# Long-form Question Answering on GroundBench (test)

Source: [Long-form Question Answering on GroundBench (test)](<https://www.sota2.com/research/sota/long-form-question-answering-on-groundbench-test>)

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

Data updated: 2026-02-26T05:45:10.755Z

Evaluation of Long-form Question Answering performance on the test split of the GroundBench dataset, focusing on faithfulness and completeness metrics.

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 | Long-form Question Answering |
| Dataset (including any stated split) | GroundBench (test) |
| Metric | Faithfulness (Full) |
| 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

- [Faithfulness (Full)](<https://www.sota2.com/research/sota/long-form-question-answering-on-groundbench-test.md?metric=Faithfulness%20(Full)>): Higher is better; unit: PERCENT.

- [Faithfulness (Partial)](<https://www.sota2.com/research/sota/long-form-question-answering-on-groundbench-test.md?metric=Faithfulness%20(Partial)>): Higher is better; unit: PERCENT.

- [Faithfulness (No)](<https://www.sota2.com/research/sota/long-form-question-answering-on-groundbench-test.md?metric=Faithfulness%20(No)>): Lower is better; unit: PERCENT.

- [Completeness (Rate)](<https://www.sota2.com/research/sota/long-form-question-answering-on-groundbench-test.md?metric=Completeness%20(Rate)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Faithfulness (Full) | Result date | Report paper |
| --- | --- | --- | --- |
| RHIO-13B | 87.5 | Not recorded | 68fa76ab1e209131cea6e52d |
| GPT-4o | 86.5 | Not recorded | 68fa76ab1e209131cea6e52d |
| Llama-3.1-70B-Instruct | 82.6 | Not recorded | 68fa76ab1e209131cea6e52d |
| SFT-13B | 76.8 | Not recorded | 68fa76ab1e209131cea6e52d |

### RHIO-13B

Faithfulness (Full): 87.5. Result ID: 69945d5b3a6c3148491cc58f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ab1e209131cea6e52d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76ab1e209131cea6e52d.md>)

[Implementation](<https://github.com/LuckyyySTA/faithful-LFQA>)

Recorded evaluation configuration: {"Parameters":"13B"}. This record does not assert independent reproduction.

### GPT-4o

Faithfulness (Full): 86.5. Result ID: 69945d5b3a6c3148491cc58c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ab1e209131cea6e52d.md>)

Recorded evaluation configuration: {"Model Type":"Proprietary"}. This record does not assert independent reproduction.

### Llama-3.1-70B-Instruct

Faithfulness (Full): 82.6. Result ID: 69945d5b3a6c3148491cc58d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ab1e209131cea6e52d.md>)

Recorded evaluation configuration: {"Backbone":"Llama-3.1","Parameters":"70B"}. This record does not assert independent reproduction.

### SFT-13B

Faithfulness (Full): 76.8. Result ID: 69945d5b3a6c3148491cc58e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ab1e209131cea6e52d.md>)

Recorded evaluation configuration: {"Parameters":"13B","Type":"Supervised Fine-Tuned Baseline"}. This record does not assert independent reproduction.

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