# Helpfulness Evaluation on Figstep-audio Harmful-Safe Source: [Helpfulness Evaluation on Figstep-audio Harmful-Safe]() Collection index: [llms.txt]() Data updated: 2026-05-08T08:27:23.117Z Evaluation of model helpfulness performance using the Figstep-audio Harmful-Safe dataset, measured by the BRR metric. 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 | Helpfulness Evaluation | | Dataset (including any stated split) | Figstep-audio Harmful-Safe | | Metric | BRR | | 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 - [BRR](): Higher is better; unit: PERCENT. ## Recorded results | Method | BRR | Result date | Report paper | | --- | --- | --- | --- | | SARSteer | 88.8 | Not recorded | 68fa7c5d1e209131cea8dc06 | | SARSteer | 79.95 | Not recorded | 68fa7c5d1e209131cea8dc06 | | SARSteer | 79.95 | Not recorded | 68fa7c5d1e209131cea8dc06 | | MDSteer-c2r | 79.68 | Not recorded | 68fa7c5d1e209131cea8dc06 | | No Defense | 70.2 | Not recorded | 68fa7c5d1e209131cea8dc06 | | No Defense | 70.2 | Not recorded | 68fa7c5d1e209131cea8dc06 | | RRS | 70 | Not recorded | 68fa7c5d1e209131cea8dc06 | | AdaShield | 69.8 | Not recorded | 68fa7c5d1e209131cea8dc06 | | MDSteer-h2s | 68.8 | Not recorded | 68fa7c5d1e209131cea8dc06 | | FSD | 63.2 | Not recorded | 68fa7c5d1e209131cea8dc06 | | No Defense | 61.4 | Not recorded | 68fa7c5d1e209131cea8dc06 | | FSD | 61.2 | Not recorded | 68fa7c5d1e209131cea8dc06 | | MDSteer-h2s | 60.8 | Not recorded | 68fa7c5d1e209131cea8dc06 | | MDSteer-c2r | 54.2 | Not recorded | 68fa7c5d1e209131cea8dc06 | | AdaShield | 52.6 | Not recorded | 68fa7c5d1e209131cea8dc06 | ### SARSteer BRR: 88.8. Result ID: 69fd9df447f34ce924e09055. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction. ### SARSteer BRR: 79.95. Result ID: 69fd9df447f34ce924e0904f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### SARSteer BRR: 79.95. Result ID: 69fd9df447f34ce924e09070. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense runtime (s)":"266"}. This record does not assert independent reproduction. ### MDSteer-c2r BRR: 79.68. Result ID: 69fd9df447f34ce924e09054. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction. ### No Defense BRR: 70.2. Result ID: 69fd9df447f34ce924e0904a. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### No Defense BRR: 70.2. Result ID: 69fd9df447f34ce924e0906e. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### RRS BRR: 70. Result ID: 69fd9df447f34ce924e0906f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense runtime (s)":"1145"}. This record does not assert independent reproduction. ### AdaShield BRR: 69.8. Result ID: 69fd9df447f34ce924e0904b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### MDSteer-h2s BRR: 68.8. Result ID: 69fd9df447f34ce924e09053. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction. ### FSD BRR: 63.2. Result ID: 69fd9df447f34ce924e0904c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### No Defense BRR: 61.4. Result ID: 69fd9df447f34ce924e09050. [Reporting paper]() Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction. ### FSD BRR: 61.2. Result ID: 69fd9df447f34ce924e09052. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Kimi-Audio"}. This record does not assert independent reproduction. ### MDSteer-h2s BRR: 60.8. Result ID: 69fd9df447f34ce924e0904d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### MDSteer-c2r BRR: 54.2. Result ID: 69fd9df447f34ce924e0904e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2-Audio"}. This record does not assert independent reproduction. ### AdaShield BRR: 52.6. Result ID: 69fd9df447f34ce924e09051. [Reporting paper]() [Method origin paper]() 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.