# AI-generated text detection on MixSet 1.0 (test)

Source: [AI-generated text detection on MixSet 1.0 (test)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test>)

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

Data updated: 2026-02-26T05:35:49.451Z

Evaluation of AI-generated text detection models on the MixSet 1.0 test split, utilizing metrics derived from various LLM and human revision strategies.

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 | AI-generated text detection |
| Dataset (including any stated split) | MixSet 1.0 (test) |
| Metric | Average Score |
| 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

- [Average Score](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

- [LLM Complete (Llama2)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Complete%20(Llama2)>): Higher is better; unit: PERCENT.

- [LLM Complete (GPT-4)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Complete%20(GPT-4)>): Higher is better; unit: PERCENT.

- [LLM Rewrite (Llama2)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Rewrite%20(Llama2)>): Higher is better; unit: PERCENT.

- [LLM Rewrite (GPT-4)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Rewrite%20(GPT-4)>): Higher is better; unit: PERCENT.

- [LLM Polish (Token) (Llama2)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Polish%20(Token)%20(Llama2)>): Higher is better; unit: PERCENT.

- [LLM Polish (Token) (GPT-4)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Polish%20(Token)%20(GPT-4)>): Higher is better; unit: PERCENT.

- [LLM Polish (Sentence) (Llama2)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Polish%20(Sentence)%20(Llama2)>): Higher is better; unit: PERCENT.

- [LLM Polish (Sentence) (GPT-4)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=LLM%20Polish%20(Sentence)%20(GPT-4)>): Higher is better; unit: PERCENT.

- [Humanize (Llama2)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=Humanize%20(Llama2)>): Higher is better; unit: PERCENT.

- [Humanize (GPT-4)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=Humanize%20(GPT-4)>): Higher is better; unit: PERCENT.

- [Human Adapt (Sentence)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=Human%20Adapt%20(Sentence)>): Higher is better; unit: PERCENT.

- [Human Adapt (Token)](<https://www.sota2.com/research/sota/ai-generated-text-detection-on-mixset-1-0-test.md?metric=Human%20Adapt%20(Token)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| GigaCheck | 99 | Not recorded | 68fa75991e209131cea68dbf |
| Radar | 87.6 | Not recorded | 68fa75991e209131cea68dbf |
| GPT-sentinel | 71.3 | Not recorded | 68fa75991e209131cea68dbf |
| DistilBERT | 66.4 | Not recorded | 68fa75991e209131cea68dbf |
| Entropy | 64.8 | Not recorded | 68fa75991e209131cea68dbf |
| DetectGPT | 63.5 | Not recorded | 68fa75991e209131cea68dbf |
| log likelihood | 62.4 | Not recorded | 68fa75991e209131cea68dbf |
| log-rank | 61.5 | Not recorded | 68fa75991e209131cea68dbf |
| GLTR | 58.8 | Not recorded | 68fa75991e209131cea68dbf |

### GigaCheck

Average Score: 99. Result ID: 6993f17c3a6c31484919aa6e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

Recorded evaluation configuration: {"Backbone":"Mistral-7B"}. This record does not assert independent reproduction.

### Radar

Average Score: 87.6. Result ID: 6993f17c3a6c31484919aa6b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f81e209131cea4e73e.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GPT-sentinel

Average Score: 71.3. Result ID: 6993f17c3a6c31484919aa6c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f521e209131cea45074.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### DistilBERT

Average Score: 66.4. Result ID: 6993f17c3a6c31484919aa6d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67221e209131cea14884.md>)

[Implementation](<https://github.com/huggingface/transformers>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Entropy

Average Score: 64.8. Result ID: 6993f17c3a6c31484919aa6a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65e11e209131cea0d18b.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### DetectGPT

Average Score: 63.5. Result ID: 6993f17c3a6c31484919aa69.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa701a1e209131cea49064.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### log likelihood

Average Score: 62.4. Result ID: 6993f17c3a6c31484919aa67.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66ae1e209131cea11b7a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### log-rank

Average Score: 61.5. Result ID: 6993f17c3a6c31484919aa66.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa701a1e209131cea49064.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GLTR

Average Score: 58.8. Result ID: 6993f17c3a6c31484919aa68.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75991e209131cea68dbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65e11e209131cea0d18b.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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