JusperLee/

AudioTrust

🎧 AudioTrust: Benchmarking the Multifaceted Trustworthiness of Audio Large Language Models

📜 Paper | 🎶 Demo | 💻 Code | 🤗 Dataset

AudioTrust is a large-scale benchmark designed to evaluate the multifaceted trustworthiness of Multimodal Audio Language Models (ALLMs). It examines model behavior across six critical dimensions:

	📊 Benchmark Tasks




	

Task Metric Description

Hallucination Detection Accuracy / Recall Groundedness of response in audio

Robustness… See the full description on the dataset page: https://huggingface.co/datasets/JusperLee/AudioTrust.