# Audio Hallucination Evaluation on Clotho-1K Source: [Audio Hallucination Evaluation on Clotho-1K]() Collection index: [llms.txt]() Data updated: 2026-04-13T08:39:55.031Z Evaluation of model performance on audio hallucination detection using the Clotho-1K dataset, measured across metrics like HR, Acoustic Attribute, Source, Prior-Driven, and Fabricated errors. 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 | Audio Hallucination Evaluation | | Dataset (including any stated split) | Clotho-1K | | Metric | HR | | Unit | PERCENT | | Direction | Lower is better | | Results loaded | 14 of 14 | | 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 - [HR](): Lower is better; unit: PERCENT. - [Acoustic Attribute Error](): Lower is better; unit: PERCENT. - [Source Error](): Lower is better; unit: PERCENT. - [Prior-Driven Error](): Lower is better; unit: PERCENT. - [Fabricated Error](): Lower is better; unit: PERCENT. ## Recorded results | Method | HR | Result date | Report paper | | --- | --- | --- | --- | | Qwen2.5-Omni-7B | 16.98 | Not recorded | 69dc78617d0172140ec7891c | | Gemini-2.5-pro | 19.42 | Not recorded | 69dc78617d0172140ec7891c | | GPT-audio | 21.81 | Not recorded | 69dc78617d0172140ec7891c | | Audio-Flamingo-3 | 23.72 | Not recorded | 69dc78617d0172140ec7891c | | Qwen2-Audio-7B-Instruct | 24.25 | Not recorded | 69dc78617d0172140ec7891c | | StepAudio2 | 25.55 | Not recorded | 69dc78617d0172140ec7891c | | Qwen-audio-chat | 25.93 | Not recorded | 69dc78617d0172140ec7891c | | MiMo-Audio-7B-Instruct | 26.43 | Not recorded | 69dc78617d0172140ec7891c | | Qwen2.5-Omni-7B | 26.53 | Not recorded | 69dc78617d0172140ec7891c | | Gemini-2.5-flash | 26.85 | Not recorded | 69dc78617d0172140ec7891c | | Qwen2.5-Omni-3B | 29.75 | Not recorded | 69dc78617d0172140ec7891c | | Kimi-Audio-7B-Instruct | 36.48 | Not recorded | 69dc78617d0172140ec7891c | | GPT-audio-mini | 37.69 | Not recorded | 69dc78617d0172140ec7891c | | SALMONN-7B | 40.5 | Not recorded | 69dc78617d0172140ec7891c | ### Qwen2.5-Omni-7B HR: 16.98. Result ID: 69dcaa6647f34ce924c3a06e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"NAICL":"true"}. This record does not assert independent reproduction. ### Gemini-2.5-pro HR: 19.42. Result ID: 69dcaa6647f34ce924c3a06a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-audio HR: 21.81. Result ID: 69dcaa6647f34ce924c3a06c. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Audio-Flamingo-3 HR: 23.72. Result ID: 69dcaa6647f34ce924c3a067. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2-Audio-7B-Instruct HR: 24.25. Result ID: 69dcaa6647f34ce924c3a064. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### StepAudio2 HR: 25.55. Result ID: 69dcaa6647f34ce924c3a062. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen-audio-chat HR: 25.93. Result ID: 69dcaa6647f34ce924c3a063. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MiMo-Audio-7B-Instruct HR: 26.43. Result ID: 69dcaa6647f34ce924c3a065. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-Omni-7B HR: 26.53. Result ID: 69dcaa6647f34ce924c3a06d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-2.5-flash HR: 26.85. Result ID: 69dcaa6647f34ce924c3a069. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-Omni-3B HR: 29.75. Result ID: 69dcaa6647f34ce924c3a061. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Kimi-Audio-7B-Instruct HR: 36.48. Result ID: 69dcaa6647f34ce924c3a066. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-audio-mini HR: 37.69. Result ID: 69dcaa6647f34ce924c3a06b. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### SALMONN-7B HR: 40.5. Result ID: 69dcaa6647f34ce924c3a068. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.