# Sentencing Bias Evidence Evaluation on 30 RJA Sentencing Bias Case Reports 1.0

Source: [Sentencing Bias Evidence Evaluation on 30 RJA Sentencing Bias Case Reports 1.0](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0>)

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

Data updated: 2026-03-06T08:00:23.683Z

Evaluation of model performance on assessing evidence related to sentencing bias using the 30 RJA Sentencing Bias Case Reports 1.0 dataset, utilizing specialized metrics like OR CI Score and Attribution Score.

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 | Sentencing Bias Evidence Evaluation |
| Dataset (including any stated split) | 30 RJA Sentencing Bias Case Reports 1.0 |
| Metric | OR CI Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [OR CI Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=OR%20CI%20Score>): Higher is better; unit: SCALAR.

- [RR CI Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=RR%20CI%20Score>): Higher is better; unit: SCALAR.

- [OR Sample Size Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=OR%20Sample%20Size%20Score>): Higher is better; unit: SCALAR.

- [RR Sample Size Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=RR%20Sample%20Size%20Score>): Higher is better; unit: SCALAR.

- [Cross-Method Comparison Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=Cross-Method%20Comparison%20Score>): Higher is better; unit: SCALAR.

- [Chi-square Sample Size Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=Chi-square%20Sample%20Size%20Score>): Higher is better; unit: SCALAR.

- [p-Value Contextualization Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=p-Value%20Contextualization%20Score>): Higher is better; unit: SCALAR.

- [Limitations Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=Limitations%20Score>): Higher is better; unit: SCALAR.

- [Attribution Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=Attribution%20Score>): Higher is better; unit: SCALAR.

- [Overall Score](<https://www.sota2.com/research/sota/sentencing-bias-evidence-evaluation-on-30-rja-sentencing-bias-case-reports-1-0.md?metric=Overall%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | OR CI Score | Result date | Report paper |
| --- | --- | --- | --- |
| LLM-powered interpretive layer | 0.8 | Not recorded | 69aa5f6b2d603e6de3e41781 |
| LLM-powered interpretive layer | 0.79 | Not recorded | 69aa5f6b2d603e6de3e41781 |

### LLM-powered interpretive layer

OR CI Score: 0.8. Result ID: 69aa89956dbb7075b9541dd3.

[Reporting paper](<https://www.sota2.com/research/paper/69aa5f6b2d603e6de3e41781.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69aa5f6b2d603e6de3e41781.md>)

Recorded evaluation configuration: {"Evaluator Type":"Human (H)"}. This record does not assert independent reproduction.

### LLM-powered interpretive layer

OR CI Score: 0.79. Result ID: 69aa89956dbb7075b9541dd2.

[Reporting paper](<https://www.sota2.com/research/paper/69aa5f6b2d603e6de3e41781.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69aa5f6b2d603e6de3e41781.md>)

Recorded evaluation configuration: {"Evaluator Type":"Avg. Machine (L)","Context":"15 evaluation runs per report"}. This record does not assert independent reproduction.

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