# One-Time Edit on CounterFact

Source: [One-Time Edit on CounterFact](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact>)

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

Data updated: 2026-05-27T08:50:43.407Z

Evaluation of one-time model editing performance on the CounterFact dataset, measured using Efficacy, Generalization, Specificity, Fluency, and Consistency.

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 | One-Time Edit |
| Dataset (including any stated split) | CounterFact |
| Metric | Efficacy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 20 of 20 |
| 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

- [Efficacy](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact.md?metric=Efficacy>): Higher is better; unit: PERCENT.

- [Generalization](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact.md?metric=Generalization>): Higher is better; unit: PERCENT.

- [Specificity](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact.md?metric=Specificity>): Higher is better; unit: PERCENT.

- [Fluency Score](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact.md?metric=Fluency%20Score>): Higher is better; unit: SCALAR.

- [Consistency](<https://www.sota2.com/research/sota/one-time-edit-on-counterfact.md?metric=Consistency>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Efficacy | Result date | Report paper |
| --- | --- | --- | --- |
| MEMIT (aligned) | 99.88 | Not recorded | 6a167a7edbd7150506f7ff10 |
| AlphaEdit (aligned) | 99.88 | Not recorded | 6a167a7edbd7150506f7ff10 |
| MEMIT (aligned) | 99.83 | Not recorded | 6a167a7edbd7150506f7ff10 |
| PRUNE (aligned) | 99.82 | Not recorded | 6a167a7edbd7150506f7ff10 |
| MEMIT (aligned) | 99.8 | Not recorded | 6a167a7edbd7150506f7ff10 |
| PRUNE (aligned) | 99.8 | Not recorded | 6a167a7edbd7150506f7ff10 |
| AlphaEdit (aligned) | 99.77 | Not recorded | 6a167a7edbd7150506f7ff10 |
| PRUNE (aligned) | 99.57 | Not recorded | 6a167a7edbd7150506f7ff10 |
| AlphaEdit (aligned) | 99.57 | Not recorded | 6a167a7edbd7150506f7ff10 |
| PRUNE (aligned) | 99.23 | Not recorded | 6a167a7edbd7150506f7ff10 |
| MEMIT (aligned) | 99.22 | Not recorded | 6a167a7edbd7150506f7ff10 |
| RECT (aligned) | 99.12 | Not recorded | 6a167a7edbd7150506f7ff10 |
| RECT (aligned) | 98.27 | Not recorded | 6a167a7edbd7150506f7ff10 |
| AlphaEdit (aligned) | 95.3 | Not recorded | 6a167a7edbd7150506f7ff10 |
| RECT (aligned) | 95.12 | Not recorded | 6a167a7edbd7150506f7ff10 |
| RECT (aligned) | 83.23 | Not recorded | 6a167a7edbd7150506f7ff10 |
| Pre-edited | 22.23 | Not recorded | 6a167a7edbd7150506f7ff10 |
| Pre-edited | 16.22 | Not recorded | 6a167a7edbd7150506f7ff10 |
| Pre-edited | 13.95 | Not recorded | 6a167a7edbd7150506f7ff10 |
| Pre-edited | 7.85 | Not recorded | 6a167a7edbd7150506f7ff10 |

### MEMIT (aligned)

Efficacy: 99.88. Result ID: 6a16b035fbc373592c0ff2b9.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5"}. This record does not assert independent reproduction.

### AlphaEdit (aligned)

Efficacy: 99.88. Result ID: 6a16b035fbc373592c0ff2ba.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5"}. This record does not assert independent reproduction.

### MEMIT (aligned)

Efficacy: 99.83. Result ID: 6a16b035fbc373592c0ff2af.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

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

### PRUNE (aligned)

Efficacy: 99.82. Result ID: 6a16b035fbc373592c0ff2bb.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5"}. This record does not assert independent reproduction.

### MEMIT (aligned)

Efficacy: 99.8. Result ID: 6a16b035fbc373592c0ff2aa.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"LLaMA3"}. This record does not assert independent reproduction.

### PRUNE (aligned)

Efficacy: 99.8. Result ID: 6a16b035fbc373592c0ff2b1.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

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

### AlphaEdit (aligned)

Efficacy: 99.77. Result ID: 6a16b035fbc373592c0ff2b0.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

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

### PRUNE (aligned)

Efficacy: 99.57. Result ID: 6a16b035fbc373592c0ff2ac.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"LLaMA3"}. This record does not assert independent reproduction.

### AlphaEdit (aligned)

Efficacy: 99.57. Result ID: 6a16b035fbc373592c0ff2b5.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"GPT-2 XL"}. This record does not assert independent reproduction.

### PRUNE (aligned)

Efficacy: 99.23. Result ID: 6a16b035fbc373592c0ff2b6.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"GPT-2 XL"}. This record does not assert independent reproduction.

### MEMIT (aligned)

Efficacy: 99.22. Result ID: 6a16b035fbc373592c0ff2b4.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"GPT-2 XL"}. This record does not assert independent reproduction.

### RECT (aligned)

Efficacy: 99.12. Result ID: 6a16b035fbc373592c0ff2bc.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5"}. This record does not assert independent reproduction.

### RECT (aligned)

Efficacy: 98.27. Result ID: 6a16b035fbc373592c0ff2b2.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

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

### AlphaEdit (aligned)

Efficacy: 95.3. Result ID: 6a16b035fbc373592c0ff2ab.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"LLaMA3"}. This record does not assert independent reproduction.

### RECT (aligned)

Efficacy: 95.12. Result ID: 6a16b035fbc373592c0ff2b7.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"GPT-2 XL"}. This record does not assert independent reproduction.

### RECT (aligned)

Efficacy: 83.23. Result ID: 6a16b035fbc373592c0ff2ad.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"LLaMA3"}. This record does not assert independent reproduction.

### Pre-edited

Efficacy: 22.23. Result ID: 6a16b035fbc373592c0ff2b3.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"GPT-2 XL"}. This record does not assert independent reproduction.

### Pre-edited

Efficacy: 16.22. Result ID: 6a16b035fbc373592c0ff2ae.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

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

### Pre-edited

Efficacy: 13.95. Result ID: 6a16b035fbc373592c0ff2b8.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5"}. This record does not assert independent reproduction.

### Pre-edited

Efficacy: 7.85. Result ID: 6a16b035fbc373592c0ff2a9.

[Reporting paper](<https://www.sota2.com/research/paper/6a167a7edbd7150506f7ff10.md>)

Recorded evaluation configuration: {"Model":"LLaMA3"}. This record does not assert independent reproduction.

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