# Harmfulness Evaluation on Figstep-audio Harmful

Source: [Harmfulness Evaluation on Figstep-audio Harmful](<https://www.sota2.com/research/sota/harmfulness-evaluation-on-figstep-audio-harmful>)

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

Data updated: 2026-05-08T08:27:22.954Z

Evaluation of model performance on the Figstep-audio Harmful dataset for Harmfulness Evaluation, measured using metrics like ASR.

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 | Harmfulness Evaluation |
| Dataset (including any stated split) | Figstep-audio Harmful |
| Metric | ASR |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [ASR](<https://www.sota2.com/research/sota/harmfulness-evaluation-on-figstep-audio-harmful.md?metric=ASR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ASR | Result date | Report paper |
| --- | --- | --- | --- |
| MDSteer-c2r | 90.8 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| MDSteer-h2s | 84 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| MDSteer-h2s | 72.4 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| RRS | 52.4 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| No Defense | 51.6 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| No Defense | 51.6 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| MDSteer-c2r | 30.71 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| AdaShield | 30 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| FSD | 19.6 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| No Defense | 15.6 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| FSD | 12 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| SARSteer | 10.8 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| SARSteer | 10.8 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| SARSteer | 10 | Not recorded | 68fa7c5d1e209131cea8dc06 |
| AdaShield | 0 | Not recorded | 68fa7c5d1e209131cea8dc06 |

### MDSteer-c2r

ASR: 90.8. Result ID: 69fd9df447f34ce924e0901e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88272.md>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### MDSteer-h2s

ASR: 84. Result ID: 69fd9df447f34ce924e0901d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### MDSteer-h2s

ASR: 72.4. Result ID: 69fd9df447f34ce924e09023.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

### RRS

ASR: 52.4. Result ID: 69fd9df447f34ce924e09063.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c11e209131cea7a528.md>)

Recorded evaluation configuration: {"Defense runtime (s)":"1145"}. This record does not assert independent reproduction.

### No Defense

ASR: 51.6. Result ID: 69fd9df447f34ce924e0901a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### No Defense

ASR: 51.6. Result ID: 69fd9df447f34ce924e09062.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### MDSteer-c2r

ASR: 30.71. Result ID: 69fd9df447f34ce924e09024.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88272.md>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

### AdaShield

ASR: 30. Result ID: 69fd9df447f34ce924e0901b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727e1e209131cea56bc0.md>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### FSD

ASR: 19.6. Result ID: 69fd9df447f34ce924e09022.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76b01e209131cea6e7cf.md>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

### No Defense

ASR: 15.6. Result ID: 69fd9df447f34ce924e09020.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

### FSD

ASR: 12. Result ID: 69fd9df447f34ce924e0901c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76b01e209131cea6e7cf.md>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### SARSteer

ASR: 10.8. Result ID: 69fd9df447f34ce924e0901f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction.

### SARSteer

ASR: 10.8. Result ID: 69fd9df447f34ce924e09064.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: {"Defense runtime (s)":"266"}. This record does not assert independent reproduction.

### SARSteer

ASR: 10. Result ID: 69fd9df447f34ce924e09025.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

### AdaShield

ASR: 0. Result ID: 69fd9df447f34ce924e09021.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c5d1e209131cea8dc06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727e1e209131cea56bc0.md>)

Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction.

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