# Knowledge-Grounded Dialog on OR-QUAC

Source: [Knowledge-Grounded Dialog on OR-QUAC](<https://www.sota2.com/research/sota/knowledge-grounded-dialog-on-or-quac>)

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

Data updated: 2026-04-23T09:54:10.430Z

Evaluation of model performance on the OR-QUAC dataset for Knowledge-Grounded Dialog tasks, typically measured using generation metrics like BLEU and F1.

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 | Knowledge-Grounded Dialog |
| Dataset (including any stated split) | OR-QUAC |
| Metric | BLEU-4 |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [BLEU-4](<https://www.sota2.com/research/sota/knowledge-grounded-dialog-on-or-quac.md?metric=BLEU-4>): Higher is better; unit: SCALAR.

- [BLEU-1](<https://www.sota2.com/research/sota/knowledge-grounded-dialog-on-or-quac.md?metric=BLEU-1>): Higher is better; unit: SCALAR.

- [F1 Score](<https://www.sota2.com/research/sota/knowledge-grounded-dialog-on-or-quac.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | BLEU-4 | Result date | Report paper |
| --- | --- | --- | --- |
| JSA-RAG | 7.76 | Not recorded | 68fa7ae31e209131cea85b05 |
| VRAG | 6.71 | Not recorded | 68fa7ae31e209131cea85b05 |
| RAG | 6.57 | Not recorded | 68fa7ae31e209131cea85b05 |

### JSA-RAG

BLEU-4: 7.76. Result ID: 69e9ec3047f34ce924d18e86.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ae31e209131cea85b05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ae31e209131cea85b05.md>)

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

### VRAG

BLEU-4: 6.71. Result ID: 69e9ec3047f34ce924d18e85.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ae31e209131cea85b05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d3a1e209131cea3892e.md>)

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

### RAG

BLEU-4: 6.57. Result ID: 69e9ec3047f34ce924d18e84.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7ae31e209131cea85b05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

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.