# Machine-generated text detection on MGTBench 2.0

Source: [Machine-generated text detection on MGTBench 2.0](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0>)

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

Data updated: 2026-02-26T05:43:53.655Z

Evaluation of machine-generated text detection models using the MGTBench 2.0 dataset, reporting F1 scores for various LLMs including Llama 3.1, Mixtral, GPT-4o-mini, GPT-3.5, and MoonShot-8k.

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 | Machine-generated text detection |
| Dataset (including any stated split) | MGTBench 2.0 |
| Metric | F1 Score (Llama-3.1-70b - ST) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [F1 Score (Llama-3.1-70b - ST)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Llama-3.1-70b%20-%20ST)>): Higher is better; unit: PERCENT.

- [F1 Score (Llama-3.1-70b - Hu)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Llama-3.1-70b%20-%20Hu)>): Higher is better; unit: PERCENT.

- [F1 Score (Llama-3.1-70b - So)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Llama-3.1-70b%20-%20So)>): Higher is better; unit: PERCENT.

- [F1 Score (Mixtral-8x7b - ST)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Mixtral-8x7b%20-%20ST)>): Higher is better; unit: PERCENT.

- [F1 Score (Mixtral-8x7b - Hu)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Mixtral-8x7b%20-%20Hu)>): Higher is better; unit: PERCENT.

- [F1 Score (Mixtral-8x7b - So)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(Mixtral-8x7b%20-%20So)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-4o-mini - ST)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-4o-mini%20-%20ST)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-4o-mini - Hu)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-4o-mini%20-%20Hu)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-4o-mini - So)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-4o-mini%20-%20So)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-3.5 - ST)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-3.5%20-%20ST)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-3.5 - Hu)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-3.5%20-%20Hu)>): Higher is better; unit: PERCENT.

- [F1 Score (GPT-3.5 - So)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(GPT-3.5%20-%20So)>): Higher is better; unit: PERCENT.

- [F1 Score (MoonShot-8k - ST)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(MoonShot-8k%20-%20ST)>): Higher is better; unit: PERCENT.

- [F1 Score (MoonShot-8k - Hu)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(MoonShot-8k%20-%20Hu)>): Higher is better; unit: PERCENT.

- [F1 Score (MoonShot-8k - So)](<https://www.sota2.com/research/sota/machine-generated-text-detection-on-mgtbench-2-0.md?metric=F1%20Score%20(MoonShot-8k%20-%20So)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | F1 Score (Llama-3.1-70b - ST) | Result date | Report paper |
| --- | --- | --- | --- |
| DistillBert-F | 99.7 | Not recorded | 69608b6553fdd5c361744159 |
| Roberta-F | 99.5 | Not recorded | 69608b6553fdd5c361744159 |
| Binoculars | 89.7 | Not recorded | 69608b6553fdd5c361744159 |
| Binoculars | 84.3 | Not recorded | 69608b6553fdd5c361744159 |
| Rank\_GLTR | 83 | Not recorded | 69608b6553fdd5c361744159 |
| ENTROPY | 81 | Not recorded | 69608b6553fdd5c361744159 |
| LL | 80.9 | Not recorded | 69608b6553fdd5c361744159 |
| Rank | 79.7 | Not recorded | 69608b6553fdd5c361744159 |
| LL | 68.9 | Not recorded | 69608b6553fdd5c361744159 |
| ENTROPY | 66.8 | Not recorded | 69608b6553fdd5c361744159 |
| Roberta-F | 66.7 | Not recorded | 69608b6553fdd5c361744159 |
| Rank\_GLTR | 65.5 | Not recorded | 69608b6553fdd5c361744159 |
| DistillBert-F | 62.2 | Not recorded | 69608b6553fdd5c361744159 |
| Rank | 53.8 | Not recorded | 69608b6553fdd5c361744159 |

### DistillBert-F

F1 Score (Llama-3.1-70b - ST): 99.7. Result ID: 6996f3f15698665ec44096b4.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### Roberta-F

F1 Score (Llama-3.1-70b - ST): 99.5. Result ID: 6996f3f15698665ec44096b6.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### Binoculars

F1 Score (Llama-3.1-70b - ST): 89.7. Result ID: 6996f3f15698665ec44096b2.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### Binoculars

F1 Score (Llama-3.1-70b - ST): 84.3. Result ID: 6996f3f15698665ec44096b1.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### Rank\_GLTR

F1 Score (Llama-3.1-70b - ST): 83. Result ID: 6996f3f15698665ec44096ae.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### ENTROPY

F1 Score (Llama-3.1-70b - ST): 81. Result ID: 6996f3f15698665ec44096b0.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### LL

F1 Score (Llama-3.1-70b - ST): 80.9. Result ID: 6996f3f15698665ec44096aa.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### Rank

F1 Score (Llama-3.1-70b - ST): 79.7. Result ID: 6996f3f15698665ec44096ac.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"DP-MGTD (Ours)"}. This record does not assert independent reproduction.

### LL

F1 Score (Llama-3.1-70b - ST): 68.9. Result ID: 6996f3f15698665ec44096a9.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### ENTROPY

F1 Score (Llama-3.1-70b - ST): 66.8. Result ID: 6996f3f15698665ec44096af.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### Roberta-F

F1 Score (Llama-3.1-70b - ST): 66.7. Result ID: 6996f3f15698665ec44096b5.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### Rank\_GLTR

F1 Score (Llama-3.1-70b - ST): 65.5. Result ID: 6996f3f15698665ec44096ad.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### DistillBert-F

F1 Score (Llama-3.1-70b - ST): 62.2. Result ID: 6996f3f15698665ec44096b3.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

### Rank

F1 Score (Llama-3.1-70b - ST): 53.8. Result ID: 6996f3f15698665ec44096ab.

[Reporting paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69608b6553fdd5c361744159.md>)

Recorded evaluation configuration: {"Setting":"Base"}. This record does not assert independent reproduction.

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