# Medical Question Answering on Medical QA offline evaluation

Source: [Medical Question Answering on Medical QA offline evaluation](<https://www.sota2.com/research/sota/medical-question-answering-on-medical-qa-offline-evaluation>)

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

Data updated: 2026-02-26T05:46:40.408Z

Evaluation of model performance on the 'Medical QA offline evaluation' dataset for Medical Question Answering, assessed using quality metrics including Honesty, Relevance, and Completeness.

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 | Medical Question Answering |
| Dataset (including any stated split) | Medical QA offline evaluation |
| Metric | Honesty |
| 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

- [Honesty](<https://www.sota2.com/research/sota/medical-question-answering-on-medical-qa-offline-evaluation.md?metric=Honesty>): Higher is better; unit: SCALAR.

- [Relevance](<https://www.sota2.com/research/sota/medical-question-answering-on-medical-qa-offline-evaluation.md?metric=Relevance>): Higher is better; unit: SCALAR.

- [Completeness](<https://www.sota2.com/research/sota/medical-question-answering-on-medical-qa-offline-evaluation.md?metric=Completeness>): Higher is better; unit: SCALAR.

- [Length](<https://www.sota2.com/research/sota/medical-question-answering-on-medical-qa-offline-evaluation.md?metric=Length>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Honesty | Result date | Report paper |
| --- | --- | --- | --- |
| Uni-Normalize Scheme | 0.83 | Not recorded | 698eaffec97a0b3e3b5e66a1 |
| Uni-Normalize Scheme + Like/Dislike Helpful RM | 0.779 | Not recorded | 698eaffec97a0b3e3b5e66a1 |
| Baseline | 0.77 | Not recorded | 698eaffec97a0b3e3b5e66a1 |

### Uni-Normalize Scheme

Honesty: 0.83. Result ID: 6995f3785698665ec42695f2.

[Reporting paper](<https://www.sota2.com/research/paper/698eaffec97a0b3e3b5e66a1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698eaffec97a0b3e3b5e66a1.md>)

Recorded evaluation configuration: {"Alignment Scheme":"Uni-Normalize"}. This record does not assert independent reproduction.

### Uni-Normalize Scheme + Like/Dislike Helpful RM

Honesty: 0.779. Result ID: 6995f3785698665ec42695f3.

[Reporting paper](<https://www.sota2.com/research/paper/698eaffec97a0b3e3b5e66a1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698eaffec97a0b3e3b5e66a1.md>)

Recorded evaluation configuration: {"Alignment Scheme":"Uni-Normalize + Like/Dislike Helpful RM"}. This record does not assert independent reproduction.

### Baseline

Honesty: 0.77. Result ID: 6995f3785698665ec42695f1.

[Reporting paper](<https://www.sota2.com/research/paper/698eaffec97a0b3e3b5e66a1.md>)

Recorded evaluation configuration: {"Alignment Scheme":"Baseline"}. This record does not assert independent reproduction.

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