AuTAgent: A Reinforcement Learning Framework for Tool-Augmented Audio Reasoning

TL;DR

AuTAgent is a reinforcement learning framework that intelligently selects external audio tools to enhance large audio language models' reasoning, significantly improving accuracy on benchmark tasks.

Contribution

The paper introduces AuTAgent, a novel RL-based method for dynamic tool invocation in audio reasoning, addressing the challenge of integrating external tools effectively.

Findings

Improves accuracy by up to 9.80% on benchmarks.

Learns to filter irrelevant tools, reducing information overload.

Demonstrates strong transferability across models.

Abstract

Large Audio Language Models (LALMs) excel at perception but struggle with complex reasoning requiring precise acoustic measurements. While external tools can extract fine-grained features like exact tempo or pitch, effective integration remains challenging: naively using all tools causes information overload, while prompt-based selection fails to assess context-dependent utility. To address this, we propose AuTAgent (Audio Tool Agent), a reinforcement learning framework that learns when and which tools to invoke. By employing a sparse-feedback training strategy with a novel Differential Reward mechanism, the agent learns to filter out irrelevant tools and invokes external assistance only when it yields a net performance gain over the base model. Experimental results confirm that AuTAgent complements the representation bottleneck of LALMs by providing verifiable acoustic evidence. It improves accuracy by 4.20% / 6.20% and 9.80% / 8.00% for open-source and closed-source backbones on the MMAU Test-mini and the MMAR benchmarks, respectively. In addition, further experiments demonstrate exceptional transferability. We highlight the complementary role of external tools in augmenting audio model reasoning.