All That Glitters Is Not Audio: Rethinking Text Priors and Audio Reliance in Audio-Language Evaluation

• URL: http://arxiv.org/abs/2604.24401v1
• Date: Mon, 27 Apr 2026 12:25:18 GMT
• Title: All That Glitters Is Not Audio: Rethinking Text Priors and Audio Reliance in Audio-Language Evaluation
• Abstract summary: Large Audio-Language Models show consistent performance gains across speech and audio benchmarks, yet high scores may not reflect true auditory perception.<n>We present a diagnostic framework using two axes: text prior, which measures answerability from text and general knowledge alone, and audio reliance, which assesses actual dependency on the acoustic signal.
• Score: 45.45465533352999
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Audio-Language Models show consistent performance gains across speech and audio benchmarks, yet high scores may not reflect true auditory perception. If a model can answer questions without processing the acoustic signal, the benchmark fails as a measure of auditory understanding. We present a diagnostic framework using two axes: text prior, which measures answerability from text and general knowledge alone, and audio reliance, which assesses actual dependency on the acoustic signal. Evaluating eight LALMs across three benchmarks, we find that models retain 60-72% of their full audio scores even without any audio input. Moreover, among items that require audio, only 3.0-4.2% need the complete audio clip; the majority can be resolved using localized fragments. These findings challenge the assumption that benchmark performance equals robust audio understanding, and we conclude with practical guidelines for improving evaluation reliability and benchmark design.

Related papers

• PRISM-Bench: An Audio-Centric Diagnostic Benchmark for Text-to-Audio-Video Generation [23.127789678422698]
  Text-to-audio-video (T2AV) generation has advanced rapidly, but its evaluation still underestimates the audio modality.<n>We present PRISM-Bench, the first audio-centric diagnostic benchmark for T2AV generation.
  arXiv  Detail & Related papers  (2026-09-04T08:27:10Z)
• Improving Text-to-Audio Instruction Following via Fine-Grained Feedback from Audio-Aware Large Language Models [51.75139461584678]
  Recent text-to-audio models generate high-quality audio, but often fail to follow instructions involving multiple sound events and temporal order.<n>We propose an instruction-level framework that uses audio-aware large language models as fine-grained judges to verify target event presence and temporal relations.<n>Experiments show that our method improves event completeness, temporal ordering, and joint instruction-following accuracy across existing benchmarks and S3Bench, while maintaining audio quality.
  arXiv  Detail & Related papers  (2026-07-15T03:13:06Z)
• HalluAudio: A Comprehensive Benchmark for Hallucination Detection in Large Audio-Language Models [30.18524844766061]
  We introduce HalluAudio, the first large-scale benchmark for evaluating hallucinations across speech, environmental sound, and music.<n>HalluAudio comprises over 5K human-verified QA pairs and spans diverse task types, including binary judgments, multi-choice reasoning, attribute verification, and open-ended QA.<n>Our results reveal significant deficiencies in acoustic grounding, temporal reasoning, and music attribute understanding, underscoring the need for reliable and robust LALMs.
  arXiv  Detail & Related papers  (2026-04-21T10:05:28Z)
• How Auditory Knowledge in LLM Backbones Shapes Audio Language Models: A Holistic Evaluation [97.0235251827591]
  Large language models (LLMs) have been widely used as knowledge backbones of Large Audio Language Models (LALMs)<n>We study how much auditory knowledge they encode through text-only pre-training and how this affects downstream performance.<n>Our findings reveal that auditory knowledge varies substantially across families, and text-only results are strongly correlated with audio performance.
  arXiv  Detail & Related papers  (2026-03-19T17:50:07Z)
• SCENEBench: An Audio Understanding Benchmark Grounded in Assistive and Industrial Use Cases [27.340743922132067]
  SCENEBench targets a broad form of audio comprehension across four real-world categories: background sound understanding, noise localization, cross-linguistic speech understanding, and vocal characterizer recognition.<n>The purpose of this benchmark suite is to assess audio beyond just what words are said - rather, how they are said and the non-speech components of the audio.<n>We assess five state-of-the-art LALMs and find critical gaps: performance varies across tasks, with some tasks performing below random chance and others achieving high accuracy.
  arXiv  Detail & Related papers  (2026-03-10T16:15:12Z)
• SAM Audio Judge: A Unified Multimodal Framework for Perceptual Evaluation of Audio Separation [52.468945848774844]
  This paper addresses the need for automated systems capable of evaluating audio separation without human intervention.<n>The proposed evaluation metric, SAM Audio Judge (SAJ), is a multimodal fine-grained reference-free objective metric.<n>SAJ supports three audio domains (speech, music and general sound events) and three prompt inputs (text, visual and span), covering four different dimensions of evaluation.
  arXiv  Detail & Related papers  (2026-01-27T15:29:02Z)
• UltraEval-Audio: A Unified Framework for Comprehensive Evaluation of Audio Foundation Models [36.71750531005594]
  UltraEval-Audio is a unified evaluation framework for audio foundation models.<n>It supports 10 languages and 14 core task categories, while seamlessly integrating 24 mainstream models and 36 authoritative benchmarks.<n>It adopts a novel comprehensive evaluation scheme for audio codecs, evaluating performance across three key dimensions.
  arXiv  Detail & Related papers  (2026-01-04T04:54:12Z)
• AudioMarathon: A Comprehensive Benchmark for Long-Context Audio Understanding and Efficiency in Audio LLMs [53.248502396225724]
  AudioMarathon is a benchmark designed to evaluate both understanding and inference efficiency on long-form audio.<n>We evaluate state-of-the-art LALMs and observe clear performance drops as audio length grows.<n>The results show large gaps across current LALMs and highlight the need for better temporal reasoning.
  arXiv  Detail & Related papers  (2025-10-08T17:50:16Z)
• AHELM: A Holistic Evaluation of Audio-Language Models [78.20477815156484]
  multimodal audio-language models (ALMs) take interleaved audio and text as input and output text.<n>AHELM is a benchmark that aggregates various datasets -- including 2 new synthetic audio-text datasets called PARADE and CoRe-Bench.<n>We also standardize the prompts, inference parameters, and evaluation metrics to ensure equitable comparisons across models.
  arXiv  Detail & Related papers  (2025-08-29T07:40:39Z)
• AudioBench: A Universal Benchmark for Audio Large Language Models [41.46064884020139]
  We introduce AudioBench, a universal benchmark designed to evaluate Audio Large Language Models (AudioLLMs)<n>It encompasses 8 distinct tasks and 26 datasets, among which, 7 are newly proposed datasets.<n>The evaluation targets three main aspects: speech understanding, audio scene understanding, and voice understanding (paralinguistic)
  arXiv  Detail & Related papers  (2024-06-23T05:40:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.