# Open-domain Conversation on Chinese open-domain conversation Self-chat (test)

Source: [Open-domain Conversation on Chinese open-domain conversation Self-chat (test)](<https://www.sota2.com/research/sota/open-domain-conversation-on-chinese-open-domain-conversation-self-chat-test>)

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

Data updated: 2026-02-26T05:39:29.523Z

Evaluation of Open-domain Conversation performance on the test split of the Chinese open-domain conversation Self-chat dataset, utilizing human-centric metrics such as Coherence, Informativeness, Safety, and Engagingness.

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 | Open-domain Conversation |
| Dataset (including any stated split) | Chinese open-domain conversation Self-chat (test) |
| Metric | Coherence |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Coherence](<https://www.sota2.com/research/sota/open-domain-conversation-on-chinese-open-domain-conversation-self-chat-test.md?metric=Coherence>): Higher is better; unit: PERCENT.

- [Informativeness](<https://www.sota2.com/research/sota/open-domain-conversation-on-chinese-open-domain-conversation-self-chat-test.md?metric=Informativeness>): Higher is better; unit: PERCENT.

- [Safety](<https://www.sota2.com/research/sota/open-domain-conversation-on-chinese-open-domain-conversation-self-chat-test.md?metric=Safety>): Higher is better; unit: PERCENT.

- [Engagingness](<https://www.sota2.com/research/sota/open-domain-conversation-on-chinese-open-domain-conversation-self-chat-test.md?metric=Engagingness>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Coherence | Result date | Report paper |
| --- | --- | --- | --- |
| PLATO-XL (Diamante) | 194.8 | Not recorded | 68fa6d4d1e209131cea38f33 |
| PLATO-XL | 178.8 | Not recorded | 68fa6d4d1e209131cea38f33 |
| EVA 2.0 | 150.8 | Not recorded | 68fa6d4d1e209131cea38f33 |
| CDial-GPT | 48.4 | Not recorded | 68fa6d4d1e209131cea38f33 |

### PLATO-XL (Diamante)

Coherence: 194.8. Result ID: 698f85f63a6c314849f9ad72.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38f33.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38f33.md>)

[Implementation](<https://github.com/PaddlePaddle/Knover/tree/develop/projects/Diamante>)

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

### PLATO-XL

Coherence: 178.8. Result ID: 698f85f63a6c314849f9ad71.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38f33.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dbd1e209131cea3bc04.md>)

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

### EVA 2.0

Coherence: 150.8. Result ID: 698f85f63a6c314849f9ad70.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38f33.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f41e209131cea4e5ca.md>)

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

### CDial-GPT

Coherence: 48.4. Result ID: 698f85f63a6c314849f9ad6f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d4d1e209131cea38f33.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c661e209131cea33d70.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.