When Scaling Fails: Mitigating Audio Perception Decay of LALMs via Multi-Step Perception-Aware Reasoning

• URL: http://arxiv.org/abs/2603.02266v1
• Date: Sat, 28 Feb 2026 12:54:51 GMT
• Title: When Scaling Fails: Mitigating Audio Perception Decay of LALMs via Multi-Step Perception-Aware Reasoning
• Abstract summary: Post-training models for structured reasoning trajectories results in marginal or even negative gains compared to post-training for direct answering.<n>We propose MPAR$2$, a paradigm that encourages dynamic perceptual reasoning and decomposes complex questions into perception-rich sub-problems.
• Score: 49.28390630057018
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Test-Time Scaling has shown notable efficacy in addressing complex problems through scaling inference compute. However, within Large Audio-Language Models (LALMs), an unintuitive phenomenon exists: post-training models for structured reasoning trajectories results in marginal or even negative gains compared to post-training for direct answering. To investigate it, we introduce CAFE, an evaluation framework designed to precisely quantify audio reasoning errors. Evaluation results reveal LALMs struggle with perception during reasoning and encounter a critical bottleneck: reasoning performance suffers from audio perception decay as reasoning length extends. To address it, we propose MPAR$^2$, a paradigm that encourages dynamic perceptual reasoning and decomposes complex questions into perception-rich sub-problems. Leveraging reinforcement learning, MPAR$^2$ improves perception performance on CAFE from 31.74% to 63.51% and effectively mitigates perception decay, concurrently enhancing reasoning capabilities to achieve a significant 74.59% accuracy on the MMAU benchmark. Further analysis demonstrates that MPAR$^2$ reinforces LALMs to attend to audio input and dynamically adapts reasoning budget to match task complexity.

Related papers

• A Closer Look at Failure Modes in Temporal Understanding of Large Audio-Language Models [64.68800440221138]
  Large Audio Language Models (LALMs) achieve strong performance on a variety of audio understanding tasks but struggle with temporal reasoning.<n>We introduce a benchmark with 1,657 questions across three foundational tasks designed for mechanistic analysis.<n>Comparing attention upweighting against scaling, we find that redistributing attention across audio tokens is more effective than increasing audio attention.
  arXiv  Detail & Related papers  (2026-06-16T01:57:56Z)
• Balancing Faithfulness and Performance in Reasoning via Multi-Listener Soft Execution [79.98699884805636]
  Reasoning Execution by Multiple Listeners (REMUL) is a multi-party reinforcement learning approach.<n>REMUL builds on the hypothesis that reasoning traces which other parties can follow will be more faithful.<n>Speakers are rewarded for producing reasoning that is clear to listeners.
  arXiv  Detail & Related papers  (2026-02-18T02:55:55Z)
• SEE: Signal Embedding Energy for Quantifying Noise Interference in Large Audio Language Models [49.313324100819955]
  Signal Embedding Energy (SEE) is a method for quantifying the impact of noise intensity on LALM inputs.<n>SEE exhibits a strong correlation with LALM performance, achieving a correlation of 0.98.<n>This paper introduces a novel metric for noise quantification in LALMs, providing guidance for robustness improvements in real-world deployments.
  arXiv  Detail & Related papers  (2026-01-12T08:57:55Z)
• Addressing Overthinking in Large Vision-Language Models via Gated Perception-Reasoning Optimization [56.59356959631999]
  Gated Perception-Reasoning Optimization (GPRO) is a meta-reasoning controller that dynamically routes computation among three decision paths.<n>GPRO substantially improves both accuracy and efficiency, outperforming recent slow-thinking methods.
  arXiv  Detail & Related papers  (2026-01-07T23:05:17Z)
• Incentivizing Consistent, Effective and Scalable Reasoning Capability in Audio LLMs via Reasoning Process Rewards [24.40159537923851]
  We develop a principled method for developing robust and scalable reasoning in Audio Large Language Models.<n>We achieve state-of-the-art results on MMAU Test-mini, substantially outperforming Gemini 2.5 Pro and GPT-4o Audio, and near-human-level performance on MMSU reasoning tasks.
  arXiv  Detail & Related papers  (2025-10-23T06:18:10Z)
• Unleashing Perception-Time Scaling to Multimodal Reasoning Models [60.578179197783754]
  Recent advances in inference-time scaling have substantially enhanced the reasoning capabilities of Large Vision-Language Models (LVLMs)<n>Inspired by this success, similar strategies have been applied to multimodal reasoning, yet their impact on visual perception remains unclear.<n>We propose Perception-Time Scaling (PTS), a novel paradigm that encourages token-rich perception and decomposes complex perception problems into intermediate tractable sub-problems.
  arXiv  Detail & Related papers  (2025-10-10T03:17:52Z)
• AU-Harness: An Open-Source Toolkit for Holistic Evaluation of Audio LLMs [8.918587474371321]
  Large Audio Language Models (LALMs) are rapidly advancing, but evaluating them remains challenging.<n>We introduce AU-Harness, an efficient and comprehensive evaluation framework for LALMs.<n>Our system achieves a speedup of up to 127% over existing toolkits through optimized batch processing and parallel execution.
  arXiv  Detail & Related papers  (2025-09-09T15:30:40Z)
• PixelThink: Towards Efficient Chain-of-Pixel Reasoning [70.32510083790069]
  PixelThink is a simple yet effective scheme that integrates externally estimated task difficulty and internally measured model uncertainty.<n>It learns to compress reasoning length in accordance with scene complexity and predictive confidence.<n> Experimental results demonstrate that the proposed approach improves both reasoning efficiency and overall segmentation performance.
  arXiv  Detail & Related papers  (2025-05-29T17:55:49Z)
• Audio-CoT: Exploring Chain-of-Thought Reasoning in Large Audio Language Model [26.20569269005708]
  Large Audio-Language Models (LALMs) have demonstrated remarkable performance in tasks involving audio perception and understanding.<n>However, their reasoning capabilities - critical for solving complex real-world problems - remain underexplored.<n>We conduct the first exploration into integrating Chain-of-Thought (CoT) reasoning into LALMs to enhance their reasoning ability across auditory modalities.
  arXiv  Detail & Related papers  (2025-01-13T11:54:40Z)
• On the Self-Verification Limitations of Large Language Models on Reasoning and Planning Tasks [17.329365493094542]
  We present a principled empirical study of the performance of GPT-4 in three domains: Game of 24, Graph Coloring, and STRIPS planning. We observe significant performance collapse with self-critique and significant performance gains with sound external verification.
  arXiv  Detail & Related papers  (2024-02-12T23:11:01Z)

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.