# Social Support Dialogue Generation on EmpatheticDialogues (test)

Source: [Social Support Dialogue Generation on EmpatheticDialogues (test)](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test>)

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

Data updated: 2026-04-21T13:20:51.379Z

Evaluation of models generating social support dialogues on the test split of the EmpatheticDialogues dataset, measured using specific metrics including D-2, Empathy (Emp.), Helpfulness (Hel.), and Safety (Saf.) scores.

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 | Social Support Dialogue Generation |
| Dataset (including any stated split) | EmpatheticDialogues (test) |
| Metric | D-2 Score |
| Unit | SCALAR |
| 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

- [D-2 Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=D-2%20Score>): Higher is better; unit: SCALAR.

- [Empathy Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Empathy%20Score>): Higher is better; unit: SCALAR.

- [Helpfulness Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Helpfulness%20Score>): Higher is better; unit: SCALAR.

- [Pre. Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Pre.%20Score>): Higher is better; unit: SCALAR.

- [Informativeness Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Informativeness%20Score>): Higher is better; unit: SCALAR.

- [Fluency Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Fluency%20Score>): Higher is better; unit: SCALAR.

- [Safety Score](<https://www.sota2.com/research/sota/social-support-dialogue-generation-on-empatheticdialogues-test.md?metric=Safety%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | D-2 Score | Result date | Report paper |
| --- | --- | --- | --- |
| GPT-Soc. | 12.23 | Not recorded | 69e705247d0172140ec7f2e1 |
| ComPASS | 11.35 | Not recorded | 69e705247d0172140ec7f2e1 |
| GPT-Emp. | 9.39 | Not recorded | 69e705247d0172140ec7f2e1 |
| Sibyl | 6.42 | Not recorded | 69e705247d0172140ec7f2e1 |

### GPT-Soc.

D-2 Score: 12.23. Result ID: 69e7799b47f34ce924cf58db.

[Reporting paper](<https://www.sota2.com/research/paper/69e705247d0172140ec7f2e1.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5.1","Dialogue Style":"Social Support","Profile-based setting":"true"}. This record does not assert independent reproduction.

### ComPASS

D-2 Score: 11.35. Result ID: 69e7799b47f34ce924cf58dc.

[Reporting paper](<https://www.sota2.com/research/paper/69e705247d0172140ec7f2e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e705247d0172140ec7f2e1.md>)

Recorded evaluation configuration: {"Model":"ComPASS-Qwen","Tool-augmented":"true","Dialogue Style":"Social Support","Profile-based setting":"true"}. This record does not assert independent reproduction.

### GPT-Emp.

D-2 Score: 9.39. Result ID: 69e7799b47f34ce924cf58dd.

[Reporting paper](<https://www.sota2.com/research/paper/69e705247d0172140ec7f2e1.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5.1","Dialogue Style":"Empathetic","Profile-based setting":"true"}. This record does not assert independent reproduction.

### Sibyl

D-2 Score: 6.42. Result ID: 69e7799b47f34ce924cf58de.

[Reporting paper](<https://www.sota2.com/research/paper/69e705247d0172140ec7f2e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76b11e209131cea6e7f2.md>)

Recorded evaluation configuration: {"Fine-tuned":"true","Training Data":"EmpatheticDialogues","Dialogue Style":"Empathetic","Profile-based setting":"true"}. This record does not assert independent reproduction.

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