# Human Preference (Realism) on ImageNet (val)

Source: [Human Preference (Realism) on ImageNet (val)](<https://www.sota2.com/research/sota/human-preference-realism-on-imagenet-val>)

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

Data updated: 2026-03-25T14:52:51.450Z

Evaluation of model performance regarding realism preferences based on human judgment using the validation split of the ImageNet dataset.

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 | Human Preference (Realism) |
| Dataset (including any stated split) | ImageNet (val) |
| Metric | Preference Score (Ours) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [Preference Score (Ours)](<https://www.sota2.com/research/sota/human-preference-realism-on-imagenet-val.md?metric=Preference%20Score%20(Ours)>): Higher is better; unit: SCALAR.

- [Preference Tie Count](<https://www.sota2.com/research/sota/human-preference-realism-on-imagenet-val.md?metric=Preference%20Tie%20Count>): Higher is better; unit: SCALAR.

- [Preference Score (Pretrained)](<https://www.sota2.com/research/sota/human-preference-realism-on-imagenet-val.md?metric=Preference%20Score%20(Pretrained)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Preference Score (Ours) | Result date | Report paper |
| --- | --- | --- | --- |
| Ours (LlamaGen-L backbone) | 318 | Not recorded | 69c36c480636e5bf4afc671a |
| Ours (LlamaGen-L backbone) | 117 | Not recorded | 69c36c480636e5bf4afc671a |

### Ours (LlamaGen-L backbone)

Preference Score (Ours): 318. Result ID: 69c3f6b347f34ce924adf3b3.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c480636e5bf4afc671a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c36c480636e5bf4afc671a.md>)

Recorded evaluation configuration: {"Setting":"No CFG"}. This record does not assert independent reproduction.

### Ours (LlamaGen-L backbone)

Preference Score (Ours): 117. Result ID: 69c3f6b347f34ce924adf3b2.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c480636e5bf4afc671a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c36c480636e5bf4afc671a.md>)

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

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