. Result ID: 6996db375698665ec427a91b.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Llama-3-Instruct","N":100}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 44. Result ID: 6996db375698665ec427a91a.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Llama-3-Instruct","N":50}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 43. Result ID: 6996db375698665ec427a91e.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Qwen-2.5-Instruct","N":100}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 36. Result ID: 6996db375698665ec427a91d.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Qwen-2.5-Instruct","N":50}. This record does not assert independent reproduction.

### AlphaEdit

Top-1 Accuracy: 30. Result ID: 6996db375698665ec427a91c.

[Reporting paper](<https://www.sota2.com/research/paper/698d5e62c97a0b3e3b5e3c5a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77370.md>)

Recorded evaluation configuration: {"Model":"Qwen-2.5-Instruct","N":10}. This record does not assert independent reproduction.

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