# Harmful Compliance Analysis on Paired safety frame contrast items n=20 paired items per model (test)

Source: [Harmful Compliance Analysis on Paired safety frame contrast items n=20 paired items per model (test)](<https://www.sota2.com/research/sota/harmful-compliance-analysis-on-paired-safety-frame-contrast-items-n-20-paired-items-per-model-test>)

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

Data updated: 2026-05-08T11:01:57.697Z

Evaluation of model performance on a specific test set consisting of 20 paired safety frame contrast items, measuring harmful compliance using delta and p-value statistics.

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 | Harmful compliance analysis |
| Dataset (including any stated split) | Paired safety frame contrast items n=20 paired items per model (test) |
| Metric | Harm Delta (eval-dep) |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 5 of 5 |
| 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

- [Harm Delta (eval-dep)](<https://www.sota2.com/research/sota/harmful-compliance-analysis-on-paired-safety-frame-contrast-items-n-20-paired-items-per-model-test.md?metric=Harm%20Delta%20(eval-dep)>): Lower is better; unit: SCALAR.

- [Harm p-value (eval-dep)](<https://www.sota2.com/research/sota/harmful-compliance-analysis-on-paired-safety-frame-contrast-items-n-20-paired-items-per-model-test.md?metric=Harm%20p-value%20(eval-dep)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Harm Delta (eval-dep) | Result date | Report paper |
| --- | --- | --- | --- |
| OLMo-3-Instruct (7B) | -0.036 | Not recorded | 69fd6e555497b3b670c89695 |
| Llama-3.1-70B | 0 | Not recorded | 69fd6e555497b3b670c89695 |
| Phi-3.5-mini (3.8B) | 0.093 | Not recorded | 69fd6e555497b3b670c89695 |
| Mistral-Small-3.2 (24B) | 0.121 | Not recorded | 69fd6e555497b3b670c89695 |
| Llama-3.1-8B | 0.175 | Not recorded | 69fd6e555497b3b670c89695 |

### OLMo-3-Instruct (7B)

Harm Delta (eval-dep): -0.036. Result ID: 69fdc29947f34ce924e18ab1.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

Recorded evaluation configuration: {"Model family":"OLMo-3","Model size (parameters)":"7B","Protocol":"paired-prompt protocol"}. This record does not assert independent reproduction.

### Llama-3.1-70B

Harm Delta (eval-dep): 0. Result ID: 69fdc29947f34ce924e18ab5.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

Recorded evaluation configuration: {"Model family":"Llama","Model size (parameters)":"70B","Protocol":"paired-prompt protocol"}. This record does not assert independent reproduction.

### Phi-3.5-mini (3.8B)

Harm Delta (eval-dep): 0.093. Result ID: 69fdc29947f34ce924e18ab3.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

Recorded evaluation configuration: {"Model family":"Phi","Model size (parameters)":"3.8B","Protocol":"paired-prompt protocol"}. This record does not assert independent reproduction.

### Mistral-Small-3.2 (24B)

Harm Delta (eval-dep): 0.121. Result ID: 69fdc29947f34ce924e18ab2.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

Recorded evaluation configuration: {"Model family":"Mistral","Model size (parameters)":"24B","Protocol":"paired-prompt protocol"}. This record does not assert independent reproduction.

### Llama-3.1-8B

Harm Delta (eval-dep): 0.175. Result ID: 69fdc29947f34ce924e18ab4.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e555497b3b670c89695.md>)

Recorded evaluation configuration: {"Model family":"Llama","Model size (parameters)":"8B","Protocol":"paired-prompt protocol"}. This record does not assert independent reproduction.

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