# Dialogue Annotation on Eedi (test)

Source: [Dialogue Annotation on Eedi (test)](<https://www.sota2.com/research/sota/dialogue-annotation-on-eedi-test>)

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

Data updated: 2026-04-06T08:57:23.424Z

Evaluation of dialogue annotation performance on the test split of the Eedi dataset, measured using Cohen's Kappa and Accuracy.

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 | Dialogue Annotation |
| Dataset (including any stated split) | Eedi (test) |
| Metric | Cohen's Kappa |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 12 of 12 |
| 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

- [Cohen's Kappa](<https://www.sota2.com/research/sota/dialogue-annotation-on-eedi-test.md?metric=Cohen%27s%20Kappa>): Higher is better; unit: SCALAR.

- [Accuracy](<https://www.sota2.com/research/sota/dialogue-annotation-on-eedi-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Cohen's Kappa | Result date | Report paper |
| --- | --- | --- | --- |
| Sonnet 4.6 (RAG\_FINETUNED\_UTT) | 0.743 | Not recorded | 69d33df57d0172140ec73651 |
| Qwen3-32b (RAG\_FINETUNED\_UTT) | 0.694 | Not recorded | 69d33df57d0172140ec73651 |
| GPT-5.2 (RAG\_FINETUNED\_UTT) | 0.659 | Not recorded | 69d33df57d0172140ec73651 |
| Sonnet 4.6 (RAG\_NO\_FINETUNE) | 0.632 | Not recorded | 69d33df57d0172140ec73651 |
| Sonnet 4.6 (RAG\_FINETUNED\_CHUNK) | 0.632 | Not recorded | 69d33df57d0172140ec73651 |
| Qwen3-32b (RAG\_NO\_FINETUNE) | 0.59 | Not recorded | 69d33df57d0172140ec73651 |
| Qwen3-32b (RAG\_FINETUNED\_CHUNK) | 0.545 | Not recorded | 69d33df57d0172140ec73651 |
| GPT-5.2 (RAG\_FINETUNED\_CHUNK) | 0.53 | Not recorded | 69d33df57d0172140ec73651 |
| GPT-5.2 (RAG\_NO\_FINETUNE) | 0.505 | Not recorded | 69d33df57d0172140ec73651 |
| Sonnet 4.6 (NO\_RAG) | 0.41 | Not recorded | 69d33df57d0172140ec73651 |
| GPT-5.2 (NO\_RAG) | 0.351 | Not recorded | 69d33df57d0172140ec73651 |
| Qwen3-32b (NO\_RAG) | 0.16 | Not recorded | 69d33df57d0172140ec73651 |

### Sonnet 4.6 (RAG\_FINETUNED\_UTT)

Cohen's Kappa: 0.743. Result ID: 69d3757147f34ce924bb70bb.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Sonnet 4.6","Condition":"RAG\_FINETUNED\_UTT","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Qwen3-32b (RAG\_FINETUNED\_UTT)

Cohen's Kappa: 0.694. Result ID: 69d3757147f34ce924bb70bf.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Qwen3-32b","Condition":"RAG\_FINETUNED\_UTT","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### GPT-5.2 (RAG\_FINETUNED\_UTT)

Cohen's Kappa: 0.659. Result ID: 69d3757147f34ce924bb70b7.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"GPT-5.2","Condition":"RAG\_FINETUNED\_UTT","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Sonnet 4.6 (RAG\_NO\_FINETUNE)

Cohen's Kappa: 0.632. Result ID: 69d3757147f34ce924bb70b9.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Sonnet 4.6","Condition":"RAG\_NO\_FINETUNE","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Sonnet 4.6 (RAG\_FINETUNED\_CHUNK)

Cohen's Kappa: 0.632. Result ID: 69d3757147f34ce924bb70ba.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Sonnet 4.6","Condition":"RAG\_FINETUNED\_CHUNK","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Qwen3-32b (RAG\_NO\_FINETUNE)

Cohen's Kappa: 0.59. Result ID: 69d3757147f34ce924bb70bd.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Qwen3-32b","Condition":"RAG\_NO\_FINETUNE","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Qwen3-32b (RAG\_FINETUNED\_CHUNK)

Cohen's Kappa: 0.545. Result ID: 69d3757147f34ce924bb70be.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Qwen3-32b","Condition":"RAG\_FINETUNED\_CHUNK","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### GPT-5.2 (RAG\_FINETUNED\_CHUNK)

Cohen's Kappa: 0.53. Result ID: 69d3757147f34ce924bb70b6.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"GPT-5.2","Condition":"RAG\_FINETUNED\_CHUNK","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### GPT-5.2 (RAG\_NO\_FINETUNE)

Cohen's Kappa: 0.505. Result ID: 69d3757147f34ce924bb70b5.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"GPT-5.2","Condition":"RAG\_NO\_FINETUNE","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Sonnet 4.6 (NO\_RAG)

Cohen's Kappa: 0.41. Result ID: 69d3757147f34ce924bb70b8.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Sonnet 4.6","Condition":"NO\_RAG","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### GPT-5.2 (NO\_RAG)

Cohen's Kappa: 0.351. Result ID: 69d3757147f34ce924bb70b4.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"GPT-5.2","Condition":"NO\_RAG","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

### Qwen3-32b (NO\_RAG)

Cohen's Kappa: 0.16. Result ID: 69d3757147f34ce924bb70bc.

[Reporting paper](<https://www.sota2.com/research/paper/69d33df57d0172140ec73651.md>)

Recorded evaluation configuration: {"Model":"Qwen3-32b","Condition":"NO\_RAG","Retrieval depth (k)":"3"}. This record does not assert independent reproduction.

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