# Compound AI System Alignment on Coherent and Progressive Image Sequence Generation (test)

Source: [Compound AI System Alignment on Coherent and Progressive Image Sequence Generation (test)](<https://www.sota2.com/research/sota/compound-ai-system-alignment-on-coherent-and-progressive-image-sequence-generation-test>)

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

Data updated: 2026-03-09T06:39:14.723Z

Evaluation of Compound AI System Alignment performance on the test split of the Coherent and Progressive Image Sequence Generation dataset, measured by Preference Score and OC Ratio.

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 | Compound AI System Alignment |
| Dataset (including any stated split) | Coherent and Progressive Image Sequence Generation (test) |
| Metric | Preference Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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](<https://www.sota2.com/research/sota/compound-ai-system-alignment-on-coherent-and-progressive-image-sequence-generation-test.md?metric=Preference%20Score>): Higher is better; unit: SCALAR.

- [OC Ratio](<https://www.sota2.com/research/sota/compound-ai-system-alignment-on-coherent-and-progressive-image-sequence-generation-test.md?metric=OC%20Ratio>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Preference Score | Result date | Report paper |
| --- | --- | --- | --- |
| SysDPO-Direct | 0.25 | Not recorded | 68fa790f1e209131cea7be6f |
| Only Train Language Model | 0.23 | Not recorded | 68fa790f1e209131cea7be6f |
| Best-of-Sampling | 0.16 | Not recorded | 68fa790f1e209131cea7be6f |
| Only Train Diffusion Model | -0.03 | Not recorded | 68fa790f1e209131cea7be6f |
| System Before Alignment | -0.2 | Not recorded | 68fa790f1e209131cea7be6f |

### SysDPO-Direct

Preference Score: 0.25. Result ID: 69ae6ae66dbb7075b955da7e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

Recorded evaluation configuration: {"Mode":"Proposed"}. This record does not assert independent reproduction.

### Only Train Language Model

Preference Score: 0.23. Result ID: 69ae6ae66dbb7075b955da81.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

Recorded evaluation configuration: {"Language Model":"Trained","Diffusion Model":"Frozen"}. This record does not assert independent reproduction.

### Best-of-Sampling

Preference Score: 0.16. Result ID: 69ae6ae66dbb7075b955da80.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

Recorded evaluation configuration: {"Sample Size":"4"}. This record does not assert independent reproduction.

### Only Train Diffusion Model

Preference Score: -0.03. Result ID: 69ae6ae66dbb7075b955da82.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

Recorded evaluation configuration: {"Language Model":"Frozen","Diffusion Model":"Trained"}. This record does not assert independent reproduction.

### System Before Alignment

Preference Score: -0.2. Result ID: 69ae6ae66dbb7075b955da7f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be6f.md>)

Recorded evaluation configuration: {"Backbone":"Llama-3-8B"}. This record does not assert independent reproduction.

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