# Justification Quality Evaluation on AVeriTeC Retrieved (H) 50 correctly verified claims

Source: [Justification Quality Evaluation on AVeriTeC Retrieved (H) 50 correctly verified claims](<https://www.sota2.com/research/sota/justification-quality-evaluation-on-averitec-retrieved-h-50-correctly-verified-claims>)

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

Data updated: 2026-04-06T07:09:15.280Z

Evaluation of model performance on the specific subset of 50 correctly verified claims from the AVeriTeC Retrieved (H) dataset, focusing on the quality of generated justifications.

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 | Justification Quality Evaluation |
| Dataset (including any stated split) | AVeriTeC Retrieved (H) 50 correctly verified claims |
| Metric | MOS |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [MOS](<https://www.sota2.com/research/sota/justification-quality-evaluation-on-averitec-retrieved-h-50-correctly-verified-claims.md?metric=MOS>): Higher is better; unit: SCALAR.

- [Win Rate](<https://www.sota2.com/research/sota/justification-quality-evaluation-on-averitec-retrieved-h-50-correctly-verified-claims.md?metric=Win%20Rate>): Higher is better; unit: PERCENT.

- [Tie Rate](<https://www.sota2.com/research/sota/justification-quality-evaluation-on-averitec-retrieved-h-50-correctly-verified-claims.md?metric=Tie%20Rate>): Higher is better; unit: PERCENT.

- [Loss Rate](<https://www.sota2.com/research/sota/justification-quality-evaluation-on-averitec-retrieved-h-50-correctly-verified-claims.md?metric=Loss%20Rate>): Lower is better; unit: PERCENT.

## Recorded results

| Method | MOS | Result date | Report paper |
| --- | --- | --- | --- |
| DebateCV | 3.67 | Not recorded | 68fa7a391e209131cea81e62 |
| DebateCV | 3.59 | Not recorded | 68fa7a391e209131cea81e62 |
| DebateCV | 3.07 | Not recorded | 68fa7a391e209131cea81e62 |
| HerO | 2.6 | Not recorded | 68fa7a391e209131cea81e62 |
| HerO | 2.49 | Not recorded | 68fa7a391e209131cea81e62 |
| HerO | 2.22 | Not recorded | 68fa7a391e209131cea81e62 |

### DebateCV

MOS: 3.67. Result ID: 69d35bfd47f34ce924baaf91.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

Recorded evaluation configuration: {"Criterion":"Evidence Used","Evidence condition":"Retrieved (H)","Backbone":"Llama-3.1-8B"}. This record does not assert independent reproduction.

### DebateCV

MOS: 3.59. Result ID: 69d35bfd47f34ce924baaf95.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

Recorded evaluation configuration: {"Criterion":"Reasoning Pathway","Evidence condition":"Retrieved (H)","Backbone":"Llama-3.1-8B"}. This record does not assert independent reproduction.

### DebateCV

MOS: 3.07. Result ID: 69d35bfd47f34ce924baaf93.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

Recorded evaluation configuration: {"Criterion":"Sources of Uncertainty","Evidence condition":"Retrieved (H)","Backbone":"Llama-3.1-8B"}. This record does not assert independent reproduction.

### HerO

MOS: 2.6. Result ID: 69d35bfd47f34ce924baaf90.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea6733f.md>)

[Implementation](<https://github.com/ssu-humane/HerO>)

Recorded evaluation configuration: {"Criterion":"Evidence Used","Evidence condition":"Retrieved (H)"}. This record does not assert independent reproduction.

### HerO

MOS: 2.49. Result ID: 69d35bfd47f34ce924baaf94.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea6733f.md>)

[Implementation](<https://github.com/ssu-humane/HerO>)

Recorded evaluation configuration: {"Criterion":"Reasoning Pathway","Evidence condition":"Retrieved (H)"}. This record does not assert independent reproduction.

### HerO

MOS: 2.22. Result ID: 69d35bfd47f34ce924baaf92.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a391e209131cea81e62.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea6733f.md>)

[Implementation](<https://github.com/ssu-humane/HerO>)

Recorded evaluation configuration: {"Criterion":"Sources of Uncertainty","Evidence condition":"Retrieved (H)"}. This record does not assert independent reproduction.

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